Routledge Handbook of Public Policy

Routledge Handbook of

Public Policy

This Handbook provides a comprehensive global survey of the policy process. Written by an

outstanding line-up of distinguished scholars and practitioners, the Handbook covers all aspects of

the policy process including:

 theory – from rational choice to the new institutionalism

 frameworks – network theory, the advocacy coalition framework, punctuated equilibrium

models and institutional analysis and development theory

 key stages in the process – agenda-setting, formulation, decision-making, implementation

and evaluation

 the roles of key actors and institutions

 policy learning and policy dynamics from path dependency to process sequencing

This is an invaluable resource for all scholars, graduate students and practitioners in public policy

and policy analysis.

Eduardo Araral Jr. is Assistant Professor at the Lee Kuan Yew School of Public Policy,

National University of Singapore.

Scott Fritzen is Associate Professor at the Lee Kuan Yew School of Public Policy, National

University of Singapore.

Michael Howlett is Burnaby Mountain Chair in the Department of Political Science at Simon

Fraser University, Vancouver, Canada.

M Ramesh is Chair Professor of Govenance and Public Policy at the Hong Kong Institute of

Education and Visiting Professor at the Lee Kuan Yew School of Public Policy, National

University of Singapore.

Xun Wu is Associate Professor at the Lee Kuan Yew School of Public Policy, National University

of Singapore.

Routledge Handbook of

Public Policy

Edited by

Eduardo Araral Jr.

Scott Fritzen

Michael Howlett

M Ramesh

Xun Wu

First published 2013

by Routledge

2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN

Simultaneously published in the USA and Canada

by Routledge

711 Third Avenue, New York, NY 10017

Routledge is an imprint of the Taylor & Francis Group, an informa business

Wu; individual chapters, the contributors

The right of E. Araral Jr., S. Fritzen, M. Howlett, M. Ramesh, X. Wu to be identiﬁed as

editors of this work has been asserted by them in accordance with the Copyright, Designs and

Patent Act 1988.

form or by any electronic, mechanical, or other means, now known or hereafter invented,

including photocopying and recording, or in any information storage or retrieval system,

without permission in writing from the publishers.

Trademark notice: Product or corporate names may be trademarks or registered trademarks, and

are used only for identiﬁcation and explanation without intent to infringe.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data

Routledge Handbook of Public Policy / edited by E. Araral ... [et al.].

p. cm.

Includes bibliographical references and index.

1. Policy sciences. I. Araral, Eduardo II. Title: Handbook of Public Policy.

H97.R68 2013

320.6–dc23

2012008846

ISBN: 978-0-415-78245-6 (hbk)

ISBN: 978-0-203-09757-1 (ebk)

Typeset in Bembo

by Taylor & Francis Books

Contents

List of illustrations ix

List of contributors xi

PART I

Introduction to the study of the public policy process: history

and method 1

1 Public policy debate and the rise of policy analysis

Michael Mintrom and Claire Williams

2 The policy-making process

Michael Howlett and Sarah Giest

3 Comparative approaches to the study of public policy-making

Sophie Schmitt

4 International dimensions and dynamics of policy-making

Anthony Perl

PART II

Conceptualizing public policy-making 57

5 State theory and the rise of the regulatory state

Darryl S.L. Jarvis

6 The public choice perspective

Andy Whitford

7 Institutional analysis and political economy

Michael D. McGinnis and Paul Dragos Aligica

8 Postpositivism and the policy process

Raul Perez Lejano

PART III

Modelling the policy process: frameworks for analysis 113

9 The institutional analysis and development framework

115

Ruth Schuyler House and Eduardo Araral Jr.

10 The advocacy coalition framework: coalitions, learning and policy change

125

Christopher M. Weible and Daniel Nohrstedt

11 The punctuated equilibrium theory of agenda-setting and policy change

138

Graeme Boushey

12 Policy network models

153

Chen-Yu Wu and David Knoke

PART IV

Understanding the agenda-setting process 165

13 Policy agenda-setting studies: attention, politics and the public

167

Christoﬀer Green-Pedersen and Peter B. Mortensen

14 Focusing events and policy windows

175

Thomas A. Birkland and Sarah E. DeYoung

15 Agenda-setting and political discourse: major analytical frameworks and

their application

189

David A. Rochefort and Kevin P. Donnelly

16 Mass media and policy-making

204

Stuart Soroka, Stephen Farnsworth, Andrea Lawlor and Lori Young

PART V

Understanding the formulation process 215

17 Policy design and transfer

217

Anne Schneider

18 Epistemic communities

229

Claire A. Dunlop

19 Policy appraisal

244

John Turnpenny, Camilla Adelle and Andrew Jordan

Contents

20 Policy analytical styles 255

Igor S. Mayer, C. Els van Daalen and Pieter W.G. Bots

PART VI

Understanding the decision-making process 271

21 Bounded rationality and public policy decision-making

273

Bryan D. Jones and H.F. Thomas III

22 Incrementalism

287

Michael Hayes

23 Models for research into decision-making processes: on phases, streams,

rounds and tracks of decision-making

299

Geert R. Teisman and Arwin van Buuren

24 The garbage can model and the study of the policy-making process

320

Gary Mucciaroni

PART VII

Understanding the implementation process 329

25 Bureaucracy and the policy process

331

Ora-orn Poocharoen

26 Disagreement and alternative dispute resolution in the policy process

347

Boyd Fuller

27 Governance, networks and intergovernmental systems

361

Robert Agranoﬀ, Michael McGuire and Chris Silvia

28 Development management and policy implementation: relevance beyond

the global South

374

Derick W. Brinkerhoﬀ and Jennifer M. Brinkerhoﬀ

PART VIII

Understanding the evaluation process 385

29 Six models of evaluation

387

Evert Vedung

30 Policy feedback and learning

401

Patrik Marier

Contents

vii

31 Randomized control trials: what are they, why are they promoted as the

gold standard for casual identiﬁcation and what can they (not) tell us?

415

David Fuente and Dale Whittington

32 Policy evaluation and public participation

434

Carolyn M. Hendriks

PART IX

Policy dynamics: patterns of stability and change 449

33 Policy dynamics and change: the never-ending puzzle

451

Giliberto Capano

34 Policy trajectories and legacies: path dependency revisited

462

Adrian Kay

35 Process sequencing

473

Carsten Daugbjerg

36 Learning from success and failure?

484

Allan McConnell

Author index

495

Subject index 510

Contents

viii

Illustrations

Figures

9.1 A framework for institutional analysis 118

10.1 Flow diagram of the advocacy coalition framework 128

11.1 Incremental and punctuated policy change distributions 148

14.1 Comparative attention to terrorism 180

14.2 Dominant problem frames in school violence, by venue 182

16.1 Coverage of natural disasters/weather and pollution/climate change by year 210

18.1 Bibliographic analysis of 638 articles, book chapters and books citing Haas’s

1992 introductory article to the IO special edition 231

18.2 Bibliographic analysis of 638 articles, book chapters and books citing Haas’s

1992 introductory article to the IO special edition (subject areas receiving 11

citations or more) 232

20.1 Policy analysis tasks 258

20.2 Policy analysis styles 259

20.3 The underlying values and criteria of policy analysis 262

20.4 Overview of the complete hexagon model of policy analysis 266

23.1 A depiction of four models for the analysis of decision-making processes 302

23.2 The concept of decision-making used in the phase model 304

23.3 The concept of decision-making used in the stream model 305

23.4 The concept of decision-making used in the rounds model 306

23.5 The concept of decision-making used in the tracks model 307

29.1 Goal-attainment model 389

29.2 Side-eﬀects model with speciﬁed pigeonholes for side-eﬀects 391

29.3 Potential stakeholders in local social welfare interventions 395

Tables

2.1 A taxonomy of substantive policy instruments 22

2.2 A resource-based taxonomy of procedural policy instruments 23

3.1 Subdisciplines of comparative public policy research 38

4.1 Supranational governance dynamics 52

5.1 General approaches to political phenomena and illustrative theoretical examples 60

5.2 A comparative typology of the interventionist and regulatory state 68

14.1 The substance of the agenda when events are or are not mentioned in testimony 181

14.2 Key issues addressed in aviation security legislation 183

16.1 Coverage of natural disasters/weather and pollution/climate change 211

20.1 Positive and negative images of the policy analyst 265

22.1 Four quadrants of policy change 289

22.2 A typology of decision environments 291

23.1 Comparative perspectives on the phase model, the stream model and the

rounds model 308

23.2 Events in decision-making on the Betuwe line and a fourfold research result 314

28.1 Competing perspectives on development management 378

36.1 The three realms of policy success 488

Box

20.1 Translation of values into quality criteria 264

List of illustrations

Contributors

Camilla Adelle is a Senior Research Associate in the Science, Society and Sustainability Research

Group, School of Environmental Sciences, University of East Anglia, Norwich, UK.

Robert Agranoﬀ is an Emeritus Professor in the School of Public and Environmental Aﬀairs,

Indiana University, Bloomington, USA and Catedrático in the Government and Public

Administration Program, Instituto Universitario Ortega y Gasset, Madrid, Spain. He joined

Indiana University in 1980 from Northern Illinois University and started at Ortega y Gasset in

1990. He continues to be professionally active, specializing in public administration, inter-

governmental relations and management, and public network studies.

Paul Dragos Aligica is a Senior Research Fellow at the Mercatus Center, a Senior Fellow on

the F.A. Hayek Program for Advanced Study in Philosophy, Politics and Economics, a Faculty

Fellow at the James Buchanan Center for Political Economy at George Mason University,

USA, and an Adjunct Fellow at the Hudson Institute, Washington DC, USA. He studies

institutional analysis and institutional theory.

Eduardo Araral Jr. is an Assistant Professor at the Lee Kuan Yew School of Public

Policy, National University of Singapore. He is interested in the study of institutions for

collective action from local to international levels and covering a variety of areas from the

commons, international security, climate adaptation, bureaucracy and aid and water resources

management.

Thomas A. Birkland is William T. Kretzer Distinguished Professor at the School of

International and Public Aﬀairs, North Carolina State University, USA where he teaches

courses on the public policy process, environmental politics and policy, and disaster policy

and management. His research has focused on how and to what extent major ‘focusing

events’ inﬂuence policy agendas. His current research seeks to connect focusing events and

policy failure to policy change and learning. He is the author of An Introduction to the Policy

Process (2nd edn, M.E. Sharpe), and After Disaster and Lessons of Disaster (both Georgetown

University Press), and is co-editor with Todd Schaefer of The Encyclopedia of Mass Media

(CQ Press).

Pieter Bots is an Associate Professor in the Faculty of Technology, Policy and Management

(TPM), Delft University of Technology, Netherlands. His expertise lies mainly in conceptual

models and computer-based tools to support problem formulation and multi-actor systems

analysis, preferably in combination with serious gaming/simulation.

Graeme Boushey is an Assistant Professor in the Department of Political Science at the Uni-

versity of California, Irvine, USA, where he teaches courses in American politics, Californian

and state politics, public policy, and research methodology. His research focuses on public

policy processes and political decision-making in American federalism. His teaching and research

are organized around practical and theoretical questions of state and federal policy-making.

Derick W. Brinkerho ﬀ is a Distinguished Fellow in International Public Management, at the

Research Triangle Institute and has more than 30 years of experience with public management

issues in developing and transitioning countries, focusing on policy analysis, programme imple-

mentation and evaluation, participation, institutional development, democratic governance, and

change management. He is co-editor of the journal Public Administration and Development, and

serves on the editorial boards of Public Administration Review and International Review of Adminis-

trative Sciences. He also holds an associate faculty appointment at the Trachtenberg School of

Public Policy and Public Administration, George Washington University, USA.

Jennifer M. Brinkerhoﬀ is Professor of Public Administration and International Aﬀairs and

Co-Director, GW Diaspora Program at the Elliott School of International Aﬀairs at George

Washington University, USA. She teaches courses on public service, international development

policy and administration, development management and organizational behaviour. She is par-

ticularly keen on encouraging people to pursue service careers thoughtfully, grounding their

commitment to change in self-awareness and working in community. She is the author

of Digital Diasporas: Identity and Transnational Engagement (Cambridge University Press, 2009)

and Partnership for International Development: Rhetoric or Results? (Lynne Rienner, 2002); the

editor of Diasporas and Development: Exploring the Potential (Lynne Rienner, 2008); and co-editor

of NGOs and the Millennium Development Goals: Citizen Action to Reduce Poverty (Palgrave

Macmillan, 2007).

Giliberto Capano is Professor of Political Science and Public Policy at the University of

Bologna, Italy. He specializes in comparative higher education, public administration and the-

ories of policy-making. He has recently edited (with Elisabetta Gualmini) La Pubblica amminis-

trazione in Italy (Il Mulino, 2011) and European and North American Policy Change (Routledge,

2009) with Michael Howlett.

Carsten Daugbjerg is a Professor at the Institute of Food and Resource Economics, University

of Copenhagen, Denmark. His research area is comparative public policy with a particular

interest in historical institutitionalism, policy networks, policy instruments and ideas in public

policy. His empirical research has focused on agricultural policy reform, trade negotiations in

the World Trade Organization, global food regulation, government interest group relations and

environmental policy.

Sarah E. DeYoung is a doctoral student at North Carolina State University, USA, in the

Psychology in the Public Interest programme. Her research interests are disaster policy and

community disaster preparedness.

Kevin P. Donnelly is Assistant Professor of Political Science and Public Administration at

Bridgewater State University, USA. His previous work includes Foreign Remedies: What the

Experience of Other Nations Can Tell Us about Next Steps in Reforming U.S. Health Care (2012),

co-authored with David A. Rochefort.

List of contributors

xii

Claire A. Dunlop teaches politics and administration at the University of Exeter, UK. Pre-

viously she worked for the Scottish Consumer Council and the Scottish Executive. She has worked

and published widely on issues related to knowledge use in governments and policy appraisal.

Her recent work has appeared in Political Studies, Public Management Review and Regulation and

Governance.

Stephen J. Farnsworth is Professor of Political Science and International Aﬀairs at the George

Washington University, USA, where he directs the university’s Center for Leadership and

Media Studies. He is the author or co-author of four books on media and politics and a former

Fulbright Scholar at McGill University, Canada.

David Fuente is a PhD student in the Department of City and Regional Planning at the

University of North Carolina at Chapel Hill, USA.

Boyd Fuller is an Assistant Professor in the Lee Kuan Yew School of Public Policy at the

National University of Singapore. His research focuses on the resolution of diﬃcult water,

environmental and land conﬂicts when stakeholders have apparently irreconcilable diﬀerences in

values, identities or culture. He is currently researching the use of traditional and innovative

dispute resolution techniques for public disputes in post-conﬂict areas of Southeast Asia. His

previous research examined the mediation of intractable environmental conﬂicts in the United

States. He has eight years’ experience designing and implementing water supply projects in

developing countries.

Sarah Giest is a PhD student in the Department of Political Science at Simon Fraser University

in Burnaby, BC, Canada. She works on innovation systems and networks with special reference to

comparative biotechnology clusters.

Christoﬀer Green-Pedersen is Professor of Political Science at Aarhus University, Denmark.

His research focuses on agenda-setting dynamics and party competition.

Michael Hayes is Professor of Political Science at Colgate University, USA, specializing in

interest group theory, incrementalism and policymaking, eﬀects of public opinion on the

legislative process. He has taught at Lawrence University and Rutgers University and pub-

lished Incrementalism and Public Policy; Lobbyists and Legislators; various book chapters and

articles (Journal of Politics, Polity) on the inﬂuence of interest groups and public opinion on

policy-making.

Carolyn M. Hendriks’s work examines the democratic practices of contemporary governance,

particularly with respect to public deliberation, inclusion and political representation. She

has taught and published widely on the application and politics of inclusive and deliberative

forms of citizen engagement. With a background in both political science and environmental

engineering, Carolyn has a particular interest in the governance of the environment, as well as

science and technology issues. She has recently published a book on the practice of delib-

erative democracy entitled The Politics of Public Deliberation: Citizen Engagement and Interest

Advocacy (Palgrave Macmillan, 2011).

Ruth Schuyler House is a PhD student in the Lee Kuan Yew School of Public Policy at the

National University of Singapore. Her research concentrates on the ﬁelds of water and

List of contributors

xiii

sanitation policy. She is also interested in the changing face of governance and emerging

opportunities for the private sector to assume responsibility for and lend valuable knowledge

and resources to addressing poverty alleviation.

Michael Howlett is Burnaby Mountain Professor in the Department of Political Science at

Simon Fraser University in Vancouver, Canada. He has published widely, most recently

Designing Public Policies: Principles and Instruments (Routledge, 2011) and is co-editing with David

Laycock Regulating Next Generation Agri-Food Biotechnologies, Lessons from European, North Amer-

ican and Asian Experiences (Routledge, 2012)

Darryl S.L. Jarvis is Vice-Dean of Academic Aﬀairs and Associate Professor at the Lee Kuan

Yew School of Public Policy, National University of Singapore and specializes in risk analysis

and the study of political and economic risk in Asia, including investment, regulatory and

institutional risk analysis. He has been a consultant to various government bodies and business

organizations, and for two years was a member of the investigating team and then chief

researcher on the Building Institutional Capacity in Asia (BICA) project commissioned by the

Japanese Ministry of Finance.

Bryan D. Jones’s research interests centre on the study of public policy processes, American

governing institutions, and the connection between human decision-making and organizational

behaviour. Before joining the Department of Government at the University of Texas, Austin,

USA, in 2008, Professor Jones was the Donald R. Matthews Distinguished Professor of Amer-

ican Politics at the University of Washington. Previously, he was Distinguished Professor and head

of department at Texas A&M University, and also taught at Wayne State University. His books

include Politics and the Architecture of Choice (2001) and Reconceiving Decision-Making in Democratic

Politics (1994), both winners of the APSA Political Psychology Section Robert Lane Award; The

Politics of Attention (co-authored with Frank Baumgartner, 2005); Agendas and Instability in

American Politics (co-authored with Frank Baumgartner, 1993), winner of the 2001 Aaron

Wildavsky Award for Enduring Contribution to the Study of Public Policy of the American

Political Science Association’s Public Policy Section and The Politics of Bad Ideas (co-authored

with Walt Williams).

Andrew Jordan is Professor of Environmental Politics at the University of East Anglia, Nor-

wich, UK. He is interested in the governance of environmental problems in many diﬀerent

contexts, but especially that of the European Union.

Adrian Kay is an Associate Professor in the Crawford School at the Australian National Uni-

versity and Director of its Policy and Governance Program. His current research interests are in

the relationships between international and domestic policy processes, with a particular empiri-

cal focus on health and agriculture.

David Knoke is a Professor in the Department of Sociology at University of Minnesota, USA,

where he teaches a graduate network analysis seminar that attracts students from diverse dis-

ciplines. He co-authored Network Analysis (1982) and has published 15 books and more than

100 articles and book chapters, primarily on organizations, networks, politics and social statistics.

He was principal co-investigator on several National Science Foundation-funded projects on

voluntary associations, lobbying organizations in national policy domains, and organizational

surveys of diverse establishments.

List of contributors

xiv

Andrea Lawlor is a PhD student the Department of Political Science McGill University, Canada

specializing in media and public policy. Other research interests include political parties, party

ﬁnancing and Quebec politics. Her research can be found in Canadian Public Administration.

Raul Perez Lejano is an Associate Professor of Planning, Policy and Design in the School of

Social Ecology at the University of California, Irvine, USA. His research is currently in envir-

onmental planning, new decision models for explaining non-state/non-market institutions,

relational models of collective action, planning theory, institutional models of care and justice

research.

Patrik Marier holds a Tier-II Canada Research Chair (CRC) in Comparative Public Policy. His

research focuses broadly on the impact of changing demographic structures on reforms to the

welfare state in comparative contexts. His earlier work examined the politics of pension reform in

a number of countries including Sweden, Canada, Mexico and the United States. His current

research looks more broadly at the impact of ageing populations on a number of public policy

ﬁelds including education, health care and labour policy across comparative cases.

Igor S. Mayer is a Senior Associate Professor in the Faculty of Technology, Policy and Man-

agement (TPM) at Delft University of Technology, Netherlands. He is also the director of the

TU-Delft Centre for Serious Gaming.

Allan McConnell is Professor in the Department of Government and International Relations,

University of Sydney, Australia. He has published extensively on policy and political issues

surrounding crises, disasters, failures and ﬁascoes, as well as the more upbeat topic of policy success.

Michael D. McGinnis is Professor and Director of Graduate Studies in the Department of

Political Science at Indiana University, Bloomington, USA. He studies the unique contributions

that faith-based organizations make to the design and implementation of public policy related to

humanitarian relief, development assistance, peace-building and reconciliation in troubled

regions of the world, as well as standard public services in education, health care and welfare

assistance in societies less directly challenged by the ravages of war.

Michael McGuire is Professor in the School of Public and Environmental Aﬀairs, Indiana

University, Bloomington, USA. He works on intergovernmental and interorganizational colla-

boration and networks and has published widely on public management topics. His recent

publications on the subject have appeared in Public Administration Review and Public Administra-

tion among others.

Michael Mintrom is Professor of Public Sector Management at Monash University and the

Australia and New Zealand School of Government. He is the author of Contemporary Policy

Analysis (Oxford University Press, 2012), People Skills for Policy Analysts (Georgetown University

Press, 2003), and Policy Entrepreneurs and School Choice (Georgetown University Press, 2000).

Peter B. Mortensen is Associate Professor of Political Science at Aarhus University. His

research focuses on agenda-setting and public policy.

Gary Mucciaroni is a member of the Political Science Department at Temple University,

USA and was Department Chair between 2005 and 2010. The focus of his research and

List of contributors

publication is the politics of public policy-making in the United States – examining the forces

and actors that impinge upon the policy choices of government. He has published The Political

Failure of Employment Policy, 1945–1982, Reversals of Fortune: Private Interests and Public Policy,

Deliberative Choices: Debating Public Policy in Congress and Same Sex, Diﬀerent Politics: Issues and

Institutions in the Struggle for Gay and Lesbian Rights.

Daniel Nohrstedt is an Associate Professor of Political Science at the Department of Gov-

ernment, Uppsala University, Sweden. His research is primarily focused on crisis and learning,

intelligence issues, counter-terrorism policy and the public policy process.

Anthony Perl is Professor and Director of the Urban Studies Program and a member of the

Department of Political Science at Simon Fraser University in Vancouver, Canada. His current

research crosses disciplinary and national boundaries to explore the policy decisions that aﬀect

transportation, cities and the environment. His latest book, co-authored with Richard Gilbert,

is Transport Revolutions: Moving People and Freight Without Oil published by New Society Pub-

lishers in 2010.

Ora-orn Poocharoen is an Assistant Professor at the Lee Kuan Yew School of Public Policy

at the National University of Singapore. Her research interests include public management

reform, public administration theory, organization theory, comparative public administration

and public policy analysis.

David A. Rochefort is Arts and Sciences Distinguished Professor of Political Science at

Northeastern University, USA. His previous works include Foreign Remedies: What the Experience

of Other Nations Can Tell Us about Next Steps in Reforming U.S. Health Care (2012), written with

Kevin P. Donnelly, and co-editor with Roger W. Cobb of The Politics of Problem Deﬁnition (1994),

among other titles.

Sophie Schmitt is an Assistant Professor in the Department of Comparative Public Policy

and Administration, University of Konstanz, Germany, and specializes in comparative public

policy with a particular focus on environmental and social policy and regulation. She has

recently published articles in the Journal of European Public Policy and European Integration Online

Papers.

Anne Schneider is a Professor in the School of Justice at the University of Arizona, USA and

specializes in public policy and the role of policy in democracy. She and long-time co-author

Helen Ingram are the authors of Policy Design for Democracy (Kansas University Press), and

Deserving and Entitled (SUNY Press), and ‘The social construction of target populations,’ American

Political Science Review, June 1993.

Chris Silvia looks at interorganizational approaches to service delivery through the lens of

emergency management. His research aims to help public managers, who are more accustomed

to leading in hierarchical organizations, successfully contribute to leading networks. His recent

research has appeared in Journal of Public Aﬀairs Education, State and Local Government Review and

Public Administration Review.

Stuart Soroka is an Associate Professor in the Department of Political Sciences, McGill University,

Montreal, Canada.

List of contributors

xvi

Geert R. Teisman is currently Professor of Public Administration and Complex Decision

Making and Process Management at the Erasmus University Rotterdam, Netherlands, and was

previously Professor of Spatial Planning at the University of Nijmegen. He is member of the

board of Habiforum, a joint venture between science, private sector and public authorities which

explores possibilities of multiple land use. He is Scientiﬁ c Director of Living with Water, a

foundation which governs a variety of projects in which water system improvement is one of its

important aims. He is member of the executive board of Netlipse, a European knowledge

exchange network in the ﬁeld of infrastructure project management. He regularly advises gov-

ernments and private organizations on complex decision-making, strategic planning, public-

private partnerships, process management, intergovernmental co-operation in metropolitan areas

and policy evaluation.

H.F. Thomas III is a PhD student at Pennsylvania State University, USA, working on information,

science and technology policy issues.

John Turnpenny is a Senior Research Associate at the University of East Anglia, Norwich, UK.

His research focuses on the relationship between science, evidence and public policy-making.

Arwin van Buuren is an Associate Professor at the Department of Public Administration at

the Erasmus University Rotterdam, Netherlands. His research focuses on the domain of

water and climate governance. He is especially interested in the way in which complex

governance processes are organized and how the relationship between knowledge and govern-

ance can be optimized. He combines fundamental with applied research. His recent articles

have been published in International Journal of Project Management, International Journal of Public

Management, Public Management Review, Climatic Change and Environmental Impact Assessment

Review.

Els van Daalen is an Associate Professor at the Faculty of Technology, Policy and Manage-

ment (TPM), Delft University of Technology, Netherlands. Her research interests include

system dynamics modelling and the use of models in policy-making.

Evert Vedung is an Emeritus Professor of Political Science and a specialist in housing policy at

Uppsala University, Sweden. He has held posts at Mälardalen University College and the

University of Uppsala. He has been a board member of the Swedish Evaluation Society since its

foundation.

Christopher M. Weible is an Assistant Professor in the School of Public Aﬀairs at the Uni-

versity of Colorado Denver, USA. He teaches courses in environmental policy and politics,

public policy, policy processes, policy analysis, and research methods and design. His area of

research is in policy change, learning, political behaviour, coalitions and networks, institutions and

collective action, collaborative governance, and the role of scientiﬁc and technical information

in the policy process.

Andy Whitford is a Professor of Public Administration and Policy in the School of Public and

International Aﬀairs at the University of Georgia, USA. He concentrates on research and

teaching in organizational studies and public policy, with speciﬁc interest in organization theory,

models of decision-making and adaptation, and the political control of the bureaucracy. His

interests in public policy include environmental, regulation and public health policy. He has

List of contributors

xvii

particularly strong interest in the use of simulation and experimental methods for understanding

organizational behaviour and individual choice.

Dale Whittington is a Professor of Environmental Sciences and Engineering, City and

Regional Planning, and Public Policy, at the University of North Carolina at Chapel Hill,

USA. Since 1986 he has worked for the World Bank and other international agencies on the

development and application of techniques for estimating the economic value of environmental

resources in developing countries, with a particular focus on water and sanitation and vaccine

policy issues.

Claire Williams is a policy advisor at Victoria University of Wellington, New Zealand. She

has previously published in Policy and Society.

Chen-Yu Wu is a graduate student in the Department of Sociology at the University of

Minnesota, USA, and was a Teaching Assistant for the courses ‘Introduction to Sociology’ and

‘Basic Social Statistics.’

Lori Young is a PhD student in the Annenberg School of Communication at the University of

Pennsylvania, USA. Her published work includes studies of foreign news reporting in Canada

and the United States and has appeared in journals such as The International Journal of Press/Pol-

itics.

List of contributors

xviii

Part I

Introduction to the study of

the public policy process

History and method

Public policy debate and the rise

of policy analysis

Michael Mintrom and Claire Williams

Since the mid-1960s, an increasingly large number of people, trained in the humanities and

social sciences, have come to devote their professional lives to producing policy advice. This is a

global phenomenon, although until recently the intensiﬁcation of activities associated with

policy advising has been most pronounced in the United States. As government demand for

well-trained policy analysts and advisors has risen, universities have sought to provide relevant

graduate-level training.

The rise of policy analysis is usefully construed as a movement. Use of this term implies a

deliberate eﬀ ort on the part of many people to reconceive the role of government in society

and renegotiate aspects of the relationships that exist between individuals, collectivities and

governments. However, the claim that there is a policy analysis movement should not be taken

to imply either consistency of purpose or a deliberate striving for coordination among producers

of policy analysis. While not directly comparable in a political sense with other social movements,

the policy analysis movement has been highly inﬂuential. It has served to transform the advice-

giving systems of governments, altered the nature of policy debate, and, as a consequence,

challenged informal yet long-established advising practices through which power and inﬂuence

ﬂow. The profundity of this transformation has often eluded the attention of social and political

commentators. That is because the relevant changes have caused few immediate or obvious rup-

tures in the processes and administrative structures typically associated with government or,

more broadly speaking, public governance.

Early representations of policy analysis tended to cast it as a subset of policy advising. As such,

policy analysis was seen primarily as an activity conducted inside government agencies with the

purpose of informing the choices of a few key people, principally elected decision-makers

(Lindblom 1968; Wildavsky 1979). Today, the potential purposes of policy analysis are under-

stood to be much broader. Many more audiences are seen as holding interests in policy and as

being open to – indeed demanding of – appropriately presented analytical work (Radin 2000).

Beyond people in government, people in business, members of non-proﬁt organizations, and

informed citizens all constitute audiences for policy analysis. While policy analysts were once

thought to be mainly located within government agencies, today policy analysts also can be found

in most organizations that have direct dealings with governments, and in many organizations

where government actions signiﬁcantly inﬂuence the operating environment. In addition, many

university-based researchers, who tend to treat their peers and their students as their primary

audience, conduct studies that ask questions about government policies and that answer them

using forms of policy analysis. Given this, an appropriately encompassing deﬁnition of con-

temporary policy analysis needs to recognize the range of topics and issue areas policy analysts

work on, the range of analytical and research strategies they employ, and the range of audiences

they seek to address. In recognizing the contemporary breadth of applications and styles of

policy analysis, it becomes clear that eﬀective policy analysis calls for not only the application of

sound technical skills (Mintrom 2012), but also deep substantive knowledge, political percep-

tiveness and well-developed interpersonal skills (Mintrom 2003). Although producing high-

quality, reliable advice remains a core expectation for many policy analysts, advising now

appears as a subset of the broader policy analysis category. The transition from policy analysis as

a subset of advising to advising as a subset of analysis represents a signiﬁcant shift in orientation

and priorities from earlier times.

In what follows, we ﬁrst review the sources of increasing demand for policy analysis. We then

review the growth and adaptation in the supply of policy analysis that has occurred in response

to this demand. We conclude by discussing the current state of public policy debate and the

likely future trajectory of both the practice of policy advising and the training of policy analysts.

Throughout, we take the United States as our main point of reference. However, we have also

sought to demonstrate the comparative relevance of our argument. We have done this by discussing

the international embrace of New Public Management orthodoxy (from the late 1970s to the

mid-1990s) and government responses to the global ﬁnancial crisis (from 2007 into 2012).

Before proceeding, it is useful to deﬁne terms. For the purposes of this discussion, public policies

are considered to be any actions taken by governments that represent previously agreed responses

to speciﬁed circumstances. Governments design public policies with the broad purpose of

expanding the public good (Howlett 2011; Mintrom 2012). Policy studies refers to research on

policy topics and analytical work conducted primarily by university-based researchers with the

goal of critically assessing past, present and proposed policy settings. Policy studies can be under-

taken by researchers in many disciplinary or cross-disciplinary settings. They can draw upon a

range of analytical and interpretative methodologies. Policy studies may be historical and com-

parative in their scope. The term policy sciences refers to the subset of analytical techniques

devised in the social sciences that have been applied to understanding the design, implementa-

tion and evaluation of public policies. Policy analysis is here discussed as work intended to

advance knowledge of the causes of public problems, alternative approaches to addressing them,

the likely impacts of those alternatives, and trade-oﬀs that might emerge when considering appro-

priate governmental responses to those public problems. Policy advising is deﬁned here as the

practice of providing information to decision-makers in government, with the intention of

improving the base of knowledge upon which decisions are made. While policy advising need

not be based upon rigorous policy analysis, over recent decades such policy analysis has come to

play a more central part in the development of advice for decision-makers.

The evolving demand for policy analysis

Demand for policy analysis has been driven mostly by the emergence of problems and by

political conditions that have made those problems salient. Early in the development of policy

analysis techniques, the people who identiﬁed the problems that needed to be addressed tended

to be government oﬃcials. They turned to academics for help. Frequently, those academics

deemed to be most useful, given the problems at hand, were economists with strong technical

skills, who had the ability to estimate the magnitude of problems, undertake statistical analyses,

Mintrom and Williams

and determine the costs of various government actions. During the twentieth century, as transpor-

tation, electriﬁcation and telecommunications opened up new opportunities for market exchange,

problems associated with decentralized decision-making became more apparent (McCraw 1984).

Meanwhile, as awareness grew of the causes of many natural and social phenomena, calls emerged

for governments to establish mechanisms that might eﬀectively manage various natural and

social processes. Many matters once treated as social conditions, or facts of life to be suﬀered,

were transformed into policy problems (Cobb and Elder 1983). Together, the increasing scope

of the marketplace, the increasing complexity of social interactions, and expanding knowledge

of social conditions created pressures from a variety of quarters for governments to take the lead

in structuring and regulating individual and collective action. Tools of policy analysis, such as

the analysis of market failures and the identiﬁcation of feasible government responses, were devel-

oped to guide this expanding scope of government. Yet as the reach of policy analysis grew,

questions were raised about the biases inherent in some of the analytical tools being applied. In

response, new eﬀorts were made to account for the eﬀects of policy changes, and new voices

began to contribute in signiﬁcant ways to policy development. To explore the factors prompt-

ing demand for policy analysis, it is useful to work with a model of the policy-making process.

A number of conceptions of policy-making have been developed in recent decades. Here, we

apply the ‘stages model’, where ﬁve stages are typically posited: problem deﬁnition, agenda setting,

policy adoption, implementation, and evaluation (Eyestone 1978).

Initial demands for policy analysis were prompted by growing awareness of problems that

governments could potentially address. Questions inevitably arose concerning the appropriateness

of alternative policy solutions. Thus, in the United States in the 1930s, as the federal government

took on major new roles in the areas of regulation, redistribution and the ﬁnancing of infra-

structural development, the need arose for high quality policy analysis. With respect to regulatory

policy, concerns about the threat to the railroads of the emerging trucking industry prompted

the expansion of the Interstate Commerce Commission (Eisner 1993). This body employed lawyers

and economists who helped to devise an expanding set of regulations that eventually covered

many industries. Although the American welfare state has always been limited compared with

welfare states elsewhere, especially those within Europe, its development still required concerted

policy analysis and development work on the part of a cadre of bureaucrats (Derthick 1979).

Initially, much of the talent needed to ﬁll these positions was drawn from states like Wisconsin,

where welfare policies had been pioneered by Robert LaFollette. As the role of the United States

government in redistribution expanded, policy analysts swelled the ranks of career bureaucrats in

the Treasury, the Oﬃce of Management and Budget, and the Department of Health and Human

Services. Meanwhile, beneﬁt-cost analysis, a cornerstone of modern policy analysis and a core

component of public economics, was developed to help in the planning of dam construction in

the Tennessee Valley in the 1930s. Politicians at the time worried that some dams were being

built mainly to perpetuate the ﬂow of cash to construction companies rather than to meet growing

demand for electricity and ﬂood control (Eckstein 1958). The broad applicability of the tech-

nique soon became clear and its use has continually expanded. At the same time, eﬀorts to improve

the sophistication of the technique, and to develop variations on it that are best suited to diﬀerent

sets of circumstances have also continued to occur (Boardman et al. 2006; Carlson 2011).

Several features of policy development, the politics of agenda setting, and the policymaking

process have served over time to increase demand for policy analysts. The dynamics at work

have been similar to those through which an arms race generates ongoing and often expanding

demand for military procurements. In Washington, DC, the growing population of policy analysts

employed in the federal government bureaucracy led to demand elsewhere around town for

policy analysts who could verify or contest the analysis and advice emanating from government

Public policy debate

agencies. A classic example is given by the creation of the Congressional Budget Oﬃce (CBO).

This oﬃce was established as an independent resource for Congress that would generate analysis

and advice as a check on the veracity of the analyses prepared and disseminated by the execu-

tive-controlled Oﬃce of Management and Budget (Wildavsky 1992). The General Accounting

Oﬃce – renamed in 2004 the Government Accountability Oﬃce – was also developed to

provide independent advice for Congress. The purview of this Oﬃce has always been broader

than that of the Congressional Budget Oﬃce.

As the analytical capabilities available to elected politicians grew, groups of people outside

government but with signiﬁcant interests in the direction of government policy began devoting

resources to the production of high-quality, independent advice. Think tanks, like the Brook-

ings Institution and the American Enterprise Institute, established in the 1920s and 1940s

respectively, and both still going strong, represent archetypes of many independent policy shops

now located in Washington, DC (Smith 1991).

Today, due to the knowledge base of think tanks and their entrepreneurial capacity, their

staﬀ and aﬃliates are often regarded as members of a select few who dominate policy. Networks

between non-governmental policy experts and politicians in the United States are strengthened

through appointments into government departments – often described as a “revolving door”–

in which policy experts frequently move into and out of key positions in the government over

successive presidential elections. Therefore, United States think tanks are often viewed as

trustworthy sources of advice to the administration because of their deep connections, the star

status of some of their people, and their sophisticated inﬂuence eﬀorts.

It is useful to note that the roles, character, and eﬀectiveness of think tanks are shaped by

country-speciﬁc institutional and political factors. In the United States, most think tanks tend to

serve both as ‘watch dogs’ and as ‘idea brokers’; how much they exhibit either characteristic

depends largely on their alignment or lack of alignment with the political colours of the current

Administration. In contrast, most think tanks operating in Asian countries tend to view their

role as ‘regime-enhancing’ (Stone 2000; Stone and Denham 2004; Abelson 2004).

Parliamentary systems, found most conspicuously in Commonwealth countries, tend to

exhibit greater centralization of legislative power and accountability than systems like that of the

United States where the separation of powers is a fundamental characteristic of governance.

Centralization of power gives those in leadership positions greater control over policy. The

electoral system can serve to either enhance or diminish the power that those in government

exercise over policy development. This has been clearly exhibited in the case of New Zealand,

where a switch in 1996 from ﬁrst-past-the-post to mixed member proportional representation

in parliament changed the political landscape. Due to the increased diversity and number of

actors in the policy-making process, policy deliberation now tends to be longer and more con-

sultative, requiring greater input from a wider range of political actors and interest groups (Boston

et al. 2003). This change in the electoral system has resulted in greater demand for policy ana-

lysis. The potential for inﬂuence among political parties, members of parliament and interest

groups outside the government has led them to build analytical strengths that had previously

been almost the exclusive preserve of government departments.

The American constitution has an enshrined checks and balances system including a federal

and presidential system which is highly decentralized and fragmented. Establishment of the

Congressional Policy Advisory Board in 1998 has provided a further channel for experts to

share their advice with members of Congress as three-quarters of the policy experts on the

board are from think tanks. Combined with weak parties and considerable employee turnover,

the United States government is open to a host of government and non-government experts

engaging in policy-making (McGann and Johnson 2006).

Mintrom and Williams

Although policy analysis is often understood as work that occurs before a new policy or

programme is adopted, demand for policy analysis has also been driven by interest in the

eﬀectiveness of government programmes. The question of what happens once a policy idea has

been adopted and passed into law could be construed as too operational to deserve the attention

of policy analysts. Yet, during the past few decades, elements of implementation have become

recognized as vital for study by policy analysts (Bardach 1977; Lipsky 1984; Pressman and

Wildavsky 1973). In part, this has been led by results of programme evaluations, which have

often found policies not producing their intended outcomes, or worse, creating whole sets of

unintended and negative consequences. One important strain of the work devoted to assessing

implementation has contributed to what is now known as the ‘government failure’ literature

(Niskanen 1971; Weimer and Vining 2005; Wolf 1979). The possibility that public policies

designed to address market failures might themselves create problems led to a greater respect for

market processes and a degree of scepticism on the part of policy analysts towards the remedial

abilities of government (Rhoads 1985). In turn, this required policy analysts to develop more

nuanced understandings of the workings of particular markets. The government failure litera-

ture has led to greater interest in government eﬀorts to simulate market processes, or to reform

government and contract out aspects of government supply that could be taken up by private

ﬁrms operating in the competitive marketplace (Osborne and Gaebler 1993). Where eﬀorts to

reform government have been thorough-going, an interesting dynamic has developed. As the

core public sector has shrunk in size, the number of policy analysts employed in the sector has

increased, both in relative and in absolute terms. This dynamic is indicative of governments devel-

oping their capacities in the management of contracts rather than in the management of services

(Savas 1987). In these new environments, people exhibiting skills in mechanism design and

beneﬁt-cost analysis have been in demand. Therefore, employment opportunities for policy ana-

lysts have tended to expand, even as the scope of government has been curtailed. Fiscal con-

servatism, which was a hallmark of governments in many jurisdictions that embraced the New

Public Management practices added to this trend (Yergin and Stanislaw 1998; Williamson 1993).

When budgets are squeezed, it becomes critical for all possible eﬃciency gains to be realized.

More than any other trained professionals, policy analysts working in government are well-

placed to undertake the kind of analytical work needed to identify cost-saving measures and to

persuade elected decision-makers to adopt them.

The exact inﬂuence policy advisors have been able to exert over political decision-makers is

diﬃcult to measure because of the range of factors that inﬂuence policy outcomes. Moreover,

the changing nature of political and institutional environments throughout periods in history

have meant that not only has the deﬁnition of policy analyst changed over time but the extent

of their inﬂuence has also ﬂuctuated. ‘Windows of opportunity’ for policy inﬂuence appear

more frequently in some circumstances than in others (Kingdon 1995). For example, the move

towards addressing large national debts led to wide-scale reform of government policy settings

in many jurisdictions during the 1980s and 1990s.

By the 1980s the world had become intensely global: economically, socially and politically.

Global problems often resulted in similar policy outcomes. In New Zealand at that time, poli-

tical leaders became aware that the country’s economic position was deteriorating. Faced with

economic stagnation, social disruption and ideological shifts away from dependence upon gov-

ernment for the allocation of many resources in society, the New Zealand government became

open to the inﬂuence of institutions and agencies with perceived expertise in the dominant

problems of the day (Oliver 1989).

The strength of central agencies, endowed with economic expertise, such as the New Zealand

Treasury and organizations such as the New Zealand Business Round Table (NZBRT), were

Public policy debate

intensiﬁed by individual networks intertwining ministers with key bureaucrats and non-elected

political actors. Not unlike other countries such as Norway, Sweden, Britain and France, in the

1980s, a single economics ministry in New Zealand meant that there were few opportunities

for countervailing power. The system in New Zealand at that time was thus described by one

commentator as an ‘elective dictatorship’ (Boston 1989).

In New Zealand the Treasury was the Government’s most important advisory body. By

recruiting a senior manager from the Treasury, Roger Kerr, to lead the NZBRT in 1986,

business leaders recognized the value of networks and the strength of the Treasury department

(Mintrom 2006). The NZBRT ensured that the views of business leaders were fully articulated

in policy discussions. The Minister of Finance was reported at the time to have a relationship

with the Round Table that was ‘so close you couldn’t slide a Treasury paper between them’

(Murray and Pacheco 2001).

The conditions that allowed a few advisors broad inﬂuence over political decision-makers can

partly explain why the economic restructuring undertaken in New Zealand in the 1980s and

1990s was so extensive. As with ‘Reaganism’ in the United States, and at least the early days of

‘Thatcherism’ in Britain, ‘Rogernomics’ in New Zealand followed a pattern of international

policy transfer which included commercialization, liberalization and deregulation. These ideas

were not new. They began in North America and Europe a decade earlier and ‘represented a

global spread of neo-liberal politics’ (Dolowitz and Marsh 2000; Heﬀernan 1999).

Over the past few decades, technological advances have improved communication and made

it easier for policy advisers and decision-makers to assess diﬀerent policy decisions, thus increasing

the occurrences of policy transfer. While it is relatively simple to spot broad trends in policy

development, individual countries always exhibit speciﬁc, even idiosyncratic, policy design ele-

ments. Liberalization in New Zealand, the United States and Britain during the 1980s and 1990s

included rescission of the existing regulatory regime and a reduction in protectionist legislation

aimed at instituting competitiveness through a free market. In New Zealand, it was carried out

extensively at an unprecedented speed and has largely been unparalleled (Goldﬁnch 1998). In

the United States the focus on monetary policy as a prime instrument to shape the economy

represented a comprehensive change in policies which were up to that time not politically

viable. In contrast, liberalization and deregulation in Britain were not as quick, nor were they a

radical departure from ideas already in train, and were thus seen as more path dependant and

incremental (Niskanen 1988; Heﬀernan 1999).

The global ﬁnancial crisis emerged rapidly in mid-2007. This saw governments around the

world intervene in the operation of ﬁnancial markets in a manner that would have been

shunned during the 1980s and 1990s. The degree to which public policy settings undergird the

eﬀective workings of markets has become salient again. As a consequence, the current period

has entrenched the pervasive role of government in capitalist societies. At its best, this new era

of policy-making has seen eﬀorts being made to carefully balance responses to market failures

against risks of government failure. At its worst, the new era has oﬀered instances of those who

contributed to the crisis being cushioned, with those most harmed by the crisis receiving no

governmental relief. In short, the current period has opened space for extensive debate about

public policy and the role of government in society. The developmental role of governments in

promoting economic advancement – observed most starkly in the Asian tiger economies over

recent decades – has been eyed by some observers as a sound prescription, even for the most

advanced economies in the world. Sceptics have been much more concerned about extensive

government interventions into economic activities. Frequent voice has been given to the worry

of government failure – exhibited most obviously by special interest capture of government

subsidies.

Mintrom and Williams

Discussions during the 1960s and 1970s of the role of policy analysts in society often por-

trayed them as ‘whiz kids’ or ‘econocrats’ on a quest to imbue public decision-making with

high degrees of rigour and rationality (Self 1977; Stevens 1993). Certainly, proponents of beneﬁt-

cost analysis considered themselves to have a technique for assessing the relative merits of

alternative policy proposals that, on theoretical grounds, trumped any others on oﬀer. Likewise,

proponents of programme evaluation employing quasi-experimental research designs considered

their approach to be superior to other approaches that might be used to determine programme

eﬀectiveness (Cook and Campbell 1979). That the application of both beneﬁt-cost analysis and

quantitative evaluation techniques has continued unabated for several decades speaks to their

perceived value for generating usable knowledge. However, the limitations of such techniques

have not been lost on critics. In the case of beneﬁt-cost analysis, features of the technique that

make it so appealing – such as the reduction of all impacts to a common metric and the cal-

culation of net social beneﬁts – have also attracted criticism. In response, alternative methods for

assessing the impacts of new policies, such as environmental impact assessments, social impact

assessments and health impact assessments have gained currency (Barrow 2000; Lock 2000; Wood

1995). Similarly, widespread eﬀorts have been made to promote the integration of gender and

race analysis into policy development (Mintrom 2012; Myers 2002; True and Mintrom 2001).

In the case of evaluation studies, fundamental and drawn-out debates have occurred covering

the validity of various research methods and the appropriate scope and purpose of evaluation

eﬀorts. Signiﬁcantly for the present discussion, these debates have actually served to expand demand

for policy analysis. Indeed, government agencies designed especially to audit the impacts of

policies on the family, children, women and racial minorities have now been established in

many jurisdictions. Further, evaluation of organizational processes, which often scrutinize the

nature of the interactions between organizations and their clients are now accorded equivalent

status among evaluators as more traditional eﬀorts to measure programme outcomes (Patton

1997; Weiss 1998). Yet these process evaluations are motivated by very diﬀerent questions and

draw upon very diﬀerent methodologies than traditional evaluation studies that assessed programme

impacts narrowly.

This review of the evolving demand for policy analysis has identiﬁed several trends that have

served both to embed policy analysts at the core of government operations and to expand

demand for policy analysts both inside and outside government. These trends have much to do

with the growing complexity of economic and social relations and knowledge generation. Yet

there is also a sense in which policy analysis itself generates demand for more policy analysis.

While these trends have been most observable in national capitals, where a large amount of

policy development occurs, they have played out in related ways in other venues as well. For

instance, in federal systems, expanding cadres of well-trained policy analysts have become engaged

in sophisticated, evidence-based policy debates in state and provincial capitals.

Cities, too, are increasingly making extensive use of policy analysts in their strategy and planning

departments. At the global level, key coordinating organizations, such as the World Bank, the

International Monetary Fund, the World Trade Organization, and the Organization for Eco-

nomic Cooperation and Development have made extensive use of the skills of policy analysts to

monitor various transnational developments and national-level activities of particular relevance

and interest.

Today, the demand for policy analysis is considerable, and it comes both from inside and

outside governments. This demand is likely to keep growing as calls emerge for governments to

tackle emerging, unfamiliar problems. On the one hand, we should expect to see ongoing

eﬀorts to harness technical procedures drawn from the social sciences and natural sciences for

the purposes of improving the quality of policy analysis. On the other hand, more people are

Public policy debate

likely to apply these techniques, reinvent them, or develop whole new approaches to counteract

them, all with the purpose of gaining greater voice in policy-making at all levels of government

from the local to the global.

The evolving supply of policy analysis

To meet the growing demand for policy analysis, since the mid-1960s supply has greatly

expanded. But this expansion has been accompanied, at least around the edges of the enterprise,

by a transformation in the very nature of the products on oﬀer. Consequently it is now com-

monplace to ﬁnd policy analysts who question the questions of the past, who are politically

motivated, or who seek to satisfy intellectual curiosity rather than to oﬀer solutions to immediate

problems. Thus, the apparently straightforward question of what constitutes policy analysis cannot

be addressed in a straightforward manner. Answers given will be highly contingent on context.

For example, an answer oﬀered in the mid-1970s would be narrower in scope than an answer

oﬀered now. Likewise, the question of who produces policy analysis is also contingent. Econ-

omists have always been well represented in the ranks of policy analysts. People from other

disciplines have also contributed to policy analysis, although there has often been a sense that

other disciplines have less to contribute to policy design and the weighing of alternatives.

Today, practitioners and scholars heralding from an array of disciplines are engaged in this kind

of work, and the relevance of disciplines other than economics is well understood. In this review

of the evolving supply of policy analysis, we consider both the development of a mainstream

and the growing diversity of work that now constitutes policy analysis.

There has always been a mainstream style of conducting policy analysis. That style is more

prevalent today than it has ever been. The style is portrayed by practitioners and critics alike as

the application of a basic, yet continually expanding, set of technical practices (Stokey and

Zeckhauser 1978). Most of those practices derive from microeconomic analysis. They include

the analysis of individual choice and trade-oﬀs, the analysis of markets and market failure, and the

application of beneﬁt-cost analysis. Contemporary policy analysis textbooks tend to build on this

notion of policy analysis as a technical exercise (Bardach 2000; Mintrom 2012; Munger 2000;

Weimer and Vining 2005). For example, in his practical guide to policy analysis, Eugene Bar-

dach (2000) contends that a basic, ‘eightfold’ approach can be applied to analysing most policy

problems. The approach requires us to deﬁne the problem, assemble some evidence, construct the

alternatives, select the criteria, project the outcomes, confront the trade-oﬀs, decide, and tell our

story. As with Stokey and Zeckhauser’s approach, Bardach’s approach clearly derives from

microeconomic analysis, and beneﬁt-cost analysis in particular. None of this should surprise us

since, to the extent that a discipline called policy analysis exists at all, it is a discipline that grew

directly out of microeconomic analysis. That disciplinary linkage remains strong. Economists

comprise the majority of members of the United States-based Association for Public Policy Analysis

and Management (APPAM), the largest such association in the world. Likewise, contributions

to the Association’s Journal of Policy Analysis and Management are authored predominantly,

although certainly not exclusively, by economists.

Viewed positively, we might note that the basic approach to policy analysis derived from

microeconomics is extremely serviceable. As scholars have continued to apply and expand this

style of analysis, a rich body of technical practices has been created. Extensive eﬀorts have been

made to ensure that university students interested in careers in policy analysis gain appropriate

exposure to these approaches and receive opportunities to apply them. Most Masters pro-

grammes in public policy analysis or public administration require students to take a core set of

courses that expose them to microeconomic analysis, public economics, beneﬁt-cost analysis,

Mintrom and Williams

descriptive and inferential statistics, and evaluation methods. Sometimes the core is augmented

by courses on topics such as strategic decision-making, the nature of the policy process, and

organizational behaviour. Students are usually also given opportunities to augment their core

course selections with a range of elective courses from several disciplines. Without doubt, gradu-

ates of these programmes emerge well-equipped to immediately begin contributing in useful

ways to the development of public policy. Over recent decades, many universities in the United

States have introduced Masters programmes designed to train policy analysts along the lines

suggested here. More recently, similar programmes have been established in many other coun-

tries around the globe. Those establishing them know that there is strong demand for the kind

of training they seek to provide. These professional courses create opportunities for people who

have already been trained in other disciplines to acquire valuable skills for supporting the devel-

opment of policy analyses. Graduates end up being placed in many organizations in the public,

private and non-proﬁt sectors.

Viewed more negatively, the mainstream approach to teaching and practising policy analysis

can be critiqued for its narrowness and the privileging of techniques derived from economic

theory over analytical approaches that draw upon political and social theory. Suppose we again

describe the policy-making process as a series of stages: problem deﬁnition; agenda setting; policy

adoption; implementation; and evaluation. The mainstream approach to policy analysis has little

to contribute to our understanding of agenda setting, or the politics of policy adoption, imple-

mentation and evaluation. Steeped as it is in the rational choice or utilitarian perspective, main-

stream policy analysis is poorly suited to helping us to understand why particular problems

might manifest themselves at given times, why some policy alternatives might appear politically

palatable while others will not, and why adopted policies often go through signiﬁcant trans-

formations during the implementation stage. In addition, technical approaches to programme

evaluation, while obviously necessary for guiding the measurement of programme eﬀectiveness,

typically prescribe narrow data collection procedures that can leave highly relevant information

unexamined. Studies that dwell on assessing programme outcomes are unlikely to reveal the

multiple, and perhaps conﬂicting, ways that programme personnel and programme participants

often make sense of programmes. In turn, this might cause analysts to misinterpret the motiva-

tions that lead programme personnel and participants to redeﬁne programme goals from those

originally intended by policy-makers. Problems of programme design or programme theory

might end up being ignored in favour of interpretations that place the blame on faulty imple-

mentation (Chen 1990). More generally, mainstream approaches to policy analysis can encou-

rage practitioners to hold tightly to assumptions about individual and collective behaviour that

are contradicted by the evidence. In the worst cases, this can lead to inappropriate speciﬁcation

of proposals for policy change.

The foregoing observations might cause us to worry that mainstream training in policy ana-

lysis does not suﬃciently equip junior analysts to become reﬂective practitioners or practitioners

who listen closely to the voices of people who are most likely to be aﬀected by policy change

(Forrester 1999; Schön 1983). Certainly, strong critiques of mainstream analytical approaches

have led to some rethinking of the questions that policy analysts should ask when working on

policy problems (Stone 2002). Yet it is a fact that many people who have gone on to become

excellent policy managers and leaders of government agencies began their careers as junior

policy analysts fresh out of mainstream policy programmes. This suggests that, within profes-

sional settings, heavy reliance is placed on mechanisms of tacit knowledge transfer, whereby

narrowly trained junior analysts come to acquire skills and insight that serve them well as policy

managers. Our sense is that the key components of this professional socialization can be codiﬁed

and are teachable (Mintrom 2003). But much of what it takes to be an eﬀective policy analyst is

Public policy debate

not captured in the curricula of the many university programmes that now exist to train future

practitioners.

Beyond mainstream eﬀorts to expand the supply of policy analysts, all of which place heavy

emphasis on the development of technical skills informed primarily by microeconomic theory,

other disciplines also contribute in signiﬁcant ways to the preparation of people who eventually

become engaged in policy analysis. In these cases, the pathway from formal study to the practice

of policy analysis is often circuitous. People with substantive training in law, engineering, natural

sciences and the liberal arts might begin their careers in closely related ﬁelds, only to migrate

into policy work later. For example, individuals trained initially as sociologists might gain pro-

fessional qualiﬁcations as social workers and, after years in the ﬁeld, assume managerial roles that

require them to devote most of their energies to policy development. People pursuing these

alternative professional routes towards working as policy analysts can bring rich experiences and

diverse insights to policy discussions. The resulting multidisciplinary contributions to policy

discussions have been known to generate their share of intractable policy controversies (Schön

and Rein 1994). Yet, when disagreements that emerge from diﬀerences in training and analy-

tical perspectives are well managed, these multidisciplinary forums can produce eﬀective policy

design. Indeed, increasing eﬀorts are now being made to address signiﬁcant problems through

‘joined-up government’ initiatives (Perri 6: 2004). Through these, individuals with diverse profes-

sional backgrounds who are known to have been working on similar problems are brought

together to work out cohesive policy strategies to address those problems. For example, pedia-

tricians, police oﬃcers, social workers and educators might be asked to work together to devise

policies for eﬀective detection and prevention of child abuse.

Two somewhat contradictory trends in the evolving supply of policy analysis have been noted

here. On the one hand, the growth of university-based professional programmes designed to

train policy analysts has contributed to a signiﬁcant degree of analytical isomorphism. No matter

which university or country such programmes are based, the students are required to study

much the same set of topics and to read from a growing canon of articles and books covering

aspects of policy analysis. These programmes promote ways of thinking about public policy that

owe heavy debts to the discipline of economics. On the other hand, researchers and practi-

tioners working out of other disciplines have increasingly become involved in the production of

policy analysis. These contributions from outside the mainstream have tended to promote analytical

diversity. Taken together, these contradictory trends deﬁne the contemporary ﬁeld of policy

analysis. Often, the diﬀerences are downplayed. For example, while the Journal of Policy Analysis

and Management publishes research articles and shorter pieces on teaching practice, all of which

can be informed by a variety of disciplinary perspectives, the overall impression given is that of a

journal devoted to the furtherance of mainstream methods of policy analysis. In contrast, con-

troversies in the interdisciplinary ﬁeld of evaluation studies have left a distinctly diﬀerent

impression. As a result, courses on policy evaluation, when taught in professional public policy

programmes, can introduce perspectives that sit uncomfortably with those taught in other core

courses. Similarly, the eclecticism of contributions to organizational studies can transform other-

wise staid courses on public management and administration into eye-opening explorations of

organizational behaviour that present perspectives and analyses diﬀering starkly from mainstream

economic interpretations.

The evolving supply of policy analysis is likely to continue to be deﬁned by a mixture of

mainstream and alternative analytical perspectives. New policy problems generated by changing

social conditions, technologies and political agendas cannot be expected to readily lend them-

selves to mainstream policy analysis. Indeed, although policy analysis textbooks present speciﬁc

forms of market failure or government failure as defensible rationales for policy action, many

Mintrom and Williams

contemporary problems defy such categorization. For example, changing understandings of morally

appropriate behaviour (Mooney 2000) and diﬀering perspectives on the degree to which parents

and the state should be trusted to act in the best interests of children (Nelson 1984; Shapiro 1999)

have provided impetus for a range of recent policy disputes. Mainstream approaches to policy

analysis are not well suited to assessing the relative merits of competing arguments and per-

spectives in these areas. This suggests that, for any headway to be made in addressing disputes of

this sort, the comparative strengths of alternative disciplinary perspectives must be drawn upon

to guide policy analysis. The frontiers of policy analysis are also being advanced by a growing

awareness of the degree to which national policies hold implications for international relations,

transnational norms, and global trade or environmental concerns (Sandler 2004; Tabb 2004).

These developments will require further innovation in the design and application of policy analysis

techniques. In many instances, the mainstream perspective will need to be augmented by alter-

native perspectives that oﬀer sound analytical traction on otherwise diﬃcult conceptual and

practical problems. Thus, deﬁnitions of policy analysis are likely to keep expanding and the set

of actors having relevant and important contributions to make will remain dynamic.

Prospects for the policy analysis movement

The policy analysis movement began to emerge in the mid-1960s. Today, it is large, diverse and

global. The phenomenon of many people generating policy analysis for consumption by an array

of audiences can be claimed to be a movement for several reasons. First, policy analysts, no matter

where they are located within society, all focus their energies in one way or another on iden-

tifying, understanding and confronting public problems. As techniques of policy analysis have

been more routinely applied to investigating public problems, new problems and new approaches

to addressing them have been revealed. While a ﬁnite set of known policy problems might exist

at any given time, that set is continually being revised, as particular problems are resolved and as

others become salient. Second, while people engaged in policy analysis work out of a variety of

perspectives and often make contradictory and con ﬂicting arguments, there is a widely shared

recognition that knowledge of public problems and how they can best be addressed requires thor-

ough, theory-driven, and evidence-based investigation. This is highly signiﬁcant, because chan-

ging perceptions of what kind of claims should guide public deliberations have made it harder

for long-entrenched groups to use their informal, quiet power to inﬂuence government actions.

Third, the increasing reliance placed on this style of policy analysis has required a core of people

to routinely apply a well-established set of analytical and research methods. The result has been

the clear deﬁnition of a mainstream of policy analysis, and many university programmes have

been established to professionalize budding analysts. Fourth, the development of policy analysis

has not been restricted to those people applying mainstream techniques. Indeed, there has been

a level of openness to people from various disciplines oﬀering alternative theoretical and empirical

arguments concerning particular policy matters. Critique of mainstream methods has often resul-

ted in signiﬁcant eﬀorts to improve the analytical approaches employed. Finally, people engaged

in policy analysis, regardless of their substantive interests or their immediate purposes, have shared

an understanding that they are engaged in important, socially relevant work. They share a common

understanding that, at base, public policy involves systematic eﬀorts to change social institutions.

The seriousness of this work helps to explain why policy disputes are often prolonged and heated.

Taken together, these various features of contemporary policy work clearly represent markers of

a policy analysis movement. Those associated with it are, in their many particular ways, con-

tributing to the ongoing task of deﬁning the appropriate role of government in society, and

how governments can best mediate social and economic relations.

Public policy debate

What are the future prospects of this movement? How might it continue to evolve? Internal

and external dynamics are likely to ensure further expansion of the movement. With respect to

internal dynamics, a clear lesson of the foregoing discussion is that good policy analysis creates

its own demand. This happens because, in competitive settings like debates over policy choices,

opposing parties face strong incentives to ﬁnd ways to outsmart their opponents. If one party’s

arguments are consistently supported by strong policy analysis and this appears to give them an

edge in debate, then other parties will soon see the merit in upping their own game. This can

involve emulating, revising or critiquing the methods of opponents. But, no matter what, the

result is further production of policy analysis. Aside from this, careful policy studies, especially

evaluations of existing programmes, often reveal aspects of policy design or implementation that

require further attention. When we are forced to return to the drawing board, further policy

analysis occurs. The lesson that good policy analysis creates its own demand suggests that the

currently observed momentum and vibrancy of that movement will continue.

The external dynamics that drive expansion of the policy analysis movement derive from chan-

ging social, economic and political conditions. In the future, the increasing integration of econo-

mies and societies, referred to generally as globalization, can be expected to generate new sets of

policy problems. In this regard, the changes associated with globalization echo the dynamics that

were observed from the late nineteenth century well into the twentieth century in federal sys-

tems. During that period, increasing commerce across state and provincial borders, and the

emergence of intensive inter-jurisdictional competition prompted new considerations of the role

of government in society. Extensive eﬀort also went into determining what levels of government

were best suited to performing diﬀerent functions. New times introduce new problems and

questions. Drawing lessons from the past, it seems clear that many new policy problems will arise

on government agendas and be the focus of extensive debates in the coming decades. While

globalizing forces will be responsible for generating many of these challenges, the challenges

themselves will be manifested at all levels of government, from the local upwards. As in the past,

people both inside and outside government can be expected to show intense interest in these

policy challenges, and they will call for further supply of innovative and high-quality policy

analysis to build their knowledge and strengthen their debating points.

Bibliography

Abelson, Donald. 2004. ‘The think tank industry in the USA’, in Diane Stone and Andrew Denham.

Think Tank Traditions: Policy Research and the Politics of Ideas. Manchester: Manchester University Press.

Bardach, Eugene. 1977. The Implementation Game: What Happens After a Bill Becomes a Law. Cambridge,

MA: MIT Press.

—— 2000. A Practical Guide for Policy Analysis: The Eightfold Path to More Eﬀective Problem Solving. New

York: Seven Bridges Press/Chatham House.

Barrow, C.J. 2000. Social Impact Assessment: An Introduction. New York: Oxford University Press.

Boardman, Anthony E., David H. Greenberg, Aidan R. Vining and David L. Weimer. 2006. Cost-beneﬁt

Analysis: Concepts and Practice. Englewood Cliﬀs, NJ: Pearson/Prentice Hall.

Boston, Jonathan. 1989. ‘The Treasury and the organisation of economic advice: Some international

comparisons’, in Brian Easton (ed.) The Making of Rogernomics. Auckland: Auckland University Press, 68–91.

—— Stephen Church and Tim Bale. 2003. ‘The impact of proportional representation on government

eﬀectiveness: The New Zealand experience’, Australian Journal of Public Administration 62(21).

Carlson, Deven. 2011. ‘Trends and innovations in public policy analysis’, Policy Studies Journal 39: 13–26.

Chen, Huey-Tsyh. 1990. Theory-Driven Evaluations. Newbury Park, CA: Sage.

Cobb, Roger W. and Charles W. Elder. 1983. Participation in American Politics: The Dynamics of Agenda-Building,

2nd edn. Boston, MA: Allyn and Bacon.

Cook, Thomas D. and Donald T. Campbell. 1979. Quasi-Experimentation: Design and Analysis Issues for

Field Settings. Boston, MA: Houghton Miﬄin Company.

Mintrom and Williams

Dahl, Robert A. 1998. On Democracy. New Haven, CT: Yale University Press.

Derthick, Martha. 1979. Policymaking for Social Security. Washington, DC: Brookings Institution.

Dolowitz, D. and D. Marsh. 2000. ‘Learning from abroad: The role of policy transfer in contemporary

policy-making’, Governance 13: 5–24.

Eckstein, Otto. 1958. Water Resource Development. Cambridge, MA: Harvard University Press.

Eisner, Marc Allen. 1993. Regulatory Politics in Transition. Baltimore, MD: Johns Hopkins University Press.

Eyestone, Robert. 1978. From Social Issues to Public Policy. New York: John Wiley & Sons.

Forester, John. 1999. The Deliberative Practitioner: Encouraging Participatory Planning Processes. Cambridge,

MA: MIT Press.

Goldﬁnch, Shaun. 1998. ‘Remaking New Zealand’s economic policy: Institutional elites as radical innovators’,

Governance 11, 177–207.

Heﬀernan, Richard. 1999. New Labour and Thatcherism: Political Change in Britain. New York: St. Martin’s

Press.

Howlett, Michael. 2011. Designing Public Policies: Principles and Instruments. Abingdon: Routledge.

Kingdon, John W. 1995. Agendas, Alternatives, and Public Policies. New York: HarperCollins College Publishers.

Lindblom, Charles E. 1968. The Policymaking Process. Englewood Cliﬀs, NJ: Prentice-Hall.

Lipsky, Michael. 1980. Street-Level Bureaucracy: The Dilemmas of the Individual in Public Services. New York:

Russell Sage Foundation.

—— 1984. ‘Bureaucratic disentitlement in social welfare programs’, Social Service Review 58(1): 3–27.

Lock, Karen. 2000. ‘Health impact assessment’, British Medical Journal 320: 1395–8.

McCraw, Thomas K. 1984. Prophets of Regulation. Cambridge, MA: Belknap Press.

McGann, James and Erik Johnson. 2006. Comparative Think Tanks, Politics and Public Policy. Cheltenham:

Edward Elgar Publishing.

Mintrom, Michael. 2003. People Skills for Policy Analysts. Washington, DC: Georgetown University Press.

—— 2006. ‘Policy entrepreneurs, think tanks, and trusts’, in Raymond Miller (ed.) New Zealand Government

and Politics, 4th

edn, Melbourne: Oxford

University Press, 536–46.

—— 2012. Contemporary Policy Analysis. New York: Oxford University Press.

——and Sandra Vergari. 1996. ‘Advocacy coalitions, policy entrepreneurs, and policy change’, Policy Studies

Journal 24: 420–35.

Mooney, Christopher Z. (ed.) 2000. The Public Clash of Private Values: The Politics of Morality Policy. New

York: Chatham House Publishers of Seven Bridges Press.

Munger, Michael C. 2000. Analyzing Policy: Choices, Conﬂicts, and Practices. New York: W.W. Norton.

Murray, Georgina and Douglas Pacheco. 2001. ‘Think tanks in the 1990s’, The Australian National University.

Myers, Samuel L. Jr. 2002. ‘Presidential address - analysis of race as policy analysis’, Journal of Policy Analysis

and Management 21: 169–90.

Nelson, Barbara. 1984. Making an Issue of Child Abuse. Chicago: University of Chicago Press.

Niskanen, William A., Jr. 1971. Bureaucracy and Representative Government. Chicago: Aldine-Atherton.

—— 1988. Reaganomics: An Insider’s Account of the Policies and the People. New York: Oxford University Press.

Oliver, W. Hugh. 1989. ‘The labour caucus and economic policy formation, 1981-1984’, in Brian Easton

(ed.) The Making of Rogernomics. Auckland: Auckland University Press, 11–52.

Osborne, David and Ted Gaebler. 1993. Reinventing Government: How the Entrepreneurial Spirit is Transforming

the Public Sector. New York: Penguin Books.

Patton, Michael Quinn. 1997. Utilization-Focused Evaluation: The New Century Text. Thousand Oaks, CA:

Sage.

Perri 2004. ‘Joined-up government in the western world in comparative perspective: A preliminary lit-

erature review and exploration’, Journal of Public Administration Research and Theory 14: 103–38. Available

at: http://en.wikipedia.org/wiki/Perri_6

Pressman, Jeﬀrey L. and Aaron Wildavsky. 1973. Implementation. Berkeley, CA: University of California

Press.

Radin, Beryl A. 2000. Beyond Machiavelli: Policy Analysis Comes of Age. Washington, DC: Georgetown

University Press.

Rhoads, Steven E. 1985. The Economist’s View of the World: Government, Markets, and Public Policy. New

York: Cambridge University Press.

Rochefort, David A. and Roger W. Cobb (eds). 1994. The Politics of Problem Deﬁnition:

Shaping the Policy

Agenda.

Lawrence, KS: University Press of Kansas.

Sabatier, Paul A. 1988. ‘An advocacy coalition framework of policy change and the role of policy-oriented

learning therein’, Policy Sciences 21: 129–68.

Public policy debate

Sandler, Todd. 2004. Global Collective Action. New York: Cambridge University Press.

Savas, E.S. 1987. Privatization: The Key to Better Government. Chatham, NJ: Chatham House.

Schön, Donald A. 1983. The Reﬂective Practitioner: How Professionals Think in Action. New York: Basic

Books.

——and Martin Rein. 1994. Frame Reﬂection: Toward the Resolution of Intractable Policy Controversies. New

York: Basic Books.

Self, Peter. 1977. Econocrats and the Policy Process: The Politics and Philosophy of Cost-Beneﬁt Analysis. Boulder,

CO: Westview Press.

Shapiro, Ian. 1999. Democratic Justice. New Haven, CT: Yale University Press.

Smith, James A. 1991. The Idea Brokers: Think Tanks and the Rise of the New Policy Elite. New York: The

Free Press.

Stevens, Joe B. 1993. The Economics of Collective Choice . Boulder, CO: Westview Press.

Stokey, Edith and Richard Zeckhauser. 1978. A Primer for Policy Analysis. New York: W.W. Norton.

Stone, Deborah. 2002. Policy Paradox: The Art of Political Decision Making, revised edn. New York: W.W.

Norton.

Stone, Diane 2000. ‘Non-governmental policy transfer: The strategies of independent policy institutes’,

Governance 13: 65.

——and Andrew Denham. 2004. Think Tank Traditions: Policy Research and the Politics of Ideas. Manchester:

Manchester University Press.

Tabb, William K. 2004. Economic Governance in the Age of Globalization. New York: Columbia University

Press.

True, Jacqui and Michael Mintrom. 2001. ‘Transnational networks and policy diﬀusion: The case of

gender mainstreaming,’ International Studies Quarterly 45: 27–57.

Weimer, David L. and Aidan R. Vining. 2005. Policy Analysis: Concepts and Practice, 4th edn. Upper Saddle

River, NJ: Prentice Hall.

Weiss, Carol H. 1998. Evaluation, 2nd edn. Upper Saddle River, NJ: Prentice Hall.

Wildavsky, Aaron. 1979. Speaking Truth to Power: The Art and Craft of Policy Analysis. Boston, MA: Little-Brown.

—— 1992. The New Politics of the Budgetary Process, 2nd edn. New York: Harper Collins Publishers.

Williamson, John (ed.). 1993. The Political Economy of Policy Reform. Washington, DC: Institute for International

Economics.

Wolf, Charles, Jr. 1979. ‘A theory of nonmarket failures’, Journal of Law and Economics 22: 107–39.

Wood, Christopher. 1995. Environmental Impact Assessment: A Comparative Review. Burnt Mill: Longman

Scientiﬁc and Technical.

Yergin,

Daniel and Joseph

Stanislaw. 1998. The Commanding Heights: The Battle Between Government and the

Marketplace that Is Remaking the Modern World. New York: Touchstone.

Mintrom and Williams

The policy-making process

Michael Howlett and Sarah Giest

The policy cycle model of the policy process

The idea that policy development can be thought of as a series of steps in a decision-making

process was ﬁrst broached systematically in the work of Harold Lasswell, a pioneer in the ﬁeld

of policy research (Lasswell 1956, 1971). In most recent work, a ﬁve-stage model of the policy

process has been most commonly used. In this model, “agenda-setting” refers to the ﬁrst stage

in the process when a problem is initially sensed by policy actors and a variety of solutions put

forward. “Policy formulation” refers to the development of speciﬁc policy options within

government when the range of possible choices is narrowed by excluding infeasible ones, and

eﬀorts are made by various actors to have their favored solution ranked highly among the

remaining few. “Decision-making” refers to the third stage in which governments adopt a

particular course of action. In the fourth stage of “policy implementation” governments put their

decisions into eﬀect using some combination of the tools of public administration in order to

alter the distribution of goods and services in society in a way that is broadly compatible with

the sentiments and values of aﬀected parties. Finally “policy evaluation” refers to the ﬁfth stage in

the process in which the results of policies are monitored by both state and societal actors, often

leading to the reconceptualization of policy problems and solutions in the light of experiences

encountered with the policy in question (Howlett et al. 2009).

This idea of policy-making existing as a set of interrelated stages provides a general “frame-

work” for understanding the policy development process and points to several of the key

temporal activities and relationships that should be examined in furthering study of the issue.

However, it does not answer several key questions such as the actual substance of policy, the

number and type of relevant actors involved in the process and what motivates them, the exact

manner and sequence in which actual policy development processes occur, and whether there

exist basic patterns of development in diﬀerent issues areas, sectors or jurisdictions (Sabatier 1991).

Empirical studies aimed at answering these questions and generating more detailed models of

the policy-making process, however, have developed some knowledge about the activities of

policy actors at each stage of the cycle, summarized below, which draw attention to the limited

number of common modes, or styles, which characterize policy processes throughout the world

and the factors which drive policy-making through the various stages.

Styles of policy behavior in the policy cycle

Agenda-setting

In the scholarly literature on agenda-setting, for example, a useful distinction is often drawn

between the systemic or unoﬃcial public agenda and the institutional or formal, oﬃcial, agenda

which helps to conceptualize policy-making dynamics at this stage of the process. The systemic

agenda “consists of all issues that are commonly perceived by members of the political com-

munity as meriting public attention and as involving matters within the legitimate jurisdiction of

existing governmental authority” (Cobb and Elder 1972). This is essentially a society’s agenda for

discussion of public problems, such as crime or health care, water quality or wilderness pre-

servation. The formal or institutional agenda, on the other hand, consists of only a limited number

of issues or problems to which attention is devoted by policy elites (Baumgartner and Jones 1991;

Kingdon 1984). Each society has literally hundreds of issues which some citizens ﬁnd to be

matters of concern and would have the government do something about. However, only a small

proportion of the problems on the public or systemic agenda are actually taken up by policy

actors actively involved in policy development and understanding how and why this movement

occurs is key to understanding process dynamics at the ‘front-end’ of the policy process.

Almost 40 years ago the American political scientists Cobb, Ross and Ross developed a model

of typical agenda-setting styles based on this insight. In their analysis, they argued that three basic

patterns of agenda-setting could be discerned, distinguished by the origins of the issue as well as the

resources utilized to facilitate its inclusion on the agenda. In the outside initiation pattern “issues arise

in nongovernmental groups and are then expanded suﬃciently to reach, ﬁrst, the public [systemic]

agenda and, ﬁnally, the formal [institutional] agenda” (Cobb et al. 1976). The mobilization case is quite

diﬀerent and describes “decision-makers trying to expand an issue from a formal to a public agenda.”

In this model, issues are simply placed on the formal agenda by the government with no necessary

preliminary expansion from a publicly recognized grievance. In the third type of agenda-setting,

inside initiation,inﬂuential groups with special access to decision-makers initiate a policy and do not

necessarily want it to be expanded and contested in public. This can be due to technical as well as

political reasons. Entrance is virtually automatic due to the privileged place of those desiring a decision.

Further research conducted by John Kingdon into the dynamics of agenda-setting in the US

Congress in the early 1980s focused on the timing of these processes of agenda entrance. In his

work, he focused on the role played by policy entrepreneurs both inside and outside government

in taking advantage of agenda-setting opportunities—policy windows—to move items onto formal

government agendas (Kingdon 1984). Kingdon argued that the characteristics of issues (the pro-

blem stream) combined with the characteristics of political institutions and circumstances (the

politics stream) and the development of policy solutions (the policy stream), lead to the opening and

closing of opportunities for agenda entrance. Such opportunities can be seized upon or not, as the

case may be, by policy entrepreneurs who are able to recognize and act upon them. The policy

window types which can lead to issue entrance include routine, spill-over, discretionary and random

types (Kingdon 1984; Howlett 1998). Empirical evidence suggests that the frequency of occurrence

of the window types varies by level of institutionalization—with the most institutionalized types

occurring much more frequently than the least institutionalized (Howlett 1998).

Policy formulation

Studies of the second stage of the policy cycle, policy formulation, have also emphasized the impor-

tance of speciﬁc kinds of actors interacting to develop and reﬁne policy options for government

Howlett and Giest

(Freeman 1965; Linder and Peters 1990) and the resulting small number of styles of policy for-

mulation. But unlike agenda-setting, where the public is often actively involved, in policy formula-

tion, the relevant policy actors are restricted to those who not only have an opinion on a subject,

but also have some minimal level of knowledge of the subject area, allowing them to comment,

at least hypothetically, on the feasibility of options put forward to resolve policy problems.

The power of the ideas held by these actors and their stability in policy subsystems, in particular,

has been a subject of much attention in studies of policy formulation in recent years. Carstensen

(2011), for example, has suggested that ideas change incrementally, when new elements of meaning

are added to them, resulting in a characteristic process of policy formulation in which similar

ideas inform similar policy options over long periods of time. This shifts the focus of the analysis

towards the inﬂ uence of discourses within certain institutional settings as a key factor aﬀecting

the nature of the kinds of policy alternatives put forward in the policy formulation (Schmidt 2006,

2010, 2011). Focusing on the information or knowledge available to decision-makers has led policy

research in this vein to focus on concepts such as epistemic communities (ECs) in policy formula-

tion. These are loose groupings of experts or knowledge “providers” for the decision-making

process—opening up new opportunities for inﬂuence on policy alternatives. The suggested

mechanism underlying this is their “ability to transfer policy by assuming control over knowledge

production and in doing so guiding decision-maker learning” (Dunlop 2009: 290; Haas 1992).

Scholars over the years have developed a variety of such concepts in order to help identify who

are the key actors in these policy subsystems, what brings them together, how they interact, and what

eﬀect their interaction has on policy development (Jordan 1990a, 1990b; Jordan and Schubert 1992).

Most of these distinguish between a large set of actors with some knowledge of the policy issue in

question, such as an epistemic community and a smaller set in which actors not only have the requisite

knowledge, but also have established patterns of more or less routine interactions with each other

and with decision-makers to allow them to inﬂuence alternative formation (Knoke 1993).

Membership in knowledge-based policy communities extends to actors such as state policy-

makers (administrative, political, and judicial), members of non-governmental organizations (NGOs)

concerned with the subject, members of the media who report on the subject, academics who

follow or research the area, and members of the general public who, for whatever reason, have

taken an interest in the subject (Sabatier 1987, 1988). In many issue areas, the policy community

also involves members of other organizations such as businesses, labor unions, or various for-

malized interest groups or professional associations concerned with government actions in the

sector concerned. In some cases, international actors such as multinational corporations, inter-

national governmental or NGOs, or the governments of foreign states, can also be members of

sectoral policy communities (Haas 1992). The subset of these actors who interact within more

formalized institutions and procedures of government are deﬁned as members of policy networks

(Coleman and Skogstad 1990; Marin and Mayntz 1991; Pross 1992). These policy networks

include representatives from the community, but are “inner circles” of actors, who eﬀectively

hold the power to veto many policy options as untenable or unfeasible.

In this view, the likely results of policy formulation are contingent upon the nature and con-

ﬁguration of the policy community and network in the speciﬁc sector concerned. A key variable

that many observers have argued aﬀects the structure and behavior of policy networks, for example,

is their number of members, which inﬂuences their level of integration and the types of inter-

actions they undertake (van Waarden 1992; Atkinson and Coleman 1989, 1992; Coleman and

Skogstad 1990). What is important for policy communities, on the other hand, is not the number of

participants in the community but the number of relatively distinct “idea sets” which exist within

it. This aﬀects the nature of conﬂict and consensus which exists in the community and, as a

result, aﬀects the behavior of community actors (Schulman 1988; MacRae 1993; Smith 1993).

The policy-making process

Taken together, these conﬁgurations can deﬁne a policy formulation style, one which can be

quite long-lasting and resistant to change. In open subsystems where networks have many

members and communities share many idea sets, for example, it can be expected that a propensity

for new, radical, alternatives to the status quo may be generated in the policy formulation process.

In closed subsystems, where networks have few members and communities are dominated by a

single idea set, on the other hand, a status quo orientation will emerge in the policy options devel-

oped and put before decision-makers. In subsystems where only a few actors make up the network

but communities are open to new ideas, signiﬁcant alternatives to the status quo may emerge from

the formulation process, but usually over the opposition of network members. In subsystems

where many actors deal with few ideas, marginal or incremental options tend to develop.

Although the network framework has been criticized as being a descriptive metaphor, due to

the fact that “the independent variables, (of policy-making) are not network characteristics per se

but rather characteristics of components within the networks” (Dowding 1995: 137), it remains

helpful in that it provides a general mechanism to help organize the complex reality of multiple

actors, institutions, and ideas found in the policy formulation process (Howlett et al. 2009;

Smith 1994; Howlett and Ramesh 1998; Howlett 2002a).

Decision-making

Similar styles have been identiﬁed at the decision-making stage of the policy process. Many

early studies of policy-making in companies, governments, and organizations conducted largely

by students of public and business administration, for example, argued that decision-makers

attempt to follow a systematic method for arriving at logical, eﬃcient decisions. They argued

that policy-makers achieved superior results when they ﬁrst established a goal; explored alter-

native strategies for achieving it; attempted to predict its consequences and the likelihood of

each occurring; and then chose the option which maximized potential beneﬁts at least cost or

risk (Gawthrop 1971; Carley 1980; Cahill and Overman 1990).

This model was “rational” in the sense that it prescribed a standard set of procedures for policy-

making which were expected to lead in all circumstances to the choice of the most eﬃcient means

of achieving policy goals (Jennings 1987; Torgerson 1986). Pure “rational” models of decision-

making thought of policy-makers as neutral “technicians” or “managers,” who identify a problem

and then ﬁnd the most eﬀective or eﬃcient way of solving it (Elster 1991). Many of the latest eﬀorts

to enhance the eﬃciency and eﬀectiveness of public policy decision-making, such as, for example,

the “evidence-based policy movement” (Pawson 2006), focuses on the application of a systemic

evaluative rationality to policy problems (Sanderson 2006; Mintrom 2007) in classic rational style.

Empirical research into decision-making processes, however, has discovered that political

processes of bargaining and negotiation often outweigh “ rational” deliberations and calculations

of costs and beneﬁts are subject to substantive and procedural limitations. Policy-makers were

often found to be neither necessarily neutral nor competent and other models of the public

policy decision-making processes have argued that this is not an accidental situation but rather

an inherent and unavoidable characteristic of the policy-making exercise.

Perhaps the most noted critic of the rational model of decision-making was the American

behavioral scientist Herbert Simon. In a series of books and articles beginning in the early

1950s, he argued that there were several hurdles that prevented decision-makers from attaining

“pure” comprehensive rationality in their decisions (Simon 1955, 1957). Simon noted deﬁnite

cognitive limits to the decision-makers’ ability to consider all possible options, which forces them

to selectively consider alternatives. He concluded from his assessment of the rational model that

public decisions in practice did not maximize beneﬁts over costs, but merely tended to satisfy

Howlett and Giest

whatever criteria decision-makers set for themselves in the instance in question. This “satisﬁcing”

criterion, as he put it, was a realistic one given the “bounded rationality” with which human

beings are endowed. This analysis did not rule out the use of rationality-forcing techniques but

pointed out that rational models would only result in optimal decisions in very speciﬁc situa-

tions where full information on options existed and could be mobilized by key decision-makers

(Smith and May 1980).

The well-known incremental model of policy-making developed by Yale University political

scientist Charles Lindblom incorporated these insights into limited or “bounded” rationality into

a general model of public policy decision-making (Dahl and Lindblom 1953; Lindblom 1955,

1958, 1959; Hayes 1992). Lindblom summarized the elements of his model as consisting of the

following a “mutually supporting set of simplifying and focusing stratagems”:

 limitation of analysis to a few somewhat familiar policy alternatives … diﬀering only marginally

from the status quo;

 an intertwining of analysis of policy goals and other values with the empirical aspects of the

problem (that is, no requirement that values be speciﬁed ﬁrst with means subsequently

found to promote them);

 a greater analytical preoccupation with ills to be remedied than positive goals to be sought;

 a sequence of trials, errors, and revised trials;

 analysis that explores only some, not all, of the important possible consequences of a considered

alternative;

 fragmentation of analytical work to many (partisan) participants in policymaking (each

attending to their piece of the overall problem domain).

(Lindblom 1979)

This resulted, as he put it in his oft-cited 1959 article on “The Science of ‘Muddling Through’,” in

decision-makers working through a process of “continually building out from the current

situation, step-by-step and by small degrees” or “increments” in which policies were invariably

developed through a process of “successive limited comparisons” with earlier ones with which

decision-makers were already familiar (Lindblom 1959).

Braybrooke and Lindblom (1963) argued that diﬀerent styles of decision-making could be

discerned depending upon the amount of knowledge at the disposal of decision-makers, and the

amount of change the decision involved from earlier decisions. Other authors, like Graham

Allison, also developed similar models of distinct decision-making styles (Allison 1969, 1971),

but did not specify in any detail the variables which led to the adoption of a particular style

(Bendor and Hammond 1992).

Attempting to improve upon these models, John Forester (1984, 1989) suggested that

decision-making is aﬀected by the number of agents involved in a decision, their organiza-

tional setting, how well a problem is deﬁned, the information available on the problem, its

causes and consequences, and the amount of time available to decision-makers to consider possible

contingencies and their present and anticipated consequences. The number of agents can expand

and multiply almost to inﬁnity. The setting can include many diﬀerent organizations and can be

more or less open to external inﬂuences. The problem can be ambiguous or susceptible to

multiple competing interpretations. Information can be incomplete, misleading or purposefully

withheld or manipulated, and time can be limited or artiﬁcially

constrained and manipulated. In

this model, decision-makers

situated in complex subsystems are expected to undertake adjust-

ment strategies while those dealing with simple conﬁgurations of actors and ideas will be more

prone to undertake search-type strategies. The nature of the decision criteria, on the other hand,

The policy-making process

varies with the severity of the informational, time and other resource constraints under which

decision-makers operate. Hence decision-makers faced with high constraints will tend to favor

satisﬁcing over optimization, itself an outcome more likely to occur in situations of low con-

straint. Overall this would lead to a pattern in which most decisions would be incremental

except in speciﬁc circumstances where other kinds of decisions could emerge.

Policy implementation styles

Generally speaking, comparative implementation studies have also shown that governments tend

to develop speciﬁc implementation styles (Knill 1998; Hawkins and Thomas 1989; Kagan 1991;

Howlett 2002b) which combine various kinds of instruments into a more or less coherent whole

which is then applied in particular sectors (Howlett 2011; Wu et al. 2010). Linked to implementa-

tion, policy tools provide the substance or content to what was planned in the formulation stage

and decided upon afterwards in the decision-making stage of the policy process (Howlett 2011).

These tools fall into two types. Substantive instruments are those directly providing goods and

services to members of the public or governments. They include a variety of tools or instruments

relying on diﬀerent types of governing resources for their eﬀectiveness (Tupper and Doern 1981;

Vedung 1997; Woodside 1986; Peters and Van Nispen 1998; Salamon 1989). A useful way to

classify these (see Table 2.1) is according to the type of governing resource upon which they

rely: nodality or information; authority, treasure or ﬁnancial resources, or administrative or

organizations ones (Hood 1986).

Procedural instruments are diﬀerent from substantive ones in that their impact on policy out-

comes is less direct. Rather than aﬀect the delivery of goods and services, their principal intent is

to modify or alter the nature of policy processes at work in the implementation process (Howlett

2000; in’t Veld, 1998). A list of these instruments is provided in Table 2.2.

Why a particular combination of procedural and substantive instruments is utilized in parti-

cular policy issue areas is a key question (Gunningham and Young 1997; Dunsire 1993; How-

lett 2000; Salamon 2002; Clark and Russell 2009; McGoldrick and Boonn 2010). Some studies

suggest that policy mixes can be designed. Studies such as Gunningham, Grabosky and Young’s

work on “smart regulation,” for example, led to the development of eﬀorts to identify com-

plementaries and conﬂicts within instrument mixes or tool “portfolios” involved in more

complex and sophisticated policy designs (Barnett et al. 2009; Buckman and Diesendorf 2010).

Each tool mix decision combines advantages and disadvantages of each tool in its relationship

to other tools as well as the eﬀect it has on costs and beneﬁts for government (Howlett 2011).

Table 2.1 A taxonomy of substantive policy instruments

Principal use Governing resource

Nodality Authority Authority Organization

Effectors  Advice  Licences  Grants  Bureaucratic

training  Certiﬁcation  Loans administration

regulation  Tax expenditures  Public enterprises

user charges

Detectors  Reporting  Census-taking  Polling  Record-keeping

registration consultants policing surveys

Source: Howlett, Michael. Designing Public Policies: Principles and Instruments. New York: Routledge, 2011. Available at:

www.routledge.com/books/details/9780415781336/

Howlett and Giest

The preferences of state decision-makers and the nature of the constraints within which they

operate are key factors which many studies have highlighted as aﬀecting instrument choices and

policy implementation styles (Bressers and O’Toole 1998). A state must have a high level of

administrative capacity, for example, in order to utilize authority, treasure, and organization-

based instruments in situations in which they wish to aﬀect signiﬁcant numbers of policy targets.

When it has few of these resources, it will tend to utilize instruments such as incentives or

propaganda, or to rely on existing voluntary, community or family-based instruments (Howlett

et al. 2009). Similarly a key feature of procedural instrument choice is a government’s capacity to

manipulate policy subsystems. Undertaken in order to retain the political trust or legitimacy

required for substantive policy instruments to be eﬀective (Beetham 1991; Stillman 1974),

procedural policy instrument choice is also aﬀected by the size of the policy target. Whether a

government faces sectoral delegitimation or widespread systemic delegitimation aﬀects the types

of procedural instruments a government will employ (Habermas 1973, 1975; Mayntz 1975).

Policy evaluation styles

The last stage of the cycle is policy evaluation. For many early observers, policy evaluation was

expected to consist of assessing if a public policy was achieving its stated objectives and, if not,

what could be done to eliminate impediments to their attainment. Thus David Nachmias (1979)

deﬁned policy evaluation as “the objective, systematic, empirical examination of the eﬀects ongoing

policies and public programs have on their targets in terms of the goals they are meant to achieve.”

However, while analysts often resorted to concepts such as “success” or “failure” to conclude their

evaluation, as Ingram and Mann cautioned:

[T]he phenomenon of policy failure is neither so simple nor certain as many contemporary

critics of policy and politics would have us believe. Success and failure are slippery concepts,

often highly subjective and reﬂective of an individual’s goals, perception of need, and perhaps

even psychological disposition toward life.

(Ingram and Mann 1980)

Table 2.2 A resource-based taxonomy of procedural policy instruments

Principal use Governing resource

Nodality Authority Treasure Organization

Positive  Education  Agreements  Interest-group

funding

 Hearings

 Exhortation  Treaties  Research  Evaluations

 Advertising  Advisory-group

creation

 Intervenor-

funding

 Institutional-

bureaucratic reform

 Training

Negative  Misleading

information/

propaganda

 Banning groups

and associations

 Eliminating

funding

 Administrative delay

 Information

suppression

Source: Howlett, Michael. Designing Public Policies: Principles and Instruments. New York: Routledge, 2011. Available at:

www.routledge.com/books/details/9780415781336/

The policy-making process

That is, public policy goals are often not stated clearly enough to ﬁnd out if and to what extent

they are being achieved, nor are they shared by all key policy actors. Moreover, the possibilities for

objective analysis are also limited because of the diﬃculties involved in the attempt to develop

objective standards by which to evaluate a government’s level of success in dealing with subjective

claims and socially constructed problems.

What is signiﬁcant in the evaluative process is thus not so much ultimate success and failure, but

that policy actors and the organizations and institutions they represent can learn from the formal

and informal evaluation of policies in which they are engaged. This can lead them to modify their

positions in the direction of greater substantive or procedural policy change, or it can lead them to

resist any alteration to the status quo (Majone 1989). And, like with other stages of the policy process,

a few common or typical styles of learning have been identiﬁed in the literature on the subject.

A signiﬁcant variable in this regard is the capacity of an organization to absorb new information

(Cohen and Levinthal 1990). Only when state administrative capacity is high, for example,

would one expect any kind of learning to occur. If a relatively closed network dominates the

subsystem, then this learning is likely to be restricted to some form of “lesson-drawing,” in

which policy-makers draw lessons from past uses of policy instruments (Rose 1991; Bennett and

Howlett 1992). Open subsystems allow for “social learning” when administrative capacity is high

and more informal evaluation where the ability to adapt is rather low. “Social learning” occurs

when ideas and events in the larger policy community penetrate into policy evaluations. If the

policy subsystem is dominated by closed networks, one would expect to ﬁnd formal types of

evaluation with little substantive impact on either policy instruments or goals and lesson-drawing

attempts (Howlett et al. 2009).

Policy evaluations do not necessarily result in major policy change. That is, while the concept

of evaluation suggests that an implicit “feedback loop” is an inherent part of the policy cycle, in

many cases this loop may not be operationalized (Pierson 1993). Path dependence, in which

policies are set on “trajectories” following some “critical juncture” can hinder policy change

and learning (Pierson 2000). Organizational-institutional properties (Eising 2004; Olsen and

Peters 1996) are also often seen as barriers to learning from policy evaluations.

Conclusion: policy development as policy style

As Lasswell noted in the 1950s, envisioning policy development as a staged, sequential, and itera-

tive process is a useful analytical and methodological device. Methodologically such an approach

reduces the complexity of public policy-making by breaking down that complexity into a small

number of stages and substages, each of which can be investigated alone, or in terms of its rela-

tionship to any or all the other stages of the cycle. The policy cycle idea also helps to answer many

key questions about public policy-making regarding the eﬀectiveness of diﬀerent tools and the

identiﬁcation of bottlenecks in policy processes. The stages allow for the identiﬁcation of typical

actors and actions in diﬀerent phases of tackling a problem which makes it easier to identify

independent and dependent variables in the study of policy processes and behaviour and moves

thinking forward by helping to identify the relatively limited range of styles of activity possible

at each stage of the cycle (Freeman 1985; Coleman 1994; Tuohy 1992; Vogel 1986).

Bibliography

Allison, G. 1969. ‘Conceptual models and the Cuban missile crisis’, American Political Science Review 63(3):

689–718.

—— (1971) Essence of Decision: Explaining the Cuban Missile Crisis. Boston: Little Brown.

Howlett and Giest

Atkinson, M. and W. Coleman. 1989. ‘Strong states and weak states: Sectoral policy networks in advanced

capitalist economies’, British Journal of Political Science 19(1): 47–67.

——and W. Coleman. 1992. ‘Policy networks, policy communities and the problems of governance’,

Governance 5(2): 154–80.

Barnett, Pauline, Tim Tenbensel, Jacqueline Cuming, Clare Alyden, Toni Ashton, Megan Pledger and

Mili Burnette. 2009. ‘Implementing new modes of governance in the New Zealand health system: An

empirical study’, Health Policy 93: 118–27.

Baumgartner, Frank R. and Bryan D. Jones. 1991. ‘Agenda dynamics and policy subsystems’, Journal of

Politics 53(4): 1044–74.

Beetham, D. 1991. The Legitimation of Power. London: Macmillan.

Bendor, Jonathan and Thomas H. Hammond. 1992. ‘Re-thinking Allison’s models’, American Political Science

Review 86(2): 301–22.

Bennett, Colin J. and Michael Howlett. 1992. ‘The lessons of learning: Reconciling theories of policy

learning and policy change’, Policy Sciences 25(3): 275–94.

Braybrooke, David and Charles E. Lindblom. 1963. A Strategy of Decision: Policy Evaluation as a Social Process.

New York: Free Press of Glencoe.

Bressers, Hans Th.A. and Laurence J. O’Toole. 1998. ‘The selection of policy instruments: A network-based

perspective’, Journal of Public Policy 18(3): 213–39.

Buckman, Greg and Mark Diesendorf. 2010. ‘Design limitations in Australian renewable electricity policies’,

Energy Policy 38(7): 3365–76.

Cahill, Anthony G. and E. Sam Overman. 1990. ‘The evolution of rationality in policy analysis’,in

S.S. Nagel (ed.) Policy Theory and Policy Evaluation: Concepts, Knowledge, Causes, and Norms. New York:

Greenwood Press.

Carley, Michael. 1980. Rational Techniques in Policy Analysis.

London: Heinemann Educational

Books.

Carstensen, Martin B. 2011. ‘Ideas are not as stable as political scientists want them to be: A theory of

incremental ideational change’, Political Studies 59(3): 596–615.

Clark, C.D. and C.S. Russell. 2009. ‘Ecological conservation: The problems of targeting policies and

designing instruments’, Journal of Natural Resources Policy Research (1): 21–34.

Cobb, Roger W. and Charles D. Elder. 1972. Participation in American Politics: The Dynamics of Agenda-Building.

Boston: Allyn and Bacon.

—— J.K. Ross and M.H. Ross. 1976. ‘Agenda building as a comparative political process’, American Political

Science Review 70(1): 126–38.

Cohen, Wesley M. and Daniel A. Levinthal. 1990. ‘Absorptive capacity: A new perspective on learning

and innovation’, Administrative Science Quarterly 35: 128 – 52.

Coleman, William D. 1994. ‘Policy convergence in banking: A comparative study’, Political Studies 42:

274–92.

——and Grace Skogstad (eds). 1990. Policy Communities and Public Policy Canada: A Structural Approach.

Mississauga, ON: Copp Clark Pitman.

Dahl, Robert A. and Charles E. Lindblom. 1953. Politics, Economics and Welfare: Planning and Politico-economic

Systems Resolved into Basic Social Processes. New York: Harper and Row.

Dowding, K. 1995. ‘Model or metaphor? A critical review of the policy network approach’, Political Studies

43(1): 136–58.

Dunlop, Claire A. 2009. ‘Policy transfer as learning: Capturing variation in what decision-makers learn

from epistemic communities’, Policy Studies 30(3): 289–311.

Dunsire, Andrew. 1993. Manipulating Social Tensions: Collaboration as an Alternative Mode of Government

Intervention. Cologne: Max Plank Institute.

Eising, R. 2004. ‘Multilevel governance and business interests in the European Union ’, Governance 17(2):

211–45.

Elster, Jon. 1991. ‘The possibility of rational politics’,

in D. Held

(ed.) Political Theory Today. Oxford: Polity.

Forester, John. 1984. ‘Bounded rationality and the politics of muddling through’, Public Administration

Review 44(1): 23–31.

—— 1989. Planning in the Face of Power. Berkeley, CA: University of California Press.

Freeman, Gary P. 1985. ‘National styles and policy sectors: Explaining structured variation’, Journal of Public

Policy 5(4): 467–96.

Freeman, J. Leiper. 1965. The Political Process: Executive - Bureaucratic - Legislative Relations, revised edn. New

York: Random House.

Gawthrop, Louis C. 1971. Administrative Politics and Social Change. New York: St. Martin’s Press.

The policy-making process

Gunningham, N., P. Grabosky and D. Sinclair. 1998. Smart Regulation: Designing Environmental Policy.

Oxford: Clarendon Press.

——and M.D. Young. 1997. ‘Toward optimal environmental policy: The case of biodiversity conservation’,

Ecology Law Quarterly 24: 243–8.

Haas, Peter M. 1992. ‘Introduction: Epistemic communities and international policy coordination’, International

Organization 46(1): 1–36.

Habermas, Jürgen. 1973. ‘What does a legitimation crisis mean today? Legitimation problems in late

capitalism’, Social Research 40(4): 643–67.

—— 1975. Legitimation Crisis. Boston, MA: Beacon Press.

Hawkins, Keith and John M. Thomas (eds). 1989. Making Regulatory Policy. Pittsburgh, PA: University of

Pittsburgh Press.

Hayes, Michael T. 1992. Incrementalism and Public Policy. New York: Longmans.

Hood, Christopher. 1986. The Tools of Government. Chatham, NJ: Chatham House.

Howlett, Michael. 1998. ‘Predictable and unpredictable policy windows: Institutional and exogenous

correlates of Canadian federal agenda-setting’, Canadian Journal of Political Science 3(3): 495–524.

—— 2000. ‘Managing the “hollow state”: Procedural policy instruments and modern governance’, Canadian

Public Administration 43(4): 412–31.

—— 2002a. ‘Do networks matter? Linking policy network structure to policy outcomes: Evidence from

four Canadian policy sectors, 1990–2000’, Canadian Journal of Political Science 35(2): 235–68.

—— 2002b. ‘Policy instruments and implementation styles: The evolution of instrument choice in

Canadian environmental policy’, in D.L. Van Nijnatten and R. Boardman (eds) Canadian Environmental

Policy: Context and Cases. Toronto: Oxford University Press, 25 – 45.

—— 2009. ‘Governance modes, policy regimes and operational plans: A multi-level nested model of

policy

instrument choice and

policy design’, Policy Sciences 42(1): 73–89.

—— 2011. Designing Public Policies: Principles and Instruments. London and New York: Routledge.

——and M. Ramesh. 1998. ‘Policy subsystem conﬁgurations and policy change: Operationalizing the

postpositivist analysis of the politics of the policy process’, Policy Studies Journal 26(3): 466–82

——M Ramesh and Anthony Perl. 2009. Studying Public Policy: Policy Cycles and Policy Subsystems. Toronto:

Oxford University Press.

Ingram, Helen M. and Dean E. Mann. 1980. Why Policies Succeed or Fail. Beverly Hills, CA: Sage.

in’t Veld, Roeland J. 1998. ‘The dynamics of instruments’, in B. Guy Peters and F. K. M. Van Nispen (eds)

Public Policy Instruments: Evaluating the Tools of Public Administration . New York: Edward Elgar, 153–62.

Jennings, B. 1987. ‘Interpretation and the practice of policy analysis’, in F. Fischer and J. Forester (eds)

Confronting Values in Policy Analysis: The Politics of Criteria. Newbury Park, CA: Sage: 128–52.

Jordan, A. Grant. 1990a. ‘Policy community realism versus “new” institutionalist ambiguity’, Political Studies

38(3): 470–84.

—— 1990b. ‘Sub-governments, policy communities and networks: Reﬁlling the old bottles?’, Journal of

Theoretical Politics 2(3): 319–38.

Jordan, G. and K. Schubert. 1992. ‘A preliminary ordering of policy network labels’, European Journal of

Political Research 21(1/2): 7–27.

Kagan, Robert A. 1991. ‘Adversial legalism and American government’, Journal of Policy Analysis and

Management 10(3): 369–406.

Kingdon, John. 1984. Agendas, Alternatives and Public Policies. Boston: Little, Brown.

Knill,

Christoph. 1998. ‘European policies:

The impact of national administrative traditions’, Journal of

Public Policy 18(1): 1–28.

Knoke, David. 1993. ‘Networks as political glue: Explaining public policy-making’, in W.J. Wilson (ed.)

Sociology and Public Agenda. London: Sage, 164–84.

Lasswell, H.D. 1956. The Decision Process: Seven Categories of Functional Analysis. College Park, MD: University

of Maryland.

—— 1971. A Pre-View of Policy Sciences. New York: American Elsevier.

Lindblom, C.E. 1995. Bargaining: The Hidden Hand in Government. Los Angeles, CA: Rand Corporation.

—— 1958. ‘Policy analysis’, American Economic Review 48(3): 298–312.

—— 1959. ‘The science of muddling through’, Public Administration Review 19(2): 79– 88.

—— 1979. ‘Still muddling, not yet through’, Public Administration Review 39(6): 517 –26.

Linder, Stephen H. and B. Guy Peters. 1990. ‘Research perspectives on the design of public policy:

Implementation, formulation, and design’, in Dennis J. Palumbo and D.J. Calista. Implementation and the

Policy Process: Opening Up the Black Box . New York: Greenwood Press.

Howlett and Giest

McGoldrick, Daniel E. and Ann V. Boonn. 2010. ‘Public policy to maximize tobacco cessation’, American

Journal of Preventive Medicine 38(3) Supplement 1: S327–S32.

MacRae, Duncan, Jr. 1993. ‘Guidelines for policy discourse: Consensual versus adversarial’, in Frank

Fischer and John Forester (eds) The Argumentative Turn in Policy Analysis and Planning. Durham, NC:

Duke University Press: 291–318.

Majone, Giandomenico. 1989. Evidence, Argument, and Persuasion in the Policy Process. New Haven, CT:

Yale University Press.

Marin, Bernd and Renate Mayntz (eds). 1991. Policy Networks: Empirical Evidence and Theoretical Considerations.

Frankfurt-am-Main: Campus.

Mayntz, Renate. 1975. ‘Legitimacy and the directive capacity of the political system’, in Leon N. Lind-

berg, Robert Alford, Colin Crouch, and Claus Oﬀe (eds) Stress and Contradiction in Modern Capitalism.

Lexington, MA: Lexington Books: 261–74.

Mintrom, Michael. 2007. ‘The policy analysis movement.’, in L. Dobuzinskis, M. Howlett and D.

Laycock (eds) Policy Analysis in Canada: The State of the Art. Toronto: University of Toronto Press:

7–84.

Nachmias, David. 1979. Public Policy Evaluation: Approaches and Methods. New York: St. Martin’s Press.

Olsen, Johan P. and B. Guy Peters (eds). 1996. Lessons from Experience: Experiential Learning in Administrative

Reforms in Eight Democracies. Oslo: Scandinavian University Press.

Pawson, Ray. 2006. Evidence-Based Policy: A Realist Perspective. London: Sage.

Peters, B. Guy and F.K.M. Van Nispen (eds). 1998. Public Policy Instruments: Evaluating the Tools of Public

Administration. New York: Edward Elgar.

Pierson, Paul. 1993. ‘When eﬀect becomes cause: Policy feedback and political change’, World Politics 45:

595–628.

—— 2000. ‘Increasing returns, path dependence, and the study of politics’, American Political Science Review

94(2): 251–67.

Pross, A. Paul. 1992. Group Politics and Public Policy, Toronto: Oxford University Press.

Rose, Richard. 1991. ‘What is lesson-drawing?’, Journal of Public Policy 11(1): 3–30.

Sabatier,

Paul A. 1987. ‘Knowledge,

policy-oriented learning, and policy change’, Knowledge: Creation,

Diﬀusion, Utilization 8(4): 649–92.

—— 1988. ‘An advocacy coalition framework of policy change and the role of policy-oriented learning

therein’, Policy Sciences 21(2 & 3): 129–68.

—— 1991. ‘Toward better theories of the policy process’, PS: Political Science and Politics 24(2): 144–56.

Salamon, Lester M. (ed.) 1989. Beyond Privatization: The Tools of Government Action. Washington, DC:

Urban Institute.

—— 2002. ‘The new governance and the tools of public action’, in L.M. Salamon (eds) The Tools of

Government: A Guide to the New Governance. New York: Oxford University Press: 117–55.

Sanderson, I. 2006. ‘Complexity, “practical rationality” and evidence-based policy making’, Policy and

Politics 34(1): 115–32.

Schulman, Paul R. 1988. ‘The politics of “ideational policy”’, Journal of Politics 50: 263–91.

Schmidt, Vivien A. 2006. ‘Discursive institutionalism: The explanatory power of ideas and discourse’,

Annual Review of Political Science 1: 303–26.

—— 2010. ‘Taking ideas and discourse seriously: Explaining change through discursive institutionalism as

the fourth new institutionalism?’, European Political Science Review 2(1): 1–25.

—— 2011. ‘Speaking of change: Why discourse is key to the dynamics of policy transformation’, Critical

Policy

Studies 5: 106–26.

Simon, Herbert

A. 1955. ‘A behavioral model of rational choice’, Quarterly Journal of Economics 69(1): 99–118.

—— 1957. Administrative Behavior: A Study of Decision-Making Processes in Administrative Organization, 2nd

edn. New York: Macmillan.

Smith, Gilbert. 1994. ‘Policy networks and state autonomy’, in S. Brooks and A.-G. Gagnon (eds) The

Political Inﬂuence of Ideas: Policy Communities and the Social Sciences. New York: Praeger.

——and David May. 1980. ‘The artiﬁcial debate between rationalist and incrementalist models of decision-

making’, Policy and Politics 8(2): 147–61.

Smith, Martin J. 1993. Pressure, Power and Policy: State Autonomy and Policy Networks in Britain and the United

States. Aldershot: Harvester Wheatsheaf.

Stillman, Peter G. 1974. ‘The concept of legitimacy’, Polity 7(1): 32–56.

Torgerson, Douglas. 1986. ‘Between knowledge and politics: Three faces of policy analysis’, Policy Sciences

19(1): 33–59.

The policy-making process

Tuohy, Caroline. 1992. Policy and Politics in Canada: Institutionalized Ambivalence. Philadelphia, PA: Temple

University Press.

Tupper, Allan and G.B. Doern. 1981. ‘Public corporations and public policy in Canada’, in Allan Tupper

and G.B. Doern (eds) Public Corporations and Public Policy in Canada. Montreal: Institute for Research on

Public Policy: 1–50.

van Waarden, F. 1992. ‘Dimensions and types of policy networks’, European Journal of Political Research

21(1/2): 29–52.

Vedung, Evert. 1997. ‘Policy instruments: Typologies and theories’, in M.L. Bemelmans-Videc, R.C. Rist

and E. Vedung (eds). Carrots, Sticks and Sermons: Policy Instruments and Their Evaluation. New Brunswick,

NJ: Transaction Publishers: 21–58.

Vogel, David. 1986. National Styles of Regulation: Environmental Policy in Great Britain and the United States.

Ithaca, NY: Cornell University Press.

Weiss, Carol H. 1977. Using Social Research in Public Policy Making. Lexington, MA: Lexington Books.

Woodside, K. 1986. ‘Policy instruments and the study of public policy’, Canadian Journal of Political Science

19(4): 775–93.

Wu, Xun, M. Ramesh, Michael Howlett and Scott Fritzen. 2010. The Public Policy Primer: Managing Public

Policy. London: Routledge.

Howlett and Giest

Comparative approaches to the

study of public policy-making

Sophie Schmitt

Introduction: what is comparative public policy analysis?

Scholars of comparative public policy are interested in systematically studying public policies

and their origins in order to gain a better understanding of the causes, factors and institutional or

actor constellations that bring about diﬀerent kinds of policy decisions. They try to advance our

understanding of the processes and determinants of public policy-making by comparing the

situations and contexts that lead policy-makers to agree on similar or diverging policies. In this vein,

these analyses include comparisons over time and/or over units (comparing, for example, national,

state or local governments). In a nutshell, comparative policy researchers seek to explain why

and under what conditions policy-makers agree on what policies (see, for example, Lerner and

Lasswell 1951; Windhoﬀ-Héritier 1987; Howlett and Ramesh 2003). In an oft-cited deﬁnition,

Dye describes public policy as “whatever governments choose to do or not to do” (1976: 1). A

review of a number of subdisciplines of comparative public policy analysis shows that scholars of

diﬀerent research traditions employ diﬀerent strategies in order to probe this inclusive and vast

deﬁnition of “public policy” in both theoretical and empirical respects. While policy science

covers a broad ﬁeld that includes analyses of actors, institutions, instruments, programs, decision-

making processes, policy implementation or evaluation from various theoretical perspectives

(cf. Mayntz 1983; Howlett and Ramesh 2003), the analytical focus of the literature on

comparative policy analysis is on the policy decisions themselves and the explanatory factors.

Research on comparative policy-making can be classiﬁed into three groups according to its

analytical focus. First, while a number of approaches aim to compare public policies as such, other

subdisciplines address the patterns of policy-making or the decision-making processes from a com-

parative perspective. In this context, the ﬁrst category covers the entire policy change literature,

which is rather fragmented with regard to the diﬀerent approaches for conceptualizing and deﬁning

the subject of study (public policy).

Second, the patterns of policy-making are at the focus of scholars of policy diﬀusion, policy

convergence, policy termination or policy dismantling. Whereas the policy diﬀusion perspective

departs from the patterns of spread and the mechanisms of policy change, the policy con-

vergence literature does not try to explain the individual policy-making behavior of govern-

ments but focuses on the degree of similarity their policy decisions reach over time. Scholars of

policy termination and of policy dismantling compare decisions to completely or partially revoke

policies introduced in the past.

Third, the process-oriented strands of comparative policy research focus on particular expla-

natory factors or causes that determine the decision-making of governments. In this context,

research on policy learning or lesson-drawing analyzes how policy actors perceive and process

policy experiences made elsewhere, while the policy transfer perspective describes and explains

how policies travel from one unit (e.g. nation-state) to another. Institutional analysis is a useful

strategy to assess the impact of institutional constraints (e.g. veto players) on the actors’ policy

decision.

However, the broad ﬁeld of comparative policy research also diverges with respect to the

underlying methodological choices and to the range of causal explanations taken into account.

This has led to contradictory results and a variety of identiﬁed factors and constellations that

determine policy-making, its direction, extent and implications. Thus, in order to advance

this ﬁ eld of research as a whole and to conceptually locate related research within the broader

ﬁeld, it is essential to have an overview of its diﬀerent strands of research, approaches and

analytical foci.

This chapter provides a systematic review of the ﬁeld of comparative public policy research

with regard to the ﬁelds of environmental and social policy. The following sections address ﬁrst,

diﬀerent ways of conceptualizing and operationalizing the subject of study (public policy),

second, diﬀerent analytical perspectives of the patterns of public policies, and third, literature with

distinct foci on the processes and explanatory factors of public policy-making. Next, I will provide

a short review of the methods used to compare and analyze decision-making in the ﬁelds of

social and environmental policy. I will then conclude with a short discussion of the main strengths

and shortcomings of current approaches to comparing public policy-making over space and

time along with an outlook on the challenges for further research.

Varieties of comparative public policy research: policies, patterns

or processes

The broad ﬁeld of comparative public policy research comprises of research perspectives or

traditions, theoretical approaches, empirical scopes and research methods. Numerous sub-

disciplines and perspectives have emerged over the last decades. This development reﬂects the

variety—and probably also limited commensurability—of the diﬀerent paradigms guiding

scholars of comparative public policy. So far, no dominant approach has emerged. In the fol-

lowing sections, I identify and discuss four dimensions along which studies of comparative

public policy diverge.

Comparing public policies

There is a variety of di ﬀerent approaches in the literature to depict and measure the subject of

study, namely “public policy.” This conceptual heterogeneity delimits the comparability and

consolidation of the ﬁndings.

Scholars of comparative public policy analysis seek to systematically describe and explain the

decisions of diﬀerent governments, their timing and content. To this end, several strategies

have emerged to determine governmental behavior and decisional output—including for

instance legislative acts, executive decrees or administrative circulars. A ﬁrst group of scholars

measures the impact of policies by analyzing the policies’ eﬀects on the target of regulation. A

second group of researchers focuses on policy outcomes, namely the immediate consequences of a

Schmitt

policy decision, while the third approach refers to policy outputs, namely the content of the

decision. As an example from the social policy ﬁeld, one might consider unemployment pro-

tection policies. In this context, the decision to introduce monetary unemployment beneﬁts at a

certain level reﬂects the immediate policy output. The same decision could also be approxi-

mated by referring to total government spending on this particular social protection scheme

(i.e. the outcome) or by measuring the eﬀects of the policy decision on the regulatory target,

such as, for instance, the number or the quality of life of unemployed persons (i.e. the impact).

In the ﬁeld of environmental policy-making one could consider emission standards for industrial

plants. The deﬁnition of the standard itself, such as a limit value for the emission of sulfur

dioxide, connotes the policy output, whereas the policy impact would be measured by the

associated change in industrial emissions and air quality with respect to this particular sub-

stance. These brief examples illustrate the considerable risk of incomparability and incommen-

surability that result from the conceptualization and operationalization of the dependent variable

(government policy decisions). From a policy cycle perspective, policy output on the one

hand and policy outcome or impact on the other refer to diﬀerent stages of the policy process.

While the former is more a part of the policy formulation stage, the latter concepts pertain to

the implementation phase (see Windhoﬀ-Héritier 1987: 18f; cf. Sabatier and Mazmanian 1980,

Howlett and Ramesh 2003: 143ﬀ.). There is, however, a divide between policy outputs

and their consequences (outcome and impact) not only with respect to the temporal patterns

but also with regard to the potentially high number of intervening factors that inﬂuence the

eﬀects of policy decisions. It is almost impossible to infer governmental decisions and intentions

based on indirect outcome and impact measures. Jänicke (1992) for instance ﬁnds that envir-

onmental policy performance and success crucially depend on institutional arrangements, eco-

nomic structures and international dynamics. Esty and Porter (2005) conducted an extensive

study in which a variety of diﬀerent factors were analyzed to examine their inﬂuence on

environmental performance (outcome data). The authors ﬁnd other factors besides institu-

tional arrangements to have an impact on the policy outcome in addition to the policy

outputs themselves including socio-economic and infrastructure-related factors. Thus, as this

example shows, selecting indicators for the subject of analysis or the dependent variable is far

from trivial.

The literature on both social and environmental policy shows that the majority of large-n

comparative studies rely on outcome or impact instead of output measures due to eased data

availability. Wälti (2004), for instance, approximates the diﬀerent nation-states’ prioritization of

environmental protection by using air quality data on nitrogen and sulfur oxides levels. Others

argue that policy outcomes are the more appropriate and feasible way to measure a government’s

commitment toward environmental protection since its willingness to reduce air pollution is not

only reﬂected in single policy decisions (e.g. to lower limit values). It is rather the result of a series

of decisions including institutional ones that cannot be comprehensively assessed by comparing

single policy outputs (see Scruggs 2003: 20ﬀ.). As a consequence, many scholars have opted for

environmental performance indices that are designed to reﬂect the policy-makers ’ overall

commitment for protecting the environment. These indices are compiled from various outcome

measures. They diﬀer in their breadth in terms of the number of included pollutants (cf. Crepaz

1995; Jänicke 1996; Jahn 1998; Scruggs 1999, 2003).

Impact and outcome measures are also common in the literature on social policy. This might

also be due to the conceptual and notional focus given that most scientists belonging to this area

of research focus on the analysis of the welfare state as a whole rather than the explanation of

single social policies. Analogously to the ﬁeld of environmental policy research, policy output

studies are uncommon. Nevertheless, scholars of social policies are usually more careful to

Studying public policy-making

distinguish between approaches focusing on outcome measures and those addressing the poli-

cies’ impact. The former branch of research usually seeks to indirectly measure the politicians’

commitment with respect to certain policy ﬁelds while the latter research takes more the form

of program evaluation studies. Outcome indicators in the ﬁeld of social policy diﬀer with regard

to their conceptual closeness to the underlying policy decision.

On rather abstract levels, data on public expenditure allow for the comparison of public

commitment to policies, governments or time. Expenditure data are usually easily available proxies

for policy decisions. They are often used in quantitative comparative studies (cf. Stephens et al.

1999; Huber and Stephens 2001; Castles 2004, 2009; Clasen and Siegel 2007). However, as

with the case of environmental policy the direct linkage between policy outputs and outcomes

is questionable. A number of authors ﬁnd alternative factors that have an impact on public

spending levels (Wilensky 1974; Castles and Mitchell 1992; Pierson 1996; Castles 1999; Kit-

schelt 2001; Allan and Scruggs 2004). The comparison between the politics of pension reform

and unemployment protection, for instance, illustrates the complex interplay between socio-

economic conditions, government action and welfare expenditure and the risk of endogeneity

for the resulting analyses. On the one hand, demographic change and aging of the population

are likely to trigger policy-making by governments which then translates into (expected)

expenditure levels for this speciﬁc social policy program (Bonoli 2000). On the other hand,

socio-economic conditions might also lead decision-makers to adapt unemployment protection

policies. This is quite evident in the ﬁeld of unemployment policies, where altered rates of

unemployment directly impact on the resulting expenditure levels of the protection scheme

(Allan and Scruggs 2004). Again, these developments might motivate policymakers to take

action (Vis 2009). As these brief examples show, multidirectional eﬀects cause problems of

endogeneity which in turn pose serious risks for the validity of the ﬁndings obtained from such

analyses. Furthermore, expenditure analyses entail the risk of omissions, as stable outcome levels of

social policy might obscure underlying changes due to altered contexts (e.g. higher unemployment

replacement rates that coincide with lower unemployment rates, cf. Knill et al. 2010).

As a consequence of these problems and the goal to more closely reﬂect actual political deci-

sions, scholars have turned toward the measurement of replacement rates instead of cumulated

expenditure levels (Martin 1996; Green-Pedersen 2004; Swank 2005; Clasen and Clegg 2007;

Kühner 2007). Replacement rates refer to the amount of monetary compensation paid to

retirees, unemployed persons or those unable to work because of illness. Social generosity from

this perspective is operationalized through the beneﬁt levels (e.g. pension, unemployment or

child allowances) for such people, usually industrial workers (Allan and Scruggs 2004). This strategy

allows for consistent cross-country or cross-time comparisons for certain sub-aspects of policy-

making. A major problem with this approach is that available data on replacement levels cover

only few aspects of the policy. On the one hand, the use of replacement rates is biased toward

policy decisions with monetary implications. On the other hand, by focusing on the beneﬁts for

representative standard employees, such studies often also neglect the full range of policy deci-

sions with regard to the diﬀerent levels of beneﬁts. Thus, cross-country comparisons on the

basis of replacement rate indicators only incur serious risk of biased results due the systematic

focus on the monetary implications of social policies (Whiteford 1995; Lynch 2001; Bonoli

2007; Scruggs 2007; Knill et al. 2010).

The recent literature on social policy-making, however, shows a number of more sophisticated

ﬁne-grained analytical frames to study and compare policy-making in diﬀerent countries. Immergut

et al. (2007) as well as Häusermann (2010) systematically analyze and compare policy decisions

in the ﬁeld of pension politics in diﬀerent countries. Altogether, the decision of how to con-

ceptualize and measure the subject of study (i.e. public policy) is of crucial relevance as it impacts

Schmitt

on the ability to describe and explain public policy-making. The comparative approach

often requires a certain degree of pragmatism and simpliﬁcation with respect to the oper-

ationalization of the underlying concepts—particularly when it comes to designing quantitative

large-n studies. However, as this brief overview of diﬀerent ways of conceptualizing and oper-

ationalizing the dependent variables in comparisons of public policy-making illustrates, these

decisions might considerably inﬂuence the validity of the ﬁndings due to the choice of variables

and indicators.

In this context, policy (change) analyses with a direct focus on policy output are one way to

assess the underlying decision-making of governments—hence covering various stages of the

policy cycle including agenda-setting, policy formulation and decision-making (see Howlett and

Ramesh 2003: 162). They represent the eventual consensus that has been reached by the group

of involved policy actors. Policy outcome studies are based on rather indirect measures of the

underlying policy decisions. These include performance indices or pollution data for certain sub-

stances and expenditure data for the ﬁelds of environmental and social policy, respectively.

Given the supposedly complex processes that account for the outcomes, the direction of causal

mechanisms is far from evident, which again creates risks of endogeneity. Scholars of quantita-

tive comparative environmental policy research have partly acknowledged this challenge by

including certain policy outputs (e.g. limit values for emission) as only one set of explanatory fac-

tors for predicting policy outcomes (see above; cf. in particular Esty and Porter 2005). Nevertheless,

given the bi-directional causation between policy output and outcome, one could argue that

policy outcomes are crucial in explaining follow-up policy-making. As outcomes reﬂect the

eﬀectiveness of previous policy decisions (through records of either ﬁnancial expenditure or

pollution) they might also create incentives for future policy-making.

Furthermore, the heterogeneity of the diﬀerent approaches found in the literature impacts on the

degree of comparability of the results of diﬀerent studies themselves. Even within the subﬁeld of

outcome studies, there are currently many diﬀerent strategies to design environmental perfor-

mance indices or to compile expenditure indicators for social policies. This arbitrariness might

explain the limited comparability and inconsistency of the ﬁndings.

Comparing patterns of public policy-making

The above discussion on the diﬀerent ways to conceptualize and operationalize “public policies”

in comparative research on environmental and social policy-making illustrates that governmental

behavior and decision-making can be measured at diﬀerent levels of abstraction depending on

the scientiﬁc aim of the research endeavor. As a consequence of the diﬃculty in comparing the

policy-making behavior of governments in its entirety, various research traditions have emerged

to analyze and compare public policies from distinct perspectives. In this section, I present and

discuss three diﬀerent research traditions that analyze public policy-making from a comparative

perspective by taking a particular analytical lens and focusing on the observed patterns of policy

change or adoption.

First, studies of public policy can be classiﬁed according to the direction or range of policy

change that follows from a government’s choices. While the majority of scholars include positive

as well as negative decisions in the analysis, scholars of policy termination or dismantling focus exclu-

sively on decisions that are of reductive nature through the partial or complete abolishment of

policies.

Research on policy termination was initiated by Brewer in the early 1970s with the objective

directing scientiﬁc attention toward this ﬁnal stage of the policy cycle process. He deﬁned this

type of policy choice as those adjusting “policies and programs that have become dysfunctional,

Studying public policy-making

redundant, outmoded, and/or unnecessary” (Brewer 1978: 338). His initiative led to a vital but

short-lived scholarly reaction to analytically advance research into this aspect of policy-making

(cf. Bardach 1976; Biller 1976; Behn 1978; deLeon 1978; Hogwood and Peters 1982). Despite

persistent theoretical problems and inconsistencies, a number of empirical studies followed

(e.g. Frantz 1992; Daniels 1995, 2001; deLeon and Hernández-Quezada 2001; Geva-May 2001;

Harris 2001; Graddy and Ye 2008). The policy termination approach develops types or forms of

policy termination that are of diﬀerent scope and that are debated between groups of opponents

and proponents of diﬀerent strengths (cf. in particular Bardach 1976).

Policy dismantling is a related concept in that it focuses on policy decisions to reduce public

generosity (distributive policies like social policy) or water down the rigidity of regulatory

policies (e.g. environmental policies). Policy dismantling has initially been studied in the context

of welfare state analyses. Scholarly attention has continuously been at a high level since Pierson’s

seminal book (1994). So far, various methods have been applied to analyze such policy decisions

mainly in the ﬁeld of social or welfare state policy (e.g. Esping-Andersen 1996; Clayton and

Pontusson 1998; Bonoli 2001; Kitschelt 2001; Green-Pedersen 2002; Korpi and Palme 2003;

Hacker 2004; Starke 2006, Starke et al. 2008).

A second criterion for case selection is based on the empirically observed patterns of spread of

certain policy innovations. The broad literature of policy diﬀusion seeks to explain the dissemination

of certain policies among units at diﬀerent levels (i.e. nation-states, states or local entities) by

identifying causes and factors that promote policy diﬀusion. In order to reduce complexity in

comparing public policy-making over cases or units, policy diﬀusion scholars analyze single

policies or policy innovations by describing the underlying adoption patterns over time and

units. Assuming that policy choices are “interdependent” (e.g. Braun and Gilardi 2006: 299),

diﬀusion scholars try to explain the spread of policies by identifying the relevant domestic and

international factors at work. This way, diﬀusion scholars seek to estimate the international

inﬂuence on domestic policy decisions (cf. Frank et al. 2000; Busch et al. 2005; Elkins and

Simmons 2005; Elkins et al. 2006). Since its emergence in the 1970s (Gray 1973; Collier and

Messick 1975), policy diﬀusion research has become ever more sophisticated especially with respect

to the development of ambitious research methods to detect diﬀusion channels (cf. Tyran and

Sausgruber 2005; Brooks 2007; Volden et al. 2008; Marsh and Sharman 2009; Plümper and

Neumayer 2010).

Third, the policy convergence perspective is another way of comparing the patterns of policy-

making of diﬀerent nation-states, states or local entities by focusing on and explaining the

degree of similarity they reach over time as a consequence of passing policies of similar

contents (Bennett 1991; Knill 2005). This rather indirect analysis of public policies does not

seek to explain the actual decision of a government but how it changes the status quo in rela-

tion to other cases (e.g. nation-states). There are a number of studies dealing with the degree of

transnational convergence in the ﬁelds of environmental and social policy (cf. Heichel et al.

2005; Jordan 2005; cf. Overbye 1994; Howlett 2000; Starke et al. 2008). The analytical moti-

vation of convergence scholars has often been twofold: to describe the dynamics of policy

change over time and in relation to other units, and to explain the observed patterns. This way,

they address the question of whether and under which conditions globalization leads to “races

to the top” or “races to the bottom” in for instance the adoption of environmental policies (see

e.g. Holzinger et al. 2008; cf. Drezner 2001, 2005; Busch and Jörgens 2005; Holzinger and

Knill 2005).

This brief summary of diﬀerent analytical strands, that each address certain sub-aspects of policy-

making from a comparative perspective, highlights the complexity of the subject of investiga-

tion. Comparing public policies over space and time often requires narrowing the analytical lens

Schmitt

in favor of a more selective focus on policy-making. As a consequence of this conceptual sim-

pliﬁcation the above approaches suﬀer from certain theoretical limitations that impede on the

generalizability of the results with respect to public policy-making in general.

First, the risk of biased results cannot be excluded due to the one-directional analysis of

policy-making that is inherent to, for instance, the subdisciplines of policy dismantling, policy

termination and policy diﬀusion research. While the former two perspectives make statements

about government decisions of reductive nature, the latter have traditionally exclusively dealt

with expansion policies (although this does not occur by deﬁnition).

Second, biases are likely to aﬀect those studies that select their subject of study (i.e. the policy)

based on its empirically observed distribution. In other words, results from diﬀusion studies

might only allow for valid inference with respect to policies that have reached a certain degree

of dissemination. Policies that do not reach the stage of a “best practice” are ex ante excluded

from analysis, which is why large-n diﬀusion studies are most likely to overestimate the impact

of international factors, such as harmonization (for a critical analysis of the limitations of this

particular research perspective cf. Howlett and Rayner 2008).

Comparing processes of public policy-making

In the following paragraphs, I will provide a further criterion along which to classify the lit-

erature on comparative public policy: the processes and factors of public policy-making. While

the conceptualization and operationalization of the dependent variable (see above) crucially inﬂu-

ences the validity of the ﬁndings, the selection of policies according to their patterns of spread

(see above) primarily impacts on the generalizability of the results toward public policy-making

in general.

There are several subdisciplines of comparative research on public policies that explain policy

decisions or outputs by focusing on selective explanatory factors or processes. These include

scholars of policy learning and policy transfer as well as research on political institutions. One

common feature of these studies is the objective to analyze and compare the respective pro-

cesses of the decision-making and the behavior of the involved actors. As a consequence, the

results of these studies are often ex ante focused on speciﬁc explanations.

One way of narrowing down the empirical focus of comparative public policy analysis can be

found within the literatures on policy learning, policy transfer or emulation. Scholars belonging to

these research traditions identify certain mechanisms or factors that cause the dissemination of

policies with the aim of tracing and explaining the underlying processes. Similarly to the diﬀu-

sion perspective, they describe the way in which policies travel from one unit (e.g. nation-state) to

another by focusing on distinct mechanisms or ways of communication. As opposed to the

previously discussed research perspective, policy transfer, learning or emulation take a process

orientation. In this context, policy learning (or lesson-drawing) refers to instances in which gov-

ernments adopt a certain policy based on the evaluation of a policy in place elsewhere (for a

thorough discussion of diﬀerent forms of horizontal and vertical learning see Bennett and Howlett

1992; cf. Sabatier 1987; Rose 1991, 1993; Hall 1993; Meseguer 2006, 2009; Radaelli 2008).

Further, lesson-drawing may cause governments to decide on policies found elsewhere as “trans-

fers can be voluntary or coercive or combinations thereof ” (Stone 2001: 9). Thus, the policy

transfer literature analyzes the role of actors by pointing to the importance of interest groups,

NGOs, think tanks, non-state policy entrepreneurs (as transfer agents) and the role of policy

networks in promoting the diﬀusion of ideas and their intersubjective perceptions of solutions as

well as in providing social learning and building up epistemic communities (cf. Dolowitz and

Marsh 1996, 2000, Radaelli 2000; Stone 2000, 2001, 2004). Cases of negative lesson-drawing,

Studying public policy-making

in turn, might also lead governments to deliberately exclude policy options found elsewhere

(cf. Rose 1991, 1993).

The literature subsumed under the label of institutionalism provides another framework for the

analysis of government choices by explaining policies and their content based on the institu-

tional conditions that channel the decision-making processes. The point of departure of this

approach is veto players, “individual or collective decision-makers whose agreement is required

for the change of the status quo” (Tsebelis 2000: 442). According to this view, policy-making is

best explained by the systematic analysis of the institutional rules that delimit the actors’ leeway

and inﬂuence on the eventual policy decision (cf. Tsebelis 1995; Hall and Taylor 1996; Scharpf

1997; Jupille and Caporaso 1999). Veto player (Bonoli 2001; Immergut and Anderson 2007)

and institution-centered analysis (Swank 2001; Streeck and Thelen 2005) are common approa-

ches in the comparative analysis of, for instance, social policy change. As the “veto players

approach emphasizes the ways in which political decision-making institutions structure attempts

to change the legislative status quo” (Immergut and Anderson 2007: 7), it allows for explaining

policy diﬀerences between units (e.g. nation-states) for a range of issues. There are a number of

studies that illustrate the role of the fragmentation of political power (horizontal or vertical) for

reforming the pension systems of Western democracies (cf. Bonoli 2001, 2004; Swank 2001;

Baccaro 2002; Ferrera 2007; Baccaro and Simoni 2008; Häusermann 2010).

Research of comparative public policies that belong to the sub-disciplines of policy learning,

policy transfer or institutional analysis puts the analytical focus on understanding and describing

the processes that impact on and shape the policy decisions. As a consequence, the empirical

results of these studies might be of limited generalizability to policy-making in its complexity.

Given the individual conceptual points of departure, scholars belonging to these research traditions

are at risk to overestimate the inﬂuence of certain explanatory concepts either through poten-

tially biased case selection (policy learning or transfer) or by the analytical focus on one expla-

nation only (institutionalism). As studies pertaining to these research traditions tend to select

their cases according to speciﬁc causal processes, their ﬁndings might not hold for explaining

policy-making in general.

Altogether, the degree of speciﬁcity of the dependent variable diﬀers over the analytical per-

spectives discussed in the previous two sections, which furthermore reduces the generalizability

of the results for public policy-making.

While transfer and diﬀusion studies solely explain whether a certain policy has been passed or

not, policy learning studies allow for a more detailed conceptual calibration. Similarly, the analy-

tical precision of policy termination studies is rather low; dismantling or retrenchment studies in

turn apply reﬁned conceptual approaches by considering diﬀerent sub-aspects of policy outputs.

Policy convergence in this context, also allows for a multidimensional perspective on the depen-

dent variable (i.e. a government’s decisions with respect to a range of regulatory aspects; e.g. the

emission standards designed to assess a country’s prioritization of regulating air quality). Never-

theless, they provide only indirect and selective accounts of the governments’ actual decisions

because they measure policy-making not in absolute terms but with respect to a certain threshold

(i.e. the average policy output of the sample included in the analysis).

Research methods and causal explanations in comparative public policy

There is a large number of empirical studies that analyze social and environmental policies based

on diﬀerent methodologies (qualitative versus quantitative) and levels of analysis (micro versus

macro). Interestingly, the methodological lines of division are almost identical to the ones

separating the various research traditions discussed above.

Schmitt

More precisely, the question of whether to study policy outcomes instead of policy outputs is

predominantly an issue for scholars of quantitative research. Given the diﬃculty of constructing

and obtaining comparable measures for policy-making over large numbers of cases, macro-

quantitative studies often switch to outcome indicators due to improved data availability. How-

ever, authors of case studies (e.g. Weidner and Jänicke 2002; Jordan et al. 2003; Jordan and

Lieﬀerink 2004; Busch and Jörgens 2005; Hall 1993; Kersbergen 1995; Huber and Stephens 2001;

Taylor-Gooby 2004) seem to be less aﬀected by the “dependent variable” problem (Green-

Pedersen 2004; Clasen and Siegel 2007; cf. Howlett and Cashore 2009) by taking a micro-analytical

perspective.

The division between quantitative and qualitative studies continues also within the diﬀerent

subdisciplines of comparative public policy research as discussed earlier in this chapter. In this

context, policy learning and lesson-drawing are mostly addressed through comparative case

study designs (for an exception see Sommerer 2010), while diﬀusion is often analyzed through

large-n designs (but see, for example, Tews et al. 2003).

The policy change literature also reﬂects the methodological divide. While the policy termi-

nation literature almost exclusively consists of comparative case studies, welfare state retrench-

ment research of policy outcomes (comprising studies based on expenditure or replacement rate

data) usually involves large-n studies. In this context, analyses of the dismantling or termination

of environmental policies are still in general the exception (but see, for example, Knill et al.

2008). While research on policy dismantling or termination tends to overestimate the factors

that lead to negative policy choices, the diﬀusion and transfer literature is biased toward policy

innovations and positive policy decisions, respectively. In spite of this potential bias, there have

been hardly any eﬀorts in comparative analyses of social or environmental policies to system-

atically compare policy outputs of both directions, with the possible exception of the literature

on policy convergence. As already mentioned, output studies that provide in-depth analyses of

policies or policy change in social or environmental policy are almost exclusively limited to

qualitative case studies due to their conceptual and empirical complexity.

The literature on comparative social and environmental policy shows biases with respect to the

explanatory factors taken into account. Policy diﬀusion, convergence or welfare- state retrench-

ment studies in these policy ﬁelds, for instance, tend to take a macro perspective focusing par-

ticularly on the eﬀects of globalization and economic pressure (Castles 1998; Drezner 2001;

Sykes et al. 2001; Weidner and Jänicke 2002; Korpi and Palme 2003; Stone 2004; Swank 2005;

Holzinger et al. 2008). Looking at the role of political parties for public policy-making (from a

meso perspective), for instance, there is an interesting fragmentation between the literatures on

environmental and social policy. While this factor has received abundant attention in research

on the welfare state in general and welfare state retrenchment in particular (e.g. Huber and

Stephens 2001; Kitschelt 2001; Green-Pedersen 2002, 2007; Korpi and Palme 2003; Allan and

Scruggs 2004; Vis and Kersbergen 2007; Vis 2009), it has less frequently been studied in the

context of environmental policy-making (but see Shipan and Lowry 2001; Knill et al. 2010).

Similarly, the two ﬁelds of research diﬀer with respect to the incorporation of institutional

explanations for policy-making. The concept of veto players is often applied to analyze reforms

of the welfare state and to explain the feasibility of cutbacks in public generosity (Bonoli 2001;

Crepaz 2001; Ganghof 2003; Immergut et al. 2007; Ha 2008). However, it is a concept scarcely

addressed in comparative analyses of environmental policy research (for an exception see Jahn

and Müller-Rommel 2010). In addition, institutional factors have received limited attention in

this ﬁeld of research (but see Dryzek et al. 2003).

This brief overview on the methods used and causal explanations tested in the analysis of

social and environmental policy-making points to systematic diﬀerences between the approaches

Studying public policy-making

used in the two policy ﬁelds. Choices of methods and explanatory factors, which determine the

comparability of results, seem to be related to both the analytical perspective on the dependent

variable and the ﬁeld of research (e.g. social or environmental policy). These patterns of com-

parative policy research illustrate the internal fragmentation of this scientiﬁ c discipline and the

limits of generalizability of its ﬁndings.

Discussion and outlook

This brief presentation and discussion of diﬀerent approaches toward the study of public policies

from the comparative perspective shows that we are still far from academic consensus with

respect to the meaning of the term “public policy” itself and in view of the range of explanatory

factors to be considered. Scholars of comparative public policy usually make an implicit statement

through either the conceptualization of the dependent variable, case selection based on certain

policy patterns, by focusing on certain processes or by choosing speciﬁc explanatory concepts.

These choices entail fundamental consequences for the expected degree of generalizability of

the results and the scope of explanation (see below Table 3.1).

On the one hand, process-oriented approaches toward public policy-making reveal a ten-

dency to focus on the explanatory factors (i.e. lesson-drawing, transfer or institutional determi-

nants in this context). On the other hand, pattern-oriented studies (termination, dismantling,

and diﬀusion) tend to put more analytical emphasis on explaining and comparing certain types

of public policies themselves. Both subdisciplines share the narrow perspective on the subject of

study, i.e. public policy-making. As a consequence, the approaches are limited to policies with

certain characteristics or policy aspects. In this context, the literatures on policy learning, transfer

or institutional change consider policies that are inﬂuenced by distinct explanatory factors. Policy

termination and policy dismantling studies incorporate exclusively negative policy decisions,

while research on policy diﬀusion is biased toward policy innovations that have reached a certain

degree of dissemination.

These features might limit the generalizability of all these approaches. The policy change

literature claims to broaden the analysis of policy-making—at least theoretically—by means

of an inclusive conceptualization of the dependent variable and explanatory processes and

factors (for a discussion of the deﬁcits in current policy change research, see Howlett and

Cashore 2009). As shown, however, in the discussion on the diﬀerent ways to conceptualize

Table 3.1 Subdisciplines of comparative public policy research

Focus on generalizability

Low High

Focus on Explanans Policy termination Policy change

(outcome/impact analyses)

Policy dismantling

Policy diffusion

Explanandum Policy learning Policy change

(output analyses)

Policy transfer

Institutionalism

Source: Author

Schmitt

and measure public policies in practice, the policy change perspective also has its limits of

comparability.

This fragmentation in current public policy analysis illustrates the potential beneﬁts of more

integrated approaches that on the one hand, give primary attention to a more careful con-

ceptualization of the subject of research and comparison. On the other hand, the ﬁndings call for

more ﬂexible combinations of research perspectives, methods, and diﬀerent explanatory factors in

both ﬁelds of research. This would enhance the generalizability and the validity of the ﬁndings and

enable theoretical progress to be made based on the consolidation of the diﬀerent approaches.

Bibliography

Allan, J.P. and L. Scruggs. 2004. ‘Political partisanship and welfare state reform in advanced industrial

societies’, American Journal of Political Science 48(3): 496–512.

Baccaro, L. 2002. ‘Negotiating the Italian pension reform with the unions: Lessons for corporatist theory’

Industrial and Labor Relations Review 55(3): 413– 31.

——and M. Simoni. 2008. ‘Policy concertation in Europe: Understanding government choice’, Comparative

Political Studies 41(10): 1323–48.

Bardach, E. 1976. ‘Policy termination as a political process’ , Policy Sciences 7(2): 123–31.

Behn, R.D. 1978. ‘How to terminate a public policy: A dozen hints for the would-be terminate’, Policy

Analyses 4(3): 393-413.

Bennett, C.J. 1991. ‘What is policy convergence and what causes it?’, British Journal of Political Science 21(2):

215–33.

——and M. Howlett. 1992. ‘The lessons of learning: Reconciling theories of policy learning and policy

change’, Policy Sciences 25(3): 275–94.

Biller, R.P. 1976. ‘On tolerating policy and organizational termination: Some design considerations’, Policy

Sciences 7(2): 133–49.

Bonoli, G. 2000. The Politics of Pension Reform: Institutions and Policy Change in Western Europe. Cambridge:

Cambridge University Press.

—— 2001. ‘Political institutions, veto points, and the process of welfare state adaptation’, in P. Pierson

(ed.) The New Politics of the Welfare State. Oxford and New York: Oxford University Press, 238– 64.

—— 2004. ‘The institutionalization of the Swiss multi-pillar pension system’,inM.ReinandW.Schmähl(eds)

Rethinking the Welfare State: The Political Economy of Pension Reform. Northampton, MA: Edward Elgar, 102–21.

—— 2007. ‘Too

narrow and too

wide at once: The “welfare state” as dependent variable in policy ana-

lysis’, in J. Clasen and N.A. Siegel (eds) Investigating Welfare State Change: The ‘Dependent Variable Problem’ in

Comparative Analysis. Cheltenham: Edward Elgar, 24–42.

Braun, D. and F. Gilardi. 2006. ‘Taking “Galton’s problem” seriously: Towards a theory of Policy diﬀusion.

Journal of Theoretical Politics 18(3): 298–322.

Brewer, G.D. 1978. ‘Termination: Hard choices—harder questions’, Public Administration Review 38(4): 338–44.

Brooks, S.M. 2007. ‘When does diﬀusion matter? Explaining the spread of structural pension reforms

across nations’, The Journal of Politics 69(3): 701–15.

Busch, P.-O. and H. Jörgens. 2005. ‘International patterns of environmental policy change and convergence’,

European Environment 15(2): 80–101.

—— H. Jörgens and K. Tews. 2005. ‘The global diﬀusion of regulatory instruments: The making of a new

international environmental regime’, The Annals of the American Academy of Political and Social Science

598(1): 146–67.

Castles, F.G. 1998. Comparative Public Policy: Patterns of Post-war Transformation. Cheltenham: Edward Elgar.

—— 1999. ‘Decentralisation and the post-war political economy’, European Journal of Political Research

36(1): 27–53.

—— 2004. The Future of the Welfare State: Crisis Myths and Crisis Realities. Oxford: Oxford University Press.

—— 2009. ‘What welfare states do: A disaggregated expenditure approach’, Journal

of Social Policy 38,

45–62.

——and D. Mitchell. 1992. ‘Identifying welfare state regimes: The links between politics, instruments and

outcomes’, Governance: An International Journal of Policy and Administration 5(1): 1–26.

Clasen, J. and D. Clegg. 2007. ‘Levels and levers of conditionality: Measuring change within welfare

states’, in J. Clasen and N.A. Siegel (eds) Investigating Welfare State Change: The ‘Dependent Variable Problem’

in Comparative Analysis. Cheltenham: Edward Elgar, 166–97.

Studying public policy-making

——and N.A. Siegel (eds). 2007. Investigating Welfare State Change: The ‘Dependent Variable Problem’ in

Comparative Analysis. Cheltenham: Edward Elgar.

Clayton, R. and J. Pontusson. 1998. ‘Welfare-state retrenchment revisited: Entitlement cuts, public sector

restructuring, and inegalitarian trends in advanced capitalist societies’, World Politics 51(1): 67–98.

Collier, D. and R.E. Messick. 1975. ‘Prerequisites versus diﬀusion: Testing alternative explanations of

social security adoption’, The American Political Science Review 69(4): 1299–315.

Crepaz, M.M.L. 1995. ‘Explaining national variations of air pollution levels: Political institutions and their

impact on environmental policy-making’, Environmental Politics 4(3): 391–414.

—— 2001. ‘Veto players, globalization and the redistributive capacity of the state: A panel study of 15

OECD countries’, Journal of Public Policy 21(1): 1–22.

Daniels, M.R. 1995. ‘Organizational termination and policy continuation: Closing the Oklahoma public

training schools’, Policy Sciences 28(3): 301–16.

—— 2001. ‘Policy and organizational termination’, International Journal of Public Administration 24(3): 249–62.

deLeon, P. 1978. ‘Public policy termination: An end and a beginning’, Policy Analyses 4(3): 369–92.

——and J.M. Hernández-Quezada. 2001. ‘The case of the national solidarity program in Mexico. A study

in comparative public policy termination’, International Journal of Public Administration 24(3): 289–309.

Dolowitz, D.P. and D. Marsh. 1996. ‘Who learns what from whom: A review of the policy transfer literature’,

Political Studies 44(2): 343–57.

—— 2000. ‘Learning

from abroad: The

role of policy transfer in contemporary policy making’, Governance:

An International Journal of Policy and Administration 13(1): 5–24.

Drezner, D.W. 2001. ‘Globalization and policy convergence’, International Studies Review 3(1): 53–78.

—— 2005. ‘Globalization, harmonization, and competition: The diﬀerent pathways to policy convergence’,

Journal of European Public Policy 12(5): 841–59.

Dryzek, J.S., D. Downes, C. Hunold and D. Schlosberg. 2003. Green States and Social Movements: Envir-

onmentalism in the United States, United Kingdom, Germany, and Norway. Oxford and New York: Oxford

University Press.

Dye, T.R. 1976. Policy Analysis: What Governments Do, Why They Do it, and What Diﬀerence it Makes.

Tuscaloosa, AL: University of Alabama Press.

Elkins, Z., A.T. Guzman and B.A. Simmons. 2006. ‘Competing for capital: The diﬀusion of bilateral

investment treaties, 1960–2000’, International Organization 60(4): 811–46.

—— and B. Simmons 2005. ‘On waves, clusters, and diﬀusion: A conceptual framework’, The Annals of

the American Academy of Political and Social Science 598(1): 33–51.

EspingAndersen, G. (ed.). 1996. Welfare States in Transition: National Adaptations in Global Economies.

London and Thousand Oaks, CA: Sage.

Esty, D.C. and M.E. Porter. 2005. ‘National environmental performance: An empirical analysis of policy

results and determinants’, Environment and Development Economics 10(4): 391–434.

Ferrera, M. 2007. ‘The European welfare state. Golden achievements and silver prospects’, URGE

working paper, 4/07.

——and M. Jessoul. 2007. ‘Italy: A narrow gate for path-shift’, in E.M. Immergut, K.M. Anderson and

I. Schulze (eds) The Handbook of West European Pension Politics. Oxford, New York: Oxford University

Press, 396–453.

Frank, D.J., A. Hironaka and E. Schofer. 2000. ‘The

nation-state and the

natural environment over the

twentieth century’, American Sociological Review 65(1): 96–116.

Frantz, J.E. 1992. ‘Reviving and revising a termination model’, Policy Sciences 25(2): 175–89.

Ganghof, S. 2003. ‘Promises and pitfalls of veto player analysis ’, Swiss Political Science Review 9(2): 1–25.

Geva-May, I. 2001. ‘When the motto is “till death do us part”: The conceptualization and the craft of

termination in the public policy cycle’, International Journal of Public Administration 24(3): 263–88.

Graddy, E.A. and K. Ye. 2008. ‘When do we “just say no”? Policy termination decisions in local hospital

services’, Policy Studies Journal 36(2): 219–42.

Gray, V. 1973. ‘Innovation in the states: A diﬀusion study’, The American Political Science Review 67(4):

1174–85.

Green-Pedersen, C. 2002. The Politics of Justi ﬁcation: Party Competition and Welfare-state Retrenchment in

Denmark and the Netherlands from 1982 to 1998. Amsterdam: Amsterdam University Press.

—— 2004. ‘The dependent variable problem within the study of welfare state retrenchment: deﬁning the

problem and looking for solutions’, Journal of Comparative Policy Analysis 6(1): 3–14.

—— 2007. ‘More than data questions and methodological issues: Theoretical conceptualization and the

dependent variable “problem” in the study of welfare reform’, in J. Clasen and N.A. Siegel (eds)

Schmitt

Investigating Welfare State Change: The ‘Dependent Variable Problem’ in Comparative Analysis. Cheltenham:

Edward Elgar, 13–23.

Ha, E. 2008. ‘Globalization, veto players, and welfare spending’, Comparative Political Studies 41(6): 783–813.

Hacker, J.S. 2004. ‘Privatizing risk without privatizing the welfare state: The hidden politics of social

policy retrenchment in the United States’, American Political Science Review 98(2): 243–60.

Hall, P.A. 1993. ‘Policy paradigms, social learning, and the state: The case of economic policymaking in

Britain’, Comparative Politics 25(3): 275–96.

——and R.C.R. Taylor. 1996. ‘Political science and the three new institutionalisms’, Political Studies 44(5):

936–57.

Harris, M. 2001. ‘Policy termination: The case of term limits in Michigan’, International Journal of Public

Administration 24(3): 165–80.

Häusermann, S. 2010. The Politics of Welfare State Reform in Continental Europe: Modernization in Hard Times.

Cambridge: Cambridge University Press.

Heichel, S., J. Pape and T. Sommerer. 2005. ‘Is there convergence in convergence research? An overview

of empirical studies on policy convergence’, Journal of European Public Policy 12(5): 817–40.

Hogwood, B.W. and G.B. Peters. 1982. ‘The dynamics of policy change: policy succession’, Policy Sciences

14(3): 225–45.

Holzinger, K. and C. Knill. 2005. ‘Causes and conditions of cross-national policy convergence’, Journal of

European Public Policy 12(5): 775–96.

——, C. Knill and T. Sommerer. 2008. ‘Environmental policy convergence: The impact of international

harmonization, transnational communication, and regulatory competition’, International Organization 62

(4): 553–87.

Howlett,

M. 2000. ‘Beyond legalism?

Policy ideas, implementation styles and emulation-based convergence

in Canadian and US environmental policy’, Journal of Public Policy 20(3): 305–29.

——and B. Cashore. 2009. ‘The dependent variable problem in the study of policy change: Under-

standing policy change as a methodological problem’, Journal of Comparative Policy Analysis 11(1): 33–46.

——and M. Ramesh. 2003. Studying Public Policy: Policy Cycles and Policy Subsystems. Toronto: Oxford

University Press.

——and J. Rayner. 2008. ‘Third generation policy diﬀusion studies and the analysis of policy mixes: Two

steps forward and one step back?’, Journal of Comparative Policy Analysis 10(4): 385–402.

Huber, E. and J.D. Stephens. 2001. Development and Crisis of the Welfare State Parties and Policies in Global

Markets. Chicago: University of Chicago Press.

Immergut, E.M. and K.M. Anderson 2007. ‘Editors’ introduction: the dynamics of pension politics’,in

E.M. Immergut, K.M. Anderson and I. Schulze (eds) The Handbook of West European Pension Politics.

Oxford, New York: Oxford University Press, 1– 45.

——, K.M. Anderson and I. Schulze (eds). 2007. The Handbook of West European Pension Politics. Oxford,

New York: Oxford University Press.

Jahn, D. 1998. ‘Environmental performance and policy regimes: Explaining variations in 18 OECD-countries’,

Policy Sciences 31: 107–31.

——and F. Müller-Rommel. 2010. ‘Political institutions and policy performance: A comparative analysis

of Central and Eastern Europe’, Journal of Public Policy 30(1): 23–44.

Jänicke, M. 1992. ‘Conditions for environmental policy success: An international comparison’, The Environmentalist

12(1): 47–58.

——(ed.). 1996. Umweltpolitik der Industrieländer: Entwicklung – Bilanz–Erfolgsbedingungen. Berlin: Sigma.

Jordan, A. 2005. ‘Policy

convergence: A passing fad or a new integrating focus in European Union studies?’,

Journal of European

Public Policy 12(5): 944–53.

Jordan, A. and D. Lieﬀerink (eds). 2004. Environmental Policy in Europe: The Europeanization of National

Environmental Policy. London and New York: Routledge.

——, R.K.W. Wurzel and A.R. Zito (eds). 2003. ‘“New” Instruments of Environmental Governance? National

Experiences and Prospects. London and Portland, OR: Frank Cass.

Jupille, J. and J. Caporaso. 1999. ‘Institutionalism and the European Union: Beyond international relations

and comparative politics’, Annual Review of Political Science 2: 429– 44.

Kersbergen, K. van 1995. Social Capitalism: A Study of Christian Democracy and the Welfare State. London and

New York: Routledge.

Kitschelt, H. 2001. ‘Partisan competition and welfare state retrenchment: When do politicians choose

unpopular policies?’, in P. Pierson (ed.) The New Politics of the Welfare State. Oxford, New York: Oxford

University Press, pp. 265–302.

Studying public policy-making

Knill, C. 2005. ‘Introduction: Cross-national policy convergence: Concepts, approaches and explanatory

factors’, Journal of European Public Policy 12(5): 764–74.

——, J. Tosun and S. Heichel. 2008. ‘Balancing competitiveness and conditionality: Environmental policy

making in low-regulating countries’, Journal of European Public Policy 15(7): 1019–40.

——, M. Debus and S. Heichel. 2010. ‘ Do parties matter in internationalised policy areas? The impact of

political parties on environmental policy outputs in 18 OECD countries, 1970–2000’, European Journal of

Political Research 49(3): 301–36.

——, K. Schulze and J. Tosun. 2010. ‘Politikwandel und seine Messung in der vergleichenden Staat-

stätigkeitsforschung: Konzeptionelle Probleme und mögliche Alternativen’, Politische Vierteljahresschrift 51

(3): 409–32.

Korpi, W. and J. Palme. 2003. ‘New politics and class politics in the context of austerity and globalization:

Welfare state regress in 18 countries, 1975–95’, American Political Science Review 97(3): 425–46.

Kühner, S. 2007. ‘Country-level comparisons of welfare state change measures: Another facet of the

dependent variable problem within the comparative analysis of the welfare state’, Journal of European

Social Policy 17(1): 5–18.

Lerner, D. and H.D. Lasswell (eds). 1951. The Policy Sciences: Recent Developments in Scope and Method.

Stanford, CA: Stanford University Press.

Lynch, J. 2001. ‘The age-orientation of social policy regimes in OECD countries, Journal of Social Policy 30

(3): 411–36.

Marsh, D. and J.C. Sharman. 2009. ‘Policy diﬀusion and policy transfer’, Policy Studies 30(3): 269–88.

Martin, J.P. 1996. ‘Measures of replacement rates for the purpose of international comparisons: A note’,

OECD Economic Studies 26(1): 99–115.

Mayntz,

R. 1983. ‘The conditions

of eﬀective public policy: A new challenge for policy analysis’, Policy

and Politics 11(2): 123–43.

Meseguer, C. 2006. ‘Rational learning and bounded learning in the diﬀusion of policy innovations’,

Rationality and Society 18(1): 35–66.

—— 2009. Learning, Policy Making, and Market Reforms. Cambridge and New York: Cambridge University

Press.

Overbye, E. 1994. ‘Convergence in policy outcomes: Social security systems in perspective’, Journal of

Public Policy 14(2): 147–74.

Pierson, P. 1994. Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment. Cambridge

and New York: Cambridge University Press.

—— 1996. ‘The new politics of the welfare state, World Politics 48(2): 143–79.

Plümper, T. and E. Neumayer. (2010. ‘Model speciﬁcation in the analysis of spatial dependence’, European

Journal of Political Research 49(3): 418–42.

Radaelli, C.M. 2000. ‘Policy transfer in the European Union: Institutional isomorphism as a source of

legitimacy’, Governance: An International Journal of Policy and Administration 13(1): 25–43.

—— 2008. ‘Europeanization, policy learning, and new modes of governance’, Journal of Comparative Policy

Analysis 10(3): 239–54.

Rose, R. 1991. ‘What is lesson-drawing?’ Journal of Public Policy 11(1): 3–30.

—— 1993. Lesson-Drawing in Public Policy: A Guide to Learning Across Time and Space. Chatham, NJ: Chatham

House.

Sabatier, P.A. 1987. ‘Knowledge, policy-oriented learning, and policy change: An advocacy coalition framework’,

Science

Communication 8(4): 649–92.

——and D.

Mazmanian. 1980. ‘The implementation of public policy: A framework of analysis’, Policy

Studies Journal 8(4): 538–60.

Scharpf, F.W. 1997. Games Real Actors Play: Actor-Centered Institutionalism in Policy Research. Boulder, CO:

Westview Press.

Scruggs, L. 1999. ‘Institutions and environmental performance in seventeen Western democracies’, British

Journal of Political Science 29: 1–31.

—— 2003. Sustaining Abundance: Environmental Performance in Industrial Democracies . Cambridge: Cambridge

University Press.

—— 2007. ‘Welfare state generosity across space and time’, in J. Clasen and N.A. Siegel (eds) Investigating

Welfare State Change: The ‘Dependent Variable Problem’ in Comparative Analysis. Cheltenham: Edward

Elgar, 133–65.

Shipan, C.R. and W.R. Lowry. 2001. ‘Environmental policy and party divergence in Congress’, Political

Research Quarterly 54(2): 245–63.

Schmitt

Sommerer, T. 2010. Transnationales Lernen als Faktor von Policy-Wandel: Eine vergleichende Analyse der

Umweltpolitik in 24 OECD-Staaten. Wiesbaden: VS Verlag für Sozialwissenschaften.

Starke, P. 2006. ‘The politics of welfare state retrenchment: A literature review’, Social Policy and Administration

40(1): 104–20.

——, Obinger, H. and F.G. Castles. 2008. ‘Convergence towards where: In what ways, if any, are welfare

states becoming more similar?’ Journal of European Public Policy 15(7): 975–1000.

Stephens, J.D., E. Huber and L. Ray, L. 1999. ‘The welfare state in hard times’, in H. Kitschelt, P. Lange,

G. Marks and J.D. Stephens (eds) Continuity and Change in Contemporary Capitalism. Cambridge: Cambridge

University Press.

Stone, D.L. 2000. ‘Non-governmental policy transfer: The strategies of independent policy institutes’,

Governance: An International Journal of Policy and Administration 13(1): 45–62.

—— 2001. ‘Learning lessons, policy transfer and the international diﬀusion of policy ideas’ (Electronic

Version). Warwick Centre for the Study of Globalisation and Regionalisation (CSGR) Working Paper

No. 69/01. Available at www2.warwick.ac.uk/fac/soc/csgr/research/workingpapers/2001/wp6901.pdf

—— 2004. ‘Transfer agents and global networks in the “transnationalisation” of policy’, Journal of European

Public Policy 11(3): 545–66.

Streeck, W. and K.A. Thelen (eds). 2005. Beyond Continuity: Institutional Change in Advanced Political

Economies. Oxford and New York: Oxford University Press.

Swank, D. 2001. ‘Political institutions and welfare state restructuring: The impact of institutions on social

policy change in developed democracies’, in P. Pierson (ed.) The New Politics of the Welfare State. Oxford

and New York: Oxford University Press, pp. 197–237.

—— 2005. ‘Globalisation, domestic politics, and welfare state retrenchment in capitalist democracies’,

Social Policy and Society 4(2): 183–95.

Sykes, R., B. Palier and P.M. Prior (eds). 2001. Globalisation and European Welfare States: Challenges and

Change. Basingstoke and New York: Palgrave MacMillan.

Taylor-Gooby, P. (ed.). 2004. New Risks, New Welfare: The Transformation of the European Welfare State.

Oxford and New York: Oxford University Press.

Tews, K., P.-O. Buschand and H. Jörgens. 2003. ‘The diﬀusion

of new environmental

policy instruments’ ,

European Journal of Political Research 42(4): 569– 600.

Tsebelis, G. 1995. ‘Decision making in political systems: Veto players in presidentialism, parliamentarism,

multicameralism and multipartyism’, British Journal of Political Science 25(3): 289–325.

—— 2000. ‘Veto players and institutional analysis’, Governance: An International Journal of Policy and

Administration 13(4): 441–74.

Tyran, J.-R. and R. Sausgruber. 2005. ‘The diﬀusion of policy innovations: An experimental investigation’,

Journal of Evolutionary Economics 15(4): 423–42.

Vis, B. 2009. ‘Governments and unpopular social policy reform: Biting the bullet or steering clear?’, European

Journal of Political Research 48(1): 31–57.

——and K. van Kersbergen. 2007. ‘Why and how do political actors pursue risky reforms?’ Journal of

Theoretical Politics 19(2): 153–72.

Volden, C., M.M. Ting and D.P. Carpenter. 2008. ‘A formal model of learning and policy diﬀusion’, The

American Political Science Review 102(3): 319–32.

Wälti, S. 2004. ‘How multilevel structures aﬀect environmental policy in industrialized countries’, European

Journal of Political Research 43(4): 599–634.

Weidner, H. and M. Jänicke (eds). 2002. Capacity Building in National Environmental Policy: A Comparative

Study of 17 Countries. Berlin: Springer-Verlag.

Whiteford, P. 1995. ‘The use of replacement rates in international comparisons of beneﬁt systems’, International

Social Security Review 48(2): 3–30.

Wilensky, H.L. 1974. The

Welfare State and

Equality: Structural and Ideological Roots of Public Expenditures.

Berkeley, CA: University of California Press.

Windhoﬀ-Héritier, A. 1987. Policy-Analyse: Eine Einführung . Frankfurt and New York: Campus Verlag.

Studying public policy-making

International dimensions and

dynamics of policy-making

Anthony Perl

The days when policy researchers could count upon domestic politics and society to contribute

suﬃcient data for a satisfactory analysis are now a memory. Whether policy researchers are

prepared to enter another analytical universe or not, the accelerating ﬂow of ideas, information,

goods and money across national borders has aﬀected the nature of policy problems, reshaped

the attempts to engage these problems and thus reoriented the way in which explanations of

policy-making can be productively pursued. The big questions that animate policy studies may

not have changed, but the available data and the concepts needed to analyze them have been

shifting. This chapter will seek to connect these emerging global dynamics to long recognized

drivers of policy-making and present a conceptual framework that can help in understanding

the resulting interactions. Enhancing the linkage between theoretical frameworks that have

informed international relations and public policy concepts promises a better understanding of

policy-making in a volatile universe.

Despite the growing awareness of globalization as an inﬂuence on many aspects of policy-

making, there remains a knowledge gap among the ﬁndings produced by those who seek to

explain policy-making on its own terms, and those who are more interested in interrogating the

transnational forces that have evolved in recent years. While international relations scholars

are perhaps more experienced in assessing how interactions across boundaries aﬀect politics, they

are also less inclined to extend these eﬀorts to elucidating the internal workings of a policy

subsystem. Such sectoral interpretations can appear mundane to those who are focused on

matters of global conﬂict and governance.

Conversely, students of public policy may be quite motivated to consider how global dynamics

aﬀect policy-making within a subsystem, in search of a better understanding about how exo-

genous inﬂuences can shape the authoritative dynamics within that domain. But researchers

who have been trained to explore policy-making in a single government or administration, or

perhaps compare those eﬀorts, are likely to lack the conceptual tools that are required to inte-

grate a global interplay of policy inputs into the workings of a particular subsystem. The most

promising approach to taking the measure of global inﬂuences on policy subsystems thus appears

to extend the policy researcher’s ability to assess these exogenous forces from an international

perspective. Extending the capacity of policy researchers to better interpret international interac-

tions could enhance the prospects for a fruitful dialogue between policy studies and international

relations. As policy studies have more to say about the increasing global inﬂuences and interactions,

those engaged in international political research could become more attracted to engaging such

ﬁndings about how global forces inﬂuence policy subsystems.

This examination of how to account for international forces in policy studies will draw upon

concepts that have been developed to examine meso-level political relationships between the

state and society. This tradition includes identifying conﬁgurations of actors and classifying them

into a typology of policy communities (i.e. ranging from the collaborative partnerships of cor-

poratism to the competitive advocacy of pluralism). More sophisticated classiﬁcations of authority

can be found in the literature on policy networks, and as the scope of such relations has broadened,

the term governance network has been put forward. These approaches to analyzing policy dynamics

that cross organizational and jurisdictional boundaries can be particularly eﬀective in assessing global

inﬂuences on policy-making because they draw upon relational measures of authority. Assessing

these relationships in the international arena is a natural conceptual progression. Furthermore, the

analytical skills that are developed through analyzing policy networks and communities could be

readily scaled up to examining global forces aﬀecting policy, unlike some conceptual approaches.

We will consider the value of applying a network perspective on understanding global inﬂuences

through four stages of consideration. Initially the chapter will examine the concept of globali-

zation and brieﬂy assess the implications that it raises for studying policy-making. Next, we will

turn to the literature on policy communities and policy networks to highlight tools that can be

used in assessing global inﬂuences on policy subsystems. Then, the chapter will consider the con-

cept of policy paradigms and contemplate the role that ideational inﬂuences play in modulating

global impacts on policy. Finally, a fourfold typology of internationalized policy environments

will be presented, in order to illuminate how particular conﬁgurations of policy communities

and networks can refract the global inﬂuences on governance. In conclusion, the dynamic role

of policy community mediators in trying to steer subsystem responses to global forces will brieﬂy

be considered.

Seeking policy explanations in a time of increasingly permeable

political boundaries

The political boundaries that have structured many examinations of policy-making are less deﬁ-

nitive than at any time since the mid-twentieth century. And as a result, political factors and

forces that could once be safely ruled out of bounds in studying policy, because they were seen

to have little impact on the policy subsystems within a given jurisdiction, have to be considered

in ways that take account of contemporary reality without hopelessly cluttering the scope of a

given investigation. Cerny (1995) contended that global forces have reshaped the way that states

function when they make policy by providing policy actors with cues of equal or greater inﬂuence

than the domestic interests and institutions which previously determined both the political cal-

culus and the substantive assessment of policy options. His claims were provocative when ﬁrst

asserted, but appear increasingly self-evident in the early twenty-ﬁrst century.

For example, American auto manufacturers had successfully resisted changes to fuel economy

regulations for decades when the US surface transportation subsystem was governed by a cohesive

domestic cluster of interests and institutions (Perl and Dunn 2007). It is not a coincidence that

the US government has established a much higher fuel economy standard for motor vehicles

following the 2008 global ﬁnancial crisis. A multi-billion dollar bailout that saved both Chrysler

and General Motors from liquidation opened the window for changing technical standards in

ways that had previously been anathema to industry leaders in the domestic subsystem. In this

instance, the international drivers of policy change dramatically, obviously because the inﬂuence

International dimensions and dynamics

of these forces was concentrated in a global ﬁnancial crisis, but many subsystems have been

reshaped by international inﬂuences in a less dramatic fashion.

Even cultural characteristics which once informed the values behind distinct traditions of the

state in policy-making (Dyson 1980) have become harder to operationalize as independent

variables. From the social welfare traditions of Scandinavian states to the neoliberalism embraced

by Anglo-American polities, state-speciﬁc ideas on the role of government no longer provide a

clear and consistent policy orientation. Instead, virtual communities of interest spread ideas and

values across social networks that span the globe at the same time that supranational institutions

develop explicitly global values that foster convergence in policy-making.

Not surprisingly, scholars in international relations have been drawing attention to the global

inﬂuences on policy-making for some time already. During the 1970s, Keohane and Nye (1977)

had identiﬁed “interdependence” as a driving force behind international economic and political

relationships. Gourevitch (1978) proposed an explanation of how these international forces could

directly aﬀect domestic politics, which would then reverberate by inﬂuencing foreign policies in

a formulation that he labeled the “second image-reversed.” This dynamic was shown to extend

from domestic politics into policy-making by Robert Putnam (1998) through his “two-level games”

perspective on policy negotiations in the international arena. While this attention to international

policy inﬂuences grew out of a focus on the workings of formal governing arrangements across

political boundaries, another stream of thinking that is closer to home in policy studies has sought

to understand how societal forces can inﬂuence policy when they cross organizational boundaries.

In this broader consideration of transnational political forces, “globalization” has become a

commonly used term to encompass economic, political and cultural inﬂuences that cut across

borders, not only those of formal state structures but also societal organizations from corporations

to virtual networks. Held has characterized such globalization as

the stretching and deepening of social relations and institutions across space and time such

that, on the one hand, day-to-day activities are increasingly inﬂuenced by events happening

on the other side of the globe and, on the other, the practices and decisions of local groups

or communities can have signiﬁcant global reverberations.

(1995: 20)

The eﬀects of such globalization have reshaped the roles and routines of policy subsystems. Policy

inputs from subnational, national, regional, and international levels are likely to expand the range

of ideas and interests that are interacting within the subsystem, yielding less stable relationships

among actors and less predictable actions at diﬀerent stages of the policy cycle.

This expansion of global inﬂuences can be seen following the liberalization of capital move-

ments in the 1980s and the subsequent impact of global ﬁnancial considerations that has played

out within just about every policy subsystem. Facing the need to engage with these global

ﬁnancial dynamics, most governments have sought to cooperate at scales ranging from informal

dialogue and information sharing (e.g. the Asia-Paciﬁc Economic Cooperation) to creating

supra-national currencies and reserve banks (e.g. the Euro and European Central Bank). These

initiatives have further ampliﬁed global inﬂuences on policy subsystems, since the economic

“fundamentals” shaping many policy contexts have become increasingly transnational.

And while the eﬀects of globalization vary across subsystems, there are few, if any, domains

that have been left unaﬀected by this erosion of political boundaries. Even subsystems supporting

“core” areas of state sovereignty, such as the military and national security, have been aﬀected

by the globalization of combat and terrorism. The multinational military intervention that

determined

the outcome of

Libya’s civil war in 2011 illustrates the inﬂuence of global forces on

Perl

at least two diﬀerent levels. On the one hand, global communications and social networks sparked

rebellions in autocratic regimes across the Arab world. And on the other hand, nations that had

not previously allied in Middle East military intervention jointly took a side in the Libyan conﬂict,

responding partly to domestic appeals that democratic ambitions and human rights must not be

extinguished by the Qaddaﬁ counteroﬀensive.

Other policy subsystems have experienced less dramatic, but no less important, changes from

beyond their borders as states have more or less explicitly pooled their sovereignty to address

environmental, ﬁnancial, trade, and other policy challenges. The common characteristic in each of

these diverse instances where global forces have modiﬁed the interaction of ideas and interests

within a policy subsystem is that more permeable political boundaries have let in ideas, information,

and interests that had not previously exerted a direct aﬀect on the actors and organizations.

This augmenting encroachment of global inﬂuences into a policy subsystem, mirrors and

often stimulates another kind of change that has impacted the nature of policy-making in recent

years. The rules and norms that once demarcated the duties of public oﬃcials from the actions

of civil society participants in a policy subsystem have also changed, with a tendency toward

blurring the distinction between public and private spheres of responsibility. The relationship

between the public realm and the private sector can no longer be counted upon to conform to

Weberian administrative norms. Scholars in public policy have thus had to develop their own

concepts and analytical tools that could better interpret this reconﬁguration of activities and

relationships between the public realm and the private sector, leading into a post-Weberian

universe where states make policy through governance (Kooiman 2003).

Assessing the potentials and the pitfalls of policy-making that unfold beyond the bureau-

cracy’s formal accountability for translating democratically attained decisions into policy out-

comes has taken policy studies into the more interactive and contingent deliberations occurring

in governance networks. Scholars such as Rhodes (1996) claimed that societies were experien-

cing a new mode of policy-making in which governance provided the capacity for “governing

without government”. Examining political dynamics that span the boundary between the state

and society is thus increasingly familiar to scholars who look at how policy communities and

policy networks have changed the way that policy gets developed. This approach can develop the

skills needed to extend such analysis to explain the political eﬀects of globalization on policy.

When the interactions between state and society have grown more convoluted as their bound-

aries have blurred, the analysis of how these relationships are managed requires some robust

guiding principles. Coleman and Perl (1999) proposed two conceptual axes that can orient navi-

gation through the many intricacies of governance and globalization. First, they draw attention

to the degree and patterns of integration of relationships among policy actors. Next, they highlight

the extent and manner in which public power is shared between state and civil society actors.

Since these dimensions are interdependent, the potential for confusion is best avoided by linking

each axis to an established analytical viewpoint in the study of meso-level policy-making. By

focusing on participants, the policy community perspective can capture the interaction of actors

and organizations within a given governance structure. (Coleman and Skogstad 1990) And similar

to van Waarden’s (1992) usage, the concept of policy network will characterize the institutional

dimension in which public power can be shared by members of the policy community.

Policy communities can be conceptualized as the cognitive and discursive space where issues

are problematized as appropriate, or inappropriate, objects of public engagement. Where such

debates used to emphasize the merits and disadvantages of placing a problem onto the public

policy agenda, thus triggering a government’s subsequent engagement with formulating and

implementing a course of action, the trend toward governance has led policy communities to

remain continuously, although variably, engaged in eﬀorts to pursue policy options. The degree of

International dimensions and dynamics

integration within policy communities is determined by their boundary rules, and represents one

of two deﬁning characteristics. Boundaries can be more or less restrictive depending on the nature

and enforcement of rules that specify size (Jordan and Schubert 1992), openness (Hassenteufel

1995) and stability of membership (Le Gales 1995). These rules also have a bearing on the quantity

and nature of information that is exchanged (Laumann and Knoke 1987).

Another signiﬁcant aspect of the boundary deﬁning a policy community is the extent of shared

values and norms among its membership. The most well developed conceptualization of such

ideational ties can be found in Sabatier and Jenkins-Smith’s (1993) “Advocacy Coalition Frame-

work” in which core beliefs about the causes and eﬀects of policy intervention focus participants

on a common understanding of problems and draw them into shared political pursuit of a pre-

ferred solution. The number of advocacy coalitions and the relationship between their preferred

policy solutions can greatly aﬀect the character of a policy community. A similar concept has been

put forward from international relations scholarship, the “epistemic community” (Haas 1992).

Here, it is ostensibly objective expertise, rather than more subjective beliefs, that draw policy actors

together into shared knowledge and cooperative action that can transcend national boundaries.

By specifying the boundaries of policy communities, and then analyzing what goes on inside

these spaces, policy scholars have been able to link the participants’ state of mind to their

interactions both within and outside the policy community, and then correlate these relation-

ships with actual policy orientations. One important ﬁnding from this line of analysis has been

that the greater the degree of integration of actors within a policy community, the higher the

likelihood of collaborative approaches to policy-making. Scharpf (1997) claims that instead of a

short-term focus, and zero sum calculus of competition, more integrated policy communities

can pursue policy options which go beyond the lowest common denominator and address the

needs of a broader cross-section of policy actors than would be able to work together in less inte-

grated contexts. Another ﬁnding has been that once a particular advocacy coalition or epistemic

community has become widely recognized for its inﬂuence or expertise, then that group’s values

and norms will convey legitimacy on some actors while denying it to others (Smith 1995).

Some policy preferences will thus become more legitimate than others for the time that con-

sensus on values persists. But when consensus within a policy community erodes, political

conﬂict over options at each stage of policy-making can be expected.

When it comes to assessing the ways in which political authority plays out in a policy sub-

system, the policy network perspective looks beyond the cognitive and discursive attributes of

policy community participants and focuses on a scarce resource that is essential to overcoming

the policy-making constraints that are posed by organizational interdependencies, technical com-

plexities, and factual uncertainties. That resource is power, in particular the power that ﬂows

either directly or indirectly from state sovereignty.

During the strategic manoeuvring of policy actors, the sharing of information and ideas, and

the negotiation of competing interests and perspectives, outcomes will be shaped considerably

by who has the power to make decisions binding across all of society based on the sovereign

authority of government. This authority may well be delegated and shared across the policy com-

munity, but in one form or another, power inﬂuences the outcome of policy-making. And it is

through those diﬀerent conﬁgurations of power that inﬂuence how particularly conﬁgured

policy networks can make a signiﬁcant impact on the policy process.

Policy network typologies have made it possible to move analysis of governance beyond the

blunt distinction between pluralism and corporatism (van Waarden 1992: 30). With the blending

of public and private roles in policy formulation and delivery, more precise network conﬁg-

urations such as “state corporatism” and “clientelism” are able to capture the eﬀects of delega-

tion of public authority to coalitions or to particular entities within the policy community. This

Perl

precision can be even more helpful in sorting out the authoritative dynamics of transnational

policy communities in which public actors operating at multiple levels of government interact

and share their authority over policy-making, either with one another or with a subset of societal

actors and organizations. The more hybridized such relationships become, and the further that

policy-making dynamics blur the distinction between public and private spheres of account-

ability, the greater value that more nuanced policy network categories can oﬀer in highlighting

both the eﬃcacy and the accountability of particular governance modes.

In a world where some of the boundaries that used to matter most for policy-making—those

between government and the private sector, and those demarcating one state’s sovereignty from

another—are no longer preeminent, other structures that can integrate the ideas and interests of

policy communities with the inﬂuence and power deployed in policy networks merit greater

analytical attention. Instead of concentrating on the forces of cohesion within policy subsystems

that emanate from formal state structures, the inﬂuence of ideas that cut across state and societal

boundaries need to be given greater weight in attempts to explain policy-making.

These ideas about what is to be done, and how it is to be done, comprise the essence of a

policy paradigm. As Peter Hall has illustrated, such globally established paradigms as the role of

monetary policy and the ﬁnancial institutions that implement it can be more inﬂuential than

any government’s formal economic policy capacity. Policy paradigms thus oﬀer an intellectual

center of gravity that can serve some of the role that state actors and structures have traditionally

played in shaping the dynamics of policy subsystems. The following section will consider the

policy paradigm as a conceptual construct that can oﬀer guidance in interpreting a policy universe

containing more varied international forces than ever before.

How policy paradigms guide governance networks through an

accelerating policy universe

The growing permeability of boundaries, both among nations and between the state and society,

has increased the volume and velocity of forces from the policy universe that are transmitted

into policy communities, raising the level of instability in policy-making. But this growing

exposure to exogenous forces has also enhanced the dissemination of a particular type of ideas

that can bring, or return, coherence to volatile governance networks by fashioning a common

point of view among policy actors. Numerous scholars have noted the inﬂuence that a coherent

set of ideas can exert in aligning the vision of policy actors toward a shared set of goals (Jobert

and Muller 1987; Hall 1993; Shon and Rein 1994). Oﬀering advice on which problems are

critical and which ones are less important, which policy instruments provide appropriate means

to address these problems and which do not, and what instrument settings could best resolve a

problem, are the stock in trade of policy practitioners’ work. Hall has labeled the guiding principles

that support such deliberations as a “policy paradigm. ” Shon and Rein characterize these ideas

as “policy frames,” while Jobert and Muller speak of a reference system that guides French

policy-making. Each of these concepts highlights the shared vision that can be created by a set

of norms which focus policy communities on commensurate goals and build conﬁdence in the

causal relationships that will guide actors within those communities to agree that particular instru-

ments should be deployed in speciﬁc ways. These policy paradigms exert inﬂuence by evoking

“images” or “generative metaphors” that help policy actors to make sense of a complex and

contingent policy universe. Examples of such inﬂuential images have included “Energy Security,”

“Universal Health Care,” and “Free Trade.” These ideas and images provide certain actors in

the policy community, and their beliefs, with greater legitimacy in the policy process than other

groups and individuals can muster.

International dimensions and dynamics

The acceleration in speed and volume of communication across boundaries has often aided

the dissemination of consistent, if not always convergent, policy paradigms. One such vector has

been the rise of a global cadre of policy consultants who perceive material incentives to sell

similar solutions across a range of policy jurisdictions (Perl and White 2002). Another driving

force for consistent thinking about how to approach problems has been the role of public and

private ﬁnancial institutions from the International Monetary Fund to ﬁnancial rating agencies

that apply analogous measures of risk and eﬃcacy to public policies and programs across the

globe. These and other inputs present policy actors with consistent signals, along with incentives

to follow them and potential penalties for straying too far from the global consensus on

appropriate policy options and eﬀective solutions.

When these transmission mechanisms are working well, a policy paradigm will be widely

accepted, and the policy subsystem settles into a period of political quiescence. Policy commu-

nities abridge the discourse and engagement in their given policy domain, often corresponding

with a delegation of power from state actors to certain policy community participants. These

terms of engagement can enable a particular policy community to dominate all stages of the

policy cycle, from agenda-setting through policy evaluation. More often than not, the actors or

advocacy coalitions granted legitimacy by the policy paradigm become the principal civil society

actors who participate in key decisions, conduct informal or formal oversight, and even deliver

program outputs. Policy discourse becomes more technical as issues of day-to-day management

and ﬁne tuning of policy instruments are deemed relevant subjects for discourse.

But the same transmission of ideas and information that can encourage consistency in outlook

across political boundaries, and among a critical mass of policy actors throughout civil society,

can also undermine the consensus that had been fostered by a policy paradigm. From the sudden

revelation of conﬁdential information through sources disseminated by Wikileaks through the

ﬂash mobs that are mobilized to protest policy options on Facebook and Twitter, experiences

and ideas that challenge conventional wisdom can be shared more immediately, and more exten-

sively, than ever before. When faced with such challenges, policy-makers’ ﬁrst instinct is to turn

inward toward the policy community participants they know best and seek a modest revision of

the recipes that are most familiar to them. The second level of response to a global perturbation

would see the state changing its relationship with actors in the policy community. Such reconﬁ-

guration could go in either of two directions. Either new actors could be admitted into the

corridors of power and with them, the range of outcomes that could be considered acceptable

would expand. Or the number of actors with a recognized role in the governance network

could be pruned back in order to make space for a new paradigm to emerge eventually. Such a

redeﬁnition of boundary rules and a reallocation of inﬂuence may reequip the policy community

to better “manage” the process of policy change.

When neither of these tactics to adjust the policy paradigm from within the policy commu-

nity succeeds, the debate is likely to shift from resolvable diﬀerences to deeper disputes where

an existing policy paradigm is called into question, and directly challenged with a contending

paradigm. Faced with conﬂicting opinions, political leaders will have to choose who to recog-

nize as authoritative, especially on matters of technical credibility. Under these circumstances,

the policy community is most likely to engage in a political contest over the issues at hand (Hall

1993: 281). In such times of widespread contention over goals and values, engagement spreads

beyond the policy community itself engaging political actors from unfamiliar, and less pre-

dictable, corners of the policy universe. Contributions from unfamiliar sources of expertise,

among them the “sub-elites” that Etzioni-Halevy (1993: 194) has identiﬁed as playing an

inﬂuential role by developing alternative policy options in between points of political decision-

making (e.g. elections), will broaden the discourse to include new approaches and ideas. And

Perl

without the paradigmatic consensus established to screen out nonconforming perspectives, the

policy community will experience more wide-ranging debates about what is to be done. At

such times, political ideologies—explicit meaning systems that profess universal principles to

guide policy-making (Swidler 1986)—will contend to sway the hearts and minds in the policy

community, and the wider engaged public, toward embracing a new paradigm.

During periods of such widespread conﬂict over policy, changes in policy communities and

networks will be more profound. Levels of integration within policy communities will decline as

the rules for appropriate discourse and action become relaxed, if not entirely suspended. As conﬂict

levels increase, state actors may pull back their delegation of authority to societal actors, or they

may shift such a delegation of authority from one set of actors to another. Corporatist policy

networks would thus be most susceptible to restructuring, while more loosely organized issue net-

works would see less change in their structure, as the content of their discourse shifted. In the event

that the level of conﬂict surpasses some threshold, which might vary across jurisdictions, the state

may take the lead in reorganizing a policy community, creating a state-directed policy network.

Broadly speaking, the more unsettled the diﬀerences of opinion regarding policy options, the

less likely it is that the policy community will be capable of managing the entire spectrum of

change. Multiple sources of input into policy formulation, implementation and evaluation will

arise, with growing potential for transnational engagement in the stages of policy development.

There may even be an international ratiﬁcation of policy, such as when the European Central

Bank prescribes and proscribes ﬁscal options for countries like Greece, Portugal and Ireland that

are struggling to reﬁnance their public debt. Jobert’s (1995) analysis of French policy paradigm shifts

suggested that during times of heightened conﬂict, the deference to expert consensus in a policy

community evaporates and competing interests and ideologies tend to pull policy actors apart. When

this happens, the disjointed inputs and unpredictable combinations of ideas and interests can take

on the character of Cohen, March and Olsen’s (1972) “garbage can” model of policy-making.

As the velocity and volume of global communication accelerates, policy paradigms can thus

provide a structure for managing the ﬂood of ideas and information that inundates policy domains

around the clock, 365 days a year. But the same channels of discourse and sources of expertise that

often manage to interpret and ﬁlter the policy inputs in a way that focuses policy communities

on a common approach, can also transmit information and interpretations that will undermine the

consensus behind a policy paradigm. In this schizophrenic twist to policy formulation in an

unpredictable world, the same policy actors who have embraced open borders and global norms can

suddenly be disrupted by these same sources of input. Structures such as policy communities,

and modes of governance practiced in policy networks, which had appeared fully functional

within a particular policy paradigm will quickly degenerate into confusion and conﬂict under the

inﬂuence of transnational ideas and information. An explicit focus on supranational governance

arrangements, or lack thereof, can better make sense of tendencies and trajectories when the global

inﬂuences on policy-making shift from promoting consensus to destabilizing it. The following

section will oﬀer such a schema for considering the inﬂuence of supranational governance forces.

The mediating inﬂuence of internationalized policy environments

The relational focus of policy community and policy network analytical approaches make them

well suited to examining the forces that cross the borders between states as well as traversing the

boundary between state and society. Whether policy paradigms are in force or in ﬂux, the levels

of shared understanding, common discourse and the locus and distribution of authority that are

gauged by taking account of policy communities and networks enables an appreciation of how

policy-making now occurs in open environments where exogenous inﬂuences can ﬂow freely.

International dimensions and dynamics

Some have expressed skepticism about the utility of scaling up concepts that were developed

to examine policy-making at the national and subnational levels to gain insight into the policy

dynamics of supranational governance (Kassim 1994). Policies made in the European Union or

eﬀorts by the United Nations to develop binding agreements on climate change, for example,

can pose challenges for meso-level analysis because of their transnational dynamics. When exo-

genous sources give rise to borderless bursts of disruptive information, the functioning of policy

communities and networks can be aﬀected in unpredictable ways. But analogous instabilities

within state–society relations have not undermined policy community and network analyses of

volatile political contexts. Risse-Kappen (1995) has suggested that contemporary transnational

governance arrangements include the same type of horizontal coordinating networks that cross

national borders, as well as bridging the boundaries between state and society.

In extending the utility of policy community and network concepts into policy-making

contexts that experience high levels of global inﬂuence, Coleman and Perl (1999) advanced the

concept of internationalized politics to explore a context in which at least some stages of the

policy cycle take place at a more encompassing level than the nation-state. Thus deﬁned, and illu-

strated in Table 4.1, internationalized policy environments may be diﬀerentiated according to

their level of public sector activism and by the degree to which supranational governing arrange-

ments are institutionalized. Public sector activism refers to the direct involvement of politicians

and senior civil servants in managing governance dynamics within and among policy commu-

nities. Societal actors will take on a dominant role in managing governance through policy

communities and networks when public sector activism is low, and politicians and senior civil

servants will steer such governance when public sector activism is high.

The second dimension of internalized politics is expressed in the degree to which international

institutions have a developed role in policy-making. In some policy domains, international institu-

tions play a role that is mandated in supranational law. With a legal mandate, international

institutions gain a capacity to manage governance in policy communities and networks. Where

supranational laws do not establish an explicit role for international institutions, national policy

actors will maintain greater autonomy in responding to international inﬂuences. Applying these two

dimensions of internationalized politics to the ways in which policy communities and networks

are aﬀected by global inﬂuences, one can distinguish between the four ideal-typical internationalized

policy environments which are depicted in Table 4.1. These are brieﬂy described below.

When the levels of public sector activism and institutionalization of supranational governing

arrangements are both high, policy-making will be characterized by a dynamic of multilevel governance

(MLG). In places where MLG is common, like the European Union, both national and supra-

national governments are well developed institutionally. As a result, politicians, bureaucrats, and

Table 4.1 Supranational governance dynamics

Public sector activism in governance

Institutionalization

of supranational

governing

arrangements

High Low

High Multilevel governance Self-regulatory and

private regimes

Low Intergovernmental negotiations Loose couplings

Source: Adapted from Coleman and Perl (1999)

Perl

civil society actors and organizations engage in a multitude of cooperative working arrangements

that cross organizational boundaries and lead to well-integrated policy communities. Public

authority for policy-making is more likely to be delegated under such conditions. While MLG

is most advanced in Europe, the activity of international institutions in trade, ﬁnance, and the

environment extends well beyond Europe and has advanced the conditions in which this mode

of internationalized politics is inﬂuential in the workings of policy communities and networks.

When they are well integrated, national level policy communities are likely to take on consider-

able responsibilities in multilevel policy implementation. There are also likely to support trans-

national policy communities, composed of actors from both national and international levels that

link these diﬀerent national policy communities while connecting them to international institu-

tions. Transnational policy communities can be very inﬂuential in the formulation of policy options,

and transnational experts are likely to work closely with national policy communities during the

implementation stage. Over time, the relative level of integration in domestic and transnational

policy communities will shape the trajectory of how MLG adapts to global drivers of change.

The more integrated that the transnational policy community is, the more it will be able to

focus inputs entering the domestic policy arena, as well as the more coherent its proposals for

policy options to address those inputs are likely to be. The more integrated the domestic policy

community is, the more discretion it will exert in ﬁltering those inputs and (re)interpreting those

policy prescriptions. Given the diversity of interests and the ﬂuidity of participation, transna-

tional policy communities are often likely to be less integrated than their domestic counterparts.

Considering the ways in which these diﬀerent layers interact to either reinforce a policy paradigm

or to undermine it is a research area that promises interesting ﬁndings as examinations of the

2008 global ﬁnancial crisis and its ongoing eﬀects begin to unravel the intricacies of transnational

policy community activity and inﬂuence.

When governments remain active in steering policy communities, but do so in an interna-

tional environment that is lacking in supranational governance structures, policy eﬀects will

likely be seen through the mode of intergovernmental negotiations. This is the well-explored

domain of international relations scholarship, and one of the most inﬂuential models of how

such negotiations relate to domestic policy can be found in Putnam’s (1988) “two-level games

metaphor,” as elaborated by Evans et al. (1993). Here, heads of state or their delegated nego-

tiators work to craft agreements at the international level that can be accepted, either through

formal ratiﬁcation or some other form of political enactment, at the domestic level. The

anticipated degree of diﬃculty in such ratiﬁcation, or acceptance, creates smaller or larger win

sets (i.e. possible agreements that can be accepted) for the negotiators.

Intergovernmental negotiations will thus reinforce a strong role for national policy commu-

nities in the agenda-setting, ratiﬁcation, and implementation stages of policy-making. When

intergovernmental negotiations become regularized, as in the regularly scheduled conferences of

parties to the United Nations Framework Convention on Climate Change or meetings of the

Group of Eight (G8) heads of state, ad hoc coalitions of political actors will be drawn beyond

domestic policy communities to participate in “people’s summits” that are held alongside, yet

outside, the oﬃcial negotiations. Here, opposition groups de ﬁne and publicize policy alternatives

to the options being discussed in oﬃcial negotiations. These alternatives are not envisioned as

replacing the options being considered by authorized negotiators. But they are expected to inﬂuence

these options by building support back home for alternative positions, or stirring up opposition

to the negotiations, and thus narrowing the set of what the oﬃcial negotiators can expect to be

accepted by their respective domestic political jurisdictions.

In many policy subsystems, the rules of the game were established by private actors who have

maintained a prominent role in both elaborating norms and enforcing them. A prime example

International dimensions and dynamics

of such a regime can be found in markets for equities, bonds, and the arcane derivative instru-

ments that have weighed so heavily on the global ﬁnancial system since 2008. These privately

led governance systems emerged for various reasons. In some circumstances, private organiza-

tions held a monopoly, or oligopoly, over the technical knowledge needed to design and

operate the policy subsystem. In other situations, jurisdictional or ﬁscal constraints (i.e. feder-

alism) fostered gaps in public sector capacity that private organizations moved in to ﬁll. These

self-regulating policy communities rarely remained purely private.

States would give assent to these privately established rules and practices by incorporating

them into law or by delegating the authority to pursue self-regulation in the public interest.

Private entities, usually non-proﬁt industry associations, would police a policy sector under such

delegated authority, leading to the evolution of “private interest governments” (Streeck and

Schmitter 1985). Such self-regulatory and private regimes have played important roles in policy

domains that experienced the greatest volume and velocity of global interchange. These formal

and informal institutions have become the epicentre of governance in important elements of the

architecture of globalization, such as ﬁnancial transactions that reach beyond the borders of most

states and create new norms of behavior and obligation.

In the context of self-regulatory and private regimes, transnational policy communities are

likely to be much more tightly integrated than those operating in multilevel governance or

intergovernmental negotiations contexts. The leading role played by private actors in the

domestic nodes of these transnational policy communities (e.g. domestic ﬁnancial organizations)

will favor the establishment and maintenance of clientelist policy network relationships. Private

regimes will derive authority from the imbalance of resources, especially knowledge, between

state and private actors, giving them strong incentives to avoid transparency in their activities.

The internationalized policy environment that gives rise to a context of loose couplings is

deﬁned largely by the absence of institutionalized structures and the lack of any obvious dele-

gation of public authority. High levels of scienti ﬁc and technical uncertainty associated with

new domains of innovation (e.g. social networks) may discourage state actors from intervention,

while no stable private organization exists to support a privately led policy regime. In some

cases, physical or political crises can undermine previous structures, leading to new conﬁgura-

tions of global and local initiative to restructure rogue states (e.g. Bosnia, Iraq) or rebuild

devastated local jurisdictions (e.g. New Orleans, Port-au-Prince).

Implementing policy under conditions of pragmatic alliances among policy actors can bring

about what Browne (1990) has termed “issue niches” in which a transient policy community

coalesces around particular understandings of an issue and the options to address it. The vener-

able “garbage can model” of Cohen et al. (1972) can be helpful in conceptualizing the arbitrary

outcomes that arise from such unstructured interaction between policy problems, solutions,

participants, and choice opportunities. Under such circumstances, policy community structure

will be quite ﬂuid, with state and societal actors drifting in and out of diﬀerent stages of policy-

making. The conﬁguration of authority will resemble an issue network in which participants are

more likely to share information and ideas, as opposed to values and interests.

Conclusion

In seeking to better understand the dynamic eﬀects of global inﬂuences on public policy, policy

researchers and international relations scholars can cumulate their insights by focusing on the rela-

tional dynamics of authority found in policy community and policy network analytical perspec-

tives. Both ﬁelds have pursued such relational measures in drawing attention to the importance

of horizontal coordination eﬀorts in decision-making. From international accords to impromptu

Perl

“coalitions of the willing,” these policy actor groupings can generate rules and norms that ﬂow

across boundaries in the wake of information and resources that move beyond borders ever

more quickly.

As further research is pursued on the governance of globalized policy domains, insights will

accumulate about various conﬁgurations of transnational policy communities from both the

policy studies and international relations avenues of enquiry. These ﬁndings can be expected to

illuminate the role of policy community mediators who are active in several transnational policy

communities and could thus be expected to inﬂuence the interactions between global and

national, or local, politics at their points of intersection and overlap. Such mediators could

function as pragmatic policy brokers, akin to Sabatier and Jenkins-Smith’s (1993) understanding

of a “policy broker” or Kingdon’s (2003) conceptualization of a “policy entrepreneur” who

connects problems with solutions by opening windows of political opportunity.

Another mode of mediation that may be even more signiﬁcant in an expanding policy uni-

verse occurs through the translation of policy paradigms between policy communities that become

connected through ﬂows of information or interests that cross established boundaries. Kuhn

anticipated that in situations where scientiﬁc paradigms conﬂict, some experts would function as

translators between the diﬀerent linguistic communities. Schön and Rein (1994) posit that such

translation can oﬀer the key to resolving conﬂict between diﬀerent paradigms or frames. To

overcome such conﬂict, the policy community mediator must construct a view of the world that

can engage the ideas and interests of multiple policy communities and open a dialogue among

them. Such transversal policy deliberations may provide a means to accommodate the disruptive

forces generated by unfamiliar, and often unintelligible, information and ideas that ﬂow freely

across the boundaries of contemporary policy subsystems.

Bibliography

Browne, William P. 1990. ‘Organized interests and their issue niches: A search for pluralism in a policy

domain’, Journal of Politics 52(2): 477–509.

Cerny, Philip. 1995. ‘Globalization and the changing logic of collective action,’ International Organization

49(4): 595–625.

Cohen, Michael James March, and Johan Olsen. 1972. ‘A garbage can model of organizational choice’,

Administrative Science Quarterly 17(1): 1–25.

Coleman, William and Anthony Perl. 1999. ‘Internationalized policy network environments and policy

network analysis’, Political Studies 47(4): 69–709.

——and Grace Skogstad (eds). 1990. Policy Communities and Public Policy in Canada: A Structural Approach,

Toronto: Copp Clark Pittman.

Dyson, Kenneth. 1980. The State Tradition in Western Europe: A Study of An Idea and an Institution. Oxford:

M. Robertson.

Etzioni-Halevy, Eva. 1993. The Elite Connection: Problems and Potentials of Western Democracy. Cambridge,

MA: Polity Press.

Evans, Peter, Harold Jacobson and Robert Putnam. 1993. Double-edged Diplomacy: International Bargaining

and Domestic Politics. Berkeley, CA: University of California Press.

Gourevitch, Peter. 1978. ‘The second image reversed: The international sources of domestic politics’,

International Organization 32(4): 881–912.

Haas, Peter. 1992. ‘Introduction: epistemic communities and international policy coordination,’ International

Organization 46(1): 1–35.

Hall, Peter, 1993. ‘Policy paradigms, social learning and the state: The case of economic policymaking in

Britain’, Comparative Politics 25(3): 275–96.

Hassenteufel, Patrick. 1995. ‘Do policy networks matter?’, in P. Le Gales and M. Thatcher (eds) Les réseaux

de politique publique: Débat autour des policy networks. Paris: L’Harmattan, 91–108.

Held, David. 1995. Democracy and the Global Order: From the Modern State to Cosmopolitan Governance,

Stanford, CA: Stanford University Press.

International dimensions and dynamics

Jobert, B. 1995 ‘Rhétorique politique, Controverses scientiﬁques et construction des normes institutio-

nelles: Esquisse d’un parcours de recherche’, in A. Faure, G. Pollet and R. Warin (eds) La construction du

sens dans les politiques: Débats autour de la notion de référentiel. Paris: L’Harmattan, 13–24.

——and Pierre Muller. 1987. L’État en action: Politiques publiques et corporatismes. Paris: Presses Universités

de France.

Jordan, G. and K. Schubert. 1992. ‘A preliminary ordering of policy network labels’, European Journal of

Political Research 21(1–2): 7–27.

Kassim, Hussein. 1994. ‘Policy networks and European Union policy making: A skeptical view’, West

European Politics 17(4): 15–27.

Keohane, Robert and Joseph Nye. 1977. Power and Interdependence: World politics in transition. Boston: Little,

Brown.

Kingdon, John. 2003. Agendas, Alternatives, and Public Policies. New York: Longman.

Kooiman, Jan 2003. Governing as Governance. Thousand Oaks, CA: Sage.

Laumann, Edward. and David Knoke. 1987. The Organizational State: Social Choice in National Policy

Domains. Madison, WI: University of Wisconsin Press, 226–48.

Le Gales, P. and M. Thatcher. 1995. Les réseaux de politique publique. Paris: Editions L’Harmattan.

Perl, Anthony and Donald White. 2002. ‘The changing role of consultants in Canadian policy analysis’,

Policy, Organisation and Society 21(4): 49–73.

Perl, Anthony and James A. Dunn Jr. 2007. ‘Reframing automobile fuel economy policy in North

America: The politics of punctuating a policy equilibrium,’ Transport Reviews 27(1): 1–35.

Putnam, Robert. 1998. ‘Diplomacy and domestic politics: The logic of two-level games’, International

Organization 42(3): 427–60.

Rhodes, R.A.W. 1996. ‘The new governance: Governing without government’, Political Studies 44(3):

652–67.

Risse-Kappen,

T. 1995. ‘Bringing transnational

relations back in: Introduction’ and ‘Governance and

transnational relations: What have we learned,’ in Thomsa Risse-Kappen (ed.) Bringing Transnational

Relations Back In. Cambridge: Cambridge University Press, 3–36, 280 – 313.

Sabatier, Paul A. and Hank C. Jenkins-Smith. 1993. Polity Change and Learning: An Advocacy Coalition

Approach. Boulder: CO: Westview Press.

Scharpf, Fritz W. 1997. Games Real Actors Play: Actor Centred Institutionalism in Policy Research. Boulder,

CO: Westview Press.

Schön, D and M. Rein, 1994. Frame Reﬂection, Toward the Resolution of Intractable Policy Controversies. New

York: Basic Books.

Smith, A. 1995. ‘Réintroduire la question du sens dans les réseaux d’action publique’, in P. Le Gales and

M. Thatcher (eds) Les réseaux de politique publique: Débat autour des policy networks. Paris: L’Harmmattan,

109–20.

Streeck, Wolfgang and Philippe Schmitter. 1985. ‘Community, market and state associations?’, in Wolfgang

Streeck and Philippe Schmitter (eds) Private Interest Government, London: Sage, 1–29.

Swidler, Ann. 1986. ‘Culture in action: Symbols and strategies’, American Sociological Review 51(2): 273–86.

van Waarden, Frans 1992. ‘Dimensions and types of policy networks’, European Journal of Political Science

21(1): 29–52.

Perl

Part II

Conceptualizing public

policy-making

State theory and the rise of the

regulatory state

Darryl S.L. Jarvis

Theories in the social sciences fall into diﬀerent types or orders depending upon the type and

range of social phenomena they attempt to explain and the basic method they use to derive their

insights and hypotheses. That is, social scientiﬁc theories diﬀer according to their ‘level of analysis,’

‘method of analysis’ and ‘unit of analysis’ (Almond and Genco 1977: 489–522). With respect to their

level of analysis, some social scientiﬁc theories are ‘general’ or macro-level social theories that

attempt to explain all phenomena within their purview. Others are less wide-ranging and focus

only on a few very speciﬁc subsets of social life, either at a micro-ormeso-level of analysis (Ray

2001: 355–88). Similarly, social theories also diﬀer according to their method of analysis: some

are ‘deductive’ theories developed largely on the basis of the application of general presuppositions,

concepts or principles to speciﬁc phenomena. Others are less deductive and more ‘inductive’,

developing generalizations only on the basis of careful observation of empirical phenomena and

subsequent testing of these generalizations against other cases (Lundquist 1987; Przeworski 1987:

31–49; Hawkesworth 1992: 291–392). And, with respect to their units of analysis, some social

theories focus attention on individuals as the basic social actor whose behaviour and actions must be

explained, while some view aggregate collections of individuals, or groups, as the relevant analytical

unit. Still, others consider larger social structures to have an independent impact on individual

and collective actions (Hay and Wincott 1998: 951–7; Clark 1998: 245–70; Tilly 1984).

If all the permutations within these three variations are considered, the list of policy-relevant

social theories would be almost inﬁnite. However, for most purposes, this task can be simpliﬁ ed

somewhat by focusing only on general, or macro-level, social theories. This is acceptable

because policy-relevant academic disciplines such as economics and political science are inter-

ested in all social behaviour and activities and tend to view public policy-making as only a

subset of such behaviour, amenable to the general theories and explanations prevalent in each

ﬁeld. As such, only a few representative cases exist, therefore, based on diﬀerences in the char-

acteristic basic unit of analysis they employ and their method of theory construction (see Table 5.1)

(Dessler 1999: 123–37).

‘State theory’ focuses on the impact and evolution of social structures and political institutions

on policy-making. Many analyses in this mould focus solely on the state, seeing it as the leading

institution in society and the key agent in the political process. Others, however, attribute

explanatory signiﬁcance to other organized social actors, such as business or labour, in addition

to the state. Both interpretations have their origin in the works of late nineteenth-century

German historical sociologists and legal theorists who highlighted the eﬀects of the development

of modern state institutions on the development of society. Rather than argue that the state

reﬂected the nature of a nation’s populace or social structure, theorists such as Max Weber and

Otto Hintze noted how the state’s monopoly on the use of force allowed it to re-order and

structure social relations and institutions (Hintze 1975; Nettl 1968: 559–92; Weber 1978).

Sociological or historical ‘neo-institutionalism’ which focuses on the role of the state in

policy-making has been summarized by Stephen Krasner as follows:

An institutionalist perspective regards enduring institutional structures as the building blocks

of social and political life. The preferences, capabilities, and basic self-identities of individuals

are conditioned by these institutional structures. Historical developments are path dependent;

once certain choices are made, they constrain future possibilities. The range of options avail-

able to policymakers at any given time is a function of institutional capabilities that were put

in place at some earlier period, possibly in response to very diﬀerent environmental pressures.

(1988: 67)

This perspective explicitly acknowledges that policy preferences and capacities are usually

understood in the context of the society in which the state is embedded (Nettl 1968: 559– 92;

Przeworski 1990; Therborn 1986: 204–31). Like its more deductive counterpart, the kinds of

actor-centred institutionalism found in the work of scholars such as Elinor Ostrom, Peter Hall

described a statist ‘institutionalist’ analysis as one focused on the impact of large-scale structures

on individuals and vice versa. As he put it (Hall 1986: 19):

The concept of institutions … refer[s] to the formal rules, compliance procedures, and

standard operating practices that structure the relationship between individuals in various

units of the polity and economy. As such, they have a more formal status than cultural

norms but one that does not necessarily derive from legal, as opposed to conventional,

standing. Throughout, the emphasis is on the relational character of institutions; that is to

say, on the way in which they structure the interactions of individuals. In this sense it is the

organizational qualities of institutions that are being emphasized.

This form of historical or sociological neo-institutionalism diﬀers from its deductive counterpart

in several critical areas. First, there is no eﬀort made in this approach to reduce institutions to

Table 5.1 General approaches to political phenomena and illustrative theoretical examples

Method of theory construction

Deductive Inductive

Unit of

Analysis

Individual Rational choice theories

(Public choice)

Sociological individualism

(Welfare economics)

Collectivity Class analysis (Marxism) Group theories

(pluralism/corporatism)

Structure Actor-centred Institutionalism

(Transaction cost analysis)

Socio-historical

neo-institutionalism (Statism)

Source: Howlett, Michael, M. Ramesh, and Anthony Perl. Studying Public Policy: Policy Cycles and Policy Subsystems.

Oxford University Press (2009)

Jarvis

less organized forms of social interaction, such as norms, rules or conventions. Second, there is

no attempt made to reduce institutions to the level of individuals and individual activities such

as economic or social transactions. And, third, institutions are simply taken as ‘givens’, that is, as

observable social entities in themselves, with little eﬀort made to derive the reasons for their

origins from a priori principles of human cognition or existence (March and Olsen 1994).

Using such a socio-historical line of analysis yields, to use Theda Skocpol’s terms, a ‘state-

centric’ as opposed to ‘society-centric’ explanation of political life, including public policy-

making (Skocpol 1985: 343). In a ‘strong’ version of the statist approach, as Adam Przeworski

put it in a pioneering book:

states create, organize and regulate societies. States dominate other organizations within a

particular territory, they mould the culture and shape the economy. Thus the problem of

the autonomy of the state with regard to society has no sense within this perspective. It

should not even appear. The concept of ‘autonomy’ is a useful instrument of analysis only

if the domination by the state over society is a contingent situation, that is, if the state

derives its eﬃcacy from private property, societal values, or some other sources located

outside it. Within a true ‘state-centric’ approach this concept has nothing to contribute.

(1990: 47–8)

In the statist version of neo-institutional analysis the state is viewed as an autonomous actor

with the capacity to devise and implement its own objectives, not necessarily just to respond to

pressure imposed upon it by dominant social groups or classes. Its autonomy and capacity

are based on its staﬃng by oﬃcials with personal and agency interests and ambitions and the

fact that it is a sovereign organization with unparalleled ﬁnancial, personnel, and – in the

ﬁnal instance – coercive resources. The proponents of this perspective claim that this emphasis

on the centrality of the state as an explanatory variable enables it to oﬀer more plausible

explanations of long-term patterns of policy development in many countries than do other

types of political theory (Krasner 1984: 223–46; Skowronek 1982; Orren and Skowronek

1998–9: 689–702).

It is diﬃcult to accept statism in the ‘strong’ form described above, however. It cannot easily

account for the existence of social liberties and freedoms or explain why states cannot always

enforce their will, as in times of rebellion, revolution, or civil disobedience. In fact, even the most

autocratic governments make some attempt to respond to what they believe to be the popula-

tion’s preferences. It is, of course, especially impossible for a democratic state to be entirely

autonomous from a society with voting rights. And, as Lindblom and others pointed out, in

addition to eﬀorts to maintain and nurture support for the regime among the population,

capitalist states, both democratic and autocratic, need to accommodate the imperatives of the

marketplace in their policies. Second, the statist view suggests implicitly that all ‘strong’ states

respond to the same problem in the same manner because of their similar organizational fea-

tures. This is obviously not the case, as diﬀerent states (both ‘strong’ and ‘weak’)

often have

diﬀerent policies dealing

with the same problem. To explain the diﬀerences, we will need to

take factors into account other than the features of the state (Przeworski 1990).

To be fair, however, few subscribe to statism in the ‘strong’ form described above.

Instead of replacing the pluralist notion of the societal direction of the state with the statist

notion of the state’s direction of society, most inductively oriented institutionalist theorists

merely want to point out the need to take both sets of factors into consideration in their ana-

lyses of political phenomenon (Hall and Ikenberry 1989; McLennan 1991). As Skocpol herself

has conceded:

State theory and the rise of the regulartory state

In this perspective, the state certainly does not become everything. Other organizations and

agents also pattern social relationships and politics, and the analyst must explore the state’s

structure in relation to them. But this Weberian view of the state does require us to see it

as much more than a mere arena in which social groups make demands and engage in

political struggles or compromises.

(1985: 7–8)

This milder version of statism thus concentrates on the links between the state and society in

the context of the former’s pre-eminence in pluralist group theory. To that extent, statism

complements rather than replaces society-centredness and restores some balance to social and

political theorizing which had lost its equilibrium (Orren and Skowronek 1993; Almond 1988:

853–901; Cortell and Peterson 2001: 768–99).

Theorizing the regulatory state: power, structural orientations,

capacities – towards a typology

Many more speciﬁc policy-related theories ﬁnd their basis in statist thinking. One such con-

temporary mode of thinking is the theory of the ‘ regulatory state’. The rise of the regulatory

state is more often asserted than theorized. What precisely constitutes a regulatory state remains

a vexed question, and what forms, functions, modalities, operational and institutional mechanisms

deﬁne its parameters tend to be inferred rather than systematically outlined.

Part of the explanation for this state of aﬀairs rests in the multiple discourses that have con-

tributed to conceptions of the regulatory state. Rather than a singular school of thought or a

compact literature, the regulatory state emerges from a conﬂation of debates as much about the

rise of transnational capital, globalization, and perceptions of the decline of the state as it does

concerns with regulation and the growth of state power. Liberal internationalist perspectives on

foreign policy, for example, conﬂate the emergence of the regulatory state with the rise of a

neo-liberal order, seeing the regulatory state as part reaction to the loss of ﬁscal authority, and

part reaction to the rising power of markets. Waves of tomes since the 1970s have thus declared

the decline of the state. Susan Strange, for example, proclaims that the state is in full retreat, its

authority and absolute power shrinking. Heads of governments, she notes, ‘may be the last to

recognize that they and their ministers have lost the authority over national societies and

economies that they used to have’ (Strange 2000: 3). For Strange, this ‘progressive loss of real

authority’ masks the emergence of transnational actors, international ﬁnance, and the rise of

market dominance, each of which are evolving non-state authority and legitimacy over their

functional domains (ibid., 91–9; see also van Creveld 1999: 336–414).

More consequential for many theorists has been the ascent of markets combined with glo-

balization. As markets have become transnational and capital mobility heightened through

ﬁnancial liberalization, the power of the state to tax and control its economic domain has

been seen as increasingly imperiled, imposing ﬁscal constraints on the state or, at worse, ‘ hol-

lowing out’ the state and its capacity for governance (see Rhodes 1994: 138–40; Holliday 2000:

167–8). In this view, states are now disciplined by market sentiment and neo-liberal ration-

alism, forcing nation-states to conform to the demands of capital lest capital migrates to more

attractive jurisdictions. The decline of the welfare state is thus explained as a combination of

diminishing state ﬁscal capacity due to the pressures of globalization, mobile capital and labour

migration – too high a tax regime and mobile, highly qualiﬁed labour will migrate (see Razin

and Sada 2005). Similarly for Ulrich Beck, the advent of increasing capital mobility forces

Western nation-states to abandon the very tools that for so long made them successful: the

Jarvis

ability to pool economic, social and individual risk through state-provisioned health and

unemployment insurance, state ownership of key resources and utilities, and state-guaranteed

entitlements in respect of education and social security (see Beck 1999; Jarvis 2007; Strange

2000; 83). For Beck, the absolute power of the state relative to capital is now inverted,

forcing nation-states in a ‘race to the bottom’ (one of the few exceptions to the ‘decline of

the state’ thesis is the work of Linda Weiss (1999)). The regulatory state thus represents the

triumph of capital, with the state forced to retreat to managerialism – a hollow shadow of its

former self.

Still others proﬀer the decline of the state as a process of the globalization of regulatory

norms and standards as power is transferred between agential actors. Cobden et al., along with

other liberal internationalists, see the power of the state being systematically transferred to

international organizations and global rule regimes, depriving the state of absolute political and

economic sovereignty because of the exigencies of globalization and the transnationalization of

an increasing spectrum of economic, political and social activity – everything from growing

international trade, investment and the movement of people that require the formation of global

standards, codes and practices to facilitate a global political-economy (Cobden et al. 2005; see

also Braithwaite and Draos 2000; Scott 2004). Global governance thus transposes the functional

imperatives of state-based governance, systematically diminishing the propinquity of state agential

authority and the raison d’être of the state itself.

All these approaches share a common conceptual framework, assuming state power to be

predominantly located in the ﬁscal capabilities of the state and derived from its taxing authority

over markets, where the power of each is inversely related to the other; a kind of zero-sum

continuum – as markets rise, states decline, and vice versa. Such approaches have a particularly

narrow conceptualization of the sources of state power, however, perhaps unfairly characteriz-

ing the regulatory state as weak, eviscerated and powerless. But as Majone (1996: 54) observes,

the sources of state power are more diﬀuse and spread across several functional domains:

1 Redistributive function where resources are transferred between groups to correct social

inequalities, or public goods provisioned to groups who are then compelled to consume

them (elementary education, public transportation, public health care, for example), and

ﬁnanced through taxation, borrowing and the spending power of the state.

2 Stabilization function in which the state manages employment, inﬂation and interest rates

through a determination of industrial and labour policy and the manipulation of ﬁscal and

monetary policy.

3 Regulatory function in which the state sets rules that deﬁne the allocative and settlement

mechanisms of markets and the requirements for market participation; deﬁne standards,

procedures and practices, and enunciate codes that order social, economic and political

engagement.

In this schema, state power is essentially dichotomized between ﬁscal authority; that is the

ability of the state to tax, borrow and spend, and between regulatory authority; that is the ability

of the state to set and make rules, enforce compliance and delegate authority (Majone 1997: 13;

Majone 1999: 4–6). The importance of this distinction for the state is that ﬁscal constraints or a

diminished legitimacy to tax and spend does not imply a diminished capacity to make rules and

regulate. Rather than a reduction in state power the means by which the state exercises its

authority is simply transposed from direct to indirect forms of government. More importantly,

as Majone observes, rule-making is largely free and imposes few ﬁscal burdens on the state apart

from the time, eﬀort and paper needed to make and print rules: ‘the public budget is a soft

State theory and the rise of the regulartory state

constraint on rule makers because the real cost of regulatory programs is borne not by the

regulators but by those who have to comply with the regulation’ (Majone 1997: 13). Measuring

the extent, reach or impact of the state simply in terms of its interventionist or ﬁscal capacities is

thus a poor proxy of state power since states can govern and exercise authority equally as

eﬀectively through rule making and regulation. As the US Oﬃce of Management and Budget

observes:

Budget and revenue ﬁgures are good summaries of what is happening in welfare, defence

or tax policy, and can be used to communicate eﬀectively with the general public over the

fray of program-by-program interest group contention … In the world of regulation,

however, where the government commands but nearly all the rest takes place in the private

economy, we generally lack aggregate numbers to describe what is being ‘taxed’ and ‘spent’

in pursuit of public policies.

(quoted in Majone 1997: 13)

These twin sources of state power, however, are not always reconciled. For Susan Strange, it

represents a paradox; what she observed as an obvious ‘decline of state power’ but at the same

time the increasing ‘intrusion of governments into our daily lives’ in a quantum that is palpably

greater than at any time before in history:

Statutory or administrative law now rules on the hours of work, the conditions of safety in

the work-place and in the home, the behavior of citizens on the roads. Schools and uni-

versities are subject to more and more decisions taken in ministries of education. Planning

oﬃcials have to be consulted before the smallest building is started or a tree is cut down.

The government inspector … has become a familiar and even fearful ﬁgure.

(Strange 2000: xi)

Yet for theorists like Majone this paradox lies at the heart of the rise of the regulatory (rule-

making) state and the decline of the interventionist (tax and spend) state. It produces both a

reduction in the size of government while expanding its powers of governance. At one and the

same time we thus observe the implementation of a neo-liberal agenda (‘downsizing’ the state,

shedding bureaucracies, cutting taxes, reducing ﬁscal expenditures) simultaneously with the emer-

gence of greater regulatory authority (more rule-making, and more indirect forms of state

control). In the United Kingdom this transpired into a 25 per cent reduction in the number of

civil servants between 1976 and the early 1990s, but a relative explosion in staﬃng levels in

regulatory bodies, growing by over 90 per cent (Hood et al. 1999: 29–31; Levi–Faur 2005: 20).

Indeed, a casual glance at the composition of the uniﬁed civil service in the UK in the late

1980s compared to the mid-1990s might indeed lead one to assume government and the state

had shrunk. The ‘Next Steps’ programme commenced in 1988, for example, announced a ratio-

nalization of the number of civil servants in ministerial departments, preserving only a small ‘core

engaged in the function of servicing ministers and managing departments’ (quoted in Dowding

1995: 72). By 1994, ministerial departments had staﬃng levels only about a third of levels at the

commencement of the programme. Yet the sense in which government shrank or its power to

govern diminished is problematic. The profusion of statutory bodies and regulatory agencies

witnessed fully 62 per cent of civil servants in ministerial departments transfer directly into

statutory and delegated agencies charged with regulatory oversight (Dowding 1995: 72–3;

Majone 1997: 10). At the same time, rule-making and the depth of regulatory direction over

domain speciﬁc areas, increased enormously. In the last year of its administration, for example,

Jarvis

the Brown government issued 2,500 pages of directives to the UK police forces concerning

protocols of conduct, governance, and directives about policing, without any changes to the

ﬁscal expenditures on policing (BBC 15 September 2010).

For Majone, the regulatory state is thus not necessarily a weaker, less powerful state, but a recon-

ﬁgured state that uses alternative modalities of governance to eﬀect its power. Indeed, for many

proponents the regulatory state strikes the right balance and modality of governance. In the UK,

the regulatory state became synonymous with the Third Way, New Labour and the premier-

ship of Tony Blair, and was constructed around a ‘range of governance programs’ that relied on

managerial and institutional arrangements to enhance market operation and eﬃciency for the

broader social good (Jayasuriya 2005: 12).

Towards a typology of the regulatory state: modalities of governance

These images of the regulatory state produce mutually reinforcing and contradictory theoriza-

tions as to its rise. On the one hand, the regulatory state is seen as an outcome of the

decline of traditional forms of statist power amid the rise of markets, and on the other, the

outcome of changing modalities of governance that preserve the centrality of the state but in

ways that conﬁne it to new, less interventionist instruments of government. Both acknowl-

edge the rise of markets, the globalization of rule-governed behaviour and the formation of

global rule regimes, and thus both accept these new modalities as legitimate and, in a sense,

optimal given the new political economy of markets. For these theorists, the adoption of reg-

ulatory modes of governance is thus seen as a ‘ necessary condition for the functioning of mar-

kets’ and not just a ‘compromise between economic imperatives and political and social values’

(Levi-Faur 2005: 19).

As a typology of the composite elements of the regulatory state, however, ‘rule-making’

does not get us very far. So governments are making rules, and perhaps more of them, and

exercise power through the issue of rules and directives. But governments have always made

rules, issued decrees and directives, and exerted power through doing so (Hood and Scott 1996:

323). How does this constitute the emergence of a fundamentally new state entity – the reg-

ulatory state? Again, the answer to this question lies across multiple literatures, suggesting a

composite set of images. For Majone, one of the major theorists of regulatory governance, its

distinctive modalities are situated in increasing levels of administrative decentralization, the

break-up of uniﬁed forms of administrative control (central bureaucracies), the creation of

single-purpose regulatory units with budget autonomy, delegation of public service delivery to

proﬁt/not-for proﬁt agencies, competitive tendering and the introduction of contractual/

quasi-contractual relationships where ‘budgets and decision-making powers are devolved to

purchasers who, on behalf of their client group, buy services from the supplier oﬀering the

best value for money’ (Majone 1997: 10, 1999: 3–9; Levi–Faur 2010). Essentially, the reg-

ulatory state is thus distinctive because of the reorganization of how the state does business: who

provides services, how tendering, contracts and quality assurance is administered, and through

what instrumentalities this is achieved. It is this latter element that is perhaps most important:

‘the rise of a new breed of specialized agencies and commissioners operating at arm’s length

from central government’ that represents the ‘most obvious structural consequence of the shift

to a regulatory mode of governance’ (Majone 1999: 17). The delegation of authority to statu-

tory, independent agencies marks a fundamental change in how rules are made, and, in turn, a

fundamental reallocation of power among government instrumentalities, moving it progres-

sively towards decentralized administrative units (Majone 1997: 21; Blankart 1990: 230–6;

Legaspi 2006: 139).

State theory and the rise of the regulartory state

For proponents, the agency model oﬀers a series of distinctive advantages over previous

modalities of governance. First, it allows specialized agencies to develop domain-speciﬁc expert

knowledge, improving governance capacity especially in domains where technical complex-

ities operate (ﬁnancial services, for example). Second, it depoliticizes governance, moving

decision-making to technical and expert domains, where decisions are more likely to be ren-

dered via evidence-based assessment and determination, balancing social and economic objec-

tives. Third, it provides technical-expert decision-makers with autonomy, creating technocratic

policy spaces that are not subject to short-termism or political pressures, but able to plan and

design policy in support of the longer-term sustainability of the sector. Fourth, freed of short-

termism or political pressures, the agency model improves the prospects for policy continuity,

increasing policy certainty and the eﬃciency of governance in the sector. Fifth, the agency

model enhances the credibility of regulatory commitments, reducing uncertainty by removing

the prospects for ‘ devastating ministerial interference’ and thus, in turn, helping to mobilize

private capital into the sector. Finally, agency-based modalities of governance are seen to engineer

high levels of legitimacy: ‘[F]aith in the power of expertise as an engine of social improvement’,

notes Majone, which ‘neither legislators, courts nor bureaucratic generalists’ possess provides ‘an

important source of legitimization for regulators’ (Majone 1997: 17, 1999: 12; see also Cook

and Mosedale 2007: 45–8). The twin pillars of expertise combined with independence thus provide

the cornerstone that cements regulatory governance as an eﬀective, if not superior modality of

governance.

For others, the regulatory state is more than just a modality of governance: it is also a means

of reform and suggests an alternative, depoliticized agency through which to achieve market opera-

tion, eﬃciency and thus development. Indeed, for many it represents a modality able to over-

come obstacles to reform, reform blockages and transform the whole of government incentive

structures in developing countries, where reforms have historically been ‘bogged down’ by the

operation of perverse incentives, ineﬃcient bureaucracies, poor institutional design, account-

ability and oversight systems. For such proponents, while the regulatory state is thus about the

design and construction of new regulatory institutions and regulatory instruments, more fun-

damentally it is also about the realization of state-market outcomes (International Finance

Corporation 2010: 21–2; see also Hira et al. 2005; United Nations 2006: 128–9). As the IFC

observes, ‘Reforms that increase quality in regulatory procedures and requirements – and more

importantly, in regulatory institutions, capacities and incentives – can simultaneously improve a

country’s quality of social life and the conditions for economic activity’ (2010: 1). In the eyes of

the IFC, the regulatory state model is thus seen as a means to:

 making public policy more eﬃcient by allocating national resources to higher value users, by

reducing the risk of policy failures, and by ﬁnding eﬀective policy designs that respect market

principles;

 lowering policy costs and barriers to market entry for ﬁrms, goods, and services, which in

turn boosts foreign direct investment (FDI) and trade, increases the returns on participation

in formal markets, speeds the uptake of new technologies and other innovations, and frees

resources for other uses;

 reducing policy risks for market actors by increasing transparency in the design and use of

policy and by involvement of stakeholders in shaping policies important to them;

 improving business security and market neutrality of policy by increasing accountability for

policy implementation and results, and lowering corruption and vulnerability to capture

government functions.

(International

Finance Corporation 2010:

13)

Jarvis

Conclusion

Clearly, this image of the regulatory state is laden with objectives that go beyond a modality of

governance and encompass forms of policy transfer designed to construct markets and a series of

speciﬁc institutional types deﬁned by neo-liberal market rationalism. The regulatory state thus

assumes a larger political project, one designed to embed developing states in a speciﬁc eco-

nomic and political order. As Julia Black notes, the focus on regulatory techniques (agency-

based regulation, stakeholder engagement and transparency practices, etc.) ‘can result in a radical

rethinking of the ways in which societal ends can be achieved. However, it can [also] divert

attention from the issue of how those ends should be deﬁned, and by whom’ (Black 2000: 598).

The danger, as Black observes, is that in pursuit of an increasingly ‘proceduralized’ approach to

regulation, the literature and practices of regulation become ‘technicized’; the predominant

concern being the implementation of regulation rather than the values that are pursued: that the

focus on the epistemological character of regulation is obscuring issues of its moral form (ibid.;

see also Levi-Faur 2005: 14; North 1990).

While Black is correct to suggest that ‘technicized’ discourses can conceal the values that

underlie them, it remains the case that constructing regulatory states in the global South of

whatever ‘moral form’ rests on a series of technical instrumentalities. These fall into three main

areas: ﬁrst, design of regulatory instruments, including institutional composition, functional

structure and rule deployment; second, capacity and operational requirements, including

resource, technical, administrative and analytical capacities; and third, institutional technologies

for normalizing and proceduralizing the various dimensions of regulation, including the instan-

tiation of legitimacy, trust, and compliance regimes. In the most visceral sense, these require-

ments speak to the displacement and redesign of entire institutional landscapes in a process that

involves new rules, new ways of making and enforcing rules, new incentive systems for

engendering compliance and distributing costs and economic gains among sectoral actors, and

new accountability, participation, and transparency instruments that serve as functional

mechanisms to sustain governance and eﬃciency in the sector. While, of course, much atten-

tion focuses on the institutional design elements of regulatory governance, the greater and

more signiﬁcant quantum rests in evolving a series of highly complex reﬂexive relationships

between agential actors, formal institutions, procedural authority and norms that instantiate

the new institutional and rule environment (Cook and Mosedale 2007: 45). It is this latter

series of institutional–socio-political technologies that suggests a much greater, more compli-

cated, problematic, and costly set of relationships to construct and a political space where,

potentially, errors, possibilities for maleﬁcence, regulatory capture, corruption and less than

optimal sector outcomes, ultimately rest. Constructing governance regimes that are legitimate

and perceived to be so, that are observed to be transparent and free from special interest

capture, and function in a way that is seen to balance public and private sector interests while

delivering enhanced social and economic outcomes, is a highly complex regulatory exercise.

These dimensions of regulatory governance thus suggest a much greater series of costs,

capacities and institutional technologies across a wide spectrum of socio-political sites (the

judiciary, administrative review systems, tribunals and appeals processes, enforcement and com-

pliance regimes, consultation and engagement systems, etc.), than might ﬁrst appear to be

the case. Indeed, while proponents of regulatory modes of governance launder their cost-

eﬀectiveness and suggest they impose few ﬁscal burdens on the state, in reality the acquisition

and realization of the soft-institutional technologies necessary to ensure their eﬃcient func-

tioning represent extensive acquisition, set up, implementation, and maintenance costs (see

Minogue 2004).

State theory and the rise of the regulartory state

Table 5.2 A comparative typology of the interventionist and regulatory state

Attributes Quasi-patrimonial state Interventionist

state

Regulatory

state

Required capacities and attributes

Functional

roles

Reproduction and maintenance of social,

political and economic order

Redistribution Constructing markets Institutional technologies for the

collection, ordering and

dissemination of information

Servicing socio-political-economic networks Macroeconomic stabilization

(economic growth, employment,

inﬂation and interest rates)

Enhancing market

efﬁciency

Technical and institutional

platforms to overcome

information asymmetries

Preserving and enhancing existing authority

structures

Enhancing access to social,

economic and political resources

Facilitating capital

mobilization

Access to information provisions

and state-based information

mechanisms of disclosure

Protecting vested interests Providing credible

commitments

Participatory processes in

decision-making

Controlling dissent Consultative and review

mechanisms

Effective accountability

mechanisms

Institutional capacity for third

parties to enforce/seek redress to

enforce government

commitments

Market-based clearing and

settlement systems across various

sectors

Institutional/market design

capacity

Table 5.2 (continued)

Attributes Quasi-patrimonial state Interventionist

state

Regulatory

state

Required capacities and attributes

Instruments Patron-client-based access to/distribution of

resources

Taxation Rule-making Institutional mechanisms to

ensure information

transparencies

Indirect coercion through access/denial of

patronage

Borrowing Compliance and

enforcement

Effective, functioning and

impartial judiciary

Dispensation of access/denial to state

resources/revenue streams

Fiscal expenditures Administrative review

and adjudication

Judicial legitimacy and

recognized authority

Control and access to markets/business/

governance domains

Budget allocations and resource

transfers between groups

Competitive tendering Negligible to low levels of judicial

corruption

Monetary policy Issuance of contracts Adequate judicial capacity

Fiscal policy Licenses Administrative review/tribunals

proceduralization

Industrial policy Setting standards and

codes

Compliance and audit capacities

across various institutional

spectrums

Deﬁning and controlling

procedural mechanisms

Functional property rights

Enforcement and punitive

mechanisms across various

institutional spectrums

Probity monitoring and

enforcement mechanisms

(Continued on next page)

Table 5.2 (continued)

Attributes Quasi-patrimonial state Interventionist

state

Regulatory

state

Required capacities and attributes

Key actors Oligarchs Political parties Regulators Adequate and independent

resources for regulators and

regulatory affairs

Political elites Civil servants Industry/private sector

groups

Platform capacity for stakeholder

engagement/review in decision-

making

Business/economic elites Corporate groups Civil society groups Sufﬁcient analytical and human

capacity to populate regulator

Nominated mandarins Trade unions Technocrats and experts Adequate compensation to

attract and retain personnel with

sufﬁcient analytical and expert

knowledge capacity

Administrative tribunals Sufﬁcient capacity and resources

to operationalize transparent

administrative review processes

Judiciary/Judges Operational accountability

mechanisms to ensure regulator

is held accountable for decisions

Disclosure, transparency and

freedom of information

mechanisms to ensure against

regulatory capture by sectional

interests

Realized legitimacy of the

regulator in discharge of

regulatory mandate

Table 5.2 (continued)

Attributes Quasi-patrimonial state Interventionist

state

Regulatory

state

Required capacities and attributes

Conﬂict arenas Relationships between oligarchs Budgetary allocations Competition for control

over rule-making

Review and disputation

procedures are in place and

operative

Elite competition for access to oligarchs Entitlements Disputes/inter-agency

competition for rule

ownership

Enforcement mechanisms for

compensation

Competition/disputes over patronage

entitlements

Budget transfers Disputes over domain

authority, reach and

extensity

High-capacity administrative

review

Factionalism and disputes between political

and social networks

Ministerial control over resource

allocation

Disputes over rule

interpretation

Compliance to and respect for

administrative proceduralism

Inter-ministerial/ministry

competition

Key Institutions and

governance modalities

Oligarch/elite control of key

decision-making institutions/apparatus

Parliament Parliamentary

committees

Ability to reallocate power from

centralized bureaucracies to

independent administrative units

Nominated senior political mandarins Civil service/bureaucracy Independent agencies Ability to mediate inter-agency

resource competition

Elite-controlled executive branch Ministerial departments Commissions Ability to coordinate among

polycentric nodes of governance

State-owned enterprises Tribunals

Command and control Public hearings

Polycentric decision-

making structures

Policy style Top-down, elite-dominated, command and

control, low levels of accountability or

transparency

Discretionary, populist, political Rule-bound, mandated,

legalistic

(Continued on next page)

Table 5.2 (continued)

Attributes Quasi-patrimonial state Interventionist

state

Regulatory

state

Required capacities and attributes

Political culture Oligarch/elite-based power Corporatist, hierarchical,

centralized, top-down; statist

Pluralist, diffuse,

administrative, technical,

specialist, domain speciﬁc,

market-orientated

personal/family power networks

Dominance of oligarchic/elite political/

social/economic networks

Circulation of power positions among

elites

Political accountability Nominal accountability Direct/representative

democracy

Indirect/agency-based

Intra-elite informal accountability through

patron-client networked based consent

Source: Adapted from Majone 1997: 12–15. Jarvis, D. S. L., Regulatory states in the South: Can they exist and do we want them? The case of the Indonesian power sector (1 December

2010). Lee Kuan Yew School of Public Policy Research Paper No. LKYSPP10-11. Available at: http://ssrn.com/abstract=1738189 or http://dx.doi.org/10.2139/ssrn.1738189

Bibliography

Almond, Gabriel A. 1988. ‘The return of the state’, American Political Science Review 82(3): 853–74.

Almond, Gabriel A. and Stephen J. Genco. 1977. ‘Clouds, clocks, and the study of politics’, World Politics

29 (4): 489–522.

Ananta, Aris and Yohanes Eko Riyanto. 2006. ‘Riding along a bumpy road: Indonesian economy in an

emerging democratic era’, ASEAN Economic Bulletin 23(1): 1–10.

BBC World Service, September 15, 2010.

Besant-Jones, John E. 2006. ‘Reforming power markets in developing countries: What have we learned?’

Energy and Mining Sector Board Discussion Paper no. 19, World Bank, Washington, DC.

Beck, Ulrich. 1999. World Risk Society. Cambridge: Polity Press.

Black, Julia. 2000. ‘Prodecuralizing regulation: part 1, Oxford Journal of Legal Studies 20(4): 597–614.

Blankart, Charles B. 1990. ‘Strategies of regulatory reform: An economic analysis with some remarks on

Germany’, in Giandomenico Mojone (ed.) Deregulation or Reregulation? Regulatory Reform in Europe and

the United States. London: Pinter Press, 211–22.

Braithwaite, John and Peter Draos (eds). 2000. Global Business Regulation.Cambridge:CambridgeUniversityPress.

Carroll, Toby. 2010. Delusions of Development: The World Bank and the Post-Washington Consensus in

Southeast Asia. London: Palgrave Macmillan.

Clark, William Roberts. 1998. ‘Agents and structures: Two views of preferences, two views of institutions,

International Studies Quarterly 42(2): 245– 70.

Cobden, Richard, David Mitrany and Kenichi Ohmae. 2005. Liberal Internationalism and the Decline of the

State. London: Palgrave Macmillan.

Cook, Paul and Sarah Mosedale. 2007. Regulation, Markets and Poverty. Cheltenham: Edward Elgar.

Cortell, Andrew P. and Susan Peterson. 2001. ‘Limiting the unintended consequences of institutional

change’, Comparative Political Studies 34(7): 768 – 99.

Dessler, David. 1999. ‘Constructivism within a positivist social science’, Review of International Studies 25:

123–37.

Dowding, Keith. 1995. The Civil Service. London: Routledge.

Hall, John A. and G. John Ikenberry. 1989. The State .

Milton Keynes: Open University Press.

Hall, Peter. 1986. Governing

the Economy: The Politics of State Intervention in Britain and France. New York:

Oxford University Press.

Hawkesworth, Mary. 1992. ‘Epistemology and policy analysis’, in William N. Dunn and Rita M. Kelly

(eds) Advances in Policy Studies since 1950. New Brunswick, NJ: Transaction Publishers.

Hay, Colin and Daniel Wincott. 1998. ‘Structure, agency and historical institutionalism’, Political Studies

46(5): 951–7.

Held, David. 1995. Democracy and the Global Order: From the Modern State to Cosmopolitan Governance.

Stanford, CA: Stanford University Press.

Hintze, Otto. 1975. The Historical Essays of Otto Hintze. New York: Oxford University Press.

Hira, A., D. Huxtable, and A. Leger. 2005. ‘Deregulation and participation: An international survey of

participation in electricity regulation’, Governance 18(1): 53–88.

Holliday, Ian. 2000. ‘Is the British state hollowing out?’, The Political Quarterly 71(2): 167–76.

Hood, Christopher and Colin Scott. 1996. ‘Bureaucratic regulation and new public management in the

United Kingdom: Mirror-image developments?, Journal of Law and Society 23(3): 321–45.

——, Colin Scott, Oliver James, George Jones and Tony Travers. 1999. Regulation Inside Government.

Oxford, Oxford University Press.

International Finance Corporation. 2010. ‘Regulatory governance in developing countries’, Investment

Climate Advisory Services, Washington, DC, World Bank.

Jarvis, Darryl S.L. 2007. ‘ Risk, globalization and the state: A critical appraisal of Ulrich Beck and the world

risk society thesis’, Global Society, 21(1): 23–46.

—— 2010. ‘Institutional processes and regulatory risk: A case study of the Thai energy sector’, Regulation

and Governance 4: 175–202.

Jaswal, Pragya and Mitali Das Gupta. 2006. ‘Energy demands and sustaining growth in Southeast Asia’,

Asia 2015: Promoting Growth, Ending Poverty. Available at: www.asia2015conference.org

Jayasuriya, Kanishka. 2005. Economic constitutionalism, liberalism and the new welfare governance’,

Working Paper no. 121, June, Asia Research Centre, Murdoch University, Perth, Western Australia.

Jordana, Jacint, David Levi-Faur and Xavier Fernandez Marin. 2009. ‘The global diﬀusion

of regulatory

agencies: Channels

of transfer and stages of diﬀusion’, IBEI Working Papers, 2009/28, Barcelona.

State theory and the rise of the regulartory state

Kapur, Devash, John P. Lewis and Richard Webb, 1997. The World Bank: Its First Half Century.

Washington, DC: The Bookings Institution.

Krasner, Stephen D. 1984. ‘Approaches to the state: Alternative conceptions and historical dynamics,

Comparative Politics 16(2): 23–246.

—— 1988. ‘Sovereignty: An institutional perspective’, Comparative Political Studies 21(1): 66–94.

Legaspi, Perla E. 2006. ‘Reform and practices in local regulatory governance: The case of the Philippines’,

in Martin Minogue and Ledivina Cariño (eds) Regulatory Governance in Developing Countries. Cheltenham:

Edward Elgar, 138–56.

Levi-Faur, David. 2005. ‘The global diﬀusion of regulatory capitalism’, Annals of the American Academy of

Political and Social Science 598: 12–32.

—— 2010. ‘Regulation and regulatory governance’, Jerusalem Papers in Regulation and Governance,

Working Paper 10, February, Jerusalem Forum on Regulation and Governance, Hebrew University, Israel.

——and Jacint Jordana. 2005. ‘Regulatory capitalism: Policy irritants and convergent divergence’, Annals of

the American Academy of Political and Social Science 598: 191–7.

Lundquist. 1987. Implementation Steering: An Actor-Structure Approach.

McLennan, Gregor. 1991. Marxism, Pluralism and Beyond: Classic Debates and New Departures. Cambridge:

Polity Press.

Majone, Giandomenico. 1996 ‘The rise of statutory regulation in Europe’, in Giandomenico Majone and

Jeremy Richardson (eds) Regulating Europe . London: Routledge, 47–60.

—— 1997 ‘From the positive to the regulatory state: Causes and consequences of changes in the mode of

governance’, Estudio/Working paper 1997/93, Centro de Estudios Avanzados en Ciencias Socales,

Instituto Juan, Madrid, Spain.

—— 1999 ‘The regulatory state and its legitimacy problems. West European Politics 22(1): 1–24.

—— 2006 ‘The internationalization of regulation: Implications for developing countries’,

in Martin Min-

ogue and Ledivina

Cariño (eds) Regulatory Governance in Developing Countries. Cheltenham: Edward

Elgar, 39–60.

March, J.G. and J.P. Olsen. 1979a. ‘Organizational choice under ambiguity’, in J.G. March and J.P. Olsen

(eds) Ambiguity and Choice in Organizations. Bergen: Universitetsforlaget, 10–23.

——and J.P. Olsen. 1979b. Ambiguity and Choice in Organizations. Bergen: Universitetsforlaget.

——and Johan P. Olsen. 1994. Institutional Perspectives on Political Institutions.

Mason, Edward S. and Robert E. Asher. 1973. The World Bank since Bretton Woods. Washington, DC: The

Bookings Institution.

Minogue, Martin. 2004. ‘Public management and regulatory governance: Problems of policy transfer to

developing countries’, in Paul Cook, Colin Kirkpatrick, Martin Minogue and David Parker (eds) Leading

Issues in Competition, Regulation and Development. Cheltenham: Edward Elgar, 165–81.

—— 2006. ‘Apples and oranges: Comparing international experiences in regulatory reform’, in Martin

Minogue and Ledivina Cariño (eds) Regulatory Governance in Developing Countries. Cheltenham: Edward

Elgar, 61–81.

Nettl, J.P. 1968. ‘The state as a conceptual variable’, World Politics 20(4): 559–92.

North, Douglas. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge

University Press.

Orren, K. and S. Skowronek (eds). 1993. The Dynamics of American Politics: Approaches and Interpretations.

Boulder, CO: Westview Press.

——and S. Skowronek. 1994. ‘Beyond the iconography of order: Notes for a “new2 institutionalism’,in

L. Dodd and C. Jillson (ed.) Dynamics of American Politics. Boulder, CO: Westview Press.

——and Stephen Skowronek. 1998–1999. ‘Regimes and regime building in American government: A

review of literature on the 1940s’, Political Science Quarterly 113(4): 689–702.

Phillips, Nicola. 2006. ‘States and modes of regulatory governance in the global political economy’,in

Martin Minogue and Ledivina Cariño (eds) Regulatory Governance in Developing Countries. Cheltenham:

Edward

Elgar, 17–38.

Przeworski, Adam.

1987a. Methods of Cross-National Research, 1970–83: An Overview,31–49.

——(ed.). 1987b. Comparative Policy Research: Learning from Experience. Aldershot: Gower.

—— 1990. The State and the Economy under Capitalism. London: Routledge.

Ray, James Lee. 2001. ‘Integrating levels of analysis in world politics’, Journal of Theoretical Politics 13: 355–88.

Razin, Assaf and Efraim Sadka. 2005. The Decline of the Welfare State. Cambridge, MA: MIT Press.

Rhodes, R.A.W. 1994. ‘The hollowing out of the state: The changing nature of the public service in

Britain’, The Political Quarterly 65(2): 138–51.

Jarvis

Robison, Richard and Andrew Rosser. 2000. ‘Surviving the meltdown: Liberal reform and political oli-

garchy in Indonesia’, in Richard Robison, Mark Beeson, Kanishka Jayasuriya and Hyuk-Rae Kim (eds)

Politics and Markets in the Wake of the Asian Crisis. London and New York: Routledge, 171 – 91.

Sari, Agus P. n.d. ‘Power sector restructuring and public beneﬁts’, World Resources Institute. Available at:

http://pdf.wri.org/power_politics/indonesia.pdf

Scott, Colin. 2004. ‘Regulation in the age of governance: the rise of the post regulatory state’, in Jacint

Jordana and Davi Levi-Faur, (eds) The Politics of Regulation: Institutions and Regulatory Reforms for the Age of

Governance. Cheltenham: Edward Elgar, 145–74.

Skocpol, Theda. 1985. ‘Bringing the state back in: Strategies of analysis in current research’, in Peter B. Evans,

Dietrich Rueschemeyer and Theda Skocpol (eds) Bringing the State Back In. Cambridge: Cambridge

University Press, 3–43.

Skowronek, Stephen. 1982. Building a New American State: The Expansion of National Administrative Capacities

1877–1920. Cambridge: Cambridge University Press.

Smith, David. 1999. Will Europe Work? London: Proﬁle Books.

Spiller, Pablo T. 1996. ‘Institutions and commitment’, Industrial & Corporate Change 5(2): 421–52.

——and Mariano Tommasi. 2005. ‘The institutions of regulation: An application to public utilities’,in

C. Menard and M.M. Shirley (eds) Handbook of New Institutional Economics. Dordrecht: Springer, 515–43.

Stern, Jon. 1997. ‘What makes an independent regulator independent’, Business Strategy Review 8(2): 67–74.

—— 2000. ‘Electricity and telecommunications regulatory institutions in small and developing countries’,

Utilities Policy 9: 131–57.

——and Stuart Holder. 1999. ‘Regulatory governance: Criteria for assessing the performance of regulatory

systems. An application to infrastructure industries in the developing countries of Asia’, Utilities Policy 8:

33–50.

Strange,

Susan. 2000. The Retreat

of the State: The Diﬀusion of Power in the World Economy. Cambridge:

Cambridge University Press.

Thomas, Steve, David Hall and Violeta Corral. 2009. ‘Electricity privatization and restructuring in Asia-Paciﬁc’.

Available at: www.psiru.org (accessed 14 June 2010).

Therborn, Göran. 1986. ‘Neo-Marxist, pluralist, corporatist, statist theories and the welfare state’,inA.Kazancigil

(ed.) The State in Global Perspective , Northampton, NH: Gower, 204–31.

Tilly, Charles. 1984. Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage Foundation.

United Nations. 2006. ‘Diverging growth and development: World economic survey 2006’, Department

of Economic and Social Aﬀairs, United Nations, New York.

van Creveld, Martin. 1999. The Rise and Decline of the State. Cambridge: Cambridge University Press.

Weber, Max. 1978. Economy and Society: An Outline of Interpretive Sociology. Berkeley, CA: University of

California Press.

Weiss, Linda. 1999. The Myth of the Powerless State. Ithaca, NY: Cornell University Press.

Williams, J.H and N.K. Dubash. 2004. ‘Asian electricity reform in historical perspective’, PaciﬁcAﬀairs 77:

411–36.

Wu, Xun and Priyambudi Sulistiyanto. 2006. ‘Independent power producer (IPP) in Indonesia and the

Philippines’, in Michael Howlett M Ramesh (eds) De-regulation and Its Discontents: Rewriting the Rules in

Asia. Cheltenham: Edward Elgar, 109–23.

Yoo, S.H. 2006. ‘The casual relationship between electricity consumption and economic growth in the

ASEAN countries’, Energy Policy 34: 3575–82.

Zhang, Yin-Fang and Margo Thomas. 2009. ‘Regulatory reform and governance: A survey of selected

developing and transition economies’, Public Administration and Development 29(4): 330–9.

State theory and the rise of the regulartory state

The public choice perspective

Andy Whitford

Introduction

Over the last ﬁve decades, few research agendas in the social sciences have brought about more

controversy about the “proper” role of government than public choice theory. On the one hand,

public choice researchers have built theories and empirical studies that start with the core intuitions

of economics but extend from that to describe a range of activities in government. On the other

hand, political scientists and those working in the policy sciences often point to public choice

theory as a primary example of the excesses and pathologies of the rational choice paradigm.

This chapter describes some of the contributions to the study of public policy processes of

those who have helped build that agenda. I start with two assumptions: that public choice oﬀers

some fundamental insights about how policy is made and implemented, and that some of the

insights drawn from public choice theory are fundamentally ﬂawed. Public choice, like many other

research agendas, is a collection of imperfect attempts to understand complex and dynamic phe-

nomena. One reason for this imperfection is that unlike in physics where the particles rarely

learn from the researchers, those who make or implement policy—or at least want to bend policy

for their own purposes—have often used public choice as a way of justifying those purposes.

Public choice is an important argument about the proper role of government (Majone 1989).

In brief, public choice theory uses the tools of economics to understand why politicians and

the people who elect them do what they do when they make decisions about what governments

will or should do. The ﬁngerprints of economists are all over public choice theory, and this is

for good reason: many of those who founded public choice theory have gone on to receive

great acclaim for their contributions, including winning the Nobel Prize. Names like Kenneth

Arrow, Duncan Black, James Buchanan, Gordon Tullock, Anthony Downs, and Mancur Olson

are known largely in economics because of the contributions they made to the theoretical infra-

structure of this ﬁeld. Other names like William Riker and Elinor Ostrom are known—even

though most of their footprints have been in political science—because of their contributions to

public choice theory.

Speciﬁcally, public choice has changed how we understand the inner workings of democratic

decision-making. For scholars like James Buchanan, who as much as (and probably more) than

any other person put public choice on the intellectual map, public choice is diﬀerent from how

political science (at least political science pre-public choice) sees the world, because public choice

is “politics without romance.” For many people, public choice is largely associated with one set

of people and in fact one state of the Union: Virginia. What public choice brought to the table

was a willingness to ignore most of what we thought we knew about politics by situating

every analysis in a logical framework built on the assumption that oﬃcials (either elected or

unelected) are not public-spirited: that they do not necessarily pursue the “public interest” as

“servants of the people.” Indeed, many debates in public choice have raged around the question

of whether we can ever know what “the people” want—whether there is a knowable “social

welfare function.”

In traditional economics, a person wants to maximize his or her own utility, and while one

of the signature results of public choice is that this assumption does not require that people fail

to care about others, outsiders often see it as a restrictive assumption that people are mostly

guided by their own self-interest. Of course, people outside government, living in markets,

sometimes seem to follow their own interests, so public choice theorists did not see much of a

problem assuming the same about people in government—that they are just like the rational

actors that populate traditional economic theories of markets.

Economists see the world through the lens of “methodological individualism” (a focus on

individuals over groups), which is diﬀerent from how political science circa the 1950s largely

saw politics. But another signature contribution of public choice—one not always recognized in

either political science or the policy sciences—is that the agenda changed how economists saw

the world: that in government, the costs and beneﬁts of choice are often collective, rather than

individual. So public choice theorists had to throw out many of the intuitions of the traditional

economic model because this means that an individual is aﬀected by the decisions they make in

complicated ways.

This chapter will tell the story of public choice theory by walking through the canon: the

insights that most people who have added to the literature would count among the foremost

contributions, the things that they can “hang their hats on.” I engage these insights by considering

six basic questions that public choice theory has sought to answer about the electors, the elec-

ted, and the governments they create and maintain. I want to be clear that there are other ways

to recount these insights. For instance, Shughart sees the contributions as revolving around the

diﬀerent ways political preferences are aggregated and/or expressed—of how theorists have

come to understand elections, legislatures, bureaucracies, and other institutions such as the courts

(Shughart 2008). Many people have tried to summarize the ﬁeld, with varying degrees of thick

description for example (Mueller 2003; Rowley and Schneider 2004; Shughart and Razzolini

2001). The approach taken in this chapter has as its inspiration Mueller (1976).

Public choice tells us that:

electing better people will not, by itself, lead to much better government. Adopting the

assumption that all individuals, be they voters, politicians, or bureaucrats, are motivated

more by self-interest than by public interest evokes a Madisonian perspective on the pro-

blems of democratic governance. Like that founding father of the American constitutional

republic, public choice recognizes that men are not angels and focuses on the importance

of institutional rules under which people pursue their own objectives.

(Shughart 2008)

As such, the agenda is similar to other, aﬃliated theories in economics, such as new institu-

tional economics, that also worry about institutional design and eﬀectiveness in a variety of

economic and political settings. That topic is the focus of the next chapter in this volume.

The public choice perspective

Six questions and the canon

This overview of public choice theory centers on six important questions this literature has

sought to answer over the past ﬁve decades. The ﬁ rst question can be written as “why is there

collective choice?” One way we can write the second question is “do we need politicians?”

The third question follows from the second: “what happens when we have politicians?” While the

fourth question is narrower than the ﬁrst three, it shows how important public choice theory

has been to the evolution of both economics and the policy sciences. The fourth question is

“what we do about those pesky public goods?” While the ﬁfth question is not unique to public

choice theory, public choice theory has helped to answer this question in ways that are both

unique and novel. The ﬁfth question this review will discuss is “what should society want?”

While the sixth question has involved only a select subgroup of public choice theorists and

researchers, this question is particularly important for a review that describes the importance of

public choice theory for those interested in public policy processes. The sixth question is “what

about those who implement public policy?”

One reason to evaluate the canon of public choice theory from the perspective of these six

questions is that those outside the canon often see it as a monolith. Speciﬁcally, readers often see

public choice theory as solely centered on not using the state to improve public policy outcomes.

This is understandable if only because of the aforementioned conservative bias sometimes expres-

sed by those who founded and have continued to expand the ﬁeld. Yet, as this review hope-

fully will make clear, public choice is sometimes less a well-constructed and synthetic paradigm

than it is a starting point for evaluating a wide array of institutions we observe every day in

governments around the world. Together these questions show the breadth, reach, and even

internal disputes that are hallmarks of public choice theory.

It is abundantly clear that a full review of these questions is beyond the scope of this chapter.

Interested readers are directed to the other reviews noted above, and those interested in a full

discussion of certain topics are directed to classics such as (Arrow 1963; Black 1958; Buchanan

and Tullock 1962; Downs 1957; Niskanen 1971; Olson 1965; Riker 1962). For instance, Shughart

and Razzolini (2001) provide a range of perspectives—including showing public choice theory’s

fractures.

Why is there collective choice? As noted above, public choice theory builds on the axioms of

traditional microeconomic theory (e.g. atomistic competition, having many buyers and sellers,

particular information conditions, and other attributes of perfect competition), which build on

simple assumptions about individuals being both rational and self-interested. In markets, while

all the attributes of traditional microeconomic theory lead to outcomes that economists would

call socially preferable (namely, Pareto eﬃcient outcomes), that is not the case in some important

situations.

The simplest context to see this is the well-known prisoner’s dilemma, which shows that two

individuals, both pursuing their self-interest and acting rationally, may achieve outcomes that

beneﬁt neither one. As Dennis Mueller puts it in his classic survey of public choice theory,

“problems of collective choice exist in all but a purely Hobbesian, anarchistic society and are

coterminous with the existence of recognizable groups and communities” (Mueller 1976: 397).

For public choice theorists, these situations happen when people debate and select diﬀerent levels

or attributes of the provision of public goods. Examples include national defense, public safety,

and other attributes of free and civilized societies. Theorists and empiricists have debated diﬀerent

solutions to the problem of public goods provision. Because of these eﬀorts, we now know that

how society solves these kinds of problems depends in part on the size of the group and how

long individuals expect to interact (Buchanan 1965; Olson 1965).

Whitford

As Mueller puts it, “Thus, democracy, formal voting procedures for making and enforcing

collective choices, is needed by communities of only a certain size and impersonality” (Mueller

1976: 398). This shows that in some situations individuals choose to make decisions collectively

and these decisions must be governed by “rules of the game,” the purpose of which is to reg-

ularize the interactions to reduce the chance that people will make decisions individually that

lead to socially suboptimal outcomes.

In one sense, public choice theory became an attempt inside the discipline of economics to

understand and perhaps formulate diﬀerent rules, given that individuals are rational, self-serving,

and cognizant of the need for collective choice. This was diﬀerent from what political scientists

were doing at that time because of a modeling technology that ﬂowed from traditional micro-

economics. Essentially, public choice theory had “pure” microfoundations. Yet, inside economics,

those foundations had led most economists to ignore collective choice. An example of this is

the Keynesian approach to macroeconomic analysis, which included government as almost an

omniscient and pure dictator/planner. Other views, such as those expressed within “rational

expectations theory,” also modeled the politician and/or bureaucrat in naïve ways—in this case,

the planner is largely inept (Mitchell 2001).

In sum, public choice theory sees a need for collective choice but also envisions collective

choice as fraught with opportunities for individuals to subvert the collective process for their

own ends.

Do we need politicians? Recognizing the need for collective choice does not answer the ques-

tion of what kind of collective choice we need. For this, public choice theorists turned to the

tools of formal modeling, and in doing so they encountered an array of subsidiary questions.

First, what happens if people try to make decisions under unanimity rule? The answer largely

has been “it isn’t easy.” Speciﬁcally, public choice theory asked whether people, making decisions

under a unanimity rule, could decide to support the provision of public goods. Writers argued

for processes that essentially auctioned oﬀ goods (using a variant of Walrasian tâtonnement, with

prices “discovered” through a series of bids) (the so-called Lindahl equilibrium) (Lindahl 1919;

Milleron 1972). The diﬃculty with this outcome is that it must be discovered, over time, through

interactions that are themselves costly (Black 1958; Buchanan and Tullock 1962). Also, the

process is susceptible to strategic behavior.

A consequence is that theorists then tried to understand optimal decision rules in a demo-

cratic society. If unanimity increased costs, was another rule less costly? For Mueller, studies like

Buchanan and Tullock (1962) and Breton (1974) show us that “the optimal rule is thus the one for

which the expected gain in utility from redeﬁning the bill to gain one more supporter just equals

the expected loss in time from doing so” (Mueller 1976: 402). Mueller goes on to say that the

problem is that the optimal rule will vary across issues.

Many societies use majority rule, and while a full discussion of its advantages and disadvantages is

beyond the scope of this chapter, a few key aspects warrant emphasis. Essentially, Rae (1969)

and Taylor (1969) show that majority rule is the only rule that reduces the potential bad out-

comes, or what we might call regret, that a supporter would feel if her proposal does not pass

(or one that she opposes does pass). The diﬃculty with majority rule, though, is the prospect of

cycling. Discussing cycling can be particularly technical but the intuition is quite straightfor-

ward. In some cases, when individuals have certain types of preferences, there is a chance that

no one proposal can gain a majority that is impervious to the introduction or consideration of

an alternative proposal. Debates have raged over what this means in practice, even though the

technical literature is fairly consistent on this possibility. A reason for this consistency is that we

have studied cycling for at least 200 years, and technical work has conﬁrmed long-held intui-

tions (McKelvey 1976; Satterthwaite 1975; Schoﬁeld 1978). Some models suggest cycling is less

The public choice perspective

likely to happen in some situations, usually involving large numbers of voters or unusual deci-

sion rules (Saari 2006; Tovey 2010).

Of course, political scientists have also helped us to extend our early understanding of the

diﬃculties of direct representation. Economists contributed much of the early legwork in this

area, but political scientists and others working in “positive political theory” have helped to

broaden those early intuitions. This is now mostly associated with the arena of “social choice and

welfare” than with public choice theory, although the roots are similar. Instead, public choice

theorists are associated with their development of the consequences of cycling such as log-

rolling. Buchanan and Tullock described some of the issues involved in logrolling extensively in

their classic 1962 book. Essentially, logrolling ﬂows from diﬀerent individuals ﬁnding issues dif-

ferently salient. Views vary on logrolling and its usefulness in voting systems, partly due to some

beliefs that logrolls may reduce eﬃciency rather than increase it. Commonly cited examples

include tax policy and pork barreling.

What happens when we have politicians? Most people think of logrolling in the context of legislatures,

and since we have elected politicians in most democracies, public choice theorists sought to under-

stand how individuals, considering their own preferences and their options, choose representatives.

Moreover, how do politicians, given those preferences and options, compete for votes?

The central result in this area is the “median voter theorem,” which says that if preferences

lie in a single (liberal-conservative) dimension, and if people vote for the candidate closest to

their own position on that dimension, then the winning candidate will pick the position of the

median voter. This result, ﬁrst oﬀered in Hotelling (1929) but elaborated most notably by

Downs (1957) and Black (1958), is considered “the pioneering contribution in public choice”

(Mueller 1976: 408, fn. 22). The power of the theorem has led to a host of reﬁnements, some

of which are particularly important for understanding public policy processes. For instance,

when candidates are elected in stages (e.g. primaries followed by general elections), the candi-

dates are pulled apart in terms of the policy positions they select (e.g., Aronson and Ordeshook

1972). When the model moves to two policy dimensions, the situation becomes even more

complicated, if only because of the importance of the cycling theorems noted above.

The literature on the median voter theorem then moved on to even more reﬁned versions.

For instance, what happens when some voters are alienated? What happens when the distributions

of policy preferences are multi-modal? What happens when we take into account logrolling

(especially when a group with minority policy preferences supports a candidate)? In all of these

and other situations, the constant worry about cycling remains. As Downs (1957) notes, there

are real implications of cycling if it leads to the regular defeat of incumbents since policy is essen-

tially “swinging” in the two-dimensional space. While in some systems (like the US) incumbents

seem to enjoy regular advantages at the point of re-election, in other situations (e.g. municipal

elections, school boards) incumbents often lose.

In Hotelling’s original model, voters choose from among two candidates, but this is not so

for much of the world since multiple parties often compete for votes. The canonical position in

public choice theory recognizes that single-member plurality, ﬁrst-past-the-post (SMP) systems

are a speciﬁc rule chosen to govern the operation of democratic systems, and that other systems,

such as proportional representation (PR) can lead to qualitatively diﬀerent outcomes. Following

pioneering work by political scientists like Riker (1962), which showed that diﬀerent systems

converged to coalitions of roughly the same sizes, scholars like Schoﬁeld added numerous

wrinkles to the study of multiparty democracies. Much of this follows from the search for stability

in systems where there are two policy dimensions.

The public choice canon also sought to understand what happens to voters when politicians

are asked to represent constituents. For instance, Mueller describes how public choice studies

Whitford

broke the vote decision into components that depended on comparing the costs (of voting),

with the potential beneﬁts (e.g. the risky public beneﬁts and the expected private beneﬁts). This

focus, found in writings like Downs (1957) and Tullock (1967), is important if only because it

led to one of the classic statements about politics held by those working in public choice theory:

that the costs of gathering political information or voting, as they grow, will swamp the beneﬁts

of information acquisition or voting—with the eﬀect that people may not vote or acquire

information. This “rational ignorance” or “rational abstention” argument is particularly attractive

because it helps to explain the lack of attention or involvement by some people in (what others may

consider) important policy decisions. The upshot: “For most people the outcome of an election

is a public good and political participation is vulnerable to free-riding” (Mueller 1976: 411).

Another important contribution is a statement about how politicians try to control those who

implement public policy. Below, I will brieﬂy discuss one position in this literature (the “budget-

maximizing bureaucrat”), but much of the work in the public choice canon contributed to

what we know now as the “Congressional dominance ” perspective. As Shughart notes:

In that model, government bureaucracies are not free to pursue their own agendas. On the

contrary, agency policy preferences mirror those of the members of key legislative com-

mittees that oversee particular areas of public policy … that constrain bureaucratic discre-

tion by exercising their powers to conﬁrm political appointees to senior agency positions,

to mark up bureau budget requests, and to hold public hearings.

(2008)

The congressional dominance model has received serious challenge in political science (Miller

2005), but remains signiﬁcantly inﬂuential in public choice theory.

What do we do about those pesky public goods? As is apparent above, public choice theorists have

made signiﬁcant contributions to the study of public goods. The purpose of this discussion is to

describe three important basic ﬁndings that economists have discovered about public goods in

practice. Recent advances in the theory of public goods provision—theoretically, empirically,

and (especially) experimentally—are also relevant here, although space constraints limit thorough

coverage (Scotchmer 2002).

In 1956 Tiebout described a key point in public choice theory about public goods provision

(Tiebout 1956). Note that there are many diﬀerent kinds of goods and that many variations in

publicness are possible. On the one hand, markets are eﬃcient at delivering private goods; on

the other hand, pure public goods require true collective choice. Tiebout essentially showed that

individuals could select from diﬀerent levels of public goods provision by “voting with their

feet” and moving from one community to another. In this decentralized model, individuals

compare and select, and Pareto optimality is achieved “by grouping individuals together in politics

of homogenous tastes” (Mueller 1976: 412), though having a large number of public goods neces-

sitates having a large number of communities to select from. It is hard to overstate the impact of

the Tiebout paradigm (Mieszkowski and Zodrow 1989). Increasingly, we understand Tiebout

sorting as part of a broader, more complex dynamic process.

If we can exclude people from partaking in the public good, then a host of other strategies

are available for ﬁguring out what goods people want, how much they want, and how to get

them

to reveal their

demand for the goods. In the public choice canon, these situations are

treated as access to “club goods.” The basic issue here is how, as we add individuals to the club, the

average cost of providing the good changes. When the average cost falls throughout, the opti-

mal group size is the entire population. However, if at some point in adding individuals the cost

of provision starts to rise, then the club will be smaller.

The public choice perspective

Buchanan (1965) laid the groundwork for understanding situations involving club goods with

the result that a host of studies have followed trying to understand settings as diverse as social

clubs and churches. One of the early issues encountered in the canon was how homogenous the

preferences were of those who wanted to join the club. In those situations, it is easier to form

the club and decide the optimal level of membership. The key is the size of the population

compared to the optimal size of the clubs, because we could see optimal provision through a

series of clubs as people select one from a set, and perhaps leave existing ones to start new ones

with people more like themselves (Mueller 1976). The real advantage of club goods, presumably,

is that Tiebout-type solutions require some sort of geographic colocation to allow people to

“vote with their feet.”

Since the advent of the Tiebout and club goods perspectives on the public goods problem, a

whole host of studies have sought to extend and enrich our understanding of these situations in

which traditional microeconomic perspectives rooted in perfect competition seem deﬁcient.

Some have been theoretical, while others have been empirical or experimental techniques. For

the purposes of this chapter, the reader is pointed to broader treatments of the development of

this literature. However, it is important to describe the impact of one strand of this literature for

understanding public policy processes generally.

Elinor Ostrom was awarded the Nobel Prize in Economics in 2009 (jointly with Oliver

Williamson, the progenitor of transactions costs economics), largely for her studies in new insti-

tutional economics and other new ﬂavors of political economy. Her signature contributions

have come from providing a uniquely political science-inﬂuenced view of how people manage

common pool resources (CPRs) (which is one way the public goods literature has evolved). CPRs

are public goods involving natural or human-made resource systems (e.g. ﬁsheries or water

sources) for which it is either too diﬃcult or costly to exclude individuals, so club goods theory

mostly does not apply. CPRs can suﬀer from congestion because adding an additional user will

damage the current or future usage of this good by others. In her Governing the Commons: The

Evolution of Institutions for Collective Action, Ostrom identiﬁed “design principles” that societies

(perhaps governments but often autonomous collectivities of individuals) might use to manage

access to the CPR and reduce damage from congestion (Ostrom 1991). Widely known as the

Institutional Analysis and Design (IAD) framework, her approach is diﬀerent from traditional

public choice theory, but can be seen as a logical continuation of debates central to the public

choice canon. Two themes warrant emphasis. First, individuals often select the solutions; they

are not imposed from outside (sometimes referred to as “covenants without the sword”

(Ostrom et al. 1992). Second, this research stream is diﬀerent from broader public choice because it

relies heavily on thousands of case studies gleaned from CPRs around the world, along with

traditional theory and experimental research.

What should society want? As is apparent from the above discussion, it is an open question

whether public choice is mostly positive or mostly normative. It is impossible to discern this—

and silly to group normative approaches as a separate question in this chapter—mainly because

contributors to any research paradigm bring their own concerns and preferences whenever they

focus on a speciﬁc research question. Many associate the public choice paradigm with con-

servative movements and think tanks (Cato or the Liberty Fund) if only because many active

researchers are also policy commentators.

Consider public goods research. The Tiebout model is a positive statement of mobility and

goods selection by individuals in a geographic space populated by competing political jurisdic-

tions; the club goods model also makes positive statements. Yet, both have normative implica-

tions. Likewise, researchers use Ostrom’s IAD framework to understand how people provide

and manage CPRs. But Ostrom herself refers to distilled “rules” in this paradigm, especially in

Whitford

the case of irrigation systems, as “design principles.” Physics is useful for studying machines, just

as the positive tools of public choice theory help us to understand public goods. But just as the

use of machines implies engineers, the public choice paradigm has spawned normative claims

about “what society should do.” In this section, I brieﬂy review three important strands that

have helped us better to understand policy processes. Outsiders may not associate these with

public choice but the canon itself claims them as “founding debates” that have framed the ﬁeld’s

evolution. The three are the study of justice, the construction of constitutions, and the qualities

of social welfare functions.

First, the public choice canon claims Rawls’ theory of justice as an example of how theorists

could engage normative analysis from “ﬁrst principles” (Rawls 1971). Speciﬁcally, Rawls starts

with ideas present in public choice theory: rational, egoistic individuals, who are not altruistic.

He then overlays a game of chance on this participation: people ﬁnd themselves in segments of

society that determine their happiness, but individuals can imagine what their location in these

segments might have been had the game of chance turned out diﬀerently. Stepping behind the

“veil of ignorance,” individuals stand in the “original position” and choose the rules of the social

contract. For Rawls, justice is fairness.

Extensions, reﬁnements, and reassessments followed Rawls, but the main point—the game of

chance—appears in other normative theories in the public choice paradigm. As Mueller notes,

“What is important to the theory of public choice, however, is not the principles Rawls arrives

at, but the process by which he gets there” (Mueller 1976). For instance, Mueller points to

similarity between Rawls and Buchanan and Tullock’s theory of the constitution as a social

contract (Buchanan and Tullock 1962). If an individual is uncertain about who he will be in the

future, he will select a constitution that aﬀects his welfare over a long period of time by “placing

themselves in the envisaged positions of all future citizens” (Mueller 1976). People are uncertain

about future circumstances, to the end that “actual constitutions formed under unanimity rules

become just political contracts” (ibid.). Uncertainty also shows up in Harsanyi’s view of constitu-

tions as social welfare functions, because individuals assume that they might be any other person in

the society in the future (that they might have a diﬀerent person’s preferences) (Harsanyi 1953).

Justice and constitutions clearly have normative aspects, but economists have long taken a

positive approach to social choice, like how they study individual choice. Individuals choose to

do the best they can with reference to their own utility functions. For over 70 years, economists

have considered the possibility of a “social welfare function, ” but working with them seemed to

require making interpersonal comparisons of utility. In the case of a system of justice or a con-

stitution, the problem was deciding what we wanted from society: this debate also arises with

regard to a social welfare function. It is easier if people share an ethical belief. Arrow’s Impossibility

Theorem helps us to understand this better (Arrow 1963).

Given ﬁve basic attributes of a society, Arrow argued that no social welfare function always

satisﬁes

all ﬁve. The ﬁve are

values that show up regularly in the public choice literature as well

as literatures on democratic theory. The ﬁrst is that the welfare function should allow for any

possible set of individual preferences. The second is that the welfare function should be Pareto

eﬃcient. The third is that function should be consistent (transitive). The fourth is that it should

be non-dictatorial (no person should be able to overrule all other preferences). The last is that

preferences over irrelevant options should not be able to aﬀect the choice between two other

alternatives. Public choice theorists want society to have a decision rule that would never violate

these postulates.

The public choice and social choice and welfare literatures have debated which postulate to

relax to make it possible to get a social welfare function. Many theorists who have reﬁned or ela-

borated on Arrow’ s theorem think a way out of this trap is to relax the assumption of “universal

The public choice perspective

domain”: that individual preferences should not be able to take any possible ordering. But this means

people cannot be “free to choose” as they are in markets because allowing that makes it diﬃcult

for us to know what society as a whole wants. Or we could relax the independence assumption.

As Mueller notes:

Each in turn raises questions of what issues are to be decided, who is to decide, and of

those who decide, which preferences shall be weighted. Such choices directly or indirectly

involve interpersonal utility comparisons and must rest on some additional value postulates,

which if explicitly introduced would imply speciﬁc interpersonal utility comparisons. The

latter cannot be avoided.

(1976: 421)

What about those who implement public policy? One of the most persistent external perceptions of

the public choice canon is “bureaucrat bashing.” Speciﬁcally, Niskanen (1971) argued that those

who implement policy—the elected and (especially) the career bureaucrats—would use advan-

tages they hold from their positions (expertise and information) to extract the largest possible

budgets. Public choice theory fears the “budget-maximizing bureaucrat.”

In many ways, the problem was one of competition. Niskanen argues that competition would

allow the comparison of the relative prices of competing bureaus and shift power from bureaucrats

to politicians—that under competition spending on services would fall and technical eﬃciency

increase. Reﬁnements argued that Niskanen inaccurately characterized the interaction between

bureaucrats and legislators (e.g. Blais and Dion 1991). Migué, Bélanger, and Niskanen (1974)

note that bureaucrats may maximize other goals instead of the supply of public services. Conybeare

(1984) argues that Niskanen implicitly assumes perfect price discrimination (see also Bendor et al.

1985; Breton and Wintrobe 1975). Miller and Moe (1983) model the legislature, which changes

the Niskanen prediction. Empirical studies are mixed at best on the Niskanen proposal (Boyne

1998; Conybeare 1984; Higgins et al. 1987). Conybeare (1984) notes that even if multiple, com-

petitive bureaus are competing for funding, rather than producing equivalent goods (as in McGuire

et al. (1979)), there may be negative side eﬀects such as high monitoring costs. What, then, is

the value of competition? Miller and Moe (1983) show that competition for the public supply

of a good is valuable when it reveals information about actual supply costs, and thus places

monitors in better decisional positions and enhances their power. The advantage of competition

is how it reveals information by allowing comparison.

Conclusion

These six questions have helped to shape the public choice theory canon. The answers that have

been given for these questions—the positions taken, the theorems oﬀered—are arguments about

the rules of the game that underlie the operation of democratic systems around the world. Public

choice is an important strand in the diverse theories developed by social scientists to study policy

processes if only because the canon has changed how economists see the state. Rather than as a

benevolent dictator or a dolt, economists mostly see the state as a collection of actors, each

working for their own ends, sometimes working within the rules of the game and sometimes

changing those rules. At a minimum, economics is a richer discipline because of public choice.

The policy sciences and political science have also moved forward, if only because public choice

provides a foil against which to argue when building new theories of the policy processes.

It is important to reiterate that public choice is not a monolith, and there is an argument to

be made that public choice itself has been eclipsed by other rigorous and insightful views of the

Whitford

state that also have come from within economics. One is the competing “Chicago School” of

political economy, in which work by Gary Becker, George Stigler, Sam Peltzman and others

has fundamentally reshaped our understanding of what is eligible for formalization in terms of

economic theory. Mitchell notes that the diﬀerent approaches taken by the Virginia and Chicago

schools have led to qualitatively diﬀerent inferences about politics (e.g. as in the case of the

literature on regulatory politics). He notes: “Most Virginians believe that politics is rarely eﬃcient

while Chicagoans, apparently, believe that political institutions routinely achieve eﬃcient results.

The contending arguments are both powerful and subtle” (Mitchell 2001).

The second “new school” is that of “political economics,” whose leaders are economists like

Alberto Alesina, Torsten Persson, and Guido Tabellini (Blankart and Koester 2006). The main

thrust of the movement is to join important schools of thought, several of which were made

more important speciﬁcally because of the public choice movement. Political economics joins

together the work of public choice with that of Lucas’ rational expectations macroeconomics,

and the game theory that has become omnipresent in all of economics. Speciﬁcally, political eco-

nomics is diﬀerent from the naïve view of macroeconomics because it builds in roles for rational

politicians, voters, and parties.

The ﬁnal assessment of the public choice paradigm is not in its narrow application to the

policy processes but its overall contribution to the way social scientists study policy: enriching

political science, challenging the policy sciences, reshaping economics, and even giving politicians

and voters something to talk about.

Bibliography

Aronson, P.H. and P.C. Ordeshook. 1972. ‘Spatial strategies for sequential elections’, in R.G. Niemi and

H.F. Weiseberg (eds) Probability Models of Collective Decision-Making. Columbus, OH: Merrill.

Arrow, Kenneth Joseph. 1963. Social Choice and Individual Values 2nd edn. New York: Wiley.

Bendor, Jonathon, Serge Taylor and Roland van Gaalen. 1985. ‘Bureaucratic expertise versus legislative

authority: A model of deception and monitoring in budgeting’. American Political Science Review 79(4):

1041–60.

Black, Duncan. 1958. The Theory of Committees and Elections. Cambridge: Cambridge University Press.

Blais, André and Stéphane Dion. 1991. The Budget-Maximizing Bureaucrat. Pittsburgh, PA: University of

Pittsburgh Press.

Blankart, C.B. and G.B. Koester. 2006. ‘Political economics versus public choice’, Kyklos 59(2): 171–200.

Boyne, George A. 1998. ‘Competitive tendering in local government: A review of theory and evidence,

Public Administration 76(4): 695–712.

Breton, Albert. 1974. The Economic Theory of Representative Government. Chicago: Aldine Publishing.

——and R. Wintrobe. 1975. ‘The equilibrium size of a budget-maximizing bureau: A note on Niskanen’s

theory of bureaucracy’, The Journal of Political Economy 83(1): 195–207.

Buchanan, James M. 1965. ‘An economic theory of clubs, Economica 32(125): 1–14.

——, and Gordon Tullock. 1962. The Calculus of Consent: Logical Foundations of Constitutional Democracy.

Ann Arbor, MI: University of Michigan Press.

Conybeare, John A.C. 1984. ‘Bureaucracy, monopoly and competition: A critical analysis of the budget-

maximizing model of bureaucracy’, American Journal of Political Science 28(3): 479–502.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper.

Harsanyi, John C. 1953. ‘Cardinal utility in welfare economics and in the theory of risk-taking’, Journal of

Political Economy 61(5): 434–5.

Higgins, Richard S., William F. Shughart II and Robert D. Tollison. 1987. ‘Dual enforcement of the

antitrust laws’, in Robert J. Mackay, James Cliﬀord

Miller and Bruce

Yandle (eds) Public Choice and

Regulation: A View from Inside the Federal Trade Commission. Stanford, CA: Hoover Institution Press.

Hotelling, Harold. 1929. ‘Stability in competition’, The Economic Journal 39(153): 41–57.

Lindahl, Erik. 1919. ‘Just taxation – a positive solution’, Classics in the Theory of Public Finance (134): 168–76.

McGuire, Thomas, Michael Coiner and Larry Spancake. 1979. ‘Budget-maximizing agencies and eﬃciency

in government’, Public Choice 34(3): 333–57.

The public choice perspective

McKelvey, Richard D. 1976. ‘Intransitivities in multidimensional voting models and some implications for

agenda control’, Journal of Economic Theory 12(3): 472–82.

Majone, Giandomenico. 1989. Evidence, Argument and Persuasion in the Policy Process. New Haven, CT:

Yale University Press.

Mieszkowski, P. and G.R. Zodrow. 1989. ‘Taxation and the Tiebout model: The diﬀerential eﬀects of

head taxes, taxes on land rents and property taxes’, Journal of Economic Literature 27(3): 1098–146.

Migué, Jean-Luc, Gérard Bélanger and William A. Niskanen. 1974. ‘Toward a general theory of managerial

discretion’, Public Choice 17(1): 27–47.

Miller, Gary J. 2005. ‘The political evolution of principal-agent models’, Annual Review of Political Science 8:

203–25.

——and Terry M. Moe. 1983. ‘Bureaucrats, legislators, and the size of government’, American Political

Science Review 77(2): 297–322.

Milleron, J.C. 1972. ‘Theory of value with public goods: a survey article, Journal of Economic Theory 5(3):

419–77.

Mitchell, William C. 2001. ‘The old and new public choice: Chicago versus Virginia’, in William F.

Shughart II and Laura Razzolini (eds) The Elgar Companion to Public Choice, Northampton, MA: Edward

Elgar, 3–32.

Mueller, Dennis C. 1976. ‘Public choice: A survey’, Journal of Economic Literature 14(2): 395–433.

—— 2003. Public Choice III. Cambridge: Cambridge University Press.

Niskanen, William A. 1971. Bureaucracy and Representative Government. Chicago: Aldine.

Olson, Mancur. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA:

Harvard University Press.

Ostrom, Elinor. 1991. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge:

Cambridge University Press.

Ostrom,

E., J. Walker

and R. Gardner. 1992. ‘Covenants with and without a sword: Self-governance is

possible’, American Political Science Review 86(2): 404–17.

Rae, Douglas W. 1969. ‘Decision-rules and individual values in constitutional choice’, The American Political

Science Review 63(1): 40–56.

Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Belknap Press of Harvard University Press.

Riker, William H. 1962. The Theory of Political Coalitions. New Haven, CT: Yale University Press.

Rowley, Charles Kershaw and Friedrich Schneider. 2004. The Encyclopedia of Public Choice, 2 vols. Boston:

Kluwer Academic Publishers.

Saari, Donald G. 2006. ‘Which is better: The Condorcet or Borda winner?’, Social Choice and Welfare 26(1):

107–29.

Satterthwaite, Mark A. 1975. ‘Strategy-proofness and Arrow ’s conditions: Existence and correspondence

theorems for voting procedures and social welfare functions’, Journal of Economic Theory 10(2): 187–217.

Schoﬁeld, Norman. 1978. ‘Instability of simple dynamic games’, The Review of Economic Studies 45(3): 575–94.

Scotchmer, S. 2002. ‘Local public goods and clubs’, Handbook of Public Economics 4: 1997–2042.

Shughart, William F. 2008. ‘Public Choice’, in David R. Henderson (ed.) The Concise Encyclopedia of

Economics. Indianapolis, IN: Liberty Fund, 427–30.

——and Laura Razzolini. 2001. The Elgar Companion to Public Choice. Northampton, MA: Edward Elgar.

Taylor, M. 1969. ‘Critique and comment: Proof of a theorem on majority rule’, Behavioral Science 14(3):

228–31.

Tiebout, C.M. 1956. ‘A pure theory of local expenditures’, The

Journal of Political Economy: 416–24.

Tovey, Craig

A. 2010. ‘The almost surely shrinking yolk’, Mathematical Social Sciences 59(1): 74–87.

Tullock, Gordon. 1967. Toward a Mathematics of Politics. Ann Arbor, MI: University of Michigan Press.

Whitford

Institutional analysis and

political economy

Michael D. McGinnis and Paul Dragos Aligica

Political economy

and the analysis of non-market decision settings

The foundational idea of the Ostroms’ institutionalism is simple but powerful in its implications:

analysis of the public sector should be no diﬀerent in its basic assumptions from studies of the

market-oriented private sector. The focus of analysis is in both cases the individuals and their

actions in various institutional arrangements. Methods used to understand complex economic

industries involving large and small proﬁt-seeking enterprises competing within more or less

eﬃcient and competitive markets, should be expected to be equally functional for the study of

equally complex public structures. The large number of public enterprises operating in a modern

society (at the national level or at local, metropolitan levels) could be studied using tools transfer-

able from one domain to another. Although the standard framework needs special adjustments

in order to get adapted to a phenomenon that is in the end diﬀerent from the market-based phe-

nomena, the analysis of public governance issues requires an approach that is fundamentally similar

to the analysis of private industries (Ostrom and Ostrom 1965, 1977; E. Ostrom 1972, 1983, 1996).

The key is that public enterprises should be seen as organizations that produce goods and

services for a variety of consumers and groups of consumers. Public goods and services have

diﬀerent characteristics and require a variety of arrangements and processes to be delivered. As

in the case of private enterprises, these features determine the nature of the institutions, strate-

gies and policies needed. The functional problems of production, acquisition, distribution, and

consumption of public and common goods lead to the emergence of complex institutional

arrangements.

Once one starts to look at the public sector through the lens of competitive public economies,

the reality will never be the same again. For instance, the observation that most private enterprises

purchase many of the goods and services that they need from other enterprises, draws attention

to the notion that production (physical rendering) of public services may be considered sepa-

rately from the provision of such services (i.e. the decision how to produce goods or services, or

decisions concerning the quantity and quality of these products) and there remains the separate

question of deciding on billing and other ﬁnancial details). Issues of scale and eﬃciency,

concerned mainly with production, can be analytically considered separately from questions of

what public goods and services should be made available to members of the group that will

collectively consume them, and from questions about how producers will be paid and their

products evaluated (Ostrom and Ostrom 1965; E. Ostrom et al. 1994).

Distinctions among production, provision, ﬁnancing, and evaluation were laid out quite

explicitly in a classic article on the organization of metropolitan governance in the United States

(Ostrom et al. 1961). Urban public economies are micro-universes that display, on a small scale,

governance problems and processes that are ubiquitous at larger scales. This perspective con-

siders local government oﬃcials as service providers responsible “for making taxing and spending

decisions, determining appropriate types of service and levels of supply, and arranging for and

monitoring production.” As providers, these public oﬃcials face several options in arranging for

the actual production of public goods, including creating its own production unit, contracting

for services with an external producer or creating a joint venture with other provision units in

order to have the services produced.

To sum up, the implication is that the organization of provision and the organization of

production have diﬀerent principles of operation and should be institutionalized in diﬀerent ways.

A local government may, on the one hand, be organizing service ﬁnancing through collective

consumption units and, on the other hand, be purchasing services from alternative public or private

suppliers. Institutional diversity is thus the natural pattern in the public sector, while centralization

and uniformity may be more of an exception.

The political economy institutionalism approach has thus more than just a methodological

value. One of its important consequences is that it introduces the notion of optimum scale of

performance. A crucial question in public administration and public policy is addressed frontally:

is there “some special characteristic of governmental functions that makes large units necessary

to eﬃciency?” In the light of the above, one cannot avoid the answer that eﬃciency depends

on the nature and type of service that a governmental agency is meant to produce and not on

some special characteristics of governmental functions associated to the scale of the activity. The

question of the validity of the notion that eﬃciency is a function of large-scale centralization is

thus confronted. The thesis advanced is that the optimum scale of production is not the same

for all public goods and services and that some services may be produced “more eﬃciently on a

large scale while other services may be produced more eﬃciently on a small scale (E. Ostrom

[1972] in McGinnis 1999b; Oakerson 1999; V. Ostrom et al. 1988). Implications questioning

the basic postulates of centralized governance in the modern public system are even today

perceived by many as counterintuitive and paradoxical.

It is clear why political economy institutionalism suggests that the existence of multiple

agencies interacting and overlapping should not be considered a priori and by deﬁnition dys-

functional. This overlapping and duplication is the result of the fact that diﬀerent services

require a diﬀerent scale for eﬃcient provision and that the principles of division of labor,

cooperation and exchange function in the public sector, too. In traditional bureaucratic theory,

duplication of functions is assumed to be wasteful and ine ﬃcient. But in a market economy

multiple ﬁrms serve the same market. Overlapping service areas and duplicate facilities do not

lead to ineﬃciency. Can we expect similar forces to operate in a public economy? The Ostroms’

answer is positive. The public sector includes not only “bureaucratic command structures con-

trolled by chief executives” but also “interorganizational arrangements” with self-regulating

tendencies with eﬃciency-inducing and error-correcting behavior (Ostrom and Ostrom 1965: 3).

The discussion leads sooner or later to the very notion of competition. In what measure and

how does the public policy process incorporate the competition process? There is no pre-

sumption that competition among public agencies is necessarily eﬃcient

or ineﬃcient in all

settings,

but only that such competiton need not take the same forms as market competition

(V. Ostrom et al. 1961; Bish 1971).

McGinnis and Aligica

The bottom line is that the presence of more than a single producer of public goods within a

certain area enables citizens to make more eﬀective choices about the mix of services they prefer

to receive while enabling citizens to “vote with their feet” or by enabling individuals residing

within an existing community to choose among alternative producers of public services. Fur-

thermore, multiple producers of some public goods may nullify each other’s actions and lead to

a reduction in the net output of public goods. In other words, political economy institutional-

ism considers the practical eﬀect of competition among public agencies as an empirical question

depending upon the type of urban public good being considered and the speciﬁc circumstances

of the case (E. Ostrom [1972] in McGinnis 1999b: 148)

Toonen (1998) argues that the 1961 Ostrom, Tiebout and Warren article synthesized all of

the following components of New Public Management and related visions of network gov-

ernance: entrepreneurial leadership, contracts for service delivery, deregulation, devolution of

authority to market actors, the important contributions made by professional associations and

community groups, and an emphasis on evaluating the performance of public oﬃcials and the

quality of the products they produce or procure. Each of these topics has attracted substantial

interest in recent years, and, to be fair, the authors did not systematically lay out any of these

implications in detail. In eﬀect, these same ideas were later rediscovered and elaborated upon by

other scholars, who were in eﬀect following a path initially laid out by these founders of the

Bloomington School of Institutional Analysis.

To sum up, the Bloomington School not only oﬀers a speciﬁc set of theoretical lenses and a

new perception of the public and policy arena but also questions the assumption deeply ingrained

in the traditional view that large bureaucracies are more eﬃcient in solving problems and in

providing public goods and services than the systems based on competition or bargaining. In

doing that, this approach challenges our perspective on the very nature of the policy process.

Complexity, scale and scope, levels and diversity, overlapping institutions and functional dif-

ferentiation invite us to rethink the public governance system. Sooner or later we are forced to

reconsider the criteria we are using in policy evaluation in all its aspects: policy theory, policy

implementation, and policy impact.

Polycentricity and monocentrism

Political economy institutionalism is an eﬀort to change the paradigm. A new domain of insti-

tutional complexity and diversity needed to be charted and that requires an entirely new fra-

mework. The concept of polycentricity lies at the heart of this new paradigm. As initially

deﬁned, the term “polycentric” connotes many centers of decision-making which are formally

independent of each other.

To the extent that they take each other into account in competitive relationships, enter into

various contractual and cooperative undertakings or have recourse to central mechanisms to

resolve conﬂicts, the various political jurisdictions in a metropolitan area may function in a

coherent manner with consistent and predictable patterns of interacting behavior.

(Ostrom et al. 1961: 831)

Subsequently, Ostrom further clariﬁed his vision of polycentricity as constituting a “highly

federalized system” that required all of the following component parts: “a rich structure of over-

lapping jurisdictions with substantial autonomy among jurisdictions, substantial degrees of demo-

cratic control within jurisdictions, and subject to an enforceable system of constitutional law”

(V. Ostrom 1973: 229).

Institutional analysis and political economy

The alternative, and still much more inﬂuential paradigm, is to presume that eﬀective gov-

ernance requires a single ultimate authority exercising control through a uniﬁed command and

control structure. Conversely, in a polycentric order, the elements of a complex system are

allowed to make mutual adjustments to each other “within a general system of rules where each

element acts with independence of other elements” (V. Ostrom 1972). The conceptual spaces

deﬁned by the two notions are interlinked. “The possibility that a polycentric political system

can exist does not preclude the possibility that a monocentric political system can exist” (ibid.).

Above and beyond the theoretical problems and debates, is a practical issue. The key ques-

tion is whether there is any prima facie ground for expecting less eﬃcient performance from

polycentric arrangements than from a centralized system. As we have seen, before the work of

the Ostroms and their associates started to have an impact, the duplication of services (one of

the main features of polycentric systems) was assumed to be wasteful and to generate disorder.

To compare these two paradigms, an extensive empirical research agenda focused on governance

in municipal areas was initiated (Oakerson 1999; McGinnis 1999b). That empirical work “started

to identify and chart the patterns of order looming underneath the apparent chaos intrinsically

associated to the experience of polycentricity” as well as speciﬁed conditions for eﬃcient per-

formance (V. Ostrom 1972). Polycentricity raises fundamental challenges to social and political

theory that have broader ramiﬁcations that go beyond the issue of the governance of metropolitan

areas. An analysis of the class of phenomena displaying a multiplicity of decision centers shows

that there are many forms of organization that might seem analogous to a polycentric order.

However, not all of them had the attributes associated to polycentricity. For instance, a break-

down of order and rules may lead to situations where multiple decision centers coexist. Political

corruption leads to situations where various forms of bosses divide “territories” and “jurisdictions.”

Could that be called polycentric order? Therefore, it is important to specify the conditions that

could create a system of government organized in a polycentric manner. In this respect, ﬁve fea-

tures are pivotal: the legitimate exercise of coercive capabilities; an overarching system of rules;

freedom to enter and exit; freedom to orderly change the rules; and incentives alignment.

A mere review of polycentricity reveals that the problem of whether the government of a

political system can be organized in a polycentric manner has a considerable history. As V. Ostrom

(1971, 1997) argued, Alexis de Tocqueville’s classic study Democracy in America is an exploration

in polycentric governance. The challenge faced by the founding fathers of the American con-

stitution was in fact to design a polycentric order. The Federalist did not use the term poly-

centricity. However, Alexander Hamilton and James Madison’ s conception of the “principles of

federalism and separation of powers within a system of limited constitutions meets the deﬁning

conditions for polycentricity” (V. Ostrom 1972, 1973, 1971). The federal system, argues V. Ostrom,

assumes a fragmentation of authority in many centers of decision-making that implies a separa-

tion of powers. In other words, designing the American constitution could be viewed as an

experiment in polycentricity. Federalism itself could be seen as one way to capture the meaning

and operationalize one aspect of this type of order.

Last but not least, it is important to highlight the normative dimensions of polycentric forms

of organization. Polycentricity seems to be a necessary condition for political objectives such as

“liberty” and “justice.” The dispersion of decision-making capabilities and the checks and bal-

ances deﬁning polycentricity allows for “substantial discretion or freedom to individuals and for

eﬀective and regular constraint upon the actions of governmental oﬃcials.” As such, is an essential

characteristic of democratic societies. Last but not least, a polycentric arrangement has a built-in

system of self-correction. While opportunistic behavior and malfeasance are a problem for any

institutional arrangements, “a political system that has multiple centers of power at diﬀering

scales provides more opportunity for citizens and their oﬃcials to innovate and to intervene so

McGinnis and Aligica

as to correct maldistributions of authority and outcomes.” Thus, “polycentric systems are more likely

than monocentric systems to provide incentives leading to self-organized, self-corrective institutional

change” (E. Ostrom 1998b). To conclude, one of the strong theses put forward by students of

polycentrism is that the very nature and existence of democratic societies depends on the presence

of sizable elements of polycentricity in their governance systems (Aligica and Boettke, 2009: 22–23).

Coproduction: consumer producers

The perspective opened up by the political economy and polycentricity theoretical lenses has

led to a series of important empirical and theoretical insights regarding the problem of public

policies and governance in the public sector. To illustrate this point, let us use as an example the

concept of coproduction.

Empirical research of Bloomington scholars identiﬁed a series of cases in which the collaboration

between those who supplied a service and those who used it was the key factor in determining

the eﬀective delivery of the service. For instance, the health of a community depends not only

on the professional quality of health care personnel but also on the informed eﬀorts of individuals,

while the quality of an “educational product” is largely determined by the eﬀorts of the “users

of educational services,” namely the eﬀorts students make. Similarly, the quality and supply of

neighborhood security depends on the joint eﬀorts of citizens as well of the professional police

oﬃcers. These are cases in which the users of services also function as coproducers, in the sense that

the production process cannot be totally separated from consumption. Without the informed and

motivated eﬀorts of service users, “the service may deteriorate into an indiﬀerent product with

insigniﬁcant value.” Under co-production, “the resources, motivations, and skills brought to bear

by the client or consumer are much more intimately connected with the level of achieved output

than in the case of goods production.” Coproduction means more than the existence of at least two

producers” (Ostrom and Ostrom [1977] in McGinnis 1999b: 93; Aligica and Boettke, 2009: 32–33).

This phenomenon is far from marginal. It appears to be characteristic of much public service

production (Parks 1993; Parks and Oakerson 1989). Appreciation of the role of coproduction

in public service delivery requires a change of focus in our approach to public administration and

public policy. Instead of focusing exclusively on bureaucratic structures or public organizations, it

highlights the critical importance of consumer input. Many public policy puzzles may be solved

if the productive roles of consumers as coproducers of the services they receive is taken into

account (Kiser and Percy 1980; E. Ostrom 1996; Ostrom and Ostrom 1977; Whitaker 1980).

Once clearly deﬁned, coproduction problems could be identiﬁed in many areas of the service

in public sectors. A typical example is the so-called “service paradox.”

When professional personnel presume to know what is good for people rather than providing

people with opportunities to express their own preferences, we should not be surprised to

ﬁnd that increasing professionalization of public services is accompanied by a serious erosion

in the quality of those services.

(Aligica and Boettke 2009: 33)

The better services are, as deﬁned by professional criteria, and as supplied exclusively by “highly

trained professionals, the less satisﬁed citizens are with those services.” Satisfaction and eﬀec-

tiveness is in the end a function of citizens “functioning as essential co-producers” (Ostrom and

Ostrom [1977] in McGinnis 1999b: 93–4).

While usually the attention in public services management and evaluation is primarily on the

production side, the strategies of consumption are as important as the strategies of production.

Institutional analysis and political economy

Coproduction requires that “both production and consumption go hand in hand to yield

optimal results.” One should try to take into account both the “economies of consumption as

well as of production and provides for the co-ordination of the two” (Ostrom and Ostrom

[1977] in McGinnis 1999b: 94).

Political economy institutionalism from the Bloomington School demonstrates that despite

the absence of the level of ﬂexibility found in fully competitive markets, the search for solutions

in the public arena is not blind. A notion like coproduction could be used as a diagnostic

tool and with it an entire battery of concepts and theories could be developed for institu-

tional design, policy analysis and assessment. The particular nature of each good and service

plus the technological and economic conditions of its production and consumption require a

complex institutional system in which public and private, markets and hierarchies combine to

generate an institutional architecture with diverse patterns of organization and diversely orga-

nized systems of governance. To cope with this challenge new methodological tools needed to

be developed.

The IAD framework

The Institutional Analysis and Development (IAD) Framework encapsulates the collective eﬀorts

of the Bloomington School to understand the ways in which institutions operate and change

over time. The IAD framework assigns all relevant explanatory factors and variables to categories

and locates these categories within a foundational structure of logical relationships.

The IAD framework serves as a tool to simplify the analytical task confronting anyone trying

to understand policies and institutions in their full complexity. At the heart of the IAD frame-

work is the action situation, in which individuals (acting on their own or as agents of formal

organizations) observe information, select actions, engage in patterns of interaction, and evaluate

outcomes. As is typical in strategic interactions, potential outcomes are diﬀerentially valued by

those actors with partial control over the determination of actual results. The IAD framework

highlights the social-cultural, institutional, and biophysical context within which strategic deci-

sions are made. Speciﬁcally, the IAD framework helps to organize the task confronting a scholar

or policy analyst approaching a policy issue by directing their attention to ﬁrst, the rules-in-use,

rather than the rules on paper, second, the underlying biophysical nature of the good under con-

sideration, in terms of it being a private, public, or toll/club good or a common-pool resource

(CPR), as well as third, the most relevant attributes of the community, especially ambient levels

of trust and shared norms of reciprocity.

Systems typically look very diﬀerent depending on the level of aggregation being used, and

the IAD framework explicitly distinguishes three levels of analysis in which diﬀerent types of

choice processes take place: ﬁrst, operational level choices (regarding policy implementation);

second, collective level choices involving the determination of which strategies, norms and rules

are or should be available to actors fulﬁlling the speciﬁc roles deﬁned by that group (as well as

specifying who is assigned to ﬁll these roles); and third, constitutional level choices relating to

who is or should be empowered to participate in the making of collective and operational-level

decisions. The critical insight behind this framework is that the outcomes of interactions in

these diﬀerent arenas of choice are explicitly connected to each other.

The actors in any action situation are presumed to be boundedly rational. They seek to

achieve goals for themselves and for the communities with which they identify but do so within

the context of ubiquitous social dilemmas and biophysical constraints, as well as cognitive limitations

and cultural predispositions. Within this broad framework, a range of theoretical perspectives

may be employed to develop and analyze models of speciﬁc situations.

McGinnis and Aligica

Rational choice theory, grounded in methodological individualism, provides the general

inspiration for the IAD framework, but diﬀerent decision models are relevant for diﬀerent

situations. For example, the same individual might follow Simon’s (1955) satisﬁcing procedure

in one decision context while engaging in more extensive information search and evaluation in

other situations. There is no reason to presume that any one individual acts in exactly the same

way in all circumstances; still, there is merit in trying to locate the relevant range of decisional

procedures within the context of a common explanation.

In the IAD framework analysts must incorporate, in some manner, the actors’ self-understanding

of their roles and their conceptions of proper or acceptable behavior in particular contexts. By

incorporating factors not typically considered by rational choice theorists, the IAD framework

can encompass approaches to analysis that many of them would no longer recognize as rational

choice. Indeed, Ostrom (1998a) has called for a “second-generation” of rational choice theory

more ﬁrmly grounded in behavioral regularities, many of which have been discovered by critics

of the rational choice tradition.

The IAD framework is implicitly grounded in a dynamic view of policy processes. Social,

institutional, and biophysical factors are inputs to processes of individual decisions (with those

decisions presumed to be inﬂuenced by their preexisting cognitive capabilities and cultural

presuppositions), and these decisions are then aggregated to constitute policy outputs that would

then interact with exogenous factors to produce observable outcomes, with actors’ evaluations

of these outcomes by these actors (or by other observers) feeding back into all of the previous

components of this never-ending process.

Institutional arrangements shape and constrain the behavior of individuals, but there is no

need to presume that individuals will always follow the rules. In the IAD framework consider-

able importance is attached to the means by which actors (at all levels) monitor each other’s

activities and sanction undesirable or inappropriate behavior. In empirical research it often turns

out that those systems which involve local participants in monitoring and sanctioning are more

likely to be sustainable than those in which those functions are instead fulﬁlled by agents of the

national government (E. Ostrom 1990).

The IAD framework bears a family relationship to game models (E. Ostrom 1986), but it is

essential to keep in mind the extent to which actors’ preferences as well as the choice options

available to them are determined by the institutional arrangements that deﬁne their position or that

shape their perceptions and options. Concurrent games in other arenas of choice interact in subtle

ways with any ongoing process of interaction (McGinnis 2011b). The payoﬀs and menu of

choices available to participants in operational games have been deﬁned by collective choice

processes. Games over collective deliberations are in turn shaped by the positions and interests

deﬁned or manifested in the constitutional choice arena. In a strict sense, this may be no diﬀerent

from a complete speciﬁcation of a game model, but in practice this concern with simultaneous

consideration of multiple choice arenas inspires Workshop-aﬃliated scholars toward a more

inductive mode of analysis (Mitchell 1988). Yet all this research remains inspired by a common

vision of the critical importance of polycentricity in the policy world, and how critical it is for

analysts to realize the complexities inherent in the systems they are studying.

Conclusions: public policy analysis

Irrespective of school of thought, knowledge base, or analytical sophistication, most forms of policy

analysis are by their very nature dealing with a single fundamental problem: collective action.

Slowly,

an interdisciplin ary theory uniting all the many factors that inﬂuence collective action is emerging, and

members of the Bloomington School are playing critical roles in its emergence (E. Ostrom 2010a).

Institutional analysis and political economy

Adeﬁning feature of the political economy institutionalism advanced by the Ostroms and

their associates is that this general theory of collective action requires realization through very

speciﬁc and problem-oriented analyses. It refers to concrete, public aﬀairs problems confronting

speciﬁc communities or groups. Its objectives are not abstract and theoretical but applied and

aimed at problem solving (V. Ostrom 1991: 21). Policy analysis “is variable with problematical

situations” and “any diagnostic assessment must necessarily be conjectural.” Without knowledge

well grounded in the problematic circumstances of a collective action “policy analyses are likely to be

misspeciﬁed.” That may seem commonsensical. However a common source of policy failure lies

in eﬀorts to formulate and implement general solutions in total ignorance of the local conditions.

A diagnostic assessment “is complexly linked to the institutional facts,” that is to say knowledge

of the institutional environment is crucial (V. Ostrom 1991: 25–6).

We have examined a few ways in which the Bloomington School has analyzed links between

institutional diversity and the complexity of polycentric systems of governance. Conversely,

policy analysts committed to the monocentric alternative often reveal a surprising lack of

interest in real-world institutions. In other words, the way policy analysis was thought about for

most of the twentiethth century resulted in misguided eﬀort to implement universal solutions,

typically centered in national level public oﬃcials (V. Ostrom 1991: 2).

This logic and its implications are not accidental but a basic paradigmatic problem. As such, it is

prone to persist as long as policy analysts continue to conceptualize social order using just two

ideal types—markets and hierarchies—and believe that all societies could be described by reference

to them, thus ignoring the “diverse nestings of institutional arrangements in time and place

speciﬁcities.” But once the theoretical lenses of polycentricity are used, policy analysis is forced

to give more attention to institutional analysis driven by analytical methods and to problem-solving

using “variable policy formulations” based upon “diﬀerent presuppositions” (V. Ostrom 1997: 104).

In his writing on public policy Vincent Ostrom tried to rescue the study of public policy

from the trap of intellectual triviality. Policy is not a purely technical exercise of adjusting means

to given ends ﬁxed in function of “obvious” problems. By their very nature, social scientists are

deeply involved in clarifying the meaning and place of values and norms in the development of

human institutions. An understanding of values, or a normative inquiry, is a prerequisite to

performing the tasks of policy design and evaluation. Moreover, if one wants to design or assess

decision rules, one needs criteria. Then the question is: what kind of criteria for those rules?

How we construct (or for that matter assess) those rules depends on the criteria used. The task

of the policy analyst is not only to use values as entry points or vehicles for analysis but to apply

them. These values are not “given.” In a word, the very notion of a “policy problem” is infused

by deeper philosophical or normative speculations related to human nature and the human

condition. The full dimensions and the complexity of the problems are fully revealed when nor-

mative concepts “collide with natural conditions of human beings” (V. Ostrom 1991: 25). The

world of policy analysis is in the end the world of implementation—namely the world of action

and decision, and therefore is the world of normative commitments.

Notes

1 As might be expected for scholars whose work extends over several decades, the terms the Ostroms

used to deﬁne their approach to the study of political institutions has undergone several changes, and

these changes can generate some confusion. For example, in Ostrom and Ostrom (1971) they use the

term “public choice” to refer to their proposed new mode of analysis, but later developments in public

choice gave a much diﬀerent meaning to that term. At the time, they were key participants in what

Vincent Ostrom famously labeled the “no-name” conference at Charlottesville, Virginia, organized by

James Buchanan in 1964 (V. Ostrom 1965). They have also used “public economy” as a generalization

McGinnis and Aligica

of market economy, intended to incorporate all relevant public, private, voluntary, and community-

based organizations active in a given area of public policy, and “public service industry” or “public

enterprise” to generalize the concept of a market sector to include non-explicitly commercial organi-

zations. The term “behavioral approach” (Ostrom and Ostrom 1965; E. Ostrom 1998a,b) has also been

used in diﬀerent contexts. Ultimately, however, they settled on the phrase “institutional analysis” as the

most accurate label for their approach, but its lack of speciﬁcity makes that also a potentially confusing

term. For this reason, and others, we will use the term Bloomington School as a shorthand label for

the approach to research pioneered by Vincent and Elinor Ostrom.

2 That is to say it was doing more than merely developing a simple market model, derived from neoclassical

economic theory, and applying it to a non-market setting in the public choice tradition. Its goals are much

bolder: “We thought an indication that quasi-market mechanisms were operable in a public service economy

would imply important new dimensions for a theory of public administration” (V. Ostrom 1972). In other

words, they were restructuring scholars’ basic conceptualization of the public sector (Ostrom et al. 1961).

3 In a monocentric political system the prerogatives for determining and enforcing the rules are “vested in a

single decision structure that has an ultimate monopoly over the legitimate exercise of coercive cap-

abilities.” In a polycentric system “many oﬃcials and decision structures are assigned limited and relatively

autonomous prerogatives to determine, enforce and alter legal relationships” (V. Ostrom [1972] in McGinnis

1999b: 55–60). In a polycentric system no one has an ultimate monopoly over the legitimate use of

force. The “rulers” are constrained and limited under a “rule of law. ” Ultimately, polycentric systems

are rule of law systems. Incentives alignment in the enforcement of general rules of conduct provides

the legal framework for a polycentric order: “If individuals or units operating in a polycentric order

have incentives to take actions to enforce general rules of conduct, then polycentricity will become an

increasingly viable form of organization” (V. Ostrom 1972). Another condition is freedom of entry and

exit in a particular system. The establishment of new decision centers under the existing rules should

not be blocked. The freedom of entry ensures the evolution and adaptation of the system. Finally,

another key condition is that spontaneity should be manifested in the reformulation and revision of the

basic rules that deﬁne the framework of a speciﬁc polycentric order. The idea is that individuals should

be free not only to play the game or have the incentives to self-enforce the rules of the game but also

to change those rules in an orderly way. Understanding and learning from experience are in fact the

vectors of an ongoing process of knowledge integration in the institutional system and the prerequisites

of subsequent adaptations to the changing environment. (V. Ostrom [1972] in McGinnis 1999b).

4 The speciﬁc form of this framework has varied considerably over time; important stages in its devel-

opment can be seen at Kiser and Ostrom (1982), E. Ostrom (1990), Ostrom et al. (1994), McGinnis

(2000), E. Ostrom (1998a, 2005, 2007b, 2010b) and Poteete et al. (2010). The IAD framework is also

related to earlier work on public service industries and local public economies (Ostrom et al. 1961;

Ostrom and Ostrom 1977; E. Ostrom 1983; Ostrom et al. 1993; McGinnis 1999b; Oakerson 1999) as

well as more recent work on social-ecological systems (E. Ostrom 2007a, 2009). For a recent overview

of all of these related frameworks, see McGinnis (2011a).

5 “Policy

analysis,” writes Vincent Ostrom,

is based on the presumption that “appropriate forms of collec-

tive choice and collective action will contribute to a more eﬀective resolution of common problems

and an advancement in the aggregate well-being experienced by members of collectivities” (V. Ostrom

in McGinnis 1999a: 394).

6 Due to the very speciﬁc and contextual nature of knowledge required for the assessment of a case, “in

ﬁnite time and place exigencies,” the need to get the local knowledge, and perceptions and ideas of

the actors involved is even more stringent. One has to derive criteria for choosing one alternative over

the other and to assess their consequences. In other words, the applied dimension presses the problem

of norms not only at an analytical level but also at the decision-making one.

7 Much of contemporary policy analysis appears to be largely institution-free except for abstract allusions

to the government, or the state (meaning nation-states).

Policy formulations presume a clean slate as though the failure to proclaim a General National

Policy implied the absence of an Energy Policy, an Urban Policy, a Housing Policy, an Edu-

cational Policy, or any other speciﬁable policy. It is as though all political systems were uni-

centric, hierarchical in order, and run from the top down. The same modes of analysis and

policy prescriptions can presumably be used anywhere in the world.

(Ostrom, V. The Meaning of American Federalism: Constituting a Self-governing

Society. ICS Press, 1991: 2)

Institutional analysis and political economy

Bibliography

Aligica, P. and P. Boettke. 2009. Rethinking Institutional Analysis and Development: The Bloomington School.

London: Routledge.

Bish, R. 1971. The Public Economy of Metropolitan Areas. Chicago: Markham Press.

Kiser, L.L. and S. Percy. 1980. ‘The concept of coproduction and its prospects for public service delivery,’

Working Paper No. W80-6, Workshop in Political Theory and Policy Analysis, Bloomington, IN:

Indiana University.

McGinnis, M.D. 1999a. Polycentric Governance and Development: Readings from the Workshop in Political Theory

and Policy Analysis. Ann Arbor, MI: University of Michigan Press.

—— 1999b. Polycentricity and Local Public Economies: Readings from the Workshop in Political Theory and Policy

Analysis. Ann Arbor, MI: University of Michigan Press.

—— 2000. Polycentric Games and Institutions: Readings from the Workshop in Political Theory and Policy Analysis.

Ann Arbor, MI: University of Michigan Press.

—— (2011a). ‘An Introduction to IAD and the Language of the Ostrom Workshop: A Simple Guide to a

Complex Framework’, Policy Studies Journal 39 (1) (March 2011), 163–77.

—— 2011b. ‘Networks of adjacent action situations in polycentric governance’, Policy Studies Journal 39(1):

45–72.

Kiser, L. and E. Ostrom. ‘The three worlds of action, ’ in Strategies of Political Inquiry, ed. E. Ostrom,

Beverly Hills, CA: Sage, 179–222.

Mitchell, W.C. 1988. ‘Virginia, Rochester, and Bloomington: Twenty-ﬁve years of public choice and

political science’, Public Choice 56(2): 101–19.

Oakerson, R.J. 1999. Governing Local Public Economies: Creating the Civic Metropolis. San Francisco: ICS Press.

——and R.B. Parks. 1988. ‘Citizen voice and public entrepreneurship: The organizational dynamic of a

complex metropolitan county’, Publius: The Journal of Federalism 18(4): 91–112.

Ostrom, E. 1972. ‘Metropolitan reform: Propositions derived from two traditions’, Social Science Quarterly

53: 474–93.

—— 1983. ‘A public choice approach to metropolitan institutions: Structure, incentives, and performance’,

Social Science Journal 20(3): 79–96.

—— 1986. ‘An

agenda for the

study of institutions’, Public Choice 48: 3–25

—— 1990. Governing the Commons: The Evolution of Institutions for Collective Action. New York: Cambridge

University Press.

—— 1996, ‘Crossing the great divide: Coproduction, synergy, and development’, World Development 24

(6): 1073–87.

—— 1998a. ‘A behavioral approach to the rational choice theory of collective action’, American Political

Science Review 92(1): 1–22.

—— 1998b. ‘The comparative study of public economies’, presented upon acceptance of the Frank

E. Seidman Distinguished Award in Political Economy, Memphis, TN, P.K. Seidman Foundation.

—— 2005. Understanding Institutional Diversity. Princeton, NJ: Princeton University Pres

—— 2007a. ‘A diagnostic approach for going beyond panaceas’, Proceedings of the National Academy of Sciences

104(39): 15181–7.

—— 2007b. ‘Institutional rational choice: an assessment of the institutional analysis and development

framework’ in P.A. Sabatier (ed.) Theories of the Policy Process, 2nd edn. Boulder, CO: Westview Press,

21–64.

—— 2009. ‘A general framework for analyzing sustainability of social-ecological systems’, Science, vol. 325,

no. 5939: 419–22.

—— 2010a. ‘Analyzing collective action’, Agricultural Economics 41(S1): 155–66.

—— 2010b. ‘Beyond markets and states: polycentric governance of Complex economic systems’, American

Economic Review 100(3):

641–72.

—— 2011. ‘Background on the

institutional analysis and development framework’, Policy Studies Journal 39

(1): 7–27.

——, Larry Schroeder and Susan Wynne. 1993. Institutional Incentives and Sustainable Development: Infrastructure

Policies in Perspective. Boulder, CO: Westview Press.

——, R. Gardner and J. Walker. 1994. Rules, Games, and Common-Pool Resources. Ann Arbor, MI: University

of Michigan Press.

Ostrom, V. 1964. ‘Editorial comment: Developments in the “no-name” ﬁelds of public administration’,

Public Administration Review 24(1): 62–3.

McGinnis and Aligica

—— 1971. The Political Theory of a Compound Republic: Designing the American Experiment, 2nd edn. Lincoln,

NE: University of Nebraska Press.

—— 1972. ‘Polycentricity’, Workshop Archives, Workshop in Political Theory and Policy Analysis,

Bloomington, IN: Indiana University. Presented at Annual Meeting of the American Political Science

Association, September 5–9, Washington, DC. Reprinted in McGinnis 1999b. Polycentry and Public

Economies, Ann Arbor, MI: University of Michigan Press.

—— 1973. The Intellectual Crisis in American Public Administration, 2nd edn. Tuscaloosa, AL: University of

Alabama Press.

—— 1991. ‘Some ontological and epistemological puzzles in policy analysis’, Working Paper No. W82-16,

Workshop in Political Theory and Policy Analysis, Bloomington, IN: Indiana University.

—— 1997. The Meaning of Democracy and the Vulnerability of Democracies: A Response to Tocqueville’s Challenge.

Ann Arbor, MI: University of Michigan Press.

——, C.M. Tiebout and R. Warren. 1961. ‘The organization of government in metropolitan areas: A

theoretical inquiry’, American Political Science Review 55: 831 – 42.

—— 1990. ‘American constitutionalism and self-governance’, in M. Rozbicki (ed.) European and American

Constitutionalism in the Eighteenth Century. Vienna: US Regional Program Oﬃce.

——, R. Bish and E. Ostrom. 1988. Local Government in the United States. San Francisco: ICS Press.

Ostrom, V. and E. Ostrom. 1965. ‘A behavioral approach to the study of intergovernmental relations’, The

Annals of the American Academy of Political and Social Science 359: 137–46.

—— 1971. ‘Public choice: A diﬀerent approach to the study of public administration’, Public Administration

Review (31): 203–16.

—— 1977. ‘Public goods and public choices’, in McGinnis 1999b. Polycentry and Public Economies, Ann

Arbor, MI: University of Michigan Press.

Parks, R.B. 1993. ‘Comparative metropolitan organization: Service production and governance structures

in St. Louis (MO) and Allegheny County (PA)’, Publius 23(1): 19–39.

——and

R.J. Oakerson. 1989. ‘Metropolitan

organization and governance: A local public economy

approach’, Urban Aﬀairs Quarterly 25(1): 18 –29.

Poteete, Amy, Marco Janssen and Elinor Ostrom. 2010. Working Together: Collective Action, the Commons,

and Multiple Methods in Practice. Princeton, NJ: Princeton University Press.

Simon, H.A. 1955. ‘A behavioral model of rational choice’, Quarterly Journal of Economics 69(1): 99–118.

Tiebout, C.M. 1956. ‘A pure theory of local expenditures’, Journal of Political Economy 64: 416–35.

Toonen, Theo A.J. 1998. ‘Networks, management and institutions: Public administration as “normal science”’,

Public Administration 76: 229–52.

Whitaker, G. 1980. ‘Coproduction: Citizen participation in service delivery’, Public Administration Review

40(3): 240–6.

Institutional analysis and political economy

Postpositivism and the

policy process

Raul Perez Lejano

Our understanding of what policy-making is, and what policy analysts do, evolved beyond a

clear conception of policy-making as a linear, rational exercise to a more unsettled but prag-

matic view of it as a complex, plurivocal, indeterminate, and often controversial act. This marked

transition has been referred to as the argumentative (or discursive, linguistic, communicative)

turn in policy (Fischer and Forester 1993). As will be discussed, this opening up to a broader

landscape has both enlivened and fractured the policy discipline (to the extent that the word dis-

cipline is itself problematic). It has moved scholars and practitioners of policy to wonder, “what

is our science?” or perhaps, more appropriately, “what is our folly?”

Turning points in the practice of policy

The notion of policy-making and analysis as a rational process is traced to the rise of systems and

decision analysis, beginning with the Second World War. This era saw the use of mathematics

and the increasingly powerful growth in computer processing toward the systematic allocation

of resources, ﬁrst for military purposes but soon thereafter to large-scale social projects. The ﬁrst

policy think tanks, of which the RAND Corporation is perhaps emblematic, arose as seats of a

grand planning endeavor, in which numerical optimization could be translated into optimal

outcomes in the battleﬁeld, in the emerging mega-cities and suburbs, in schools, hospitals, and

other sites of social activity. This heritage draws, ﬁrst and foremost, from the pioneering work of

von Neumann and Morgenstern (1944) in conceptualizing a theory of games and decisions.

This powerful analytical lens begins and ends with the notion of individual rationality as self-

utility maximization, and of society as a collection of rational individuals. Translation of this

system of rational thought to actual application would not be long, e.g. Schelling’s treatment of

war and conﬂict as multi-player games among rational actors (1980) or to Buchanan and Tul-

lock’s use of the same model to analyze political instituitons (1962). A central ﬁgure in the

translation of the rational model to the public realm of allocation of goods and services is Karl

Mannheim (1940), who provided an early vision of democratic social planning. In this modernist

vision, exempliﬁed by Le Corbusier’s utopian cities, society can be designed.

Where, then, did the turn away from the rational model occur? In part, we can trace some of

it to a turn to reﬂection occasioned by a seeping disillusionment in these grand experiments in

rational policy-making. Several key examples will suﬃce. But ﬁrst, we ask the reader to excuse

an inescapable “US-centricity” to this account. Much of the self-reﬂection one ﬁnds in the

literature occurs in US academia and media. At the same time, we occasionally bring in

examples from other regions to guard against monocentricity and, at times, as a counterpoint to

some of these observations.

Disillusionment with the state’s grand social experiments began early. From the experience of

Lyndon Johnson’s war on poverty, ending with ever-increasing disparities between rich and poor,

to the military rational planning of the Cold War, ending in national defeat to the North Vietna-

mese, United States ventures have failed in grand manner. In city planning, the rational design of

large-scale Corbusian housing complexes ends with the demolition of the grandiose Pruitt-Igoe

tenements in Chicago.

But this was not just the American experience. Indeed, one cannot look back on the great

leaps in agricultural productivity of post-feudal China without recalling, at the same time, the

social ills of the Cultural Revolution. Across the world, grand state-led social engineering, led

by institutions like the World Bank, resulted in large state agencies in developing nations that

failed to deliver goods and services eﬀectively and failed to eliminate entrenched poverty

(Rapley 1996).

In fact, much of this was foretold many decades earlier, in Weber’s description of the

sweeping rationalization of society (Weber 1904). In this account, rationalization involved the

increasingly narrow reduction of social processes and institutions into means-end technocracies in

which easily calculable or accountable outputs would be optimized in rationally designed insti-

tutions. In this type of Zweckrationalität, bureaus would maximize bureaucratic imperatives to

the exclusion of deeper, perhaps more elusive social goals, e.g. energy departments would

maximize energy output to the exclusion of environmental quality; transportation agencies would

maximize road capacity to the detriment of walkable neighborhoods, etc. In his incisive critique

of the Tennessee Valley Authority, Selznick described how the state agency itself became cap-

tured by the very entities it regulated (the power companies) because their interests coincided

(1949). Skepticism over institutions grew ever deeper during the Watergate era, to be repeated

from administration to administration, from the Iran-Contra aﬀair to Iraq. A good summary

account of the failures of the rational, modernist state is found in Scott’s description of high

modernism (Scott, J. 1998).

In contrast to the rationalist goal of clearly deﬁned objectives and alternatives, Rittel and

Webber describe pressing social issues as wicked problems—namely those for which no right or

wrong exists, with ends for which there is no consensus in society, with alternatives and con-

sequences that cannot be determined by technical analysis, and no good way to even ascertain if

the problem has been solved or not (Rittel and Webber 1973). In such a milieu, the max-

imizing impulse gives way to the question, “what is it exactly that we want to optimize?” and the

consequentialist framework gives way to the question, “how do we act when we just don’t

know what will result?”

Wicked problems are all around us. As this chapter is being written, society in the US in 2012 is

caught in an ideological divide. As ﬁscal conservatives forced the country into a near default in

pursuit of an ideological commitment to small government, low taxes, and defeating an

incumbent President, it became apparent how intractable the gulf was. The question quickly

became that of choosing between maximizing individual wealth or social equity, and society

itself was divided over the choice just as the liberals and conservatives in congress were. As to

predicting consequences of one action versus another, one side clearly believed that budget-cutting

would create growth, while the other believed that stimulus and near-term spending would

pump-prime the economy. What is right when one side’s notion of rationality is a polar

Postpositivism and the policy process

opposite of the other? At this time, there is evolving yet another turn in policy thought being

precipitated by the Depression of 2008, and it is unclear how the policy disciplines will change

from this time forward—that book is still being written.

Turning points in policy thought

At the same time as the grand experiment at social intervention was beginning to waver, a

revolution in the theoretical underpinnings of policy was evolving.

Part of the intellectual movement owes its trajectory to the “linguistic” turn in the social sci-

ences and humanities. The reference to language owes to the advancement in the ﬁeld of lin-

guistics, beginning with Saussure’s structural theory of meaning (1916). Saussure posited that the

meaning of words is not inherent in the words themselves but something that arises in the place

of the word in the constellation of other words forming the language system. Thus, the word

“poor” means what it means only in juxtaposition to words like “destitute,”“rich,”“wealthy,”

“penniless,” and all the other words that, by virtue of its position relative to each other, deter-

mine their respective meanings. Meaning is purely positional, which is why Saussure’s is said to

be a structural theory of meaning. Post-structuralism comes on the heels of Saussure’s theory by

positing further that not only is meaning relative and positional, it is not even ﬁxed this way.

Meaning is never determined and, instead, is always subject to the free play or polysemy of

language (Derrida 1978).

Parallel developments include Wittgenstein’s philosophical treatise on logic, in which, con-

trary to logicians who maintain that fundamental truths can be arrived at by beginning with

foundational facts and building logically outward, all truth claims are contingent. There is no

logical truth, only language games wherein truth is established only within a particular language

system, but never established categorically (Wittgenstein [1922]61). Truth is socially constructed,

subject to conventions established in social discourse, but always contextual, never categorical.

In the sociology of knowledge, much of the unraveling of the notion of absolute knowledge is

traced to Lyotard, who began by contrasting technical-scientiﬁc knowledge with narrative or folk

knowledge, then argued that both were language games, no one more privileged than the other

to any claim to truth (Lyotard 1979). Lyotard’s declamation against universality of knowledge

(what he called grand narratives) is thought to be the ﬁrst emphatic statement of a new, post-

modern age where there are no deﬁnitive truth claims, social norms, or teleological principles,

but only individual and particular voices in a discursive babble (or paralogy, in Lyotard’s words).

This turning away from modernity, with its pursuit of universal truths and its espousal of the

enlightenment of rationality and scientiﬁc inquiry, and a turning toward a never-ending ques-

tioning of purpose and meaning, had a profound eﬀ ect on policy thought (and, as will be argued, a

later and less dramatic eﬀect on practice). Berger and Luckmann wrote about institutions as

social constructs, subject to artiﬁce and ever-changing convention (1966). This easily lends to a

questioning of the institutional foundations of policy—the meaning of the state, the legislature,

or the legal process. Others questioned even the primacy of science as a truth—establishing

endeavor, showing that the scientiﬁc process itself is very much a cultural practice as other cultural

practices such as taboo or courtship (Latour 1987; and see Douglas 1966 for an earlier statement

of modernity and culture).

The eﬀect of these intellectual revolutions on thinking about policy, and the institutions in

which policy happens, has been deeply felt. Consider the linguistic turn and its turning away from

logocentricity, which is the notion that words point to objective truths which language accu-

rately represents. And consider the turn toward the idea that words themselves are not ﬁxed in

meaning but subject to play and debate. This extends immediately to the most physical object of

Lejano

100

policy, which are statements of policy, policy documents, and law. In this poststructural age,

even a statement like “We hold these truths to be self-evident, that all men are created equal, that

they are endowed by their Creator with certain unalienable Rights, that among these are Life,

Liberty and the pursuit of Happiness ” itself ceases to be self-evident. The words are enshrined

for all of history, but their meaning changes. Equality and liberty mean diﬀerent things today than

when ﬁrst written in Jeﬀerson’s time. And what they mean today is itself subject to debate. Does

equality mean extending the same bundle of rights and privileges to undocumented immigrants?

If not, then these persons are consituted as less than citizens, which is what the Declaration of

Independence meant by persons (which, in Jeﬀerson’s time, referred to white, male landowners).

The telos of policy is also uncertain. What is the maximand of rational policy-making if we

cannot even come to agreement over what liberty, rights, and public good mean? There is no

transcending ethic that all can come to agreement over. In the context of the ideological debates in

the US Congress, some have tried to appeal to an overweening ethical principle “Nation before

Party!” but even this is questioned, as the nation’s good has been conﬂated on all sides with the

party’s. As Sarah Palin said:

Politics isn’t just a game of competing interests and clashing parties. The people of America

expect us to seek public oﬃce and to serve for the right reasons. And the right reason is to

challenge the status quo and to serve the common good.

(http://www.npr.org/templates/story/story.php?storyId=94118910 (downloaded

December 10, 2011))

Everyone, even the political maverick, lays claim to the public good.

The deconstruction of institutions has also had a profound eﬀect, beginning with the decentering

of the state (which parallels the decentering of the author as the font of meaning of the logos)

away from the presumption that the state represents the people and acts for the greatest good of

the greatest number (Mill 1863). Devolution away from the state has been most marked in the

“government to governance” literature in which focus turns to networks of policy actors acting

in often innovative ways (Koppenjan and Klijn 2004; Edelenbos et al. 2010; Schout et al. 2010).

It is perhaps no accident that a central ﬁgure in linguistics, Noam Chomsky, would emerge as one

of the foremost critics of the new order, characterizing the state as a conglomerate of military,

industrial, government, and big business interests (Chomsky and Herman 1988). But perhaps the

most trenchant critique would come from the French sociologist, Michel Foucault, who posited

all modern institutions to be modeled after the same basic mold of the penal institution, designed to

disipline the public and channel all human activity toward the furthering of the interests of the

powerful (1977). And as the policy-making institutions go, so does the very integrity of the policy

that emerges from them. Thus, in a skeptical age, anactionsuchastheTARP(TroubledAssetRelief

Program), the US Federal Reserve’s lending instrument that kept large ﬁnancial institutions from

bankruptcy in 2008, could be so casually described by political commentators in terms such as:

The bailout money is just going to line the pockets of the wealthy, instead of helping to

stabilize the economy or even the companies receiving the bailouts: Bailout money is being

used to subsidize companies run by horrible business men, allowing the bankers to receive

fat bonuses, to redecorate their oﬃces, and to buy gold toilets and prostitutes.

(www.worldcrisisnewswatch.com/2011/05/tarp-bailout-scam-explained-government.html)

Thus, the postpositivist turn has come to mean not just a turning away from the notion that

the public good was measurable and subject to systematic analysis and optimization. In fact it

Postpositivism and the policy process

101

questioned the foundational meaning of policy itself, not only uncertainty over the correlation

between policy actions and the public’s welfare, but that the very intent of policy could not be

presumed to be aligned with such collective good.

Policymaking in an interpretive world

To sum up the state of aﬀairs thus far, if:

 policy goals cannot be set deﬁnitively, in a manner achieving broad consensus over them;

 policy situations cannot be described or analyzed in a positivist manner (i.e. their truth

determined by some deﬁnitive measures);

 policy instruments and policy texts cannot be unambiguously determined with regard to

their meaning;

 policy-making and implementing institutions cannot be assumed to have the capacity to

realize policy targets, nor assumed to be aligned with them in purpose

then what policy should be and how policy should be enacted cannot be determined in objective

manner and, instead, a matter of interpretation. This is not a trivial proposition, inasmuch as the

essence of community and political life, as Aristotle said in the Politics, is their “establishment

with a view to some good.” If we cannot agree on what “good” means for the community,

then how do we set policy and go about attaining its ends?

In an interpretive world, where meaning and good ends are subject to deliberation and debate,

policy-making is less about measuring and projecting the optimal trajectory for policy but the

crafting of a compelling and consensus-building story that captures the public’ s moral imagina-

tion and moves people and groups to action. In this account, we refer to a number of key

works in the postpositivist policy literature—at this point, a handful of useful summary treat-

ments can be found in Stone, Roe, Fischer and Forester, Fischer, Hajer, Kaplan, Schneider and

Ingram, Yanow, Schon and Rein, and Dryzek.

Making sense of the world, and policy situations encountered in it, requires seeking satisfac-

tory interpretations of these situations. For Ricoeur, people make sense of the world by crafting

coherent narratives about it (1981; also Bruner 1991). In Ricoeur’s words, “By means of the

[narrative’s] plot, goals, causes, and chance are brought together within the temporal unity of a

whole and complete action” (1984: ix). So it is, then, that by reconstructing a person’s underlying

narrative account, we can best reproduce their experience of a situation. To do this, Ricoeur

espouses a hermeneutic approach to narrative interpretation (1976, 1981).

Goﬀman said something similar when he evoked the idea of frames, which were interpretive

constructs used to create coherent meaning in an otherwise complex and incomprehensible

situation (1974). Frames, as Goﬀman introduced them, are interpretive schemes that selectively

highlight aspects of the issue and introduce a speciﬁc perspective that then allows one to capture

reality in a simpler manner (1974; also Schön and Rein 1994). But seldom found in this research is

the explicit recognition that, very often or perhaps generally, these mental models, cultural models,

and frames take on the appearance and essence of story. Bruner is one of the few to make this

explicit connection (1987). As others have expressed it: “Given the amount of uncertainty about

the world, people create cause-and-eﬀect stories to ﬁll in the blanks. Frames—also known as

mental models, schemas—are essentially such stories” (Wesselink and Warner 2010). Stone talks

about how “causal stories” are used to craft and process policy ideas through the process (1997)—

though, we might add, the causal story is just one of the genres one ﬁnds in the policy ﬁeld (i.e.

others include moral fables, epic myths, tales of chivalry, surrealist ﬁction, etc.). And so, for

Lejano

102

example, along with the proposal for TARP was the narrative of how ﬁnancial institutions like

Goldman Sachs were “too big to fail.” In this interpretive model, it is on the strength of the

policy story that policies win or lose.

As a result, over several decades now, scholars have called for the systematic use of narrative

analysis in public administration and policy research (Roe 1989, 1994; Yanow 1992, 2007; Balfour

and Mesaros 1994; Sköldberg 1994; Schram and Neisser 1997; Hajer and Wagenaar 2003;

Ospina and Dodge 2005; Hampton 2009; Lejano and Leong 2012). This is joined by an already

considerable literature on narrative approaches to studying organizational process and design (e.g.

see Martin 1982; Boje 1991; Weick 1995; Boyce 1995; Czarniawska 1998; Patriotta 2003).

The move towards narrative approaches began with the interpretive turn in the social sciences

(Berger and Luckmann 1966; Rabinow and Sullivan 1979, 1987). In the 1970s and 1980s, scholars

like Fischer and Forester (1993) argued for an interpretive approach to public policy studies.

The then prevalent positivist-instrumentalist bent of the social sciences had failed in carrying out

its promise of giving a good account of what goes on in government and society at large (Dryzek

1982). The deductive approach of technical knowledge, with its search for objective measures

and universal principles, failed to account for the fact that social phenomena mean diﬀerent things

to diﬀerent people, and that such social constructions are highly contextual (ibid.: 310; also

Fischer and Forester 1993). The primacy of technical knowledge was questioned and, increasingly,

narrative knowledge began to be seen as equally valid (Lyotard 1979; Jameson 1984).

Among the ﬁrst to use narrative analysis to study public controversies was Emery Roe (1989,

1994). His point was that narratives are one of the richest vehicles for the multiple and complex

meanings that diﬀerent stakeholders bring to a public issue. Martha Nussbaum says something

similar when she describes how complex normative positions are best expressed in narrative

form (1990). Deborah Stone speaks to the power of “causal stories” which public managers and

politicians craft, to construct reality in a way that best captures their interests and policy goals (1997).

Given this irreducible element of social construction, she argues, understanding deliberation in

the public realm requires an interpretive approach (Stone 1997). In our work, we look for

stories that are told by diﬀerent publics, including citizens’ groups and media.

In this discussion, we understand a narrative to simply mean a story, composed of a coherent

sequence of events, involving a set of characters, which are actors with deﬁnite traits (e.g. see

Ryan 2007; Bal 2009). When we reconstruct a narrative underlying an issue, we are simply com-

posing a story, with events and characters, as if a narrator were recounting it to us directly. It is

possible that no one source (whether a respondent or a text) ever gives us a complete narrative

in toto, or that it gives us only one particular account of a larger narrative, and it is the analyst’s

role to construct a story that reﬂects the diﬀerent aspects of the issue in a coherent way. The

resulting, reconstructed narrative is akin to the fabula (or storyline) of a literary piece—namely it

presents the plot even when it cannot be found in any single narrator’s particular account (Bal

2009). Stories, then, are used to capture the imagination and support of decision-makers and

public to rally around a policy initiative and form a coalition in support of it. Sometimes, in

fact, the stories themselves organize these coalitions and networks of policy actors—literally,

policy movements and networks are emplotments of characters and events into one coherent

narrative (Lejano et al. 2013). Most recently, Lejano and Leong (2012) develop a hermeneutic

approach for analyzing the multiple dimensions of public debate.

But this should not be thought of as mere political “spin doctoring”, as the stories one ﬁnds

in the policy realm can be taken at face value or critiqued, and can engender belief or incre-

dulity. So the question is: ow can we tell a good, from a bad, narrative interpretation of a public

issue? There are diﬀerent ways, and Kaplan (1986) discusses some. In Kaplan’s terms, good

narratives should be empirically veriﬁable, rich, and coherent. First, a narrative can be tested

Postpositivism and the policy process

103

against known empirical information, events, etc. Secondly, given two narratives that both meet

empirical tests, the one that is able to account for more facets of experience, empirical reality,

and people’s personal accounts, should be regarded as the better, richer one. Lastly, the test of

coherence suggests that we assess the strength by which diﬀerent accounts, diﬀerent facets of the

story, the sequence of events, etc. are connected by some logical (or other) scheme. For example,

the passage “I saw a bird ﬂying. I started weeping.” is not as coherent as this one: “I saw a bird

ﬂying above me. The majesty and utter simplicity of this reality made me feel so alive, I could

not but begin to weep.” There are other prescriptions for how one distinguishes a better story

from an inferior one, but in practice, this poses no problem for us in everyday life. To cite a trivial

example, everyone knows Rocky I was a far better story than Rocky V, without even needing

to explicitly describe the complex mode by which we come to such judgements.

The pragmatism test lies in the ability of the reconstructed narrative to explain the diﬀerent

contingencies and particularities surrounding the issue. Does it help us better to understand the

motivations of the diﬀerent actors and their actions? In the case study to be taken up herein, does it

help us to understand why opposition to the project was so strong and eﬀective, and why

attempts by the agency at knowledge transfer did not suﬃce to quell this opposition?

Why is there a need to ﬁnd meanings beyond the surface of what people say or write? First of

all, there may be processes or conditions, like ocean currents beneath a surface calm, that inﬂuence

the proceedings even though these processes may be well removed from the matter at hand.

Tradition, power, culture, and other forces of structuration act like this. Second, there may be issues

that stakeholders, consciously or subconsciously, relegate to the background. Yanow notes that

in public decision-making, some issues are verboten—namely “publicly unspeakable because there

is no explicit public consensus underlying them” (1992: 400). This is also tied up with political

necessities such as maintaining “public silences about contradictions” (1992: 418).

Policy institutions and the postmodern condition

In the previous section, we took up the task of analyzing policy ends and determining policy

directions when truth and good could not be unambiguously stated. We then, now, ask the

closely related question of what is the policy enacting institution (an agency, a community, or an

individual) to do under such uncertainty? How do policy actors proceed to act in the public realm?

If, taking a cue from Lyotard, all policy discussions were simply nothing other than language

games, then policy-making must then be simply a competition over the best discourse. This can

devolve into the classic pluralist model of competition over policy agendas. In this game, the

better, more inﬂuential, better resourced interest group wins the policy battle. But this process can

take diﬀerent forms, in part determined by how policy institutions view the policy question.

Karl Popper’s epistemological framework viewed the establishment of truth and good as

unachievable by any objective, scientiﬁc model but rather the contingent and defeasible setting of

truth claims. By defeasible, we mean a process whereby claims are constantly subject to testing

and conﬁrmation or revision in the world of practical implementation (Popper 1994). This incre-

mentalist view of knowledge leads to a pluralist model such as that of Lindblom’s mutual partisan

adjustment, wherein, through a tug of war of competing interests, society ﬁnds some reasonably

optimal trajectory that is cognizant of the diﬀering interests and their strength of preference

(1965). In this model, process determines the right.

But process can determine what is right in other ways. Habermas believed in the possibility

of a kind of rationality that had a transcendent moment that allowed the diverse and conﬂictual

set of policy actors to come to agreement over what is right (1984). In this model, what is true

and good is not available a priori but reached through a process of agonistic debate. It is still a

Lejano

104

tug of war, carried out through dialogue, but with a diﬀ erence: in this model, policy actors can

come to a mutually agreed resolution of what is right and good, through the force of better

argument. The key, in this communicative model of rationality, is the practice of the ideal speech

situation, where all have an ability to speak and be heard and communication is nondistorted.

This epistemological framework ﬁnds its application in diﬀerent practical models for public delib-

eration and consensus-building (Bobrow and Dryzek 1987; Healey 1996; Innes and Booher

2009; Susskind et al. 1999). In these situations, consensus can be reached through compromise

across ﬁxed but divergent interests (Fisher et al. 1991), or through the construction of new

understandings of a situation that move policy actors to reconsider their positions or to realize a

higher goal that exceeds the provincial limits of individual preference (Schön and Rein 1994).

The social construction of policy pertains not just to the identiﬁcation of solutions to pressing

social problems, but the active construction of the problems themselves. As Schneider and Ingram

depicted it, what the issue is, what the stakes are, who the stakeholders are, and what their roles

in the policy issue as constructed might be—that is, as targets, beneﬁciaries, victims, or agents of

policy (1997). In this framework, the key step is reached well before problem-solving begins

but in the earlier phase of issue identiﬁcation and stakeholder analysis. This view of policy lends

a constructionist interpretation to the issue-salience and policy windows model of the policy

process (see Kingdon 1984). Issues do not exist a priori to be solved—rather, they are constructed

in response to the priorities of agenda-setters. In some cases, policy solutions preexist issues and

can take the form of policy responses waiting for appropriate problems to be identiﬁed in order

for them to be employed (March and Simon 1958). From this point of view, military action (to

take one example) is seen not as a response to an emergent conﬂict, but the other way around,

namely conﬂicts are identiﬁed (or designed) as a response to an unmet demand for military action.

In this type of policy milieu, ideas (which are what social constructions are, after all) and the

rhetorical skill by which they are deployed become the focus of policy attention (Throgmorton

1991; Howlett and Rayner 2008).

Models of the policy process taken up in the ﬁrst section of this book still pertain. We are

well advised to not jettison the idea of a policy process template. In the context of the linguistic

turn, however, each of the stages in the policy cycle are now seen as forums for public delib-

eration, namely as part of an extended communicative process. As Laswell ﬁrst suggested, an initial

step still remains that of issue identiﬁcation, but since meaning of policy can no longer be ﬁxed,

interpretations of issues (and their salience in the public mind) continues all throughout the

policy process. Moreover, in the postmodern era, these processes need not necessarily lead to

outcomes, at least not idealized ones, nor to a transition from one stage to another. Habermas’

speech situation need not lead to decisions, as the focus is on the communicative process. In

Lyotard’s model of paralogy, consensus is never envisioned—it is enough to simply have debate

and an airing of disparate views. In the postpositivist age, talk is action.

But this is just a short step away from a complete rethinking of the policy process.

Deconstructing the policy process

Classic models of the policy process, involving stages or cycles of issue identiﬁcation, analysis, and

ﬁnally, implementation are actually translations of the rational model into institutional form.

The classic rational model of analysis involves just these stages, from identiﬁcation of overall

goals or values, enumeration of policy alternatives, evaluation, decision, and enaction. The idealized

policy process is a hypostatization of the model of rationality.

But just as, not merely postructural theory, but the practice and follies of real-world policy has

forced us to reconsider the rational model, so is it possible to deconstruct the entire policy process.

Postpositivism and the policy process

105

In part, this deconstructionist turn is guided by the new institutionalism that emerged from the

ﬁelds of public administration and organization theory. In this postmodern view, institutions are

themselves social constructions (e.g. March and Olsen 1989) that themselves are crafted after well-

constructed narratives (Weick 1995). When institutions take on a repeatable form, it is not because

of some objective property of these institutions, but rather, because their social constructions

become norms or standards that diﬀuse from polity to polity (DiMaggio and Powell 1991).

While the turn away from logocentricity involved an undoing of the idea that meaning is

ﬁxed in the words of policy, it also undoes the primacy of the author as the center of meaning.

From this point onward, policy will be, as Barthes put it, a “writerly” text (1975), where the

implementors and targets of policy are just as much authors as the writers of policy. This turns

the policy process inside out. For example, where and at what stage is policy designed? In a

deconstructed process, policy is not simply designed in a central location and diﬀused outward;

rather, it undergoes transformation and redesign as new policy actors take on the institutional

norm and make use of it (Howlett and Rayner 2008). Consider a policy on conditional cash

transfers for the poor, wherein payments are made to recipients so long as the latter pledge to or

prove that they use the money for medical or educational purposes. Then consider a situation

wherein ﬁeld agents who implement the policy by making the actual transfers to recipients and

collect information, decide to allow a women’s cooperative to pool their outlays and obtain credit

from a local bank. Are the ﬁeld agents not also participating in design of the policy, not as formally

laid out but as actualized in the ﬁeld?

As Pressman and Wildavsky showed, neat boundaries between the policy design and imple-

mentation become blurred in practice (1979). In practice, all such conceptualizations of the policy

process do blur in many diﬀerent ways. For this reason, one can take a phenomenological

approach to analyzing policy processes. Instead of beginning, in positivist fashion, from the uni-

versal concept and applying it to actual processes deductively, one can instead begin and end

with the actual process in place, in actual context, as Husserl, the pioneer of phenomenological

thought, put it: to the thing itself (1900). That is, do not analyze the process according to a pre-

constructed template, whether it be one of stages or cycles of policy-making; rather, describe it

as it is, as experienced and observed in place. For example, Argyris and Schön wrote about how

actual programs might be characterized by theories-in-practice rather than theories underlying

their design (1974).

An example of phenomenological description is seen in Lejano (2008), wherein habitat con-

servation policy is described not according to its formal design, but in the informal, dynamically

evolving practices that made up the policy as experienced. In such a mode of description,

formal descriptors of policy processes (e.g. designers, implementors, organizational roles and rules,

bureaucratic structure) all blur into sometimes new categories requiring new modes of description.

For example, in the aforementioned study, instead of describing the program as a set of rules, the

author modeled it as a web of relationships, all cohering with each other in dynamic fashion.

Another mode of studying a deconstructed policy process is to study it in the same way as

one might study the social construction of a narrative that is shared with others. The formation

of such a narrative is congruent with the formation of the network of actors that share it.

Latour, Callon, Law, and others have sought to describe the process by which such actor-networks

are formed (Latour 2005; Callon 1986; Law 1992). Callon describes it as the translation of the

policy story into forms that can be assimilated by (or, perhaps more accurately, can emplot)

other actors. In this translation process, he posits four stages (1986):

 Problematisation:deﬁning a problem that requires joint action, and set of actors who are

potentially enmeshed in it.

Lejano

106

 Interessement: the negotiation of actors’ involvement and roles in the initiative.

 Enrollment: actors’ acceptance and acting out of these roles.

 Mobilization: the expansion of the network to include communities of support.

It is not an accident that the above sequence also resembles the exposition of a story’s plot—

namely identiﬁcation of the conﬂict, description of the setting and cast of characters, and mobili-

zation of action on the characters’ part. The description of the network process is not far from

the act of emplotment—in fact, they are essentially one and the same.

The turn away from the positivist notion of determinate problem solutions and a functionalist

notion of the policy process leads to a turning toward, instead of the linear process of locating

solutions, the discursive process of issue construction and deliberation. The focus is not on the

location of the optimal solution, but on the gathering together of a critical coalition (or net-

work) that supports and champions a policy idea to completion. Callon and Latour took this

same road by focusing on the network formation process. Howlett and Ramesh (1998) discussed

how to delineate the characterized policy subsystems (which are almost the same as the net-

works described above) in terms of their openness to new policy ideas and disposition to policy

change. In this literature, the postpositivist focus is not so much on the narrative and its narra-

tion, but on the formation of coalitions and shifts in memberships and allegiances (Sabatier and

Jenkins-Smith 1993), which one reasons, leads to shifts in their shared narratives. By retaining

the classic notion of interest-based coalitions, these theories are able to combine frame, interests,

and knowledge base in one process theory.

There are other ways to modify the otherwise linear, rational process model. As Goﬀman had

shone light onto social and organizational processes that occurred “backstage” (e.g. on the

street, at home, in the coﬀee shop, smokey room, etc.) rather than the foreground of policy-making

(e.g. senate ﬂoor, parliament, executive boardroom, etc.), researchers began paying attention to

informal processes (e.g. Innes and Booher 2010), decentralized decision-making in the ﬁeld (e.g.

Lipsky 1980), or policy-making within complex networks (e.g. Kickert et al. 1997). The tran-

sition away from formal institutional processes ﬁnds its most ambitious form in a growing lit-

erature on practice (e.g. Bourdieu 1977; Giddens 1984). The modernist impulse would seek to

ﬁnd universal rules in the structures that govern individuals in society (as in dialectical materi-

alism or classic political institutional theory) or within the individual (as in rational choice or

behavioralist theories). In contrast, theorists of practice aim to represent meaning and action as

traversing the structure-agency divide. For example, Bourdieu discusses how individual decision-

making is an improvisation guided by sometimes rigid controls, often not transparent to the

individual and often immanent in the body’s own habitual movements (1990). In this frame-

work, enacting of reproductive health policy, as an example, is as much or more determined by

the everyday routines women and men follow subconsciously than formal rules and programs—

the functioning of political institutions if steered by similar habitual patterns, such as talk and

dress and relationships between members of an organization. Such focus on practice, while by

now well established in the organizational literature, is just beginning to have an eﬀect on

policy studies.

Furthermore, if policy-making proceeds through the construction of a policy narrative, then

perhaps it is also through narrative means by which the process is best described. This involves

soliciting stories from policy actors caught up in the process (as in Pentland 1999; also Lejano et al.

2012). Description of the process is captured by narrative. These stories can vary from place to

place—perhaps aligned with easily recognized templates (e.g. policy stages) in some respect, but

ﬁnely and richly diﬀ erent in the way it unfolds. Through a postpositivist lens, these ﬁne dif-

ferences are as important as the general commonalities. Whereas the positivist lens emphasizes

Postpositivism and the policy process

107

deductive reasoning from universal principles (e.g. policy stages), postpositivism focuses on the

particulars of the phenomenon.

To sum up: why are processes not so easily circumscribed in a postpositivist frame of analysis?

Because policy no longer can be seen as an object that can be objectively determined, and

neither can we place exactly where policy is crafted and how. Policy meanings cannot be

pinned down and, so, where there used to be analysis, we have deliberation all throughout the

policy process, and where there used to be pure implementation, we now have redesign and

reinterpretation of policy. Authorship of policy is not ﬁxed and determinate, and where a policy

is authored is similarly deconstructed to include a new constellation of policy actors.

The problem of social construction

Contrary to a radical constructionist interpretation of policy, most of the scholars cited above do

not espouse simple constructionism (Fischer 2003). Policy stories are not simply ﬁctions crafted

out of sheer artiﬁce. We discussed, earlier, certain tests of validity and pragmatic value that allow

us to discern whether one narrative is more credible than another. Just because we weaken the

notion of objective standards for truth does not mean an inability to tell better from worse policy

narratives. Social constructs have to be coherent, and the parts of their story have to be internally

logically consistent.

Just as importantly, policy stories have to, ultimately, be grounded in the real experiences of policy

actors. The last ten years have been a sobering experience for us all, but especially policy scholars.

Notwithstanding the concerted eﬀort to build momentum for the war in Iraq, at some point, policy-

makers had to reconcile their policy narrative with the nagging fact that there were no weapons

of mass destruction (WMDs) in Iraq. Despite the degree to which the homeownership narrative

captured the public psyche, at some point in the last few years, the socially constructed value of

property had to reconcile with the limitations of earning power of the would-be property owners,

and bubbles began to burst. There is not an inﬁnite array of narratives that can capture the public

agenda, but only those that cohere with how people and policy-makers understand themselves and

the world around them. Such understandings are borne by learning over lifetimes of experience.

Whether bubbles or WMDs, what confronts the public today is the turning back of the pen-

dulum away from a hapless constructionism to a sober realization that policy narratives are not

ﬁctions but rather attempts to represent reality. The failure of cherished institutions (market, state,

and community) have created a void in the way we idealize policy process.

What are needed

now are approaches that reconcile the desire for objective foundations for policy-making with

the inescapable intedeterminate nature of policy. Perhaps what we are beginning to see is a turn

toward a post-constructionist understanding of policy and its processes (Lejano 2006). Doubtless,

we perceive this in the recent focus on evidence-based policy-making (Pawson 2006; Bochel

and Duncan 2007; Khagram and Thomas 2010). But another part of it may be a reversion to pure

ideology. It seems that polities, the world over, are now caught in a divide between requiring

more and better-informed judgments from decision-makers or giving up on rational process alto-

gether and relying on exhortations from political extremists (and the latter occur in many forms—

pastors, mullahs, talk show hosts, governors and former governors). In this ironic age, the new

freedoms of digital communication and social networks seem to go hand in hand with the

increasing demise of free and democratic public discourse.

The challenge for political life is to

discover a new middle ground where polities are free to imagine and craft new futures while

grounding deliberation on the realities of everyday life.

Policy studies have to ﬁnd such middle (or higher) ground, as well (Lejano 2006). As De

Leon writes concerning positivism and postpositivism: “it would seem foolish to set the two

Lejano

108

concepts at odds, in a zero-sum game, as opposed to use them to inform and support one another”

(1998, p.157). Perhaps as a necessity, policy scholars who take the route of postpositivism have

espoused a deep critique, and often wholesale rejection, of the positivist model. But this has

created a bifurcated literature, with neither side wary of the achievements of the other. Doctoral

students in the ﬁelds of policy and public administration are asked, early in their academic

careers, to choose one or other of these alternatives. To postpositivists, quantitative models seem to

merely scratch the surface of institutions, and to positivists, discursive approaches appear to be pure

artiﬁce. If the ﬁrst decade of this new century is any indication, measurement and discourse need

a new reconciliation.

As one recalls the turn that policy studies took after the crises of the 1960s and 1970s, one

wonders what corner we are turning now.

Notes

1 As an example, the ongoing ﬁnancial crisis gripping the European Union threatens institutional models

that envision new regional modes of collaborative policy-making. Asian models of state-sponsored

interventions in the market begin to reveal wicked trade-oﬀs between productivity and quality of life

and liberty. And as for the American model, we all know too well about its crises and ideological

standoﬀs.

2 As this chapter is being written, the author’s own university in Irvine, California, is undergoing a crisis

over free speech—e.g. students arrested for writing with chalk on pavement in protest against runaway

tuition fee increases.

Bibliography

Argyris, Chris and Donald Schön. 1974. Theory in Practice: Increasing Professional Eﬀectiveness. San Francisco:

Jossey-Bass.

Bal, Mieke. 2009. Narratology: Introduction to the Theory of Narrative, 3rd edn. Toronto: University of Toronto

Press.

Balfour, Danny L. and William Mesaros. 1994. ‘Connecting the local narratives: Public administration as a

hermeneutic science’, Public Administration Review 54: 559–64.

Barthes, Roland. 1975. S/Z: An Essay. New York: Hill and Wang.

Berger, Peter and Thomas Luckmann. 1966. The Social Construction of Reality. New York: Doubleday.

Bobrow, Davis and John Dryzek. 1987. Policy Analysis by Design. Pittsburgh, PA: University of Pittsburgh Press.

Bochel, H.M. and S. Duncan. 2007. Making Policy in Theory and Practice. Bristol: Policy Press.

Boje, David M. 1991. ‘The storytelling organization: A study of story performance in an oﬃce-supply

ﬁrm’, Administrative Science Quarterly 36(1): 106–26.

Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press.

—— 1990. ‘Structure, habitus, practice ’, in Pierre Boudieu (ed.) The Logic of Practice, Richard Nice (trans.).

Palo Alto, CA: Stanford University Press.

Boyce, Mary E. 1995. ‘Collective centring and collective sense-making in the stories and storytelling of

one organization’, Organization Studies 16(1): 107–37.

Bruner, Jerome. 1987. ‘Life as narrative’, Social Research: An International Quarterly 71(3): 691–710.

—— 1991. ‘The narrative construction of reality’, Critical Inquiry 18(1): 1–21.

Buchanan, James and Gordon Tullock. 1962. The Calculus of Consent. Ann Arbor, MI: University of

Michigan Press.

Callon, Michel. 1986. ‘Some elements of a sociology of translation: Domestication of the scallops and the

ﬁshermen of St Brieuc Bay’, in John Law (ed.) Power, Action and Belief: A New Sociology of Knowledge.

London: Routledge and Kegan Paul.

Chomsky, N. and E. Herman 1988. Manufacturing Consent: The Political Economy of the Mass Media. New

York: Pantheon Books.

Czarniawska,

B. 1998. A Narrative

Approach to Organization Studies. Thousand Oaks, CA: Sage.

Derrida, Jacques. 1978. Writing and Diﬀerence (L’Écriture et la Diﬀérence), A. Bass (trans.). Chicago: University

of Chicago Press.

Postpositivism and the policy process

109

de Leon, P. 1998. ‘Models of policy discourse: Insights versus prediction’, Policy Studies Journal 26(I):

147–161.

de Saussure, Ferdinand. 1916. Cours de Linguistique Générale, C. Bally and A. Sechehaye (eds) with the

collaboration of A. Riedlinger. Lausanne and Paris: Payot; Course in General Linguistics, W. Baskin

(trans.), 1977. Glasgow: Fontana/Collins.

DiMaggio, Paul and Walter Powell. 1991. ‘The iron cage revisited: Institutional Institutional isomorphism and

collective rationality in organization ﬁelds’, in P. DiMaggio and W. Powell (eds) The New Institutionalism

in Organizational Analysis. Chicago: University of Chicago Press.

Douglas, M. 1966. Purity and Danger: An Analysis of Concepts of Pollution and Taboo, Routledge and Keegan

Paul, London.

Dryzek, J. 1982. ‘Policy analysis as hermeneutic activity’, Policy Science 14(4): 309–29.

Edelenbos, J., N. van Schie and L. Gerrits. 2010. ‘Organizing interfaces between government institutions

and interactive governance’, Policy Sciences 43(1): 73–94.

Fischer, Frank. 2003. Reframing Public Policy: Discursive Politics and Deliberative Practices. Oxford: Oxford

University Press.

——and John Forester (eds). 1993. The Argumentative Turn in Policy Analysis and Planning. Durham, NC:

Duke University Press.

Fisher, R., W. Ury, and B. Patton. 1991. Getting to Yes: Negotiating Agreement Without Giving In. New

York: Penguin.

Forester, John. 1999. The Deliberative Practitioner. Cambridge, MA: MIT Press.

Foucault, Michel. 1977. Discipline and Punish: The Birth of the Prison, Alan Sheridan (trans.). New York:

Pantheon.

Giddens, Anthony. 1984. The Constitution of Society: Outline of the Theory of Structuration. Berkley, CA:

University of California.

Goﬀman, E. 1974. Frame Analysis: An Essay on the Organization of Experience. Cambridge, MA: Harvard

University Press.

Habermas, Jürgen. 1984. Reason and the Rationalization of Society, vol. 1 of The Theory of Communicative

Action, Thomas McCarthy (trans.). Boston: Beacon Press.

Hajer, Maarten A. and Hendrik Wagenaar (eds). 2003. Deliberative Policy Analysis. Cambridge: Cambridge

University Press.

Hampton, Greg. 2009. ‘Narrative policy analysis and the integration of public involvement in decision

making’, Policy Sciences 42: 227–42.

Healey, Patsy. 1996. ‘The

communicative turn in

planning theory and its implications for spatial

strategy-making’, Environment and Planning B: Planning and Design 23(2): 217–34.

Howlett, Michael and M. Ramesh 1998. ‘Policy subsystem conﬁgurations and policy change: Operationalizing

the postpositivist analysis of the politics of the policy process’, Policy Studies Journal 26(3): 466–81.

——and Jeremy Rayner. 2008. ‘Third generation policy diﬀusion studies and the analysis of policy mixes:

Two steps forward and one step back?’, Journal of Comparative Policy Analysis: Research and Practice 10(4):

385–402.

Hysing, E. 2009. ‘From government to governance? A comparison of environmental governing in Swedish

forestry and transport’, Governance 22 (4), 647– 72.

Husserl, Edmund. 1900. Logical Investigations (Logische Untersuchungen) vol. 1, J.N. Findlay (trans.). New

York: Routledge.

Innes, Judith E. and David E. Booher. 2010. Planning with Complexity: An Introduction to Collaborative

Rationality for Public Policy, New York, NY: Routledge.

Jameson, Frederic. 1994. ‘Postmodernism, or the cultural logic of late capitalism’, New Left Review 1: 146.

Khagram, S. and C.W. Thomas. 2010. ‘Toward a platinum standard for evidence-based assessment by

2020’, Public Administration Review 70(s1): S100–S106.

Kaplan, Thomas. 1986. ‘The narrative structure of policy analysis’, Journal of Policy Analysis and Management

5(4): 761–78.

Kickert, Walter J.M., Erik-Hans Klihn and Jopp F.M. Koppenjan. 1997. Managing Complex Networks:

Strategies for the Public Sector. London and Thousand Oaks: Sage.

Kingdon, John. 1984. Agendas, Alternatives, and Public Policies. Boston: Little, Brown.

Koppenjan, J. and Klijn, E.H. 2004. Managing Uncertainties in Networks: A Network Approach to Problem

Solving and Decision Making. London: Routledge.

Latour, Bruno. 1987. Science in Action. Cambridge, MA: Harvard University Press.

——2005. Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford: Oxford University Press.

Lejano

110

Law, John. 1992. ‘Notes on the theory of the actor network: Ordering, strategy, and heterogeneity’,

Available at: www.lancs.ac.uk/fass/sociology/papers/law-notes-on-ant.pdf

Lejano, Raul. 2008. ‘The phenomenon of collective action: Modeling institutions as structures of care’,

Public Administration Review May/June: 491–504.

——and Ching Leong. 2012. ‘A hermeneutic approach to explaining and understanding public con-

troversies’, JPART Journal of Public Administration Research and Theory (http://jpart.oxfordjournals.org/

content/early/2012/04/04/jopart.mus001.abstract).

——, Helen Ingram and Mrill Ingram. 2013. The Power of Narrative in Networks: Relating Alternative Ecologies.

Cambridge, MA: MIT Press.

Lipsky, Michael. 1980. Street-level Bureaucracy: Dilemmas of the Individual in Public Services. New York:

Russell Sage Foundation.

Lindblom, Charles. 1965. The Intelligence of Democracy: Decision Making Through Mutual Adjustment. New

York: Free Press.

Lyotard, Jean-François 1979. The Postmodern Condition: A Report on Knowledge. Minneapolis, MN: University

of Minnesota Press.

Mannheim, Karl. 1940. Man and Society in an Age of Reconstruction. London: Routledge.

March, James and Herbert Simon. 1958. Organizations. New York: Wiley.

——and Johan Olsen. 1989. Rediscovering Institutions: The Organizational Basis of Politics. New York: Free Press.

Martin, J. 1982. ‘Stories and scripts in organizational settings’, in A.H. Hastorf and A.M. Isen (eds) Cognitive

and Social Psychology. Amsterdam: Elsevier, 255–305.

Mill, John Stuart. 1863. Utilitarianism. London: Parker, Son, & Bourn.

Nussbaum, Martha C. 1990. Love’s Knowledge: Essays on Philosophy and Literature. New York: Oxford

University Press.

Ospina, Sonia M. and Jennifer Dodge. 2005. ‘It’s about time: Catching method up to meaning—the usefulness

of narrative inquiry in public administration research’, Public Administration Review 65(2): 143–57.

Patriotta, Gerardo. 2003. ‘Sensemaking on the shop ﬂoor: Narratives of knowledge in organizations’,

Journal of Management Studies 40(2): 349–76.

Pawson, R. 2006. Evidence-based Policy: A Realist Perspective. London: Sage.

Pentland, Brian. 1999. ‘Building process theory with narrative: From description to explanation’, Academy

Management Review 24(4): 711–24.

Popper,

Karl. 1994. All Life is Problem Solving. New York: Routledge.

Powell, Walter and Paul DiMaggio. 1991. The New Institutionalism in Organizational Analysis. Chicago:

University of Chicago Press.

Pressman, J. and A. Wildavsky. 1979. Implementation: How Great Expectations in Washington are Dashed in

Oakland: Or, Why It’s Amazing that Federal Programs Work at All, This Being a Saga of the Economic

Development Administration as Told by Two Sympathetic Observers Who Seek to Build Morals on a Foundation

of Ruined Hopes. Berkeley, CA: University of California Press.

Rabinow, P. and W. Sullivan (eds). 1979 and 1987. Interpretive Social Science: A Reader. Berkeley: University of

California Press.

Rapley, John. 1996. Understanding Development: Theory and Practice in the Third World. Boulder, CO: Lynne

Rienner.

Ricoeur, Paul. 1976. Interpretation Theory: Discourse and the Surplus of Meaning. Texas: Christian University Press.

—— 1981. Hermeneutics and the Human Sciences: Essays on Language, Action and Interpretation, John B. Thompson

(ed. and trans.). Cambridge: Cambridge University Press.

——. 1984. Time and Narrative, vol. 1, K. McLaughlin and D. Pellaver (trans.). Chicago: Chicago University

Press.

—— 1992. Oneself as Another, Kathleen Blamey (trans.). Chicago: University of Chicago Press.

Rittel, Horst and Melvin Webber. 1973. ‘Dilemmas in a general theory of planning’ Policy Sciences, vol. 4.

Amsterdam: Elsevier Scientiﬁc, 155–69. (Reprinted in N. Cross (ed.) 1984. Developments in Design

Methodology, Chichester: John Wiley & Sons 135–44.

Roe, Emery M. 1989. ‘Narrative analysis for the policy analyst: A case study of the 1980–1982 medﬂy

controversy in California,’ Journal of Policy Analysis and Management 8: 251–73. doi:10.2307/3323382.

—— 1994. Narrative Policy Analysis: Theory and Practice. Durham, NC: Duke University Press.

Ryan, Marie-Laure. 2007. ‘Toward a deﬁnition of narrative’, in David Herman (ed.) The Cambridge

Companion to Narrative. Cambridge: Cambridge University Press, 22–35.

Sabatier, P.A. and H.C. Jenkins-Smith. 1993. Policy Change and Learning: An Advocacy Coalition Approach.

Boulder, CO: Westview Press.

Postpositivism and the policy process

111

Schelling, Thomas C. 1980. The Strategy of Conﬂict, reprinted, illustrated and revised. edn. Cambridge, MA:

Harvard University Press.

Schneider, Anne and Helen Ingram. 1997. Policy Design for Democracy . Lawrence, KS: University of Kansas.

Schön, Donald and Martin Rein. 1994. Frame Reﬂection: Toward the Resolution of Intractable Policy Controversies.

New York: Basic Books.

Schout, Adriaan, Andrew Jordan, and Michelle Twena. 2010. ‘From “old” to “new” governance in the

EU: Explaining a diagnostic deﬁcit’, West European Politics 33(1): 154–70.

Schram, Sanford and Philip T. Neisser. 1997. Tales of the State: Narrative in Contemporary U.S. Politics and

Public Policy. Langham, MD: Rowman & Littleﬁeld.

Scott, James C. 1998. Seeing like a State: How Certain Schemes to Improve the Human Condition Have Failed.

New Haven, CT: Yale University Press.

Scott, W. Richard. 1998. Organizations: Rational, Natural, and Open Systems. Englewood Cliﬀs, NJ:

Prentice Hall.

—— 2001. Institutions and Organizations. 2nd edn. Thousand Oaks, CA: Sage.

Selznick, Philip. 1949. TVA and the Grass Roots: A Study in the Sociology of Formal Organization. Berkeley:

University of California Press.

Sköldberg, Kaj. 1994. ‘Tales of change: Public administration reform and narrative mode’, Organization

Science 5: 219–38.

Stone, Deborah. 1997. Policy Paradox: The Art of Political Decision-Making. New York: Norton.

Susskind, Lawrence, Jennifer Thomas-Larmer and Sarah McKearnen. 1999. The Consensus Building Handbook:

A Comprehensive Guide to Reaching Agreement. Thousand Oaks, CA: Sage.

Throgmorton. J.A. 1991. ‘The rhetorics of policy analysis’, Policy Sciences 24: 153–79.

van Neumann, John and Oskar Morgenstern. 1944. Theory of Games and Economic Behavior. Princeton, NJ:

Princeton University Press.

Weber, Max. 1958. The Protestant Ethic and the Spirit of Capitalism. New York: Charles Scribner [1904].

Weick, K.E. 1995. Sensemaking in Organizations. Thousand Oaks, CA: Sage.

Wesselink, Anna and Jeroen Warner. 2010. Reframing Floods: Proposals and Politics, Nature and Culture 5(1):

1–14.

Wittgenstein, Ludwig. [c. 1922]

1961. Tractatus Logico-Philosophicus: Suivi

De Investigations Philosophiques.

Paris: Librairie Gallimard.

Yanow, Dvora. 1992. ‘Silences in public policy disclosure: Organizational and policy myths’, Journal of

Public Administration Research and Theory 2: 399–423.

—— 2001. Conducting Interpretive Policy Analysis. Thousand Oaks, CA: Sage.

—— 2007. ‘Interpretation in policy analysis: On methods and practice ’, Critical Policy Studies 1: 110–22.

Lejano

112

Part III

Modelling the policy process

Frameworks for analysis

The institutional analysis and

development framework

Ruth Schuyler House and Eduardo Araral Jr.

As a relatively young discipline and an intrinsically complex science, public policy has yet to con-

geal into a clear-cut school of thought. Rather, its borders are diﬀuse, both overlapping with

and comprising multiple “pure” disciplines of the social and natural sciences. Indeed, the com-

plexities of the open systems subject to study necessitate a multidisciplinary approach that pragma-

tically draws from the ﬁelds most germane to the particular issues at hand. Further, by drawing

from diﬀerent approaches, scholars can develop constructively competitive views of a context in

order to come to a more enlightened understanding about the potential eﬀects of public action.

Institutional scholars’ engagement with classical economics, political science, law, sociology,

psychology, natural science, and a host of other ﬁelds in the pursuit of understanding the policy

process has generated an incredibly rich and variant body of scholarship united by a common

focus on institutions and their roles (among other important factors) in shaping, constraining,

and enabling human behavior. By focusing on rule sets and the major physical and social constructs

characterizing a particular situation, institutional scholars seek to grapple with both macro-level

social structures and individual agency and also move beyond the market versus hierarchy division

to identify a more elemental set of components comprising any and all situations of human

interaction.

We believe that institutional analysis makes an important contribution to the study of public

policy for a number of reasons, in particular its established record of drawing from multiple

ﬁelds as discussed already. For one, institutional research over the past 25 years has advanced

Douglass North’s early challenge to progress economics from a static to dynamic theory of

human exchange (North 1990, p.107). Indeed, institutions have come to be commonly recog-

nized as a critical factor shaping political and economic behavior, but we have yet to fully

understand what types of institutions (if any) are requisite for the development and the

maintenance of functioning democracies, let alone how to manage and control processes of

“institution building” and institutional change. Moreover, institutional analysis—and more

speciﬁcally, the institutional analysis and development (IAD) framework developed largely by

Elinor Ostrom and her colleagues at the Workshop in Political Theory and Policy Analysis at

Indiana University—comes as close as any approach to oﬀering a framework that is generally

applicable to comparatively studying a wide variety of social situations. Finally, as entirely new

and sometimes blended structures of organization emerge in an increasingly complex, globalized

115

political economy, we need to utilize theoretical frameworks that are capable of encompassing

multitudinous patterns of human interaction.

Frameworks, theories, and models

Before delving into the IAD framework, we must clarify what a “framework” is, and how it

relates the many theories and models—each with their own constructs and vocabularies—used

to understand public policy. Ostrom herself oﬀers a clear explanation distinguishing between

theories, frameworks, and models (2007, p.25). Most broadly, a framework identiﬁes a set of

general variables and relationships that should be studied in order to understand a particular

phenomenon, but assigns no values to the variables and does not specify the direction of rela-

tionships between them. In essence, the framework is an organizing tool that allows the com-

parison of diﬀerent theories, many of which may be compatible with a particular framework.

According to Ostrom, frameworks “attempt to identify the universal elements that any theory

relevant to the same kind of phenomena would need to include” (2005, p.826).

A theory speciﬁes a more coherent set of assumptions about elements within the framework,

including their relationships with each other, their importance to answering particular kinds of

questions, and expectations about interactions and outcomes. For example, game theory, classical

economic theory, transaction cost theory, and a host of other theories have been extensively

used alongside the IAD framework.

Finally, a model is used to represent a speciﬁc situation and makes explicit assumptions about the

values of variables and the expected outcomes of a situation given a particular set of parameters (Ostrom

2007, p.26). Since there exist numerous ways in which a particular system of human organization or

governance might be studied and understood, the IAD framework becomes useful to engage in

comparative study across disciplines and draw nearer toward generalized understandings about the

creation, implementation, interaction, and outcomes of rules arrangement in diﬀerent contexts.

Institutions

Before jumping into the application of the IAD framework, we must ﬁrst come to a common

understanding of its primary focus: institutions. What, then, are institutions, and why do they

constitute an important unit of analysis? Perhaps the most commonly cited deﬁnition of institu-

tions is Douglas North’s: institutions are “the rules of the game of a society, or, more formally,

the humanly devised constraints that structure human interaction” (1990), including formal laws

and regulations and informal constraints, such as conventions and social norms. Emphasizing

their functionality, Menard and Shirley deﬁne institutions as the “rules, norms, and constraints

that humans devise to reduce uncertainty and control their environment” (2005). While scho-

lars of diﬀerent strands of institutionalism diﬀer in opinion about the human intentionality of

institutional forms, the rational choice or “actor-centered” institutionalists who most commonly

utilize the IAD framework see institutions as constructs intentionally devised by political or eco-

nomic actors to serve particular purposes. Ostrom describes rules as “shared prescriptions (must,

must not, or may) that are mutually understood and enforced in particular situations in a pre-

dictable way by agents responsible for monitoring conduct and for imposing sanctions” and

norms as “shared prescriptions known and accepted by most of the participants themselves involving

intrinsic costs and beneﬁts rather than material sanctions or inducements” (2005, p.824).

The varied use of multiple deﬁnitions of the term “institution” (for example, when used to

refer to physical buildings like hospitals or organizations like the UN as opposed to rule sets) is

one among many challenges of institutional research. Furthermore, institutions are invisible and are

House and Araral

116

thus very diﬃcult to describe, deﬁne, and measure. And quite often, the most inﬂuential rules

governing a system of interaction are the informal norms and conventions—the “rules-in-use ”—

rather than the “rules-in-form ” on formal record (Ostrom 2007, p.23). The eﬀects of particular

rules are also diﬃcult to study in isolation, as rules and norms consistently interact with the

constellation of other rules and exogenous factors that deﬁne a particular situation.

Finally, rules exist at multiple levels, and rules at any one level are embedded in the next

higher tier. The nestedness of lower levels in higher tiers is important to understanding how

rule changes (and thereby institutional change) at one level are inﬂuenced by rules and contextual

factors at other levels. Thus, institutional analysis becomes akin to untangling the relationships

between rules within and across levels. As Kiser and Ostrom point out, rules exist at the opera-

tional level that govern the “every day” activities and decisions of agents in an interactive situation.

Moving higher, there exists a set of collective choice level rules that constrain those at the opera-

tional level and determine how operational rules are set. And at the constitutional choice level, rules

that govern the process of rule-setting, including who may be involved and how, frame the

collective-choice and operational situations (Kiser and Ostrom 1982).

Oliver Williamson, the prominent transaction cost scholar, proposes a similar delineation of

institutional tiers, where human exchanges lie at the most immediate level. These transactions are

constrained by governance structures, which are in turn shaped by the “institutional environment”

set by property rights rules, the bureaucracy, legislative processes, etc. Finally, all of these rule struc-

tures are couched within rules at the level of embedded norms and traditions, which change

most slowly and are often studied via application of social theory (Williamson 1998). Despite

minor variations, the point is that most theorizing related to institutions incorporates the recurring

themes of nestedness, conﬁgurative relationships, and interrelatedness—characteristics that make

institutional analysis complex and prescriptions for institutional reform often quite tentative.

The IAD framework

Given the complexity of studying institutions, along with the observed diﬀerences in approa-

ches to understanding institutions, the Institutional Analysis and Development (IAD) framework

uniﬁes the multidisciplinary study of rules and norms in a common analytical framework com-

prising a number of key variables that deﬁne a particular action situation. The framework is applic-

able to both “snapshot” analyses as well as process-oriented studies of institutional change, where

feedback becomes an important part of the analysis. Furthermore, the framework allows the

researcher to focus on particular components of an action situation, for example the external

physical variables, the actors’ characteristics, or the rules-in-use. Ultimately, the IAD framework

is intended to allow researchers to model the strategies players adopt within particular rule conﬁg-

urations and to better understand how the mixture of participants, rules, strategies, community

attributes, and external variables, results in certain outcomes, and thereafter, how these outcomes

and participants’ responses to them pressure changes to the prevailing rule sets.

Like any good framework, the IAD framework maps out the broad variables applicable to

any institutional form (including markets, hierarchies, and collective action systems) that may be

studied by diﬀerent theories and models. These broad variables, which Ostrom terms action arenas,

are nested subassemblies of a broader, interrelated social structure.

Action arenas are comprised of two lower-level subassemblies: the actors and the action situa-

tion (Ostrom 2007, p.28). The actors and action situation interact and are simultaneously aﬀec-

ted by exogenous variables including physical conditions (e.g. the nature of a resource in a

common pool situation); community attributes (the social and cultural context); and the rules in

use, or institutional arrangements (ibid.; Ostrom et al. 1994).

Institutional analysis and development framework

117

The action arena

The action arena under study is “the social space where individuals interact” and is comprised of

the participants and their action situation (Ostrom 2005, p.829). Ostrom describes action situations

as collections of seven variables: ﬁ rst, the participants; second, their positions; third, the set of

potential outcomes; fourth, information available; ﬁfth, the costs and beneﬁts associated with the

set of outcomes; sixth, the degree of control participants have over choices and strategies, and

seventh, relationships between actions and outcomes (Ostrom 2005, p.828). Also germane to

describing the situation is the frequency and duration of interactions and whether the situation

is recurrent or singular.

The actors, be they conceptualized as individuals, loose groups, or formal organizations, exhibit

diﬀerent attributes aﬀecting their behaviors. The researcher must investigate and make assumptions

about participants’ preferences, goals, values, beliefs, and decision-making strategies (Ostrom 2007,

p. 30). The actors may be characterized according to values of four variables: the resources they

control; the values they assign to various actions and outcomes; the way they use information; and

the processes they use for selecting strategies (Ostrom 2007, p. 28). It is also important to note that,

in the tradition of institutional economics, neoclassical economics, assumptions of pure rationality and

full information are discarded and replaced by a conceptualization of individual actors as boundedly

rational (Simon 1957) and typically lacking full information about the situation at hand.

External variables

A set of three important external contextual variables structure the arena: the rules-in-use, the

physical conditions of the good under study and the surrounding context, and the characteristics

of the community of participants (Kiser and Ostrom 2000).

Figure 9.1 A framework for institutional analysis

Source: Ostrom (1999: 42)

House and Araral

118

Rules-in-use

Returning to the notion of rules-in-use, a critical component of institutional analysis is discovering

how the nested, interlinked rules-in-use that aﬀect potential payoﬀs and losses, demarcate the

set of allowable or feasible strategies, and inﬂuence actors’ behaviors. These rules are subject to

the limits of human cognition, understanding, and language, and can only be discovered through

rigorous research of the understood “do’s and don’ts”—rules that may, indeed, be quite diﬀerent

from or supersede the rules that exist on formal record (Ostrom 2005, p.832).

Which rules, then, are important to analyzing a particular action arena? Ostrom and Crawford

developed a typology of rules that should be speciﬁed in order to determine ﬁrst, their impact

on the action situation and actors, and second, the impact of one rule change upon another. In

cataloging the rules governing a particular situation and making note of their conﬁgurations,

scholars can thereafter better explain and predict the results of rule changes on the situation and

the likelihood of change itself. The set of important rules governing any situation includes

boundary rules that set who can participate and how easy is it for them to join or exit the arena;

position rules which assign positions of leadership, subservience, rank, etc.; authority rules which deﬁne

what actions participants in diﬀerent positions may or may not take; scope rules which restrict the

set of possible outcomes; aggregation rules which deﬁne the degree of control an actor has at a

decision point; information rules which aﬀect the availability of information for diﬀerent participants;

and payoﬀ rules which assign the costs and beneﬁts linked to diﬀerent actions. All of the many possible

conﬁgurations of these rule sets deﬁne diﬀerent institutions (Ostrom and Crawford 2005).

Beyond understanding the rules-in-use governing a situation, it is also important to understand

where the rules themselves originated, the history of their evolution, and the stability of their con-

ﬁguration. The stability of a rule set’s conﬁguration depends on multiple factors, including enfor-

cement capacity (Gibson et al. 2005; Ostrom 2005, p.833), the frequency of interactions, and the

degree to which rules are similarly understood and valued by participants (Ostrom 2005, p.833).

Attributes of the material context

Another critical factor that inﬂuences the action arena is the “state of the world” in the arena at

hand—the physical context, material conditions, and the type of good under study in the situation.

These conditions delimit what kinds of strategies and outcomes are at all possible and how particular

strategies relate to diﬀerent outcomes (Ostrom 2005, p.837). Furthermore, regarding the good

under study, it becomes important for the researcher to ask critical questions about the nature of

the good, including the ability to exclude beneﬁciaries and the “subtractability of the ﬂow”

(ibid.)—or the degree of rivalry, in neoclassical economic terms. These two attributes are used to

distinguish between public and private goods, common pool resources, and club goods, each of

which is associated with a set of assumptions and expectations related to their production and pro-

vision and the incentives of participants that utilize, control, trade, allocate, acquire, or produce

them. Additionally, it may be important to consider the size of the good, whether it can be stored

and transported, and how consistently it can be made available (Ostrom 2005, p.840).

Attributes of the community

The ﬁnal set of external variables that aﬀect the action arena is the attributes of the larger commu-

nity, sometimes referred to as the culture (Ostrom 2005, p.841). According to Ostrom, the

attributes that are important to shaping behaviors in the action arena include “the norms of behavior

generally accepted in the community, the level of common understanding potential participants

Institutional analysis and development framework

119

share about the structure of particular types of action arenas, the extent of homogeneity in the

preferences of those living in a community, and the distribution of resources among those aﬀected”

(Ostrom 2005, p.841). By speciﬁcally accounting for the role of culture and shared conventions,

the IAD framework incorporates lines of enquiry more typical of sociological and anthropological

research into understanding policy situations—a shift responsive to long-standing critique of the

strict positivist tradition in policy science.

Linking arenas and studying institutional change

In addition to studying the eﬀects of the aforementioned factors on a particular action situation,

it is also often important to uncover the linkages and relationships between factors in diﬀerent

action arenas. More speciﬁcally, the relationships between institutions at diﬀerent levels—the

operational, collective choice, and constitutional—are critical to processes of institutional change.

For the researcher, it is necessary to acknowledge that rules changes in one arena are constrained

by rules at a deeper level, and that changes in successively deeper levels are typically more costly

and diﬃcult to make (Ostrom 2005, p.842). Because there are multiple sources of institutional

change with networked eﬀects, sustained change in one component is diﬃcult without con-

current change in others (Saleth and Dinar 2004). Further, change is diﬃcult to coordinate and

depends not only on changing organization but also changing human behavior, the institutional

environment, and the institutions of governance at higher levels (Williamson 1998).

As for applying the framework to understand institutional change, exogenous variables are

assumed to be ﬁxed at the beginning of analysis, and resultant outcomes are thereafter judged

by participants according to criteria such as perceived eﬃ ciency, equity, adaptability, account-

ability, and conformance to morality and general social norms. If outcomes are unproductive,

deemed unfair, or are otherwise displeasing to the actors in the arena, those actors will strategize

to change the rule structure. In this way, the IAD framework may be applied to research the

iterative, ongoing nature of policy-making.

Applications of the IAD framework

A key test of a framework’s utility is the extent to which it has been applied to a variety of

settings and research programs. The IAD framework meets this test as it is now extensively

used in teaching, research, and policy advice worldwide and is arguably one of the most utilized

frameworks in the US and Western Europe (Sabatier 2007) as well as in East, Southeast and

South Asia.

For instance, in teaching, dozens of graduate programs in public policy, political science,

economics, sociology, and ecology around the world use the IAD framework or elements of it

used in their curriculum. Examples include universities in Asia (Tsinghua University in China, the

National University of Singapore, Hong Kong University, Seoul National University, China

University of Political Science and Law, Asian Institute of Technology), Europe (Humboldt Uni-

versity, Stockholm School of Economics, Delft University Technology (Netherlands), Erasmus

University), North America (Indiana University, Arizona State University, University of Col-

orado, Rice University, University of California, Carleton University, UNAM (Mexico), Duke

University, University of Michigan, University of Nevada,) among others.

In terms of empirical and experimental research, the IAD framework has been extensively

used in a wide range of research programs. These include the commons research agenda such as

ﬁsheries (Basurto 2008; de Castro 2000), irrigation (Bruns 2009; Araral 2006); ground water

(Steed 2010; Wang 2006), forestry (Andersson 2003; Hayes 2007; Pacheco 2007; Persha 2008;

House and Araral

120

Jagger 2009; Marquez 2011); watersheds and river basins (Myint 2005; Heikilla et al. 2011); and

land use (Donnelly 2009).

The IAD framework has also been used not just for common pool natural resources but also

research on knowledge as commons (Hess and Ostrom 2006); infrastructure in developing

countries (Schroeder et al. 1993); property rights (Kauneckis 2005; Mwangi 2003); foreign aid

and bureaucratic incentives (Araral 2008); federalism and constitutional order (V. Ostrom 1991;

Jillson and Wilson 1994; Wang 2006; Oyerinde 2006); local public economies (Oakerson and

Parks 2011); public housing (Choe 1992); nongovernmental organizations (NGOs) and non-

proﬁts (Bushouse 1999; Elliott-Teague 2007; Sabet 2005); global commons (McGinnis and

Ostrom 1996; Dolsak 2000; Allen 2005) as well as collective action problems and social

dilemmas (Dudley 1993; Ahn 2001; Coleman 2009; Araral 2009; van Laerhoven 2008).

Finally, in terms of policy advice, the IAD framework has been used in diverse policy settings such

as peace and nation-building in Liberia (Sawyer 2005), foreign aid (India, Kenya) (Gibson et al. 2005);

health care (United States), forestry policy in Bolivia (Andersson 2003), irrigation in the Philippines

(Araral 2006), among others.

Conclusion

The rich body of research used to develop the IAD framework alludes to the complexity of

institutional analysis and the problems associated with prescribing institutional reform with-

out a full study of the interrelations of the rules at work, particular social contexts, goods in

question, and underlying social norms. Nevertheless, the diverse community of institutional

scholars contributing to this line of enquiry has contributed to the growing sophistication of

institutional analysis. With luck, continued advancements in understanding the multiple rules at

play and the relationships between institutional arrangements and economic growth will allow

for more appropriate reform prescriptions, greater appreciation of the limitations of outside

intervention, and realistic expectations about changing institutional arrangements and growth

patterns.

While it is increasingly accepted that there exist no “right” institutions for any and all con-

texts, Ostrom’s and her colleagues’ studies of successful and failed institutions via application of

the IAD framework have allowed her to identify some general design principles including ﬁrst,

a clear deﬁnition of group boundaries; second, rules that are matched to local conditions; third,

active participation of individuals aﬀected by rules; fourth, respect of community rule-making

rights by external authorities; ﬁfth, monitoring systems managed by community members; sixth,

established sanction systems; seventh, low-cost conﬂict resolution mechanisms; and eighth nested

systems of governance in larger systems (Ostrom 1991). Further, increasing emphasis on the

importance of informal institutions should aﬀord policy-makers a healthy dose of modesty when

designing and implementing “institution-building” programs, for as North warns, policy tools

are “blunt instruments.” We are often only able to signiﬁcantly change formal rules, while we

can exert very limited control over informal institutions (North 2008).

Finally, Robert Putnam reminds us of the ultimate goal of institutions—that of achieving

purpose, not just agreement. Institutions shape politics by way of the rules and standard

operating procedures employed to reach political outcomes. And these institutions are inﬂu-

enced by their particular historical and social contexts (Putnam et al. 1993). So then, under-

standing the collective goals that motivate institutional formation, how those institutions shape

social action and value systems, and again how those social goals lead to further institutional

change allow for a process-oriented, evolutionary understanding of economic growth and social

organization.

Institutional analysis and development framework

121

Bibliography

Ahn, T.K. 2001. ‘Foundations for cooperation in social dilemmas’ PhD dissertation, Bloomington, IN:

Indiana University.

Allen, Linda. 2005. ‘The politics of structural choice of the commission for environmental cooperation:

The theoretical foundations of the design of international environmental institutions’, PhD dissertation,

Bloomington, IN: Indiana University.

Andersson, Krister P. 2002. ‘Can decentralization save Bolivia’s forests? An institutional analysis of municipal

forest governance’, unpublished dissertation, Bloomington, IN: Indiana University.

—— 2003. ‘What motivates municipal governments? Uncovering the institutional incentives for municipal

governance of forest resources in Bolivia’, Journal of Environment and Development 12(1): 5–27.

Araral, Eduardo K. 2006. ‘Decentralization puzzles: A political economy analysis of irrigation reform in the

Philippines’, PhD dissertation, Bloomington, IN: Indiana University.

—— 2008. ‘The strategic games that donors and bureaucrats play: An institutional rational choice analysis’,

JPART 19: 853–871.

—— 2009. ‘What explains collective action in the commons? Theory and evidence from the Phillipines’,

World Development vol. 37, no. 3, 687–697.

Basurto, Xavier. 2008. ‘Biological and ecological mechanisms supporting marine self-governance: The Seri

Callo de Hacha ﬁshery in Mexico’, Ecology and Society 13(2): 20.

Bruns, Bryan Rudolph. 2009. ‘Solving Samaritan’s dilemmas in irrigation investment’. Available at: www.

indiana.edu/~workshop/colloquia/ materials/papers/Bruns-Solving%20Sama ritan’s%20Dilemmas%20in%20Ir

rigation%20DRAFT090925.pdf

Bushouse, Brenda K. 1999. ‘The mixed economy of child care: An institutional analysis of nonproﬁt, for

proﬁt, and public enterprises’, PhD dissertation, Bloomington, IN: Indiana University.

Choe, Jaesong. 1992. ‘The organization of urban common-property institutions: The case of apartment

communities in Seoul,’ PhD dissertation, Bloomington, IN: Indiana University.

Colfer, C.J.P. and R.G. Dudley. 1993. Shifting Cultivators of Indonesia: Marauders or Managers of the Forest.

Rice Production and Forest use among the Uma’ Jalan of East Kalimantan. Rome: Food and Agriculture

Organization

of the United

Nations.

Coleman, Eric. 2009. ‘Essays on the eﬀects of institutions and trust on collective action’, PhD dissertation,

Bloomington, IN: Indiana University.

Croissant, Cynthia. 2005. ‘Uses and landscape patterns: A study of relationships between human activities

and spatial patterns of land use and land cover on private parcels in Monroe County, Indiana’, PhD

dissertation, Bloomington, IN: Indiana University.

De Castro, Fabio. 2000. ‘Fishing accords: the political ecology of ﬁshing intensiﬁcation in the Amazon’,

PhD dissertation, Bloomington, IN: Indiana University.

Dolsak, Nives. 2000. ‘Marketable permits: Managing local, regional, and global commons’, PhD dissertation,

Bloomington, IN: Indiana University.

Donnelly, Shanon. 2009. ‘Linking land use, land cover, and land ownership at the parcel scale in the

midwest United States’, PhD dissertation, Bloomington, IN: Indiana University.

Elliott-Teague, Ginger. 2007. ‘NGOs in policymaking in Tanzania: The relationships of group characteristics,

political participation and policy outcomes’, PhD dissertation, Bloomington, IN: Indiana University.

Gibson, C.C., J.T. Williams and E. Ostrom. 2005. ‘Local enforcement and better forests’, World Development

33(2): 273–84.

——, Krister Andersson, Elinor Ostrom and Sujai Shivakumar. 2005. The Samaritan’s Dilemma: The Political

Economy of Development Aid. Oxford: Oxford University Press.

Hayes, Tanya. 2007. ‘Forest governance in a frontier: An analysis of the dynamic interplay between

property rights, land-use norms, and agricultural expansion in the Mosquitia forest corridor of Honduras

and Nicaragua’, PhD dissertation, Bloomington, IN: Indiana University.

Heikkila, Tanya, Edella Schlager and Mark W. Davis. 2011. ‘The role of cross-scale institutional linkages

in common pool resource management: Assessing interstate river compact’, Policy Studies Journal 39(1):

121–45.

Hess, C. and E. Ostrom. 2006. Understanding Knowledge as a Commons. Cambridge, MA: MIT Press.

Jagger, Pamela. 2009. ‘Can forest sector devolution improve rural livelihoods? An analysis of forest income

and institutions in Western Uganda’, PhD dissertation, Bloomington, IN: Indiana University.

Jillson, Calvin C. and Rick K. Wilson. 1994. Congressional Dynamics: Structure, Coordination, and Choice in

the First American Congress, 1774–1789. Stanford, CA: Stanford University Press.

House and Araral

122

Kauneckis, Derek. 2005. ‘The co-production of property rights: Theory and evidence from a mixed-right

system in southern Mexico’, PhD dissertation, Bloomington, IN: Indiana University.

Kiser, L. and E. Ostrom. 1982. ‘The three worlds of action: A metatheoretical synthesis of institutional

approaches’, in E. Ostrom (ed.) Strategies of Political Inquiry. Thousand Oaks, CA: Sage, 179–222.

——and E. Ostrom. 2000. ‘The three worlds of action: A metatheoretical synthesis of institutional

approaches’, in Robert D. Putnam and Robert Leonardi (eds) Making Democracy Work. New York:

Greenwood Publishing Grou.

McGinnis, Michael and Elinor Ostrom. 1996. ‘Design principles for local and global commons’, in Oran

R. Young (ed.) The International Political Economy and International Institutions, Vol.2. Cheltenham:

Edward Elgar.

Marquez, Lilian. 2011. ‘The eﬀect of institutions on Guatemalan forests: Conceptual, methodological and

practical implications’, PhD dissertation, Bloomington, IN: Indiana University.

Menard, C. and M.M. Shirley. 2005. Handbook of New Institutional Economics. Dordrecht: Kluwer.

Mwangi, Esther. 2003. ‘Institutional change and politics: The transformation of property rights in Kenya’s

Maasailand’, PhD dissertation, Bloomington, IN: Indiana University.

Myint, Tun. 2005. ‘Strength of “weak” forces in multilayer environmental governance: Cases from the

Mekong and Rhine River basins’, PhD dissertation, Bloomington, IN: Indiana University.

North, D.C. 1990. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge University

Press.

—— 1996. ‘Institutions, organizations and market competition’, Economic History 9612005, EconWPA.

—— 2005. ‘Institutions and the performance of economies over time’, in C. Ménard and M. Shirley, (eds)

Handbook of New Institutional Economics. Dordrecht: Springer, 21–30.

Oakerson, R. and Parks, R. 2011. ‘The study of local public economies: Multi-organizational, multi-level

institutional analysis and development’, Policy Studies Journal, vol. 39, 1(february): 147–167.

Ostrom, E. 1991. Governing the Commons: The Evolution of Institutions for Collective Action Cambridge:

Cambridge University Press.

—— 2005. ‘Doing

institutional analysis digging

deeper than markets and hierarchies’, Handbook of New

Institutional Economics, Dordrecht: Springer, pp. 819–48.

—— 2007. ‘Institutional rational choice: An assessment of the institutional analysis and development fra-

mework’,inTheories of the Policy Process, 2nd edition, edited by P.A. Sabatier, Boulder, CO: Westview

Press, 21–34.

——, Larry Schroeder and Susan Wynne. 1993. Institutional Incentives and Sustainable Development: Infrastructure

Policies in Perspective. Boulder, CO: Westview Press.

——, R. Gardner and J. Walker. 1994. Rules, Games, and Common-pool Resources. Ann Arbor, MI: University

of Michigan Press.

——and S. Crawford. 2005. Classifying Rules. Understanding Institutional Diversity. Princeton, NJ: Princeton

University Press, 187–215.

Ostrom, V. 1991. The Meaning of American Federalism: Constituting a Self-Governing Society, San Francisco,

CA: Institute for Contemporary Studies Press.

Oyerinde, Oyebade Kunle. 2006. ‘ The constitution of order among the Yoruba of Nigeria’ PhD dissertation,

Bloomington, IN: Indiana University.

Pacheco, Diego. 2007. ‘An institutional analysis of decentralization and indigenous timber management

in common-property forests of Bolivia’s lowlands’, PhD dissertation, Bloomington, IN: Indiana

University.

Persha, Lauren. 2008. ‘Decentralized forest management, anthropogenic disturbance patterns and forest

change in the Usambara Mountains, Tanzania’, PhD dissertation, Bloomington, IN: Indiana University.

Putnam, Robert D. and Robert Leonardi. 2002. Making Democracy Work. New York: Greenwood Publishing

Group.

Reiners, Derek. 2006. ‘Institutional eﬀects on decision-making and performance in public land agencies:

The case of wildﬁre in the interior west of the United States’, PhD dissertation, Bloomington, IN:

Indiana University.

Sabatier, Paul A. (ed.). 2007. Theories of the Policy Process. Boulder, CO: Westview Press.

Sabet, Daniel M. 2005. ‘Thickening civil society: nonpro ﬁt organizations and problems of water and

sanitation along Mexico’s northern Border’, PhD dissertation, Bloomington, IN: Indiana University.

Saleth, R.M. and A. Dinar. 2004. The Institutional Economics of Water: A Cross-country Analysis of Institutions

and Performance. Cheltenham: Edward Elgar.

Sawyer, A. 2005. Beyond Plunder: Toward Democratic Governance in Liberia .

Institutional analysis and development framework

123

Schoon, Michael. 2008. ‘Building robustness to disturbance: Governance in Southern African Peace Parks’,

submitted to the faculty of the University Graduate School in partial fulﬁllment of the requirements for

the joint degree Doctor of Philosophy in the School of Public and Environmental Aﬀairs and the

Department of Political Science, Bloomington, IN: Indiana University.

Simon, H.A. 1957. Models of Man, Social and Rational: Mathematical Essays on Rational Human Behavior in a

Social Setting. New York: John Wiley.

Smith, Ronald S. 2006. ‘Discerning diﬀerences in social capital: the signiﬁcance of interpersonal network

and neighborhood association structure on citizen participation’, PhD dissertation, Bloomington, IN:

Indiana University.

Steed B. 2010. ‘Natural forces, human choices: an over time analysis of responses to ecological and human

induced disturbances in southern California water basin governance’. PhD dissertation, Bloomington,

IN: Indiana University.

van Laerhoven, Frank. 2008. ‘Local governance and the challenge of solving collective action dilemmas’,

PhD dissertation, Bloomington, IN: Indiana University.

Wang, Jianxun. 2006. ‘Political economy of village governance in contemporary China’, PhD dissertation,

Bloomington, IN: Indiana University.

Williamson, O.E. 1998. ‘Transaction cost economics: How it works: where it is headed’, The Economist

146(1): 23–58.

House and Araral

124

The advocacy coalition framework

Coalitions, learning and policy change

Christopher M. Weible and Daniel Nohrstedt

Introduction

Policy process research is the study of public policy over time and the surrounding actors, con-

texts, and events. The formal academic study of policy processes began in the 1950s and 1960s,

led by the likes of Lerner and Lasswell (1951), Freeman (1955), Simon (1957), Lindblom (1959),

Dawson and Robinson (1963), Easton (1965), Ranney (1968), and Walker (1969). Policy pro-

cess research sprung, in part, from dissatisfaction with political scientists’ focus on governing

institutions (courts, legislatures, and executives), a theoretical desire to understand broader

political systems, optimism following successes of the social sciences during World War II, and a

practical desire to beneﬁt society.

Since the 1960s, a number of complementary research programs have emerged for describing

and explaining various aspects of policy processes (e.g. Sabatier 1999, 2007). Among these

research programs is the advocacy coalition framework (ACF) created by Paul Sabatier and

Hank Jenkins-Smith in the 1980s (Sabatier 1987, 1988; Jenkins-Smith 1990; Sabatier and Jenkins-

Smith 1993, 1999).

Sabatier and Jenkins-Smith established the ACF in response to several

perceived shortcomings in policy process research: a dissatisfaction with the policy cycle or stages

heuristic as a causal theory; a need to take more seriously the role of scientiﬁc and technical

information in policy processes; dissatisfaction with the top-down and bottom-up perspectives

of the implementation literature; a need to take a long-term time perspective to understand

policy processes; and a need to develop theories that assume more realistic human agents other

than the rational actor models found in microeconomics.

The ACF resulted as an amalgam inspired partly from previous studies of issue networks (Heclo

1978), implementation (Pressman and Wildavsky 1973; Mazmanian and Sabatier 1981; Hjern

and Porter 1981), learning (Heclo 1974; Weiss 1977), policy subsystems (Griﬃth 1939; Freeman

1955), belief systems (Ajzen and Fishbein 1980; Putnam 1976; Peﬄey and Hurwitz 1985; Hurwitz

and Peﬄey 1987), scientiﬁc and technical information in policy debates (Mazur 1981; Jenkins-

Smith 1990), and a model of the individual based on bounded rationality and cognitive ﬁlters

(Simon 1957). The ACF serves as an analytical guide for answering questions principally about

advocacy coalitions, policy-oriented learning, and policy change.

125

The framework

Following Laudan (1977: 70–120) and Ostrom (2005: 27–9), frameworks serve as a platform for

groups of scholars to work together toward common understandings and explanations of phe-

nomena. Frameworks provide assumptions, specify the scope of inquiry, and establish conceptual

categories with basic de ﬁnitions and general relations. Frameworks support the development

and testing of theory that, in turn, narrow the scope of inquiry, oﬀer testable hypotheses, and

postulate causal relationships among concepts. This section introduces the ACF as an actual

“framework” by describing its assumptions, scope, and the basic concepts and the general relations

among them. The next sections describe the major theoretical emphases supported by the ACF.

Assumptions

Sabatier and Jenkins-Smith based the ACF on several foundational assumptions (1993: 17–20):

The policy subsystem is the primary unit of analysis for understanding policy processes. Policy sub-

systems are deﬁned by a substantive topic and territorial domain along with a set of people actively

involved in shaping subsystem aﬀairs. Subsystems are simultaneously semi-autonomous while also

nested and interdependent (Sabatier 1998; Nohrstedt and Weible 2010). For example, a water policy

subsystem at local level will likely be nested in a regional policy subsystem which is nested

within a national policy subsystem. This same water policy subsystem at a local level overlaps to

various extents with other subsystems (e.g. a local-level transportation policy subsystem). To

attract the attention and involvement of actors, policy subsystems usually entail some degree of

authority or potential for authority to alter behavior and shape subsystem outcomes.

The set of relevant subsystem actors include any person attempting to inﬂuence subsystem aﬀairs. As

policy process research developed in the 1950s and 1960s, the emphasis was on symbiotic rela-

tions between legislative committees, interest groups, and the government agencies (what has

been referred to as the “iron triangle”). The ACF broadens the set of subsystem actors essen-

tially to include any person attempting to inﬂuence subsystem aﬀairs, including government

oﬃcials, members of the private sector and nonproﬁts, scientists and consultants, and members

of the media.

Individuals are boundedly rational with limited ability to process stimuli, motivated by belief systems,

and prone to experience the “devil shift.” The actors involved in policy subsystems are boundedly

rational, meaning they are goal-oriented but hampered by their cognitive abilities to process

stimuli (Simon 1985). To overcome these limitations, actors simplify the world through hierarchical

belief systems consisting of normative deep core beliefs, subsystem speciﬁc policy core beliefs, and

narrow secondary beliefs (Putnam 1976; Peﬄey and Hurwitz 1985; see Sabatier and Jenkins-Smith

1999: 133). These belief systems are used as heuristics to ﬁlter and interpret stimuli (Lord et al.

1979; Munro et al. 1997; Munro et al. 2002). Furthermore, the ACF borrows from prospect theory

to assume that actors remember losses more than gains (Quattrone and Tversky 1988). The

result is what the ACF calls the “devil shift,” which is the tendency for individuals to exaggerate

both the power and maliciousness of their opponents (Sabatier et al. 1987).

Subsystems are simpliﬁed by aggregating actors into one or more coalitions. Given the large number

and diversity of actors involved in subsystem aﬀairs, there is a need to simplify while not overly dis-

torting reality. The ACF directs researchers toward aggregating actors into one or more coalitions.

These coalitions are deﬁned by their shared policy core beliefs and coordination patterns.

Policies and programs incorporate implicit theories and assumptions reﬂecting the translated beliefs of one

or more coalitions. Subsystems consist of boundedly rational actors working together in coalitions

toward shaping policy outputs and outcomes. Given the importance of beliefs systems, the ACF

Weible and Nohrstedt

126

assumes that policy outputs are merely the translations of beliefs of the winning coalition or

coalitions (Pressman and Wildavsky 1973).

Scientiﬁc and technical information is important for understanding subsystem aﬀairs. Most subsystems

involve issues that are usually diﬃcult to describe and explain in terms of problem seriousness,

causes, and potential impacts of proposals. Scientiﬁc and technical information, thus, becomes

an important resource for coalition members for a variety of uses from argumentation with

opponents to the mobilization of supporters.

Researchers should adopt a longterm time perspective (e.g, of 10 years or more) to understand policy

processes and change. Policy process research is largely the study of public policy over time. The ACF

recommends that researchers examine their phenomenon as a point in the ongoing develop-

ment of the policy subsystem. Part of the rationale for taking a long-term perspective is to allow

for ample time to evaluate the outcomes of any policy decision, to assess political reactions, and

to study learning within and between coalitions.

Scope

A framework should provide a scope of inquiry for guiding scholars toward a shared research

agenda. The traditional foci of the ACF are descriptions and explanations of advocacy coalitions,

learning within and among coalition allies and opponents, and policy change. These emphases

should be viewed as general categories for both research inquiry and theory development.

Certainly other areas of research are permitted and have been conducted (e.g. Montpetit 2011;

Shanahan et al. 2011).

General conceptual categories and relations

A framework provides a common vocabulary including the major conceptual categories and

general relations among them. Figure 10.1 shows the ACF ﬂow diagram (adapted from Sabatier

and Weible 2007) featuring a policy subsystem with coalitions, their resources and beliefs, and

the policy outputs and outcomes. Policy subsystems are embedded in a broader political system

with relatively stable parameters and external events, where the latter is more prone to change

than the former.

For researchers wanting to apply the ACF, it is important to consider the following basic

concepts: (1) public policy topic that deﬁnes the subsystem; (2) actors who are involved in

subsystem aﬀairs; (3) institutions (e.g. rules) that structure overall subsystem interactions and the

behaviors within particular venues; (4) relative stable parameters and external events; (5) inter-

dependencies with other subsystems; and, ﬁnally, (6) time to permit observations of coalition

behavior, learning, and policy change.

The next step is to articulate the theoretical emphases within the ACF. Theories serve diﬀerent

functions than frameworks in a research program. Theories narrow the scope of inquiry, link con-

cepts usually in the form of expectations, propositions, or observable implications, and establish

rationales (causal mechanisms) that explain how concepts interrelate. There are three major theoretical

emphases of the ACF.

First theoretical emphasis: advocacy coalitions

The ﬁrst area of theoretical emphasis focuses on advocacy coalitions. Advocacy coalitions are

deﬁned as groups of actors sharing policy core beliefs and coordinating their behavior in a non-

trivial manner. Advocacy coalitions emerge and persist because actors vary in their belief systems

The advocacy coalition framework

127

(e.g. normative values regarding a particular policy topic) and seek to form alliances to translate

their beliefs into actual policies before actors with diﬀerent belief systems can do the same.

The ACF oﬀers three reasons that permit subsystem actors to overcome threats to collective

action in mobilizing coalition activity (Zafonte and Sabatier 1998; Sabatier and Weible 2007: 197).

First, similar beliefs among coalition allies reduce transaction costs of coordination. Second, the

level of coordination varies with some actors engaging in weak coordination (modifying their

behavior to achieve shared goals or information exchange) and others in strong coordination (jointly

developing and executing action plans) (Fenger and Klok 2001). Third, in high conﬂict situations,

the “devil shift” occurs when actors exacerbate the power and maliciousness of opponents. When

under the devil shift, coalition members will overestimate the threats from shared opponents and

the imposed costs of inaction.

Empirical research across subsystems and political systems have conﬁrmed that policy subsystems

are regularly composed of one to ﬁve advocacy coalitions ﬁghting over resources, access to

venues, and inﬂuence on the formation of public policy (Weible et al. 2009). Occasionally, these

coalitions interact competitively showing intervals, sometimes spanning decades, of one-upmanship.

In other situations a single, “dominant” coalition will largely control a subsystem for extended

periods of time over a largely nonexistent, inactive, or disorganized opposition. Key descriptive

questions regarding coalition formation and development include:

 What is the structure of the networks and belief systems of advocacy coalition members?

 How stable is coalition membership over time?

 What is the role of diﬀerent organizations within coalitions?

Figure 10.1 Flow diagram of the advocacy coalition framework

Source: Adapted from Sabatier and Weible (2007)

Weible and Nohrstedt

128

 How much consensus is there among coalition members?

 What are the patterns of coordination among subsystem actors?

 What strategies and resources do coalitions use to achieve their policy goals?

Within the scope of the emphasis on coalitions are also a set of explanatory questions addressing

the formation of coalitions and the interaction and behavior of coalition members:

 Why do coalitions form?

 How do coalition members overcome threats to collective action?

 Why do coalition allies negotiate with opponents?

These questions are usually pursued in the form of testing and developing hypotheses. The most

important hypotheses aiming at the composition and development of advocacy coalition include:

Hyphothesis 1 On major controversies within a policy subsystem when policy core beliefs

are in dispute, the lineup of allies and opponents tends to be rather stable

over periods of a decade or so.

Hypothesis 2 Actors within an advocacy coalition will show substantial consensus on issues

pertaining to the policy core beliefs, although less so on secondary aspects.

Hypothesis 3 An actor (or coalition) will give up secondary aspects of his or her (its) belief

system before acknowledging weaknesses in the policy core beliefs.

Hypothesis 4 Within a coalition, administrative agencies will usually advocate more moderate

positions than their interest group allies.

Hypothesis 5 Elites of purposive groups are more constrained in their expression of beliefs

and policy positions than elites from material groups.

(Sabatier and Weible 2007: 220)

Hypotheses 1–5 specify some of the fundamental components of the ACF, including the

hierarchy and stability of belief systems and the distinction between diﬀerent types of organi-

zations. Important contributions in the study of coordination networks and belief stability and

change within coalitions can be found in Jenkins-Smith et al. (1991), Zafonte and Sabatier (1998),

Weible and Sabatier (2005), Weible (2005), Ingold (2011), Henry et al. (2010), Henry (2011), Matti

and Sandström (2011), and Pierce (2011). Previous empirical research provides mixed results

regarding these hypotheses by showing patterns of both stability and defection of coalition

members over time (Jenkins-Smith and St. Clair 1993; Jenkins-Smith et al. 1991; Sabatier and

Brasher 1993; Zafonte and Sabatier 2004; Andersson 1999; Munro 1993).

While coalition resources have been a regular feature within the ACF since its inception

(Sabatier and Jenkins-Smith, 1993: 29), Sabatier and Weible (2007: 201-4) and Weible (2007)

placed more emphasis on the area by oﬀering a typology of coalition resources: (1) formal legal

authority to make policy decisions; (2) public opinion; (3) information; (4) mobilizable troops;

(5) ﬁnancial resources; and (6) skillful leadership. The access and use of these resources is theoretically

argued to be important for forming and maintaining coalitions and for policy change. However,

the empirical investigation of coalition resources has only recently been examined by scholars

(Nohrstedt 2011; Ingold 2011; Albright, 2011).

The formation and stability of advocacy coalitions over time and their political behavior is

conditioned by subsystem institutions and events outside the control of subsystem actors. The

most important institutional factors that constrain subsystem actors are ﬁrst, the openness of the

political system and second, norms of consensus, which aﬀect the level of inclusiveness of

The advocacy coalition framework

129

coalitions, exchange of information across coalition boundaries, and access to policy venues

(Sabatier and Weible 2007). Institutions shape the level of coordination within a coalition, pri-

marily by setting legal impediments that constrain formation of formal alliances paving the way

for weaker forms of coordination as a viable political strategy (Zafonte and Sabatier 1998;

Fenger and Klok 2001; Nohrstedt 2010). Events are likely to aﬀect coalition behavior, parti-

cularly by providing opportunities for exploitation of new resources (including mobilization of

new coalition members) and strategy in terms of venue exploitation.

Second theoretical emphasis: policy-oriented learning

Learning has been a central focus of the ACF since its inception (Sabatier 1988). Policy-oriented

learning is deﬁned as “enduring alterations of thought or behavioral intentions that result from

experience and which are concerned with the attainment or revision of the precepts of the

belief system of individuals or of collectives” (Sabatier and Jenkins-Smith 1993: 42–56). To

study policy-oriented learning is to study changes in belief system components of coalition

members over time. Policy-oriented learning may entail better understanding of political goals,

the causal relationship among key factors in the subsystem, and eﬀective strategic behaviors,

especially as used in analytic debates.

Important descriptive questions that guide inquiry into learning are the following:

 What belief system components are changing through learning?

 To what extent is one coalition learning more than another coalition?

Explanatory questions deal with issues of when, why, and how actors learn within and between

coalitions and include:

 What contexts and events foster learning by coalition members?

 How does learning diﬀuse among allies within a coalition?

 What contexts and events foster learning by brokers?

 To what extent, if at all, does a broker facilitate learning between coalitions?

 Why does learning occur, if at all, between some members of opposing coalitions and not

others?

On the whole, the intent of the ACF’ s focus on policy-oriented learning is to understand and

explain what constitutes learning and why learning occurs within coalitions and between coa-

litions. Four general factors are important for explaining policy-oriented learning. The ﬁrst

involves the attributes of professional forums (see Jenkins-Smith 1990). Professional forums are

venues of discussion involving subsystem actors. Forums are structured by diﬀerent sets of

institutional arrangements. Some forums are structured by open participation rules (open forum)

and others by closed participation rules (closed forum). Professional forums are based on common

analytical training and norms and are postulated to increase the likelihood for learning between

coalitions (ibid.).

The second variable conditioning learning is the level of conﬂict between coalitions (Jenkins-

Smith 1990; Weible 2008). The expectation is that there is an inverted quadratic relationship

between conﬂict and cross-coalition learning. On one extreme at low levels of conﬂict, there is

little cross-coalition learning as the opposing coalitions place their attention on other, more

pressing issues. On the other extreme at high levels of conﬂict, there is also little cross-coalition

learning as opposing coalitions are motivated to defend their positions and refute claims by their

Weible and Nohrstedt

130

opponents. The conditions most conducive for cross-coalition learning would then be inter-

mediate levels of conﬂict where there is enough of a threat to attract the attention of rivals but

not too much of a threat to entrench opponents in rigid policy positions.

The third variable is attributes of the stimuli or data prompting learning. The stimuli can

come from many diﬀerent sources including scientiﬁc and technical information, actors, and

events. One important attribute of stimuli relates to the level of analytical tractability of the

phenomenon (Jenkins-Smith 1990: 97–9). Subsystems with highly intractable issues are marked

by uncertainty and high levels of disagreement about the scientiﬁc and technical aspects of the

issue (e.g. about the quality of the data or theories used). In general, highly intractable issues are

expected to be associated with lower levels of cross-coalition learning.

The fourth variable that conditions learning involves attributes of individuals, including belief

systems, resources, and network contacts. Are actors with extreme beliefs less likely to learn

from their opponents than actors with moderate beliefs? Are some actors more or less supported

by their analytical training in learning than others? Are actors with network contacts—either

within their coalitions or with opposing coalitions—more likely to learn?

Central to the arguments about cross-coalition learning is a certain type of individual: the

policy brokers. Sabatier and Jenkins-Smith (1993: 27) deﬁne brokers as “a category of actors …

whose dominant concerns are keeping with the level of political conﬂict within acceptable

limits and reaching some ‘reasonable’ solution to the problem.” Sabatier and Jenkins-Smith argue

that identifying brokers is an empirical question and is not dependent upon organizational

aﬃliation. These broker may learn, possibly prior to the coalitions, and then facilitate cross-coalition

learning later in the process (Sabatier and Jenkins-Smith 1993: 218–19).

Given the research questions and categories above, the theoretical emphasis on policy-oriented

learning within the ACF is built into ﬁve hypotheses:

Hypothesis 1 Policy-oriented learning across belief systems is most likely when there is an

intermediate level of informed conﬂict between the two coalitions. This

requires that: (a) each have the technical resources to engage in such a debate;

and that (b) the conﬂict be between secondary aspects of one belief system and

core elements of the other or, alternatively, between important secondary

aspects of the two belief systems.

Hypothesis 2 Policy-oriented learning across belief systems is most likely when there exists

a forum which is: (a) prestigious enough to force professionals from diﬀerent

coalitions to participate; and (b) dominated by professional norms.

Hypothesis 3 Problems for which accepted quantitative data and theory exist are more

conducive to policy-oriented learning across belief systems than those in which

data and theory are generally qualitative, quite subjective, or altogether lacking.

Hypothesis 4 Problems involving natural systems are more conducive to policy-oriented

learning across belief systems than those involving purely social or political

systems because in the former many of the critical variables are not themselves

active strategists and because controlled experimentation is more feasible.

Hypothesis 5 Even when the accumulation of technical information does not change the

views of the opposing coalition, it can have important impacts on policy – at

least in the short run – by altering the views of policy brokers.

A review of ACF applications in Weible et al. (2009) indicates that, of the ﬁve learning

hypotheses, the ﬁrst two have been tested the most and the last two the least. From this review,

they found evidence suggesting that learning within coalitions reinforces beliefs and mixed

The advocacy coalition framework

131

results about the likelihood of cross-coalition learning. For example, Larsen et al. (2006) ﬁnd

that cross-coalition learning occurs at the secondary level of belief systems (as expected) but also

at the policy-core level (not expected). Similar mixed results can be found with the measure-

ment and ﬁndings involving professional forums. If there is any area within the ACF needing

more attention at both the theoretical and operational levels, it is policy-oriented learning.

Third theoretical emphasis: policy change

The third theoretical emphasis in the ACF involves policy change. The ACF assumes that policies

are translations of beliefs and can be conceptualized and measured hierarchically like belief sys-

tems. Policy components that span, and are salient to, a policy subsystem represent the policy

core aspects of the subsystem. Policy components that deal with only a part of a subsystem or

technical components of a policy represent secondary aspects of the subsystem. A change in policy

core aspects indicates a major change in the direction of the subsystem and is deﬁned as “major

policy change.” A change in secondary aspects indicates a minor change in the subsystem and is

deﬁned as “minor policy change” (Sabatier and Jenkins-Smith 1999: 147–8).

The descriptive questions in the study of policy change include:

 To what extent does major and minor policy change occur within the subsystem over time?

 What is the content of minor and/or major policy change?

Explanatory questions on the topic of policy change are several:

 Why do some events and contexts lead to minor and major policy change?

 What are the mechanisms linking internal and external events to minor and major policy

change?

 How do negotiated agreements lead to policy change?

 When does policy-oriented learning lead to policy change?

As reﬂected in these questions, the ACF speciﬁes four pathways to policy change (Sabatier and

Weible 2007). The ﬁrst two paths involve events either external or internal to the policy sub-

system. External events occur outside the territorial boundaries of the subsystem and/or the

topical policy boundaries of the subsystem and are, hence, largely outside the control of sub-

system actors. Internal events occur inside the territorial and/or the topical area of the policy

subsystem and are more likely aﬀected by subsystem actors. The two types of events diﬀer in their

eﬀect on coalition behavior and policy change. Most importantly, attribution of blame or suc-

cess is higher after internal subsystem events are compared to after external subsystem events (see

Nohrstedt and Weible 2010).

Policy-oriented learning is the third path and is said to lead to policy change through altering

the beliefs of coalition members. Policies may change after a dominant coalition learns or after

learning occurs between adversarial coalitions.

The fourth path is said to occur through negotiated agreements. Sabatier and Weible (2007:

205–6) describe nine prescriptions fostering negotiation between coalitions and then policy

change. These nine prescriptions include: a hurting stalemate, broad representation, leadership,

consensus, funding, commitment by actors, importance of empirical issues, trust, and lack of alter-

native venues. From this list, the most important for instigating negotiations between coalitions

is a hurting stalemate. A hurting stalemate occurs when both coalitions view the status quo as

unacceptable and do not have access to alternative venues for achieving their objectives.

Weible and Nohrstedt

132

The original version of the ACF oﬀ ered two hypotheses involving major policy change:

Hypothesis 1 Signiﬁcant perturbations external to the subsystem are a necessary, but not suﬃcient,

cause of change in the policy core attributes of a governmental program.

Hypothesis 2 The policy core attributes of a governmental program in a speciﬁc jurisdiction

will not be signiﬁcantly revised as long as the subsystem advocacy coalition

that instituted the program remains in power within that jurisdiction—except

when the change is imposed by a hierarchically superior jurisdiction.

These two hypotheses for policy change have existed since the original version (Sabatier

1987, 1988). After the introduction of the other paths to policy change in 2007, new hypotheses

were not added to encompass this expansion in logic. To better reﬂect the theoretical arguments

c. 2007, a revised Hypothesis 1 would read as follows:

Hypothesis 1 Signiﬁcant perturbations external to the subsystem, a signiﬁcant perturbation

internal to the subsystem, policy-oriented learning, negotiated agreement, or

some combination thereof are necessary, but not suﬃcient, cause of change in

the policy core attributes of a governmental program.

The underlying logic of the ACF suggests that none of the four paths are necessary and suﬃ-

cient to produce policy change. Thus, one underdeveloped area within the ACF is the expla-

nation that traces any path or combination of paths from its occurrence to policy change or

stasis. This research should center on one or more exploitive coalitions that seek to capitalize on

the opportunity aﬀorded by one of the paths. But how a coalition will exploit an opportunity is

largely unknown from a theoretical perspective, with notable explorations into the topic taken

by Smith (2000), Ameringer (2002), Albright (2011), Ingold (2011), and Nohrstedt (2005, 2008,

2010, 2011). Developing knowledge in understanding the links from one or more of the four

paths to policy change will most likely involve the analysis of how resources are redistributed

between coalitions within the subsystem. It is likely, for example, that external events or internal

events or even learning might alter the distribution of resources or how those resources are used

leading to a more empowered coalition. Another theoretical area open for inquiry is the inter-

dependencies of the four paths or multiple occurrences within the same path. For example,

when do multiple external or internal events lead to learning and possibly negotiated agree-

ments among adversarial coalitions? Similarly, when do a number of events occurring over time

accumulate into suﬃ cient momentum for a coalition to capitalize to produce major policy change

(Smith 2000)? Finally, the role and behavior of dominant coalition actors in processes leading to

policy change must be clariﬁed (Nohrstedt 2010: 23). Under what conditions do dominant

coalition members promote stability and defend the status quo and when do they seek change?

What approaches to policy change (reformist or conservative) do dominant coalitions take fol-

lowing external and internal events and why? By deﬁnition, dominant coalition members

control key political resources—primarily formal legal authority—and thereby policy programs

(Nohrstedt 2011). Therefore, understanding the motivations of dominant coalition actors is an

important avenue for future research.

Conclusions

The ACF has continued for a quarter of a century and has made progress in understanding and

explaining policy processes across the globe. The theoretical emphasis in the ACF reﬂects the

The advocacy coalition framework

133

specialization among people applying the framework as well as natural partitions of the hypotheses.

Clearly, the hypotheses are overlapping. Whereas a theoretical emphasis on the structure and

stability of networks and beliefs might be the dependent variable for one study about coalitions,

changes in coalition structure and stability might be the independent variable for another study

about policy change (Weible et al. 2011). Researchers should not view the theoretical emphases

as sharply distinct but rather as signifying helpful partitions that permit specialized inquiry into

questions about coalitions, learning, and policy change.

Among the next steps in applying the ACF is to continue to develop the theoretical descriptions

and explanations within each of these areas of emphases as well as to develop other unexplored

areas within the framework. It is quite possible, for example, that there is another theoretical

emphasis centered on the role of scientiﬁc and technical information and scientists in policy

subsystems (Jenkins-Smith 1990; Weible 2008; Montpetit 2011). Policy narratives are another

area that oﬀers opportunities for theoretical growth based in part on the ACF (Shanahan et al.

2011). Scholars can also develop better theoretical understanding about how speciﬁc concepts

or categories of concepts interrelate within the framework; for example, the role of political

opportunity structures (Sabatier 1998; Kübler 2001). Most importantly, the ACF is increasingly

serving as a framework for guiding researchers from around the world in conducting compara-

tive public policy analysis. As such, what is needed is a clearer articulation of how subsystems

operate in diﬀerent political systems and political cultures. One likely approach for international

comparisons is, ﬁrst, to focus on political system diﬀerences (e.g. the continuum from corporatism

to pluralism) and, second, to compare and contrast subsystem institutions (e.g. subsystem speciﬁc

rules and norms) and their eﬀects on coalition formation, policy learning and change.

Given the continued and growing momentum of the ACF research program, this eﬀort must

ask the question: to what end? The immediate objective of the ACF is a better understanding

and explanation of policy processes. Such knowledge serves academia in research and teaching.

Another tradition in public policy is the service to society outside the networks of academics. In

this context, extant approaches for this aspiration can be found in various forms of engaged scho-

larship (Van de Ven 2008), advocacy, and direct service. For scholars applying the ACF and

seeking practical beneﬁts, one question should be how the framework can be used for theoretically-

guided research toward academic ends and as a tool in providing advice to subsystem actors

toward better societal outcomes.

Note

1 See Sabatier and Jenkins-Smith (1993) for a thorough description of the framework’s underpinnings.

Additionally, readers interested in gaining a deeper understanding of the ACF are encouraged to read

Sabatier (1987, 1988, 1998), Jenkins-Smith (1990), Sabatier and Jenkins-Smith (1999), Sabatier and

Weible (2007), Weible et al. (2009), and Weible et al. (2011).

Bibliography

Ajzen, Icek and Martin Fishbein. 1980. Understanding Attitudes and Predicting Social Behavior. Englewood

Cliﬀs, NJ: Prentice Hall.

Albright, Elizabeth A. 2011. ‘Policy change and learning in response to extreme ﬂood events in Hungary:

An advocacy coalition approach’, Policy Studies Journal 39(3): 484–511.

Ameringer, Carl F. 2002. ‘Federal antitrust policy and physician discontent: Deﬁning moments in the

struggle for congressional relief’, Journal of Health Politics, Policy and Law 27(4): 543–74.

Andersson, Magnus. 1999. Change and Continuity in Poland’s Environmental Policy. Dordrecht: Kluwer.

Dawson, Richard and James Robinson. 1963. ‘Interplay competition, economic variables, and welfare

policies in the American states’, Journal of Politics 25 (May): 265– 89.

Weible and Nohrstedt

134

Easton, David. 1965. A Framework for Political Analysis. Chicago: University of Chicago Press.

Fenger, Menno and Pieter-Jan Klok. 2001. ‘Interdependency, beliefs, and coalition behavior: A contribution

to the advocacy coalition framework’, Policy Sciences 34: 157–70.

Freeman, J. Leiper. 1955. The Political Process. New York: Random House.

Griﬃth, Ernest. 1939. The Impasse of Democracy. New York: Harrison Hilton Books.

Heclo, Hugh. 1974. Social Policy in Britain and Sweden. New Haven, CT: Yale University Press.

—— 1978. ‘Issue networks and the executive establishment’, in Anthony King (ed.) The New American

Political System. Washington, DC: AEI.

Henry, Adam. 2011. ‘Power, ideology, and policy network cohesion in regional planning’, Policy Studies

Journal 39(3): 361–83.

——, Mark Lubell and Michael McCoy. 2010. ‘Belief systems and social capital as drivers of policy network

structure: The case of California regional planning’, Journal of Public Administration Research and Theory

21(3); 419–44.

Herron, Kerry G. and Hank C. Jenkins-Smith. 2006. Critical Masses and Critical Choices: Evolving Public

Opinion on Nuclear Weapons, Terrorism, and Security. Pittsburgh, PA: University of Pittsburgh Press.

Hirschi, Christian and Thomas Widmer. 2010. ‘Policy change and policy stasis: Comparing Swiss foreign

policy toward South Africa (1968–94) and Iraq (1990–91) ’, Policy Studies Journal 38(3): 537–63.

Hjern, Benny and David Porter. 1981. ‘Implementation structures: A new unit of administrative analysis’,

Organizational Studies 2: 211–27.

Hurwitz, Jon and Mark Peﬄey. 1987. ‘How are foreign policy attitudes structured? A hierarchical model’,

American Political Science Review 81: 1099– 120.

Ingold, Karin. 2011. ‘Network structures within policy processes: Coalitions, power and brokerage in

Swiss climate policy’, Policy Studies Journal 39(3): 434–59.

Jenkins-Smith,

Hank. 1990. Democratic Politics

and Policy Analysis. Paciﬁc Grove, CA: Brooks/Cole.

——and Gilbert St. Clair. 1993. ‘The politics of oﬀshore energy: Empirically testing the advocacy coali-

tion framework’, in Paul Sabatier and Hank Jenkins-Smith (eds) Policy Change and Learning. Boulder,

CO: Westview Press, 149–75.

——, Gilbert St. Clair and Brian Woods. 1991. ‘Explaining change in policy subsystems: Analysis of coalition

stability and defection over time’, American Journal of Political Science 35 (November): 851–72.

Kübler, Daniel. 2001. ‘Understanding policy change with the advocacy coalition framework: An application

to Swiss drug policy’, Journal of European Public Policy 8(4): 623–41.

Larsen, Jakob Bjerg, Karsten Vrangbaek and Janine M. Traulsen. 2006. ‘Advocacy coalitions and pharmacy

policy in Denmark’, Social Science and Medicine 63(1): 212–24.

Laudan, Larry. 1977. Progress and Its Problems: Towards a Theory of Scientiﬁc Growth. Berkeley, CA: University of

California Press.

Lerner, Daniel and Harold D. Lasswell (eds). 1951. The Policy Sciences. Stanford, CA: Stanford University

Press.

Lindblom, Charles E. 1959. ‘The science of muddling through’, Public Administrative Review 59: 78–88.

Lord, Charles, Lee Ross and Mark Lepper. 1979. ‘Biased assimilation and attitude polarization: The eﬀects

of prior theories on subsequently considered evidence’, Journal of Personality and Social Psychology 37:

2098–109.

Matti, Simon and Annica Sandström. 2011. ‘The rationale determining advocacy coalitions: Examining

coordination networks and corresponding beliefs’, Policy Studies Journal 39(3): 385–410.

Mazmanian, Daniel and Paul Sabatier. 1981. Implementation and Public Policy. Lanham, MD: University

Press of America.

Mazur, Allan. 1981. The Dynamics of Technical Controversy. Washington DC: Communications Press.

Montpetit, Eric. 2011. ‘Scientiﬁc

credibility, disagreement, and error costs in 17 biotechnology policy

subsystems’, Policy Studies Journal 39(3):

513–33.

Munro, John. 1993. ‘ California water politics: Explaining change in a cognitively polarized subsystem’,in

Paul Sabatier and Hank Jenkins-Smith (eds) Policy Change and Learning. Boulder, CO: Westview Press,

105–28.

Munro, Geoﬀrey D. and Peter H. Ditto. 1997. ‘Biased assimilation, attitude polarization, and aﬀect in

reactions to stereotype-relevant scientiﬁc information’, Personality and Social Psychology Bulletin 23(6):

636–53.

——, Peter H. Ditto, Lisa K. Lockhart, Angela Fagerlin, Mitchell Gready and Elizabeth Peterson. 2002.

‘Biased assimilation of sociopolitical arguments: Evaluating the 1996 US Presidential debate’, Basic and

Applied Social Psychology 24(1): 15–26.

The advocacy coalition framework

135

Nohrstedt, Daniel. 2005. ‘External shocks and policy change: Three Mile Island and Swedish nuclear

energy policy’, Journal of European Public Policy 12(6): 1041–59.

—— 2008. ‘The politics of crisis policymaking: Chernobyl and Swedish nuclear energy policy’, Policy

Studies Journal 36(2): 257–78.

—— 2010. ‘Do advocacy coalitions matter? Crisis and change in Swedish nuclear energy policy’, Journal of

Public Administration Research and Theory 20: 309– 33.

—— 2011. ‘Shifting resources and venues producing policy change in contested subsystems: A case study

of Swedish signals intelligence policy’, Policy Studies Journal 39(3): 461–84.

——and Christopher M. Weible. 2010. ‘The logic of policy change after crisis: Proximity and subsystem

interaction’, Risks, Hazards, and Crisis in Public Policy 1(2): 1–32.

Ostrom, Elinor. 2005. Understanding Institutional Diversity. Princeton, NJ: Princeton University Press.

Peﬄey, Mark and Jon Hurwitz. 1985. ‘A hierarchical model of attitude constraint’, American Journal of

Political Science 29: 871–90.

Pierce, Jonathan J. 2011. ‘Coalition stability and belief change: Advocacy coalitions in US foreign policy

and the creation of Israel, 1922–44’, Policy Studies Journal 411–34.

Pressman, Jeﬀrey and Aaron Wildavsky. 1973. Implementation. Berkeley, CA: University of California

Press.

Putnam, Robert. 1976. The Comparative Study of Political Elites. Englewood Cliﬀs, NJ: Prentice Hall.

Quattrone, George A. and Amos Tversky. 1988. ‘Contrasting rational and psychological analysis of political

choice’, American Political Science Review 82: 719–36.

Ranney,

Austin. 1968. Political Science

and Public Policy. New York: Markham.

Sabatier, Paul, A. 1987. ‘Knowledge, policy-oriented learning, and policy change: An advocacy coalition

framework’, Knowledge: Creation, Diﬀusion, Utilization 8(4): 649–92.

—— 1988. ‘An advocacy coalition model of policy change and the role of policy-oriented learning

therein’, Policy Sciences 21 (autumn): 129–68.

—— 1991. ‘Toward better theories of the policy process’, PS: Political Science and Politics 24(2): 147–56.

—— 1998. ‘The advocacy coalition framework: revisions and relevance for Europe’, Journal of European

Public Policy 5 (March): 98–130.

—— 1999. ‘The need for better theories’,inTheories of the Policy Process. Boulder, CO: Westview Press, 3–17.

—— 2007. Theories of the Policy Process. Boulder, CO: Westview Press.

——and Anne M. Brasher. 1993. ‘From vague consensus to clearly diﬀerentiated coalitions: environmental

policy at Lake Tahoe, 1964–1985’, in Paul Sabatier and Hank Jenkins-Smith (eds) Policy Change and

Learning. Boulder, CO: Westview Press, 177–208.

——and Hank C. Jenkins-Smith. 1993. Policy Change and Learning: An Advocacy Coalition Approach.

Boulder, CO: Westview Press.

——and Hank Jenkins-Smith. 1999. ‘The advocacy coalition framework: An assessment’, in Paul Sabatier

(ed.) Theories of the Policy Process. Boulder, CO: Westview Press, 117–68.

——and Christopher M. Weible. 2007. ‘The advocacy coalition framework: Innovations and clariﬁcations’,

Paul Sabatier (ed.)Theories

of the Policy Process, 2nd edn. Boulder, CO: Westview Press, 189–222.

——and Matthew Zafonte. 2001. ‘Policy knowledge, advocacy organizations’, in Neil J. Smelser and Paul

B. Baltes (eds) International Encyclopedia of the Social and Behavioral Sciences 17. Oxford: Pergamon Press,

11563–68.

——, Susan Hunter and Susan McLaughlin. 1987. ‘The devil shift: Perceptions and misperceptions of

opponents’, Western Political Quarterly 40: 51–73.

Schlager, Edella. 1995. ‘Policy making and collective action: Deﬁning coalitions within the advocacy

coalition framework’, Policy Sciences 28: 243–70.

Shanahan, Elizabeth, Michael D. Jones and Mark K. McBeth. 2011. ‘Policy narratives and policy processes’,

Policy Studies Journal 39(3): 535–61.

Simon, Herbert A. 1957. Models of Man: Social and Rational. New York: John Wiley.

—— 1985. ‘Human nature in politics: The dialogue of psychology with political science’, American Political

Science Review 79 (June): 293–304.

Smith, Adrian. 2000. ‘ Policy networks and advocacy coalitions: Explaining policy change and stability in

UK industrial pollution policy?’ Environment and Planning C: Government and Policy 18: 95–114.

Van de Ven, Andrew H. 2008. Engaged Scholarship: A Guide for Organizational and Social Research. Oxford:

Oxford University Press.

Walker, Jack, L. 1969. ‘The diﬀusion and innovation among the American states’, The American Political

Science Review 63(3): 880–99.

Weible and Nohrstedt

136

Weible, C.M. 2005. ‘Beliefs and perceived inﬂuence in a natural resource conﬂict: an advocacy coalition

approach to policy networks’, Political Research Quarterly 58(3): 461 – 75.

—— 2007. ‘An advocacy coalition framework approach to stakeholder analysis: understanding the political

context of California marine protected area policy’, Journal of Public Administration Research and Theory 17:

95–117.

—— 2008. ‘Expert-based information and policy subsystems: A Review and synthesis’, Policy Studies

Journal 36(4): 615–35.

——and Paul A. Sabatier. 2005. ‘ Comparing policy networks: Marine protected areas in California’, Policy

Studies Journal 33(2): 181–204.

——, Paul A. Sabatier and Kelly McQueen. 2009. ‘Themes and variations: Taking stock of the advocacy

coalition framework’, Policy Studies Journal 37(1): 121–40.

——, Paul A. Sabatier, Hank C. Jenkins-Smith, Daniel Nohrstedt and Adam Douglas Henry. 2011. ‘A

quarter century of the advocacy coalition framework: An introduction to the special issue’, Policy Studies

Journal 39(3): 349–60.

Weiss, Carol. 1977. ‘Research for policy’s sake: The enlightenment function of social research’, Policy

Analysis 3 (autumn): 531–45.

Zafonte, Matthew and Paul A. Sabatier. 1998. ‘Shared beliefs and imposed interdependencies as determinants

of ally networks in overlapping subsystems’, Journal of Theoretical Politics 10 (4): 473–505.

——and Paul A. Sabatier. 2004. ‘Short-term versus long-term coalitions in the policy process: Automotive

pollution control, 1963–89’, The

Policy Studies Journal 32(1): 75–107.

The advocacy coalition framework

137

The punctuated equilibrium

theory of agenda-setting

and policy change

Graeme Boushey

Introduction

On March 23, 2010, President Barack Obama signed into law the Patient Protection and

Aﬀordable Care Act (PPACA), marking a major shift in the organization and structure of health

care in the United States. The statute outlined a series of dramatic reforms to the provision,

funding and regulation of public and private health insurance, controversially requiring that all

Americans carry qualifying health coverage, directing state governments to regulate new health

insurance exchanges, expanding the coverage of poor and uninsured children, and designating a

comparative eﬀectiveness research institute to track the eﬀectiveness of preventative and clinical

public health (Kaiser Family Foundation 2010).

Although there is currently considerable uncer-

tainty regarding the future implementation of the health care reforms, the passage of the act is

nonetheless remarkable in the size and scope of policy change. The PPACA has triggered sig-

niﬁcant policy-making attention across the legal, executive, and legislative branches of both the

state and federal governments. If the statute survives the current legal challenges, the PPACA

will radically alter the US health care system.

The sweeping policy reform highlights a familiar puzzle to students of the punctuated equi-

librium theory (PET) in public policy-making. Public policy change is typically characterized by

tremendous stability, as broad policy change usually faces considerable political and institutional

barriers. Yet policy change can occur through sweeping positive feedback cycles as radical new

innovations are adopted in a relatively short time frame.

Most changes in US health policy have

occurred through incremental adjustment, as Congress approved modest reforms to national and

state programs. However, at key moments national policy-making attention has focused squarely

on the problem of underinsurance, leading to sweeping policy reforms such as the Medicare and

Medicaid programs, the expansion of the Medicare prescription drug beneﬁt, and most recently

the passage of the PPACA. Viewed over time, public policy-making is characterized by punctuated

dynamics.

As a policy process model, punctuated equilibrium theory documents how shifts in macro poli-

tical attention result in both negative and positive feedback cycles in policy-making. Negative feedback

cycles emerge as a result of the delegation of routine policy-making to policy subgovernments, as

138

a core group of institutional actors make marginal adjustments to policy in response to a dominant

policy image. Positive feedback cycles result when mass political attention focuses system-wide

attention on a new dimension of a policy problem. When this occurs, policy change occurs

rapidly as new political actors and new jurisdictions become involved in policy-making.

These distinct patterns of incremental and non-incremental policy change result from the dis-

proportionate allocation of issue attention in public policy-making (Baumgartner and Jones 2009;

Jones and Baumgartner 2005). Governments have limited agenda space but face thousands of

competing policy problems annually (Baumgartner and Jones 2009; Jones and Baumgartner

2005). Because mass political attention is rarely focused on a single issue, policy-making is

generally delegated to policy subgovernments, where actors within political institutions work to

manage an overwhelming supply of information demanding political action. When system-wide

political attention is focused on a single issue, the pressures reinforcing equilibrium in a policy

subsystem are ruptured, leading to broad political change.

To understand forces leading to both policy stability and rapid change in a single political

system, punctuated equilibrium theory approaches the study of public policy longitudinally by

exploring how focusing events, institutional venues, policy ideas and policy entrepreneurs interact

to produce policy change. The allocation of government attention is unpredictable, yet PET

identiﬁes systematic processes that lead to the dynamic processes of policy change over time

through the study of comparative issue dynamics across institutional and historical contexts.

This chapter provides an overview of the theory and methods of the punctuated equilibrium

theory of public policy. The chapter begins with a review of the decision-making model under-

lying punctuated equilibrium theory, explaining how pressures facing individual and institu-

tional decision-making precipitate punctuated dynamics in agenda setting. The chapter then focuses

on PET as a policy process model, reviewing how focusing events, policy ideas, institutional

venues and interest groups interact to produce both stability and sudden policy change over time.

The chapter concludes with a review of recent methodological advances in comparative policy

dynamics, highlighting major empirical and methodological ﬁndings as PET has evolved as a

general model of policy-making across historical and institutional contexts. With these factors

working together, the review points to a number of new opportunities for evaluating the causes

of punctuated dynamics in public policy-making.

Negative feedback cycles and policy incrementalism

Models of political decision-making begin with an important observation about the demands of

information processing in public policy-making. Political actors are constrained by a scarcity of

time, resources and political attention. Because political decision-makers must address a multi-

tude of diﬀerent issues on a daily basis—negotiating budget appropriations, designing regulatory

policy, sitting in hearings and attending to casework—decision-makers cannot dedicate the time

and resources needed to engage in a comprehensive solution search for each and every policy

problem (Lindblom 1959; Hayes 1992).

The challenges presented by the sheer volume of issues demanding political attention are

compounded by the complexity of emerging public policy problems (Workman et al. 2009).

Policy problems are multidimensional, and policy-makers are challenged by considerable uncer-

tainty regarding the appropriate solutions and potential outcomes of new policy interventions.

For example, policy-makers faced with growing concern over about air pollution must consider

diﬃcult trade-oﬀs between protecting population health, protecting the environment, and pro-

viding for the continued competitiveness of industry. The development of appropriate regula-

tions are further complicated when the costs and expected outcomes of regulatory intervention

Punctuated equilibrium theory

139

are unknown. Policy-makers face considerable challenges when problems are complex, solutions

are costly, and outcomes are uncertain.

To compensate, policy-makers engage in incremental decision-making, limiting solution

searchers to a small set of local alternatives and making small adjustments to existing policy regimes

(Lindblom 1959; Wildavsky 1964). This implies a conservative approach to policy-making, as

risk-averse policy-makers make gradual changes to existing policies through “successive limited

comparison” rather than risking large-scale policy failure (Lindblom 1959; Robinson 2006). For

example, when faced with the considerable uncertainty of the consequences of new air quality

regulations, policymakers may prefer to make minor adjustments to the existing policy—such as

adopting slightly more stringent energy eﬃciency standards for cars or manufacturers—rather

than advocating for a large-scale change such as a cap and trade system for greenhouse gas emissions.

This approach has the advantage of reducing the uncertainty regarding the cost and eﬃcacy of

policy change. In this regard, incremental decision-making can resemble path dependency

theory, as policy and solutions become locked in over time (Robinson and Meier 2006).

Institutions and negative feedback cycles

The tendency toward incremental decision-making is reinforced by policy-making institutions,

which help decision-makers organize information, set agenda priorities, develop solutions, and

choose among alternatives. In political institutions, policy-making is generally delegated to policy

subsystems, specialized policy-making jurisdictions charged with the development of public

policy within a speciﬁc issue area.

The delegation of policy-making to subgovernments helps to overcome the challenges pre-

sented by oversupply of information by facilitating parallel processing where diﬀerent subgovern-

ments “address multiple, diverse issues simultaneously ” (Workman et al. 2009: 79). In the United

States, policy-making in Congress is channeled through the committee system, a series of specia-

lized subgovernments charged with policy development across speciﬁc issue areas. The committee

system reduces the problem of information oversupply by committees acting as gatekeepers,

selectively ﬁltering information for the larger legislative body.

The organization of policy-making institutions reinforces the bias toward incremental policy-

making by imposing strict decision costs on policy-making that pressure against sudden policy

change (Workman et al. 2009). Deliberative bodies may have supermajority requirements to

enact certain forms of legislation, or require multiple legislative bodies to approve policy prior

to enactment. Broad policy change is therefore generally diﬃcult to achieve, as it requires the

coordination and cooperation of multiple veto players with distinct preferences across policy-

making institutions (Tsebelis 2002). Actors face considerably less opposition at the subgovernment

level, where policy-making occurs through marginal, consensual adjustment to prior policy regimes.

Negative feedback in public policy-making is maintained by the stability of participants

engaged in policy subsystems. When policy subsystems are dominated by a speci ﬁc interest in

the policy process, policy-making is dictated by a policy monopoly, a narrow coalition of elected

oﬃcials, bureaucrats and interest group representatives controlling the legislative agenda at the

subgovernment level (Baumgartner and Jones 2009; True et al. 2007).

As beneﬁciaries from the

status quo, these policy monopolies have little incentive to undertake radical reforms in policy-

making. Instead, policy monopolies oﬀset outside calls for reform and policy change by making

minor or superﬁcial adjustments to policy, allowing the subsystem to appear responsive without

radically adjusting public policy.

For example, faced with the challenge of conserving natural resources and energy in the United

States, the Senate Committee on Energy and Natural Resources has historically recommended

Boushey

140

increasing the fuel eﬃciency of automobiles and heavy trucks, and consistently working to reduce

the nation’s energy dependence by regulating the fuel consumption of passenger vehicles. By

making minor adjustments to vehicle fuel eﬃciency, policy-makers incrementally move toward

the overall goal of energy conservation while simultaneously reducing the uncertainty of new

policy for industry, consumers and government. Such incremental adjustments prevail at a sub-

system level where constraints on attention and information demand that policy-makers routinely

focus on limited alternatives in developing new solutions.

Finally, negative feedback cycles dominate in policy subsystems because of the relative stabi-

lity of the policy image —the collective understanding and ideas regarding the problems and issues

at stake (Baumgartner and Jones 2009; Walgrave and Varone 2008). Although policy problems

are complex and multidimensional, decision-makers operating within a policy monopoly tend

to focus only on speciﬁc dimensions of a public policy problem. Within a policy monopoly there is

little disagreement over the nature of the policy problem. Without change in the monopoly

dynamic, policy solutions become routine over time, and policy though subsystem-level responses

to well-deﬁned policy problems become repetitive. For much of recent history, energy and

natural resources have largely been understood as a problem of conservation and eﬃciency.

Government has focused largely on regulating the individual consumption of oil through speed

limits, fuel eﬃciency standards, and highway and infrastructure investment. These solutions all

respond to a common and widely shared understanding of the challenges of energy conservation.

The tremendous policy stability that typically explains policy-making results from the constraints

that individuals working within complex political institutions face as they address complex

policy problems. Policy-makers rarely have the time, resources or incentive to engage in craft-

ing radical new solutions to each and every policy problem. Even if an actor pushes for a radical

new program, they usually face institutional barriers that reroute toward gradual and consensual

policy adjustment.

Disproportionate information processing and positive feedback cycles

Policy change generally occurs gradually, but at other times issues explode onto the policy agenda,

leading to sweeping changes in public policy that cannot be explained neatly by conservative

incremental policy-making theory. These periods of policy disequilibrium result when a focus-

ing event directs system-wide attention onto a new salient dimension of a public policy issue,

elevating it to the forefront of the institutional and mass political agendas (Baumgartner and

Jones 2009; Jones and Baumgartner 2005; Kingdon 1984; Shattschneider 1975). When elevated

issue attention expands the scope of con ﬂict to new jurisdictions, the policy monopolies that

sustained the prior equilibrium are destroyed as new participants introduce legislation responding

to the shift in governments’ understanding of the policy problem (Baumgartner and Jones 2009).

Sudden periods of agenda instability and policy disequilibrium are the result of positive feedback

cycles in public policy-making, as new pressures for policy change destabilize long-standing policy

regimes (Jones and Baumgartner 2005).

In recent years, punctuated equilibrium theory has extended the theory of bounded ration-

ality to model how individual and organizational choice can explain both incremental adjustment

and sudden shifts in policy regimes. Punctuated policy dynamics are the result of dispropor-

tionate information processing in public policy processes (Jones 2001; Jones and Baumgartner

2004, 2005; Workman et al. 2009). Decision-making is limited less by the supply of available

information and more by the scarcity of attention (Jones 1994: 87). The key insight is that

patterns of policy change result from the selective allocation of attention in decision-making.

Individuals pressed with the demands of making decisions across a multitude of issues reduce

Punctuated equilibrium theory

141

information costs through incremental decision-making; however, when a subset of issues

receives substantial political attention, it leads to more extensive problem representation and

solution search than is typical (Jones 2001). These attention shifts occur as individuals ﬁlter new

information, focusing attention squarely on decisions that elevate a sense of urgency, fear or

anxiety. While individuals may intend to make proportional, rational decisions that are appro-

priate for the problems they face, emotional cues often lead them to under-respond to some

pressures, and over-respond to others. Neither individuals nor institutions respond to information

and policy relevant signals equally (ibid.).

The bias toward incremental responses to routine problems contributes to disproportionate

information processing in decision-making. Because most problems are addressed through incre-

mental adjustment, people routinely fail to incorporate new policy-relevant information when

making daily decisions. When attention is squarely focused on an emerging policy problem, indi-

viduals then over-respond to new information as a wide range of new policy relevant infor-

mation is suddenly incorporated into choice (Jones and Baumgartner 2004). In this regard, the

status quo bias of incrementalism begets large shifts in attention exactly because it reinforces a

disjointed pattern of neglect and sudden response to information (Jones and Baumgartner 2005).

Individual preferences also shift suddenly in response to changes in the context of new policy-

relevant information. Although policy problems are multidimensional, decision-makers tend to

focus narrowly on a limited set of dimensions highlighted in the most immediate problem at

hand. As research in framing eﬀects has demonstrated, preferences for policy can shift radically

depending on the speciﬁc contextual cues (Iyengar 1996; Jones 2001). Preferences may therefore

shift dramatically when policy-makers react to information that highlights a new and previously

overlooked component of a policy problem.

In focusing on the disproportionate allocation of individual attention, punctuated equilibrium

theory demonstrates that incrementalism is not the only mechanism of preference formation

and choice implied by bounded rationality. Shifts in individual attention can lead to dispropor-

tionate responses to new information. Rather than following a single path of incremental choice,

individuals instead often shift attention and preferences suddenly.

Positive feedback cycles and institutional choice

As with the pressures leading to negative feedback cycles, the processes leading to positive feedback

result from the interaction of individuals and institutions in public policy-making. Although the

organization of policy-making in the United States generally imposes a conservative pressure

against rapid policy change, policy entrepreneurs and interest group activists across national and

sub-national jurisdictions constantly pressure for new policy reform or innovation. When a new

idea or initiative takes hold and becomes legitimized across political venues, sudden shifts in

policy follow.

Punctuated equilibrium theory contends that the formalized process of agenda-setting and

political decision-making in political institutions ampliﬁes disproportionate information proces-

sing (Jones and Baumgartner 2004, 2005). Most policy issues are handled as routine develop-

ments and are delegated to policy subgovernments (Baumgartner and Jones 2009; True et al.

2006; Baumgartner and Jones 2002). However, at any given moment a small subset of issues

receives system-wide political attention. When this occurs, widespread participation of actors across

venues and the incorporation of new information can trigger rapid and extensive policy change

(Baumgartner and Jones 2009; True et al. 2007; Jones and Baumgartner 2005).

Studies of punctuated dynamics begin by identifying the factors leading to the shift in macro-

political attention preceding radical policy reforms. Elevated issue attention caused by a focusing

Boushey

142

event that galvanizes mass public and political attention can open a “window of opportunity,”

creating conditions for rapid and non-incremental policy change (Baumgartner and Jones 2009;

Birkland 1997; Kingdon 1984). Exogenous shocks work to suddenly galvanize mass political

attention on an issue area, revealing a critical policy problem that demands political interven-

tion. The tsunami leading to the Fukushima nuclear reactor meltdown in 2011 focused inter-

national political attention of the safety of nuclear power, leading governments to suspend new

projects or adopt more restrictive regulatory regimes for monitoring safety.

Exogenous shocks are not the only mechanism that can lead to shifts in public policy images

over time. Baumgartner and Jones (2009) observe that changes in policy indicators, as well as

changes in the tone of media coverage, can both redirect public and political perceptions of policy

problems. Furthermore, policy entrepreneurs play a key role in inﬂuencing perceptions of a

policy problem. Strategic issue deﬁnition and redeﬁnition (Baumgartner et al. 2008; Baumgartner

and Jones 1993; Jones and Baumgartner 2005) and policy targeting (Boushey 2010; Donovan

2001; Schneider and Ingram 1993) can draw public attention to new dimensions of a problem,

leading to swings in public opinion and support for new policy ideas. It is not simply that

focusing events elevate government attention, but perhaps more importantly that these events

redirect mass political understandings of the underlying policy problem.

These pressures have a common impact on political decision-making, especially when a new

policy idea expands the scope of conﬂict and brings new voices into a policy arena (Shattschneider

1975). As this occurs, the policy monopoly that had sustained policy equilibrium at the sub-

government level is destroyed. In its place, government generates new legislation, regulations,

and institutions to respond to the emerging policy problem, as new stakeholders across political

jurisdictions and venues are mobilized to engage in previously neglected policy arenas. Policy

punctuations represent the output of these periods of instability, as government reorganizes to

eﬀectively respond to the new policy image.

Finally, although the allocation of political attention universally follows the rules of dispropor-

tionate information processing, diﬀerent institutions impose decision costs that lead to stick slip

dynamics in public policy-making, as pressures for policy change contend with institutional rules

that impose friction or gridlock on the policy process (Baumgartner et al. 2009). Here, institu-

tional decision rules shape long-term patterns of policy equilibrium. As pressure for change accu-

mulates, more responsive institutions will see smaller but more frequent reversals in policy-making,

updating policy more rapidly in response to shifting preferences. Institutions with higher deci-

sion costs will be generally resistant to large-scale policy change. In these intuitions, positive

feedback cycles of sudden policy change will occur more rarely, but be larger in size and scope.

This dynamic is illustrated in Jones and Baumgartner’s (2005) comparison of the historical dis-

tributions of policy change across US policy-making institutions. Where institutional decision-

costs are lower (for example executive orders), punctuated dynamics are less pronounced. Where

institutional costs are highest (for example in budgeting) the distribution of policy change suggested

rare but more extreme policy change.

These observations underpin what Jones and Baumgartner (2005) refer to as the general punc-

tuation hypothesis, as cognitive and individual constraints interact to produce punctuated dynamics

in political decision-making. Because cognitive limitations leading to punctuated dynamics are

universal, punctuated dynamics will emerge across policy-making contexts, and policy-makers

will respond disproportionately to information. However, punctuated dynamics are also con-

tingent on the degrees of friction imposed by institutional rules, leading to “a greater likelihood

of punctuations in some institutions of government than others” (ibid.: 20).

The dynamics of punctuated equilibrium have been documented across issue areas and

institutional contexts. For example, researchers have documented punctuated dynamics across

Punctuated equilibrium theory

143

US policy-making institutions (Baumgartner and Jones 2005; Jones et al. 2003) as well as state

(Breunig and Koski 2009) and local government budgeting (Jordan 2003). The patterns of policy

stasis and change have been documented in policy areas as distinct as gun control (True and

Utter 2002), crime control (Schneider 2006), environmental regulation (Repetto 2006), tobacco

control (Wood 2006), education ﬁnance (Robinson 2006) and international infectious disease

policy (Shiﬀman et al. 2002). Perhaps more impressively, research in recent years has expanded

beyond the study of American politics to evaluate the general punctuation hypotheses in other

institutional contexts. Scholars have documented punctuated dynamics across Western democracies

(Baumgartner et al. 2009; Jones et al. 2009; Baumgartner et al. 2008). This growing body of

work suggests that that punctuated dynamics emerge from pressures of political decision-making

that are generalizeable across eras and institutional contexts (Baumgartner and Jones 2009).

Policy entrepreneurs, venue shopping, issue framing and policy change

Punctuated equilibrium theory suggests that interest groups and policy entrepreneurs play a pro-

minent role in developing and introducing new ideas in the political system. Punctuated dynamics

are shaped by the mobilization of competing interest groups and activists in public policy-

making. Organized interests that assume privileged positions in policy monopolies help to

reinforce the status quo, providing expert testimony and support for maintaining a stable policy

regime at the subsystem level. Conversely, the strategic choices of marginalized interest groups

are pivotal for understanding how new ideas enter the political system.

The eﬀect of the participation of interest groups is therefore critical for understanding how

policy images evolve over time. While policy monopolies maintain policy stability, marginalized

or outsider interest groups do not simply acquiesce to the bias of policy monopolies across sub-

governments. Instead, these organizations work to shape policy change at all levels of government.

This section brieﬂy discusses how the strategies adopted by outsider or marginalized interests

lead to periods of policy change. In so doing, it highlights one major mechanism leading to

policy change across political systems. Most organized eﬀorts fail to disrupt the status quo, but

when an interest group succeeds, its activities can lead to sweeping changes across institutions.

The ﬁrst major implication of punctuated equilibrium theory is that interest groups are not

complacent when they are excluded from policy-making at the subgovernment level. Interest

groups shut out of one policy venue often strategically look for agenda access elsewhere. Organized

interest groups engage in venue shopping, strategically moving across jurisdictions to secure legislative

change (Boushey 2010; Baumgartner and Jones 2009; Cashore and Howlett 2006; Pralle 2003).

In federal systems, interest groups may pressure for reforms across municipal and state governments,

adopt litigation strategies to secure change through the courts, or employ state and local ballot

initiative campaigns to introduce legislation directly before the voters as part of the initiative

system. The logic is that institutions and policy-making jurisdictions are not equally receptive to

new appeals for policy reform. For example, Cashore and Howlett (2006) discovered diﬀerences in

jurisdictional susceptibility to widespread policy reform in federal and state forestry policy, indicating

that various institutions respond diﬀerently to nearly identical pressures for policy change.

More broadly, research in state-level policy innovation and diﬀusion has documented con-

siderable variation in the responsiveness of state governments to calls for policy change.

Researchers have developed a robust understanding of how diﬀerences in the ideology, resources

and professionalism of state and provincial governments shape their responsiveness to new reform

initiatives over time (Boushey 2010; Karch 2007; Berry and Berry 1999). Boushey (2010)

observes that venue shopping in federalism can lead to punctuated patterns of policy diﬀusion

across subnational governments in federalism. The ability of interest groups to secure policy

Boushey

144

change at select venues can result in gradual policy adoptions across jurisdictions as they succeed

in pushing reforms through at state and municipal governments over time. However, in rare

instances venue-shopping strategies can trigger positive feedback cycles as an issue becomes the

focus of broader system-wide policy-making attention (ibid.).

Organized interests not only experiment with the selection of venues but also actively engage

in the framing and reframing of legislative ideas in the hope of expanding public support for

policy change (Baumgartner et al. 2008; Smith 2007). Activists seeking to shape public respon-

siveness to their policy programs experiment with framing legislation in response to changes in

the political environment, looking to attribute new meaning or justiﬁcation for their preferred

policy proposals (Baumgartner and Jones 1993).

Interest groups adopt one of two strategies when attempting to frame a policy problem. One

is simply to attach a new justiﬁcation for an existing policy proposal. For example, interest

groups justify policy reforms in the context of the major salient “issues of the day,” such as

pushing for tax reform legislation as a solution to a major ﬁnancial crisis or advocating for a ban

on factory farming in the wake of food safety concerns. Such framing and coupling strategies

are employed to redirect public attention to a di ﬀerent dimension of a policy problem that is

more favorable to the policy agenda of a given interest group.

The second form of strategic interest group framing happens with policy targeting. In this

case, interest groups narrow or expand the conﬂict surrounding a proposal by narrowing or

expanding the scope and application of the policy proposal. (Schattschneider 1975). Marijuana

reform activists have pushed for the legalization of medical marijuana, rather than the immediate

end of marijuana prohibition (Boushey 2010). Targeting policies in this manner does not simply

oﬀer new justiﬁcation for the policy proposal; it alters the scope of the policy itself.

Of course, eﬀorts to redeﬁne or reframe a policy debate generally fail to generate broad support

for a new policy reform. Even when a new way of describing a policy problem succeeds,

opposing interest groups often succeed in diminishing the impact of an emerging frame with a

counterargument. However, when issue framing or venue-shopping strategies successfully unite

broad and distinct social interests around a cause, rapid policy change may follow. This points to

an important feature of the broader policy-making processes leading to punctuated dynamics in

public policy-making: although the formal organization of policy-making institutions works to

slow pressures for sudden policy change, the multiple access points and proliferation of interest

groups also encourages the proliferation of policy ideas.

The role that interest groups play in shifting national political attention through issue framing

and venue-shopping strategies have been well-documented in case studies of interest group

mobilization. For example, the passage of the Eighteenth Amendment prohibiting alcohol in

the United States emerged from a century of political pressure, as prohibitionists pushed for local,

state, and national alcohol reform, constantly shifting the justiﬁcation for prohibition (Boushey

2010; Aaron and Musto 1981). The Tobacco Master Settlement agreement was secured through

judicial venue shopping, as 46 states joined in a giant lawsuit against the tobacco industry to

recover state Medicaid expenditures incurred covering smoking-related illness (Baumgartner and

Jones 2009; Wood 2006). Finally, the longer-term decline in public support for the death

penalty in practice occurred as the death penalty abolitionists succeeded in shifting the focus of

the debate towards the very real possibility that governments were executing the innocent. The

movement to stay the death penalty also followed a venue-shopping strategy, as activists have

pressured for moratoriums on capital punishment in receptive states, legitimizing the policy

reform gradually (Baumgartner et al. 2008).

Although these cases highlight strategies of groups operating within federalism, there is no

reason to believe that this strategic choice is limited to this institutional context. Venue shopping

Punctuated equilibrium theory

145

can occur within single institutions, as actors petition distinct subgovernments to place an issue

on the agenda. For example, activists concerned with the dangers of genetically modiﬁed foods

may bypass the Department of Agriculture and instead petition the Food and Drug Administration

to regulate new crops. The essential insight is that organized interests and policy entrepreneurs

play a key role in developing and introducing ideas into the system. Policy entrepreneurs

selectively direct issue attention toward new dimensions of a problem in the hopes of securing a

change of policy over time. Furthermore, they labor to introduce these policies into the poli-

tical system, probing for receptive venues and petitioning for widespread policy adoption. In

these two ways, interest groups play a pivotal role in precipitating punctuated dynamics in

public policy.

Empirical approaches in punctuated equilibrium

The complex processes contributing to negative and positive feedback processes over time make

it diﬃcult to evaluate punctuated equilibrium theory. Positive feedback cycles result from sto-

chastic processes, and it is nearly impossible to anticipate when or how exogenous shocks will

galvanize macropolitical attention over time (Jones and Baumgartner 2005). Sudden shifts in

systemic issue attention can be triggered by factors as distinct as an unexpected environmental

crisis, a coordinated social movement, changes in media coverage regarding a new problem, or a

long-term shift in policy indicators leading to heightened political concern about a new pro-

blem. These unpredictable punctuations are the result of complex interactions between available

policy solutions, participants, and receptive venues at a critical moment. Studies of agenda-setting

assert that shifts in policy-making occur when factors converge to open a window of opportunity

for policy change, but these approaches have oﬀered no universal method for understanding

when such shifts in policy attention will occur.

Preliminary research in punctuated equilibrium theory focused on explaining how complex

interactions precipitate negative and positive feedback cycles in public policy-making, drawing

heavily on longitudinal case studies to illustrate how exogenous shocks, changes in the policy

image, the receptivity of institutional venues, and the participation of interest groups interact

over time to produce policy change within a speciﬁc case (Baumgartner and Jones 2009). These

studies of comparative issue dynamics conﬁrmed the patterns of attention allocation leading to

widespread policy change (with the exception of Givel 2006).

These studies provided a robust

understanding of the process leading to negative and positive feedback cycles across institutions

and issue areas, and have further permitted researchers to isolate how shifts in the policy image

(Baumgartner and Jones 2009; Baumgartner et al. 2009), the participation of interest groups and

policy entrepreneurs (Boushey 2010), venue shopping (Pralle 2003) and the decision rules of distinct

institutional venues (Baumgartner et al. 2009) contributed to punctuated dynamics in public

policy-making.

To test the broader implications of punctuated equilibrium theory, researchers have shifted

their focus to aggregate patterns of policy outputs across issues and institutions over time, exploring

longitudinal changes in budgetary outlays, law-making, executive orders, public opinion, media

attention and other agenda-setting processes. Over the last decade, researchers working to

understand the evolution of issue attention in government have gathered impressive collections

of data that captures government attention over time. In the United States, Bryan Jones, Frank

Baumgartner, and John Wilkerson have organized the Policy Agendas Project, an invaluable

resource regarding trends in the federal budget, congressional hearings, important problem public

opinion data, media coverage, executive orders and other indicators of national policy-making atten-

tion from 1946 onward.

The success of this initial project inspired the Comparative Agendas

Boushey

146

Project, a collaborative endeavor documenting parallel indicators of longitudinal policy atten-

tion across 14 Western democracies.

These resources have classiﬁed policy attention by major and

subtopic codes, allowing researchers to evaluate both aggregate patterns of policy change while also

tracking the rise and fall of speciﬁc issues and policy problems across governments over time.

Punctuated equilibrium theorists have relied on these resources to evaluate how cognitive

constraints and institutional rules shape policy outputs over time. This research focused initially

on the general punctuation hypothesis, exploring whether policy change over time could be

best explained through incrementalism or disproportionate information processing. More recently,

this research has branched out to model how institutions shape information processing.

Because of the overwhelming complexity and unpredictability of policy change in political

systems, punctuated equilibrium theory has employed stochastic process models to evaluate

punctuated dynamics in public policy-making (Breunig and Jones 2011; Breunig and Koski

2009; Baumgartner et al. 2009; Robinson and Caver 2006; Jones and Baumgartner 2005; Jordan

2003; True et al. 1999; Padgett 1980). To evaluate the general punctuated hypothesis, punctuated

equilibrium has examined the underlying probability density distributions of policy inputs and

outputs data, beginning with research on public budgeting (True et al. 1999) and expanding out

to shifts in policy-making in other institutional contexts such as elections, bill introductions and

question periods, legislative hearings, shifts, and media coverage of events over time (Baumgartner

et al. 2009; Jones and Baumgartner 2005; Jones et al. 2003). This research has followed Padgett’s

(1980) seminal application of stochastic process models in public budgeting, which demon-

strated that aggregate distributions of policy change over time could be used to distinguish

between incremental and non-incremental patterns of change.

This approach is justiﬁed by the theoretical distributions associated with incremental decision-

making and disproportionate information processing. Incrementalism implies a normal distribu-

tion, as policy change occurs through a gradual adjustment with larger shifts in policy occurring

equally around the mean (Breunig and Jones 2011; Baumgartner and Jones 2005; True et al.

1999; Padgett 1980). Punctuated equilibrium theory implies leptokurtic distributions, described

by “fat tails (indicative of internal reprioritizations and external policy punctuations), and sharp

central peaks (indicative of internal inattentiveness and external temporal stability)” (Breunig

and Jones 2011; Jones and Baumgartner 2005).

Figure 11.1 illustrates policy change distributions associated with incremental and dispropor-

tionate information processing. The incremental histogram on the left follows a normal distribu-

tion of budgetary change over time. Here, annual changes in budgets are clustered around the

mean, and larger changes are evenly distributed, with extremely large punctuation representing

extremely rare events. In the punctuations histogram budgetary change is typically very small, as

evidenced by the central cluster of data around the mean. Moderate policy adjustment occurs

less frequently and major policy change more frequently than anticipated by incremental model.

This is consistent with a process of volatile policy change that occurs through long periods of

neglect and sudden moments of extreme change.

Researchers have supplemented these ocular checks of punctuated dynamics with several

statistical measures of the distribution of policy change data.

Most commonly, researchers have

estimated the L-Kurtosis (LK) or “peakedness” of a given distribution. A normal distribution can

be described as mesokurtic, and is smooth and symmetrical around the mean. Leptokurtic distribu-

tions are indicated by sharp central peaks and fat tails. LK provides a simple method for distin-

guishing between mesokurtic and leptokurtic distributions, returning values in intervals between

0 and 1, with higher values indicating non-normality.

The incremental histogram in Figure 11.1

produces an LK score of 0.12, indicative of a normal distribution. The punctuated curve on the

right produces a value of 0.17, indicative of an exponential or power law distribution.

Punctuated equilibrium theory

147

Applied to the study of policy change over time, stochastic process models therefore provide

a relatively simple method for distinguishing between policy-making driven by strict increment-

alism and policy-making driven by punctuated equilibrium. Policy change distributions that are

normally distributed indicate a process of incremental decision-making. Policy change distribu-

tions producing sharp central peaks and fat shoulders imply decision-making processes consistent

with punctuated dynamics in policy-making (Jones and Baumgartner 2005).

Punctuated equilibrium theory has made use of stochastic process models to assess patterns of

decision-making across institutional and historical contexts. These studies have provided robust

support for the general punctuation hypothesis, as virtually all patterns of policy change over

time produce punctuated dynamics consistent with disproportionate information processing.

Distributions of policy change data consistent with punctuated equilibrium theory have been

identiﬁed in studies of public budging in federal budgeting ( Jones and Baumgartner 2005), state

budgeting (Breunig and Koski 2006), and local government budgeting ( Jordan 2003). A com-

parison of national budgeting across the United States, France, Germany, Great Britain, Denmark

and Canada conﬁrmed that budget change across nations produces punctuated dynamics ( Jones

et al. 2009). This growing body of research suggests that incrementalism is one process of

decision-making embedded in a larger model of disproportionate information processing.

More recently studies of punctuated equilibrium have employed stochastic process models to

understand how diﬀerent institutions shape punctuated dynamics in policy-making. Here,

researchers have compared the shape and kurtosis of change data to understand how the institu-

tional context shapes punctuated dynamics in policy-making. Some institutions impose consider-

able friction against policy change while others are more responsive. This research has qualiﬁed

PET’s earlier understanding of disproportionate information processing, illustrating that varia-

tion in the decision rules and transaction costs imposed across institutions shapes the frequency

and magnitude of policy punctuations. Baumgartner and his coauthors (2009) compare the

kurtosis values associated with policy change distributions across multiple policy-making insti-

tutions in the United States, Denmark and Belgium, ﬁnding that institutions that impose larger

barriers to policy change are also more susceptible to large policy punctuations. Although all

policy-making departs from normal incremental decision-making process, there is considerable

variation in the slip stick dynamics imposed by diﬀerent forms of institutions. Government activ-

ities such as elections produce smaller L-kurtosis values than more rigid policy output processes,

Figure 11.1 Incremental and punctuated policy change distributions

Source: Breunig, Christian and Chris Koski. 2009. ‘Punctuated Budgets and Governor’s Institutional

Powers’, American Politics Research 37(6): 1116–38.

Boushey

148

such as lawmaking and budgeting (Baumgartner et al. 2009). Breunig and Koski (2009) conﬁrm

these dynamics in their study in US state-level gubernatorial powers. The executive’s ability to

shape state budgets and veto undesirable legislation contributed to larger punctuations than

states with weaker executive oﬃces.

Conclusion

The renewed focus on stochastic process models points to one of the most promising areas for

future research in punctuated equilibrium. The initial research testing the general punctuation

hypothesis has demonstrated that policy-making cannot uniformly be explained as resulting

from an incremental decision-making process. However, research has only recently started to

identify distributions commonly associated with disproportionate information processing in policy-

making (Breunig and Jones 2011). Future research will allow us to diﬀerentiate between incre-

mental decision-making, sudden punctuations in response to exogenous shocks, and policy change

triggered by changes in factors endogenous to a given political system. PET researchers have

already begun to test how closely policy change distributions associated with diﬀerent policy

processes match speciﬁc outcome distributions such as the power law or exponential distribu-

tions that result from distinct decision-making processes (Jones et al. 2009). Identifying the dis-

tributions attached to speciﬁc mechanisms of policy change beyond incrementalism will help us

understand the processes leading to policy change over time. Boushey (2010) integrated research

from agenda-setting and epidemiology to distinguish between incremental and punctuated

patterns of policy diﬀusion in the United States. This research suggested that innovations that

encourage positive feedback cycles in policy-making could produce steep S-shaped distributions

representative of internal contagion, leading to rapid diﬀusion across states, or stark R-shaped

exponential distributions in response to exogenous factors.

The data resources prepared in the Policy and Comparative Agendas projects will also lead to

new discovery in the ﬁeld. These resources provide leverage not only for testing how diﬀerent

institutions shape punctuated dynamics in policy-making, but also for testing how the interac-

tion of policy-makers and policy ideas across systems shapes patterns of policy change over time.

The time series measures of attention across countries will eventually provide researchers with a

powerful resource for understanding how the serial allocation of political attention across countries

shapes patterns of agenda-setting and policy adoption over space and time. For example, research

focusing on the policy responses of domestic and international political systems to exogenous

shocks will help us to understand how diﬀerent institutions respond to common exogenous shocks.

Current research shows how aggregate patterns of policy change generally unfold across diﬀer-

ent institutional contexts. Future research will identify how diﬀerent institutions interact with

common exogenous signals in policy-making.

Over the last two decades, researchers have expanded on the initial model of punctuated

equilibrium theory presented in Agendas and Instability in American Politics. On the one hand, PET

has added a much richer account of the decision-making model leading to incremental adjust-

ment and sudden positive feedback cycles. This has grounded punctuated equilibrium theory,

illustrating how disproportionate information processing leads to shifts in attention and policy

preferences in individuals and the institutions they populate. At the same time, PET theory has

expanded on the macropolitical implications of punctuated equilibrium theory, moving from a

study of policy change in the United States to a more general model of agenda-setting and

policy-making within organizations. The assumptions of PET theory have been evaluated in case

studies across issue areas and at the aggregate level across institutions. The broad interest in the

theory has generated a strong consensus regarding how policy change occurs over time, and

Punctuated equilibrium theory

149

points to a promising research agenda as new scholars identify how the allocation of policy

attention in organizations and institutions leads to profound shifts in public policy.

Notes

1 This section highlights only the subset of the major health reforms in the PPACA. An interested reader

may wish to refer to the Kaiser Family Foundation’s (2010) “Summary of the new health reform law”

for a succinct 13-page summary of the key provisions of the PPACA.

2 To conceptualize the pressures leading to incremental and sudden policy change within a single poli-

tical system over time, Baumgartner and Jones (1993) borrowed from evolutionary biology and

paleontology, drawing on Eldredge and Gould’s punctuated equilibrium theory of speciation (1972).

For a comparison of the PET framework in biology and public policy, see Robinson (2006). For a

critique of the metaphor, see Givel (2010).

3 Punctuated equilibrium theory borrows heavily from prior research in bounded rationality in policy-

making to explain pressures leading to policy stability and change. For more extensive reviews of

bounded rationality or incrementalism in policy-making, refer to Jones and Hayes’ chapters in this

volume.

4 The stability of participants at the subgovernment level has been well-documented in the study of

American politics. Researchers have alternately labeled this organization of elected oﬃcials, interest

group activists and bureaucrats as iron triangles, policy subsystems, or issue networks.

5 For far more expansive lists of the many contributions of punctuated equilibrium theory, refer to the

publications page at the Comparative Agendas Project web page (www.comparativea gendas.org/pub lications)

or the Policy Agendas publication web page (www.policyagendas.org/biblio).

6 Givel (2006) contends that punctuated equilibrium theory fails to explain the stability of tobacco policy

in the face of widespread pressure for policy reform. Baumgartner and Jones (2009) revisit the case of

tobacco reform in the second edition of Agendas and Instability in American Politics. They argue that the

pro-industry tobacco policy monopoly has been thoroughly destroyed at the federal level, and that radical

reforms to tobacco control and regulation have occurred at the state and local levels of government.

7 These data are available at the policy agendas website at: www.policyagendas.org.

8 More detailed information on the Comparative Agendas Project can be accessed at: www.CAP.org.

9 Researchers have employed other simple tests evaluate the normality of a given probability density

function. Tests such as the Shapiro-Wilk, Anderson Darling or the Kolmogorov–Smirnov test indicate

whether a given distribution is consistent with a normal (incremental) decision-making process.

10 Prior research in punctuated equilibrium employed standard measures of kurtosis (K) to estimate the

shape of the distribution. For K, normal distributions have kurtosis values of 3, with larger values

indicating leptokurtic distributions. Breunig and Jones (2011) suggest L-Kurtosis as an alternative, as

this measure results in more stable estimates of the overall shape of the distribution.

Bibliography

Aaron, Paul and David Musto. 1981. ‘Temperance and prohibition in America: A historical overview’,in

Mark Moore and Dean Gerstein (eds) Alcohol and Public Policy: Beyond the Shadow of Prohibition.

Washington, DC: National Academy Press, 125–81.

Baumgartner, Frank and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago: University

of Chicago Press.

——and Bryan D. Jones. 2002. ‘Positive and negative feedback in politics’,inPolicy Dynamics. Chicago:

University of Chicago Press

——and Bryan Jones. 2009. Agendas and Instability in American Politics, 2nd edn. Chicago: University of

Chicago Press.

——, Suzanna De Boef and Amber Boydstun. 2008. The Decline of the Death Penalty and the Discovery of

Innocence. New York: Cambridge University Press.

——, Christoﬀer Green-Pedersen and Bryan D. Jones (eds). 2008. Comparative Studies of Policy Agendas.

New York: Routledge.

——, Christian Breunig, Christoﬀer Green-Pedersen, Bryan D. Jones, Peter B. Mortensen, Michiel

Nuytemans and Stefaan Walgrave. 2009. ‘Punctuated equilibrium in comparative perspective’, American

Journal of Political Science 53(3): 603–20.

Boushey

150

Berry, Frances S. and William D. Berry. 1999. ‘ Innovation and diﬀusion models in policy research’, in Paul

A. Sabatier (ed.) Theories of the Policy Process. Boulder, CO: Westview Press, 169– 200.

Birkland, Thomas A. 1997. After Disaster: Agenda Setting, Public Policy, and Focusing Events. Washington,

DC: Georgetown University Press.

Boushey, Graeme. 2010. Policy Diﬀusion Dynamics in America. Cambridge: Cambridge University Press.

Breunig, Christian and Chris Koski. 2006. ‘Punctuated equilibria and budgets in the American states’,

Policy Studies Journal 34(3): 363–79.

——and Chris Koski. 2009. ‘Punctuated budgets and governors’ institutional powers’, American Politics

Research 37(6): 1116–38.

——and Bryan Jones. 2011. ‘Stochastic process methods with an application to budegetary data’, Political

Analysis 19(1): 103–17.

Cashore, Benjamin and Michael Howlett. 2006. ‘Behavioral thresholds and institutional rigidities as expla-

nations of punctuated equilibrium processes in Paciﬁc Northwest Forest Policy Dynamics’, in Robert

Repetto (ed.) Punctuated Equilibrium in US Environmental Policy. New Haven, CT: Yale University Press,

pp. 137–61.

Donovan, Mark. 2001. Taking Aim: Target Populations and the Wars on Aids and Drugs. Washington, DC:

Georgetown University Press.

Eldredge, Niles and Stephen Gould. 1972. ‘Punctuated equilibria: An alternative to phyletic gradualism’,in

T. Schopf (ed.) Paleiobiology, San Francisco: Freeman, Cooper, 82–115.

Givel, Michael. 2006. ‘Punctuated equilibrium in limbo: The tobacco lobby and US state policy making

from 1990 to 2003’, Policy Studies Journal 43(3): 405–18.

—— 2010. ‘The evolution of the theoretical foundations of punctuated equilibrium theory in public

policy’, Review of Policy Research 27(2): 187– 98.

Hayes,

Michael. 1992. Incrementalism and

Public Policy. New York: Longman.

Holyoke, Thomas. 2003. ‘Choosing battlegrounds: Interest group lobbying across multiple venues’, Political

Research Quarterly 56: 325–36.

Iyengar, Shanto. 1996. ‘ Framing responsibility for political issues’, Annals of the American Academy of Political

and Social Science 546(1): 59–70.

Jones, Bryan D. 1994. Reconceiving Decision-Making in Democratic Politics: Attention, Choice, and Public Policy.

Chicago: University of Chicago Press.

—— 2001. Politics and the Architecture of Choice: Bounded Rationality and Governance. Chicago: University of

Chicago Press.

——and Frank Baumgartner. 2004. ‘Representation and agenda setting’, Policy Studies Journal 32: 1–25.

——and Frank Baumgartner. 2005. The Politics of Attention: How Government Prioritizes Problems. Chicago:

University of Chicago Press.

——, Tracy Sulkin and Heather Larsen. 2003. ‘Punctuations in American political institutions’, American

Political Science Review 97(1): 151–69.

——, Frank R. Baumgartner, Christian Breunig, Christopher Wlezien, Stuart Soroka, Martial Foucault,

Abel François, Christoﬀer Green-Pedersen, Peter John, Chris Koski, Peter B. Mortensen, Frédéric

Varone and Stefaan Walgrave. 2009. ‘A general empirical law of public budgets: A comparative analysis’,

American Journal of Political Science 53(4): 855–73.

Jordan, Meagan. 2003. ‘Punctuations and agendas: A new look at local government budget expenditures’,

Journal of Policy Analysis and Management 22: 345–60.

Kaiser Family Foundation. 2010. ‘Summary of new health reform law’, Focus on Health Reform. Available

at: www.kﬀ.org/healthreform/upload/8061.pdf (accessed 23 November 2011).

Karch, Andrew 2007. ‘Emerging issues and future directions in state policy diﬀusion research’, State Politics

and Policy Quarterly 7(1):

54–80.

Kingdon, John. 1984. Agendas,

Alternatives, and Public Policies. Boston: Little, Brown.

Lindblom, Charles. 1959. ‘The science of muddling through’, Public Administration Review 19: 79–88.

Padgett, John F. 1980. ‘Bounded rationality in budgetary research’, American Political Science Review 74:

354–72.

Pralle, Sarah. 2003. ‘Venue shopping, political strategy, and policy change: A case study of Canadian forest

advocacy’, Journal of Public Policy 23: 233–60.

Repetto, Robert (ed.). 2006. Punctuated Equilibrium and the Dynamics of US Environmental Policy. New

Haven, CT: Yale University Press.

Robinson, Scott. E. 2006. ‘Punctuated equilibrium models in organizational decision making’, in Goktug

Morcol (ed.) Handbook on Human Decision Making. Boca Raton, FL: CRC Press, 133 – 49.

Punctuated equilibrium theory

151

—— and Floun’say R. Caver. 2006. ‘Punctuated equilibrium and Congressional policymaking’, Political

Research Quarterly 59(1): 161–6.

—— and Kenneth J. Meier. 2006. ‘Path dependence and organizational behavior: Bureaucracy and

internal standards’, American Review of Public Administration, 36(3): 241–60.

Schattschneider, Elmer E. 1975. The Semi-sovereign People. Hinsdale, IL: Dryden Press.

Schneider, Anne Larason. 2006. ‘Patterns of change in the use of imprisonment in the American states: An

integration of path dependence, punctuated equilibrium and policy design approaches’, Political Research

Quarterly 59(3): 457–70.

—— and Helen Ingram. 1993. ‘The social construction of target populations: Implications for politics and

policy’, American Political Science Review 87(2): 334–47.

Shiﬀman, Jeremy, Tanya Beer and Yonghong Wu. 2002. ‘The emergence of global disease control priorities’,

Health Policy and Planning 17(3): 225–34.

Smith, Mark A. 2007. The Right Talk: How Conservatives Transformed the Great Society into the Economic

Society. Princeton, NJ: Princeton University Press.

True, James and Glenn Utter. 2002. ‘Saying “yes”, “no”, and “load me up” to guns in America’, The

American Review of Public Administration 32(2): 216–41.

——, Bryan Jones and Frank Baumgartner. 1999. ‘Punctuated-equilibrium theory: Explaining stability and

change in American policymaking’, in Paul Sabatier (ed.) Theories of the Policy Process. Boulder, CO:

Westview Press, 155–88.

——, Bryan Jones and Frank Baumgartner. 2007. ‘Punctuated equilibrium theory: explaining stability and

change

in American policymaking’,

in Paul Sabatier (ed.) Theories of the Policy Process, 2nd edn. Boulder,

CO: Westview Press, 155–88.

Tsebelis, George. 2002. Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University

Press.

Walgrave, Stefaan and Frédéric Varone. 2008. ‘Punctuated equilibrium, and agenda setting: Bringing parties

back in. Policy change after the Dutroux crisis in Belgium’, Governance: An International Journal of Policy

and Administration 21(3): 365–95.

Wildavsky, A. 1964. The Politics of the Budgetary Process. Toronto: Little, Brown.

Wood, Robert. 2006. ‘Tobacco’s tipping point: the master settlement agreement as a focusing event’,

Policy Studies Journal 34(3): 419–36.

Workman, Samuel, Bryan D. Jones and Ashley E. Jochim. 2009. ‘Information processing and policy

dynamics’, Policy Studies Journal 37(1): 75–92.

Boushey

152

Policy network models

Chen-Yu Wu and David Knoke

Deﬁning policy networks

The overarching goal of policy network analysis is to understand how relationships between

actors involved in policy-making determine the outcomes of collective policy decisions (Compston

2009; Knoke 2011). To this end, policy network analysis consists of two distinct components.

First, one has to identify the important actors involved in the policy-making process. Although

policies are proposed, debated, and passed by legislators and other governmental entities, extra-

governmental actors such as non-governmental organizations (NGOs), interest groups, or even

inﬂuential individuals can be involved in inﬂuencing policies. Second, one has to describe the

type of social interactions that occur between actors during the policy-making process. Both

components are vital to enabling the researcher to explain or predict policy outcomes.

In short, policy network analysis provides scholars with an analytical framework to describe the

social network dimensions of policy-making. However, policy network analysis can also be

used to answer the fundamental research questions of social network analysis, namely: how

these networks form, how and why they persist over time, and also how they change.

Social network theory

Social network analysis consists of theories and methods for examining structural relations among

social actors and for explaining the consequences of those connections. Network theories can

apply to structures and processes occurring either in small, closed social systems (such as artiﬁ-

cially constructed experimental groups, or natural settings such as classrooms and work teams) or

in larger, open systems (such as communities, markets, and international relations). Network the-

ories and data analyses may apply to the egocentric, interpersonal, subgroup, or complete network

levels of analysis, as well as to cross-level and time-dependent eﬀects. At the micro-level, net-

work analysts examine how people form, change, or drop ties such as friendship, giving assis-

tance and advice, exchanging information, and quarreling. Interpersonal interactions aggregate

into meso-level social structures that aﬀect collective actions and performance outcomes of small

groups such as gangs, teams, and crews. Macro-level systems of cities, economies, nations, and

transnational relations may be studied using conceptual and methodological techniques

153

developed through decades of social network theory and research. The interdisciplinary ﬁeld

originated in anthropology, small group, social psychology, and sociology in the mid-twentieth

century, and really took oﬀ around 1970 as theoretical interest in structural analysis converged with

increasingly sophisticated computer software for analyzing network data based on matrix algebra

methods. The exponential growth of social research and proliferation of academic specialties intro-

duced social network analysis to increasing numbers of basic, substantive, and applied disciplines,

including political sociology, political science, and public administration.

As a minimum a social network consists of a set of actors (people, groups, organizations, nations)

and one or more relations that connect some of them (e.g., communication, trust, purchase,

attack). A boundary- or scope-condition speciﬁes which actors are included within a particular

network; for example, by formal membership as in a social club; by residence as in a neigh-

borhood; or by informal criteria such as attendance at ﬂea markets. Network theories typically

make three core assumptions about the mutual inﬂuences among actors: (1) the social structure

of any complex system consists of the stable patterns of repeated interactions connecting actors

to one another; (2) these social relations are the primary explanatory units of analysis, rather

than the attributes and characteristics of the individual actors; and (3) the perceptions, attitudes,

and actions of actors are shaped by the multiple structural networks within which they are

embedded, and in turn their behaviors can change these networks’ structures (Knoke 2001: 63–4).

Political networks are a subset of social networks in which the primary relations among actors

involve inequalities in power. Asymmetrical power rarely depends on overt coercive force and

more often relies on taken-for-grant beliefs about the possession of authority to issue commands

and to expect compliance with orders (Max Weber’s legitimate power). Political power is

inherently relational: one or more actors seek resources and other advantages despite resistance

by others. Hence, social network theory is well-suited to investigate the diverse structures and

consequences of power, from the bedroom to the boardroom to the battleﬁeld. Political net-

works can be observed at micro-levels, where friends and kin try to persuade one another to

vote and participate in electoral campaigns. At the most macro-level, political networks in the

international system are evident in shifting patterns of military alliances and interstate aggression

(e.g. Maoz 2010). This chapter concentrates on meso-level political networks comprised mainly

of formal organizations engaged in collective eﬀorts to inﬂuence public policy decisions within

nation-states.

To explain public policymaking at various governmental levels, network theorists draw

attention to the policy network as an appropriate unit of analysis. In local governments, important

political actors are typically powerful individuals (e.g. mayor, council members). But, at national

levels, inﬂuential actors are more often formal organizations such as political parties, legislative

committees, executive agencies, ministries, and interest organizations including labor unions,

business associations, and public interest groups. Kenis and Schneider (1991: 41) argued, “Policy

networks should be conceived as speciﬁc structural arrangements in policymaking.” They

deﬁned a policy network as set of public and private corporate actors linked by communication

ties for exchanging information, expertise, trust, and other political resources:

A policy network is described by its actors, their linkages and its boundary. It includes a

relatively stable set of mainly public and private corporate actors. The linkages between the

actors serve as channels for communication and for the exchange of information, expertise,

trust and other policy resources. The boundary of a given policy network is not in the ﬁrst

place determined by formal institutions but results from a process of mutual recognition

dependent on functional relevance and structural embeddedness.

(ibid.)

Wu and Knoke

154

A closely related concept, the policy domain, is socially constructed by political actors who mutually

recognize that their preferences on policy events must be taken into consideration by other

domain participants (Laumann and Knoke 1987: 10). A policy domain is deﬁned as any political

subsystem

identiﬁed by specifying a substantively deﬁned criterion of mutual relevance or common

orientation among a set of consequential actors concerned with formulating, advocating, and

selecting courses of action (i.e., policy options) that are intended to resolve the delimited

substantive problems in question.

(Knoke and Laumann 1982: 256)

Some policy domain examples are education, agriculture, national defense, welfare (Laumann

and Knoke 1987: 10), health, energy, and transportation (Burstein 1991: 328). The policy network

and the policy domain concepts are interrelated, with a policy domain delineating a bounded

system within which its organizational participants are interconnected by one or more policy

networks.

Five basic types of interorganizational relations exhibiting distinctive policy network struc-

tures, include: resource exchange; information transmission; power relations; boundary pene-

tration; and sentimental attachments (Knoke 2001: 65). Resource exchanges in policy networks,

such as money or personnel, are usually voluntary transfers, although occasional government

mandates impose interorganizational connections. Information exchanges range from technical

and scientiﬁc data to policy advice and advocacy. Power relations include both formal and informal

inequalities in authority and dominance, with public sector entities usually controlling more

power than private sector organizations to impose their interests on a domain. Boundary penetra-

tion refers to two or more actors coordinating their actions for a common goal, such as lobbying as

discussed below. Sentimental attachments are subjective and emotional aﬃliation expressing

solidarity and political support as exempliﬁed by labor unions and social movement organiza-

tions. The ﬁve basic types of interorganizational relations can be operationalized by a multitude

of empirical indicators and jointly analyzed to reveal network structures connecting policy

domain organizations in complex patterns.

An interorganizational perspective on policy domains and policy networks comprised the

theoretical and methodological basis for the organizational state model of policymaking that

Edward Laumann, David Knoke, and Franz Pappi constructed for investigating the US national

energy and health domains (Laumann and Knoke 1987), and comparative network analysis of

the US, German, and Japanese labor domains (Knoke et al. 1996). See Knoke (1998) for a

detailed overview of these and related projects. The organizational state model views national

policy-making processes as dynamic inﬂuence actions among formal organizations, with elite

persons acting as agents of these organizational principals. Within every policy domain, the set

of core organizational players are drawn from both the public and private sectors. Some key

organizations may have intense interests in numerous policy issues and participate in eﬀorts to

aﬀect the policy outcomes of dozens of legislative, executive, and judicial decisions. Other organi-

zations focus more narrowly on a single issue and mobilize only rarely to pursue their interests.

Given the diversity of organizational interests, limited resources, and complexity of inﬂuence

dynamics, no organization or small subset has the capacity to control, let alone dominate, a policy

domain’s policy-making processes. Instead, most policy ﬁghts consist of multiorganizational

coalitions brought temporarily together to work collectively on shaping policy proposals and

advocating a speciﬁc preferred policy solution to those authorities capable of rendering a bind-

ing decision on a particular policy event (e.g. a regulatory ruling, legislative act, court decision).

Policy network models

155

A policy domain’s communication and resource exchange networks enable interested organi-

zations to identify collaborators and opponents that can be mobilized on competing sides of a

policy event. Frequently, some members of an action set—organizations that all hold the same

outcome preference on the event, are connected in a communication network, and coordinate

their policy to inﬂuence actions—pool their political resources and launch coordinated cam-

paigns to pressure governmental decision-makers into choosing the policy outcome favored by

the coalition members.

Some organizations work solo but chances of a favorable outcome are usually greater when

coalitions create an eﬃcient division of labor that maximizes their strengths and expertise.

Broad-based membership organizations such as labor unions mobilize mass constituencies to

man phone banks and email legislators. Organizations with well-funded staﬀ assemble research

evidence to present during testimony at public legislative and regulatory hearings. Political con-

sultants produce persuasive head-counts at closed-door meetings with elected oﬃcials. And deep-

pocketed political action committees always have an open door to make their cases to cash-starved

politicians. Lobbying proceeds neither by political bribery nor by striking overt quid pro quo

deals, but by making a more persuasive case in an appropriate decision-making arena (Brown

1998). Lobbyists succeed by providing policy-makers with useful information and data, sub-

stantive arguments, proposed policy language, and astute political arguments for supporting one

side of a policy ﬁght over the alternatives.

After a binding policy decision is made, the coalitions disband as new policy events attract

diﬀerent combinations of organized interest groups. Viewed over time, the policy network

structures of policy domain appear continually in ﬂux. Yet, despite much instability at the micro-

level, relatively durable macro-level cleavages may emerge and persist. Examples of such enduring

cleavages are the business-vs.-labor, rural-vs.-urban, and religious-vs.-secular divisions occurring

in many nations. Many national policy domains contain fairly stable policy networks whose

boundaries and components persist over substantial time (Burstein 1991). However, as “social

constructions, created through interactions among organizations and given meaning by culture”

(1991: 345), policy domains are susceptible to reconceptualization and evolutionary transforma-

tions. Scant theory seeks to explain the historical origins of new policy domains or the trans-

formation of existing domains. An exception is Knoke’s (2004) schematic model of six elements

(focusing events, technological innovations, political entrepreneurs, issue framing, policy networks,

and institutionalization), which he oﬀered as “a provisional framework to initiate empirical research

on the sociopolitical construction of national policy domains” (2004: 94). So far, no takers.

History and development of policy network analyses

According to Knoke (2011), the term “policy network” was ﬁrst used by Katzenstein (1976) in

his article comparing the foreign economic policies of France and the United States. Although

Katzenstein did not explicitly deﬁne policy networks, he emphasized that policy networks form

a crucial link between state and society. Additionally, he proposed a simple continuum for

describing types of policy networks. One ideal type (to use the Weberian term) is the state-

centered policy network, which is deﬁned by political power and dominance; the other ideal

type is the society-centered policy network, which in contrast is deﬁned by economic power.

In practice, Katzenstein stated that all policy networks in “advanced industrial states” would fall

somewhere between the two extremes.

Since the publication of Katzenstein’s article, the volume of work on policy networks has

grown immensely. These works encompass policy-making on all levels, from the local level (e.g.

Laumann and Pappi 1976) to comparative (e.g. Marsh 1998), regional (e.g. Adshead 2002), and

Wu and Knoke

156

transnational analyses (e.g. McGann and Sabatini 2011; Witte et al. 2000). As the ﬁeld of policy

network analysis expanded, three distinct schools of thought emerged. In the following sub-

sections, we will brieﬂy discuss the US, British, and German perspectives on policy network

analysis.

United States. The American approach to policy network analysis is rooted in Laumann and

Pappi’s 1973 study, New Directions in the Study of Community Elites. In the study, as well as

subsequent replications (e.g. Galaskiewicz 1979; Laumann et al. 1978), researchers employed the

network analysis of power structures and found that actors in central network positions were

more engaged in local civic life, and as a result, enjoyed a better chance of inﬂuencing public

policy debates in their favor.

Laumann and Knoke extended this framework to national policy domains and the formal orga-

nizations within them in The Organizational State (1987), and found that the networks among

these organizations enabled organizations with opposing interests to pool their resources and

mobilize them in pursuit of commonly-desired policy outcomes. Furthermore, the authors found

that, at the national level, it is often the case that diﬀerent organizations simultaneously have

common and opposing policy interests.

In another study, Knoke et al. (1996) found that, like policy-making at the local level, actors

in centrally located network positions could more easily inﬂuence policy decisions. In their

comparative study of US, German, and Japanese policy domains, they make the distinction between

communication networks (characterized by exchanges of information) and support networks

(characterized by resource exchanges), and found that occupying a central position in commu-

nication networks provided organizations with more political inﬂuence in the US and Germany,

while the opposite was true for Japan (1996: 120). These ﬁndings suggest that the processes

involved in policy-making are not necessarily universal, but rather culturally determined.

Germany. The German approach to policy network analysis, like the American approach, is

rooted both in the previously discussed study by Laumann and Pappi (1973), and also in

Lehmbruch’s (1984, 1989) studies of corporatist politics the Federal Republic of Germany (FRG).

Compared to other more centralized states, such as the United Kingdom, the FRG’s relative

noncentralized policy networks necessitated exchanges of resources among many organizations

involved in policy-making (Lehmbruch termed this corporatist concertation). In this policy

environment, Lehmbruch (1984) found that these interorganizational policy networks serve as

an important structural and institutional constraint moderating the outcomes of collective policy

decisions. Equally importantly, Lehmbruch (1989) concluded that the formal organizations involved

in the policy-making process can inﬂuence the structure and dynamics of policy networks.

Great Britain. Unlike the American and German approaches, which analyze policy networks

using theories of power structures, British scholars’ analytical framework involves highlighting

the contrasts between diﬀerent networks’ pluralist and corporatist characteristics (e.g. Rhodes

1985, 1990). Additionally, unlike American and German scholars, British policy network analysis

places more emphasis on more ﬂuid and unpredictable policy communities rather than entren-

ched governmental and subgovernmental units (Richardson 2000; Rhodes 1990). As such,

British scholars produced a wide range of literature on self-organizing groups involved in the

policy-making processes for policy domains such as agriculture, employment, and informational

technology.

Within the British school of thought, one of its more inﬂuential works is Marsh and Rhodes’

(1992) book, Policy Networks in British Government, which provided an analytical approach empha-

sizing the importance of structural relations between governmental entities, interest groups, and

informal actors, as opposed to interpersonal relations among individual actors. Equally impor-

tantly, their approach established a unidirectional direct link between policy networks and

Policy network models

157

policy outcomes. Critics of this approach, notably Dowding (1995, 2001), have argued that it

has two glaring weaknesses: ﬁrst, it lacks a theoretical basis; and second, it does not provide any

basis for explaining network transformation (which is, as previously mentioned, an important

objective for policy network scholars). In response, various authors (e.g. Marsh and Smith 2000;

Kisby 2007) have oﬀered explanations and models addressing these criticisms.

Key ideas in modern policy network analysis

One of the features that all of these views of policy network analysis share, to some extent, is

the idea of resource dependency (Benson 1982; Rhodes 1985). In other words, in the realm of

policy-making, every actor desires resources that other actors control. As a result, every actor is

prepared to exchange or give up some resources in order to obtain resources which they do not

have (Compston 2009: 7). These exchanges are not unregulated or arbitrary; rather, actors employ

speciﬁc strategies which conform to mutually accepted “rules of the game” (Rhodes 1985: 4–5).

These rules can either be formally codiﬁed (e.g. laws regulating campaign contributions from

interest groups) or informal (e.g. hiring former government oﬃcials to serve as consultants and

advisers to interest groups).

A common example in modern-day democracies is the exchange of resources between law-

makers and interest groups. As previously mentioned, legislators and other government oﬃcials

are the only people who can legally introduce, debate, and ultimately pass policy in the halls of

government. Such political authority is a key resource that all other actors lack, and therefore

seek access to. These actors, however, control other resources coveted by these legislators.

Namely, interest groups—and the members they represent—hold both the votes that legislators

need to win elections and the funds legislators require to wage successful campaigns. Hence,

actors like interest groups can inﬂuence or pressure legislators to push for legislation advocating

for their preferred policy outcomes.

This example highlights two key characteristics in every policy network. First is the assumption

that policy actors are rational, and that actors agree to participate in these resource exchanges if

they stand to beneﬁt from them. To illustrate, an oﬃcial with a largely conservative electorate is

unlikely to heed the demands of a liberal/progressive interest group, nor accept its donations,

since the oﬃcial does not “need” the votes of the minority to secure re-election, nor does the

oﬃcial wish to alienate his or her base constituency. This characteristic can also explain why

legislators are much less likely to compromise with colleagues with opposing views when their

political parties hold a majority in government.

The second characteristic is that all policy exchanges stem from power imbalances—political,

economic, or ideological—and that all policy exchanges result in power imbalances, although

not necessarily the same ones as before the exchange. In theory, if the playing ﬁeld was level for all

actors, there would be no need to exchange resources. Another way to approach this characteristic

is to understand that all resource exchanges result in losers as well as winners (Compston 2009).

In democracies, it is rarely the case that an individual legislator is subjected to the lobbying

eﬀorts of a single nongovernmental actor; it is more likely that individual legislators are simul-

taneously considering the requests of multiple competing actors. Hence, after every exchange,

there will be some actors which are more well-oﬀ as a result, and others who are less so.

Policy network theory and empirical research

Like every analytical framework, policy network theory has strengths and weaknesses. For instance,

early critics of policy network theory charged that policy network research largely failed to

Wu and Knoke

158

develop testable theories of both policy development and outcomes (Knoke 2011). Addition-

ally, as mentioned previously, other critics—notably Dowding (1995)—claim that the discourse

of policy network theory is not only merely descriptive, but more importantly that the precise

deﬁnitions of many concepts in policy network research have not been agreed upon universally

by the scholars conducting this research. More recent work (e.g. Mikkelson 2006) has high-

lighted some confusion among scholars about whether policy networks should be treated as

dependent or independent variables. In other words, is policy network theory a better frame-

work for understanding why and how these networks form, or is it a better tool for predicting

the outcomes and consequences of this type of collective action?

Despite these criticisms, policy network theory continues to be an extremely valuable ana-

lytical tool for understanding both social network theory in general, but also the policy-making

process. A large volume of recent work was produced by European scholars, probably due to

both the creation of new nation-states following the fall of the Soviet Union in 1991, and also

the increased prominence of the European Union in European political and civic life (Knoke

2011). The increased volume of policy network research has also seen scholars expand their

foci, including: from government to strategies of governance; from centrally focused policy-

making to multilayered policy-making (McGann and Sabatini 2011; von Winter 2001); from

top-down hierarchies to grassroots inﬂuences (von Winter 2004); and from policy formation

and outcomes to policy implementation (e.g. Greenaway et al. 2007).

Although policy network research in the United States has not kept pace with Europe,

American scholars have made some signiﬁcant contributions. For instance, some scholars

uncovered new insights into policy theory and social network theory in general through the

re-examination of older studies, while others chose to focus on and expand research on policy

domains (Knoke 2011). These new studies remain very much rooted in the American per-

spective of power structures. For instance, Carpenter et al. (2003) found that as actors’ needs for

policy-related information increases, they are much more likely to invest their eﬀorts and

resources in building stronger communication links to their allies, rather than reaching out to

their opponents. The downside of this approach, however, is that by engaging in such cliquish

behavior, network actors expose themselves to an increased risk of network failure due to their

inability to obtain complete information (2003: 433).

American scholars continue to be active in research into the social construction (and recon-

struction) of policy domains. Policy domain research—particularly research into changes in

policy domains —is a signiﬁcant component of policy network theory, as it provides researchers

with a link between culture and policy-making (Knoke 2004). In other words, changes in

policy domains “potentially transforms shared cultural meanings, taken-for-granted assumptions,

normative understandings, classiﬁcatory schemas, and tacit knowledge” (2004: 3). Its implica-

tions for policy network theory are fairly clear: new domains necessitate new policy-making

strategies to comply with new legal regulations and bureaucratic practices, as well as other

unanticipated changes.

Analyses of policy networks in non-Western states remain rare (Knoke 2011). This dearth is

most likely due to various factors, such as the lack of cultural or political traditions of civic

participation or, more basically, the inability to participate in civic life under certain forms of

government. Additionally, the few studies that have been conducted tended to be descriptive,

rather than theoretically grounded empirical research. As a result, there is a huge potential for

growth for policy network literature on non-Western countries, as well as cross-national and

comparative studies.

Another impetus for the growth of policy network analysis is the increased scholarly interest

in globalization. According to McGann and Sabatini (2011), since globalization is commonly

Policy network models

159

characterized by the global movement of capital and services, policy networks are ideal for the

study of globalization due to the ability to identify and incorporate actors across social, political,

governmental, and geographical boundaries. Additionally, two factors which occurred almost

hand-in-hand with globalization also encouraged the proliferation of policy network research:

namely, increased political and economic liberalization across the globe, and technological advances

particularly in information exchange (e.g. the Internet). These developments allow civic-minded

individuals and organizations more easily to establish transnational relationships and maintain

their channels of communication (ibid.: 67–8).

While almost all research on globalization and policy networks is relatively recent, the work

has the potential of encouraging even more growth for this analytical framework. The reason is

that global polities of governance diﬀer signiﬁcantly from national governments, or even regional

alliances. United Nations global policy-making bodies diﬀer from national and regional alliances

in that they tend to be decentralized, and that they tend to lack means of enforcing policies (for

instance, nations can choose not to ratify or implement UN resolutions usually without any fear

of retribution). Consequently, network strategies employed by actors on the global stage will

inevitably diﬀer from the strategies examined in current literature, reﬂecting the diﬀerent challenges

required to gain actors’ acquiescence and avoid disagreements.

Future directions for policy network research

We reiterate the central research questions which motivate policy network research, namely, how

these networks form, how and why they persist over time, and also how they change. Pre-

sently, most literature focuses on the second question: scholars have amassed a trove of infor-

mation about how routine activities occur within the context of already-established policy domains

and policy networks. Large gaps still exist in the literature on the origins of policy networks, as

well as how they change.

Some scholars have taken steps to address these shortcomings. For instance, Knoke (2004)

proposed an analytical framework which would enable researchers to study how policy domains—

and consequently policy networks—change. He argued that changes which are disruptive to

routine activities—which he terms “focusing events”—are central to these changes. A promi-

nent example of a focusing event, which permanently altered actors’ perceptions of the world

around them, is the September 11 terrorist attacks. Other focusing events include the exponential

growth of the Internet (Rethemeyer 2007), which enabled and empowered organizational and

individual actors alike to accomplish much more (particularly in terms of communications and

networking) than was ever possible before.

Other analytical frameworks include Marsh et al.’s (2009) attempt to merge Grant’ s (1978)

distinction between “insider” and “outsider” groups with Marsh and Rhodes’ (1992) policy

network classiﬁcation system, reasoning that each approach makes up for each others’ shortcomings.

They applied this analytical framework to a case study, and concluded that the amalgamation of

the two approaches could allow scholars to develop a better classiﬁcation system of the types of

interest groups and policy networks. Another recent framework proposed by Sandström and

Carlsson (2008) argued that while network orientation has long been a vital part of policy sci-

ence, most existing research does not focus on how a network’s structure aﬀects its perfor-

mance. Their framework proposes that an eﬀective policy network—that is, one that is both

eﬃcient and innovative—consists of a heterogeneous set of actors working very closely together.

Voets et al. (2008) noted that existing research on policy networks focuses speciﬁcally on eﬃ-

ciency and eﬀectiveness as measures of network performance, and argued that network activity

entails additional costs, such as democratic quality and capacity building. Thus, the authors

Wu and Knoke

160

employed the discourse of New Public Management to provide new assessment criteria to

measure these “other” costs of policy network activity.

Another frequently voiced criticism of the policy network approach is its lack of a theoretical

grounding. Consequently, researchers have ample opportunities for developing new ideas and

theoretical frameworks to address this gap. For instance, some researchers (e.g. Knoke 2009;

Siegel 2008) have proposed analyzing policy networks using the concept of political capital,

which is the political network equivalent of social capital (Knoke 2011).

Another possible theoretical framework for policy network research is the incorporation of

the advocacy coalition framework (ACF) to policy network analysis (Weible and Sabatier 2007).

Like policy network analysis, ACF assumes that actors cooperate with allies who share similar

policy preferences and outcomes. Additionally, ACF—unlike policy network analysis—provides

researchers with mechanisms that can explain policy changes.

Other researchers have tried incorporating new research methodology into the study of policy

networks. For example, Bevir and Richards (2009) and Pal and Ireland (2009) echo previous

criticism that policy network theory cannot be used to provide causal generalizations, and conse-

quently, legitimate advice to policy-makers. To address this shortcoming, Bevir and Richards

proposed a “decentered” approach towards the study of policy networks, arguing that research

methods such as textual analysis and ethnography would allow researchers to understand

meanings behind traditions of networks—in other words, understanding how network partici-

pants view themselves and their world. In pursuit of a similar goal, Pal and Ireland noted the

importance of creating a comprehensive map of global policy networks so that researchers can

understand broader issues aﬀecting their networks of interests, not just narrow issues and speciﬁc

organizations and individuals.

Another critique of policy network theory is that existing theory does a poor job explaining

the genesis and evolution of policy networks. Scholars have tried to address this shortcoming by

incorporating other theoretical lines of thought; for example deLeon and Varda (2009) pro-

posed a theory based on collective action theory to develop “structural signatures of exchanges”

among network participants. Others oﬀered completely new theories. For example, Provan and

Kenis (2008) proposed three models of governance with distinctive structural properties and

tried to understand why network actors choose to adopt one form over another. Ultimately,

they sought to understand the impact that each form of governance has on network outcomes,

including the evolution of both the network and of governance strategies.

Ideally, policy network research will continue to develop until it achieves characteristics we

take for granted in other more established ﬁelds of social research: a distinct and grounded

theoretical framework that explains not only how established policy networks operate, but also

how these networks form and change over time. Additionally, this theoretical framework

should enable interested scholars to generate testable hypotheses and identify distinct variables

which would enable these hypotheses to be tested empirically.

Bibliography

Adshead, M. 2002. Developing European Regions? Comparative Governance, Policy Networks and European

Integration. Burlington, VT: Ashgate.

Benson, J.K. 1982. ‘A framework for policy analysis’, in D.L. Rogers and D. Whetten (eds) Interorganizational

Coordination: Theory, Research and Implementation. Ames, IA: Iowa State University Press.

Bevir, M. and D. Richards. 2009. ‘Decentering policy networks: A theoretical agenda’, Public Administration

87: 3–14.

Browne, W.P. 1998. Groups, Interests, and US Public Policy. Washington, DC: Georgetown University

Press.

Policy network models

161

Burstein, P. 1991. ‘Policy domains: Organization, culture, and policy outcomes’, Annual Review of Sociology

17: 327–50.

Carpenter, D., K. Esterling and D. Lazer. 2003. ‘The strength of strong ties: A model of contact-making in

policy networks with evidence from US health politics’, Rationality and Society 15: 411–40.

Compston, H. 2009. Policy Networks and Policy Change . New York: Palgrave Macmillan.

deLeon, P. and D.M. Varda. 2009. ‘Toward a theory of collaborative policy networks: Identifying structural

tendencies’, Policy Studies Journal 37: 59–74.

Dowding, K. 1995. ‘Model or metaphor? A critical review of the network approach’, Political Studies 43:

136–58.

—— 2001. ‘There must be an end to confusion: Policy networks, intellectual fatigue, and the need for

political science methods courses in British Universities’, Political Studies 49: 89–105.

Galaskiewicz, J. 1979. Exchange Networks and Community Politics. Beverly Hills, CA: Sage.

Grant, W. 1978. ‘Insider groups, outsider groups and interest group strategies in Britain’, Warwick: University

of Warwick Department of Politics Working Party no. 19.

Greenaway, J., B. Salter and S. Hart. 2007. ‘How policy networks can damage democratic health: A case

study in the government of governance’, Public Administration 85: 717–38.

Katzenstein, P. 1976. ‘International relations and domestic structures: Foreign economic policies of

advanced industrial states’, International Organization 30: 1–45.

Kenis, P. and V. Schneider. 1991. ‘Policy networks and policy analysis: Scrutinizing a new analytical

toolbox’, in B. Marin and R. Mayntz (eds) Policy Networks: Empirical Evidence and Theoretical Considerations.

Boulder, CO, Frankfurt: Campus Verlag and Westview Press.

Kisby, B. 2007. ‘Analyzing policy networks: Towards an ideational approach’, Policy Studies 28: 71–90.

Knoke, D. 1998. ‘The organizational state: Origins and prospects’, Research

in Political Sociology,

8: 147–63.

—— 2001. Changing Organizations: Business Networks in the New Political Economy. Boulder, CO: Westview

Press.

—— 2004. ‘The sociopolitical construction of national policy domains’, in C.H.C.A. Henning and C. Melbeck

(eds) Interdisziplinäre Sozialforschung: Theorie und empirische Adwendungen. Frankfurt: Campus Verlag.

—— 2009. ‘Playing well together: Creating corporate social capital in strategic alliance networks’, American

Behaviorial Scientist 52: 1690–708.

—— 2011. ‘Policy networks’, in J.P. Scott and P. Carrington (eds) The Sage Handbook of Social Network

Analysis, Thousand Oaks, CA: Sage.

——and E.O. Laumann. 1982. ‘The social structure of national policy domains: an exploration of some

structural hypotheses’, in P.V. Marsden and N. Lin (eds) Social Structure and Network Analysis. Beverly

Hills, CA: Sage Publications.

——, F.U. Pappi, J. Broadbent and Y. Tsujinaka. 1996. Comparing Policy Networks: Labor Politics in the US,

Germany and Japan. New York: Cambridge University Press.

Laumann, E.O. and F.U. Pappi. 1976. Networks of Collective Action: A Perspective on Community Inﬂuence

Systems. New York: Academic Press.

——, J. Galaskiewicz, and P.V. Marsden. 1978. ‘Community structure as interorganizational linkages’,

Annual Review of Sociology 4: 455–84.

——and D. Knoke. 1987. The Organizational State: A Perspective on the Social Organization of National Energy

and Health Policy Domains. Madison, WI: University of Wisconsin Press.

——and F.U. Pappi. 1973. ‘New directions in the study of community elites’, American Sociological Review

38: 212–230.

Lehmbruch, G. 1984. ‘Concertation and the structure of corporatist networks: Order and conﬂict in

contemporary capitalism’, in J.H. Goldthorpe (ed.) Order and Conﬂict in Contemporary Capitalism. Oxford:

Oxford University Press.

—— 1989 ‘Institutional

linkages and policy

networks in the federal system of West Germany’, Publius 19:

221–35.

McGann, J.G. and R. Sabatini. 2011. Global Think Tanks: Policy Networks and Governance. New York:

Routledge.

Maoz, Z. 2010. Networks of Nations: The Evolution, Structure, and Impact of International Networks, 1816–2001.

New York: Cambridge University Press.

Marsh, D. (ed.). 1998. Comparing Policy Networks. Buckingham: Open University Press.

——and R.A.W. Rhodes (eds). 1992. Policy Networks in British Government. Oxford: Clarendon Press.

——and M. Smith. 2000. ‘Understanding policy networks: Towards a dialectical approach’, Political Studies

48: 4–21.

Wu and Knoke

162

——, D. Toke, C. Belfrage, D. Tepe. and S. McGough. 2009. ‘Policy networks and the distinction between

insider and outsider groups: The case of the Countryside Alliance’, Public Administration 87: 621–38.

Mikkelson, M. 2006. ‘Policy network analysis as a strategic tool for the voluntary sector’, Policy Studies 27:

17–26.

Pal, L.A. and D. Ireland. 2009. ‘The public sector reform movement: Mapping the global policy network’,

International Journal of Public Administration 32: 621–57.

Provan, K.G. and P. Kenis. 2008. ‘Modes of network governance: Structure, management, and eﬀectiveness’,

Journal of Public Administration Research and Theory 18(2): 229–52.

Rethemeyer, R.K. 2007. ‘Policymaking in the age of internet: Is the internet tending to make policy

networks more or less inclusive?’ Journal of Public Administration Research and Theory 17: 259–84.

Rhodes, R.A.W. 1985. ‘Power dependence, policy communities and inter-governmental networks’, Public

Administration Bulletin 49: 4–29.

—— 1990. ‘Policy networks: A British perspective’, Journal of Theoretical Politics 2: 293–317.

Richardson, J. 2000 ‘Government, interest groups and policy change’ , Political Studies 48(5): 1006–25.

Sandström, A. and L. Carlsson. 2008. ‘The performance of policy networks: The relation between network

structure and network performance’, Policy Studies Journal 36: 497– 524.

Siegel, J. 2008. ‘Contingent political capital and international alliances: Evidence from South Korea’,

Administrative Science Quarterly 52: 621–66.

Voets, J., W. Van Dooren and F. de Rynck. 2008. ‘A framework for assessing the performance of policy

networks’, Public

Management Review 10: 77–90.

—— 2004. ‘From

corporatism to lobbyism: A change of paradigm in the theory and analysis of advocacy’,

Zeitschrift für Parlamentsfragen, 35: 761–76.

von Winter, T. 2001. ‘From corporatism to etatism: Changes in the structure of the German poverty

policy network’, Zeitschrift für Politikwissenschaft 11: 1573–608.

Weible, C.M. and P.A. Sabatier. 2007. ‘ A guide to the advocacy coalition framework: Tips for researchers’,in

F. Fischer, G.J. Miller and M.S. Sidney (eds) Handbook of Public Policy Analysis: Theory, Methods, and

Politics. London: CRC Press.

Witte, J. M., W.H. Reinicke and T. Benner. 2000. ‘Beyond multilaterialism: Global public policy networks’,

Internationale Politik und Gesellschaft 2: 176–88.

Policy network models

163

Part IV

Understanding the

agenda-setting process

Policy agenda-setting studies

Attention, politics and the public

Christoffer Green-Pedersen and Peter B. Mortensen

Introduction

The agenda-setting approach has a long tradition in public policy studies. Seminal work by

Schattschneider (1960) and Bachrach and Baratz (1962) pointed to the crucial role of attention

to policy problems. The allocation of attention was the “second face of power” (Bachrach and

Baratz 1962), and the “conﬂict of conﬂicts” is crucial in politics (Schattschneider 1960). From these

seminal studies grew the literature known as policy agenda-setting studies (Baumgartner et al. 2006).

The crux of this tradition is precisely the focus on attention. Policy-making cannot be under-

stood without understanding the agenda-setting process, which draws attention to policy pro-

blems that need to be addressed through policy decisions. The ﬁrst studies in policy agenda-setting

were classics like Cobb and Elder (1983) and Kingdon (1995). They mostly applied a case study

perspective and focused on how agenda-setting leads to policy decisions. Kingdon (1995: Chapter

23) became inﬂuential in the study of policy decisions from an agenda-setting perspective.

A new tradition in the policy agenda-setting tradition took oﬀ with the publication of Baum-

gartner and Jones’ (1993) seminal book Agendas and Instability in American Politics, which pushed

policy agenda-setting studies in a diﬀerent direction in several related ways. Most notably, while

previous studies focused on the role of agenda-setting dynamics in relation to speciﬁc policy

decisions (see e.g. Zahariadis 1995), Baumgartner and Jones (1993) focused on more general

patterns of agenda-setting in political systems. Instead of focusing on individual policy decisions,

they focused on how attention to broader policy issues changed over time and how these changes

paved the way for often dramatic changes in public policy. Another eﬀect of the change in

focus was the establishment of extensive datasets to measure attention dynamics over time and

across a large number of policy issues, for instance through coding of attention to diﬀerent issues

in congressional hearings.

In recent years, the policy agenda-setting tradition has taken a new turn through a surge in

comparative work following the path laid out by Baumgartner and Jones. Part of this work has

been related to a comparative evaluation of the punctuated equilibrium model (see Chapter 12).

However, the recent interest in comparative policy agenda-setting studies goes far beyond

testing the punctuated equilibrium model to an interest in understanding attention dynamics

and how they are shaped by factors like party competition and political institutions. These

167

questions lead to new insights into how public opinion and elections aﬀect public policy. This

chapter reviews this recent trend and focuses on how the new insights into agenda-setting

dynamics inform our understanding of public policy and policy change.

Comparative studies of policy agendas

Over the past ﬁve years, a network of scholars (www.comparativeagendas.org) studying agenda-

setting across a wide range of Western countries has emerged. The network emphasizes attention

dynamics rather than policy change. A central element is the establishment of comparative large-

scale datasets on attention to policy issues in, for instance, executive speeches (e.g. speeches at the

opening of parliament), party manifestos, mass media and parliamentary activities (bills, inter-

pellations, parliamentary questions, etc.). Many of these attention dynamics studies have clear

implications for our understanding of policy dynamics and will therefore be introduced below.

One of the network ’ s research areas is attention dynamics in executive speeches. While they

do not directly represent policy decisions, they often contain a government’s policy intentions

or plans for the coming year and therefore their dynamics are of obvious interest in relation to

policy dynamics. Some of the studies test the punctuated equilibrium model (see Chapter 12)

(e.g. John and Jennings 2010); one study of the UK focuses on the responsiveness to public

opinion (Jennings and John 2009); another examines the eﬀect of factors like coalition composition

and government life-cycle in the Dutch context (Breeman et al. 2009).

A central issue for comparative investigation of executive speeches has been the role of elections

and government change (Mortensen et al. 2011). This discussion is closely linked to the “politics

matters” approach in policy studies. From this perspective, elections are an important driver of

policy change. They often lead to a change in the color of the government and a government of a

diﬀerent color will pursue diﬀerent policy goals (cf. Schmidt 1996; Imbeau et al. 2001). Based on

this theory, one would expect that elections—when they lead to a change in government color—

lead to a change in issue priorities in executive speeches. A government of a diﬀerent color

would focus on diﬀerent policy issues. However, the studies of executive speeches over the past

50 years in three West European countries (the UK, the Netherlands and Denmark) found that

elections and changes in government color had very little eﬀect. Only in the UK, the political

system where one should expect the strongest eﬀect, are there signs of a small eﬀect from a change

in government color (Mortensen et al. 2011). In other words, the color of the government has

surprisingly little eﬀect on the policy agenda of modern governments. The issues and thus policy

problems, which they attend to, seem to be driven by other factors than the color of the government.

Nevertheless, the limited eﬀect of elections and change in government color should not be

interpreted as a sign of no change in the policy priorities of government. In the long run, they

change substantially. Core issues like economy and defense have lost their dominance of executive

agendas (Jennings et al. 2011) and issues like health care, the environment and crime have

gained importance across countries. This does not imply that party competition is irrelevant for

understanding what issues governments prioritize, but rather that we learn little about the

timing of such processes if we focus on elections. To attract attention to issues, political parties

need policy problems deﬁned in a favorable way. And since policy problem developments are

independent of elections, little is gained from focusing on elections when we want to under-

stand the timing of major changes in government agendas. This comparative study indicates that

it is more important whether or not parties ﬁnd themselves in government or opposition

because policy responsibility is key to shaping governments ’ prioritization of issue attention.

This idea is further explored in Green-Pedersen and Mortensen (2010). Once a government

holds power, it is also expected to provide solutions to whatever societal problems emerge. So

Green-Pedersen and Mortensen

168

when the banking system collapses, major pollution events happen or the media report on waiting

lists for hospital treatment, it is the government ’s problem, no matter whether the present

government’s policies have caused these problems in any way. Health care is a good example.

All governments in the Western world face a diﬃcult dilemma between a huge surge in the pos-

sibilities of detecting and treating diseases and the need to keep health care costs under control

(Green-Pedersen and Wilkerson 2006). Thus, health care is a policy issue that governments in

the Western world, regardless of party color, have been more or less forced to pay increasing

attention to (Mortensen et al. 2011). Consistent with this idea of the “burden” of policy

responsibility of being in government, Green-Pedersen and Mortensen (2010) have developed a

model of issue competition that points to the agenda-setting power of opposition parties. A

central implication of the model is that opposition parties are freer to focus continually on issues

that are advantageous to them, whereas government parties are forced to respond more often to

issues brought up on the so-called party system agenda. Using data on issue competition in

Denmark covering 25 years and 23 issue categories, the empirical evaluation of this model clearly

shows how opposition parties are capable of raising issues that the government needs to address

later in executive speeches, for instance.

Paying attention is of course not the same as making policy. However, from an agenda-setting

perspective increasing attention is very likely to lead to policy measures. There are limits to how

much a government can talk about health care, for instance, without doing something in terms

of policy. A recent study of law and order in Denmark (Seeberg 2011) shows that it can lead to

signiﬁcant policy changes when opposition parties raise an issue like law and order. The gov-

ernment’s ﬁrst response is most likely to address the issue in speeches, and so on, but if the issue

does not die out in terms of attention, the government will also respond by changing policies,

in this case by introducing longer and stricter sentences. Where the above study of attention to

diﬀerent issues in executive speeches focuses on similarities in attention to issues across coun-

tries, other studies focus on explaining cross-national diﬀerences in attention to policy issues and

investigating the policy consequences. These cross-national studies often focus on cross-national

diﬀerences in how much attention political parties pay to diﬀerent policy issues. The underlying

idea is that certain issues are politicized, which implies that all political parties will pay attention

to them, whether or not they want to.

Green-Pedersen and Krogstrup (2009), for instance, study the politicization of immigration in

Denmark and Sweden. The issue has been politicized in Denmark but not in Sweden and the

explanation for this lies in the composition of the right-wing bloc of parties and the incentives

this oﬀers for mainstream right-wing parties. In Denmark the Liberals and the Conservatives

dominate the right-wing bloc in terms of parliamentary seats. After 1993, when they found them-

selves in opposition, their strategy was to gain a majority with the extreme right. This involved

a successful attempt to politicize immigration, even though it meant a conﬂict with the center-

right party, the Social Liberals, which had supported the right-wing government before 1993.

Sweden only has one mainstream right-wing party: the Conservatives. In terms of gaining a

parliamentary majority, the mainstream right in Sweden is thus very dependent on the three

center-right parties, and collaboration with the extreme right, including an attempt to politicize

immigration, has thus not been an attractive option. The policy consequences of these diﬀerences

in politicization are signiﬁcant despite considerable skepticism toward immigration in both coun-

tries. In Denmark the consequence has been a signiﬁcant tightening of the rules on immigration

and for acquiring citizenship. This has not happened in Sweden.

The same basic explanatory model, where diﬀerences in party incentives to politicize issues

lead to cross-national diﬀerences in issue politicization, which again generate cross-national

diﬀerences in public policies, has been used to explain cross-national diﬀerences in policies with

Policy agenda-setting studies

169

regard to morality issues like abortion and euthanasia. Green-Pedersen (2007) has investigated

why euthanasia is legal in the Netherlands, but not in Denmark and found that the explanation

was a conﬂict between religious and secular parties in the Dutch party system. In times of

increasing secularization, morality issues have been used by the secular parties to put pressure on

religious parties, such as the Christian Democrats. These parties have based their continuing

electoral survival on issues like the welfare state and on avoiding issues such as abortion where

their distinctive religious proﬁle has become an increasing liability. The secular parties tried to

politicize morality issues by demanding more permissive regulation of abortion and euthanasia,

for instance. In Denmark, there is no conﬂict between religious and secular parties. Political parties

are generally uninterested in morality issues. They are considered “non-political,” ethical issues,

which are not relevant for party competition. The policy consequences of this deﬁnition of the

issues are mixed. On the subject of euthanasia, there has been no drive for permissive regulation

in Denmark, whereas a permissive abortion regulation was adopted in the early 1970s. How-

ever, there has been no general move toward the more permissive regulation that is found in

the Netherlands (cf. also Engeli et al. 2012). The two countries thus diﬀer in their policy proﬁle

on morality issues despite similarities in public opinion. In both countries, there is support for

legalizing euthanasia, but it plays no role in Denmark because of the non-political character of

the issue (Green-Pedersen 2007).

These recent studies all show how important it is to study party incentives and party com-

petition in terms of understanding primarily cross-national di ﬀerences in attention to policy issues,

but also cross-national diﬀerences in public policy. The diﬀerence in regulation of euthanasia in

Denmark and the Netherlands cannot be explained without understanding the cross-national

diﬀerences in party incentives to politicization and the same goes for the diﬀerence in immigration

policies between Denmark and Sweden.

Linking attention, public opinion and policy change

While the comparative studies have focused on explaining cross-national variation in attention

to policy issues and then looked at the policy consequences of these changes, other studies have

focused on how agenda-setting actually aﬀects national policy-making. Do changes in the political

agenda foreshadow changes in public policies? And does the eﬀect of changes in the political

agenda depend on the policy preferences expressed by the mass public?

By integrating the study of political agendas with classic assumptions about re-election

oriented representation it has been hypothesized that the political agenda—deﬁned as the col-

lective attention of elected national policy-makers—matters to public policies, but the direction

of the eﬀect is crucially dependent on the policy opinions expressed by the mass public, because

the potential electoral costs of unpopular policies increase with the issues’ importance on the

political agenda (see Mortensen 2010).

Macro-political institutions like Congress or national parliaments handle hundreds of issues

every year, most of which are not very important from a re-election perspective because they,

for various reasons, are subject to relatively low or decreasing political attention. Mass media

coverage tends to ﬂuctuate with the tides of the political agenda and if an issue does not attract

much attention in macro-political venues like national parliaments or Congress it probably does

not receive much attention among a broader public audience either (see e.g. Bennett 1990;

Bennett et al. 2007). Hence, on many issues national policy-makers have some leeway to leg-

islate as they see ﬁt without much concern about being punished at the next election by ret-

rospective voting. On the other hand, debates and activities in macro-political arenas both

reﬂect and reinforce outside attention to the issue. Hence, re-election concerns about a given

Green-Pedersen and Mortensen

170

issue are assumed to rise with increased macro-political attention to that issue (see also Sulkin

2005). When that happens, the argument goes, the majority of the re-election-oriented national

policy-makers seek to change policies toward the position expressed by the mass public on those

issues. In other words, the policy eﬀects of changes in the political agenda depend upon the

policy preferences expressed by a majority of the mass public. For instance, in terms of public

spending the implication is that popular spending issues where most of the public prefers increased

spending do beneﬁt in terms of monetary appropriations from increasing attention in the

macro-political arena. Unpopular spending domains where most of the public prefers decreased

public spending, on the other hand, are expected to suﬀer from increasing macro-political atten-

tion. Conversely, in times of decreasing macro-political attention where public opinion (and

hence re-election motives) presumably will be less important to national spending decisions,

unpopular spending domains may beneﬁt relatively.

Using public spending as a measure of public policy, this agenda-policy hypothesis has been

evaluated empirically in statistical time series regressions utilizing a Danish dataset of public spending

attitudes, public spending, and the political agenda covering six issues from 1980 to 2003 (see

Mortensen 2010). Across the issues, the results show that public opinion clearly aﬀects policy

changes after taking account of other potentially relevant agenda and spending determinants. For

instance, in popular spending domains (in Denmark issues like law and order, pollution control,

health and primary education where a majority over time and in every poll has expressed a

preference for increased spending), there is a positive relationship between changes in macro-

political attention and changes in public spending on these issues. In unpopular spending domains

(in Denmark these may be issues like defense, development aid and culture) there is a negative

relationship between changes in macro-political attention and changes in public spending.

A theoretically more elaborate version of this basic agenda-policy hypothesis has been pre-

sented and evaluated in a US study of the relationship between changes in congressional attention

and federal budget appropriations across 12 spending domains and 33 years (1970–2003) (see

Mortensen 2009). The modiﬁed claim is that the policy (spending) eﬀ ects of attention shifts are

strongest when the trade-oﬀ between re-election incentives and intensive group interests is low,

which is the case when attention to popular issues peaks and attention to unpopular issues

declines. Put diﬀerently, when popular issues are subject to increased macro-political attention

congressional policy-makers can satisfy both public spending attitudes and specialized spending

advocates by increasing spending on such issues. On the other hand, when popular issues attract

decreasing attention policy-makers may become more hesitant to follow a narrow re-election

incentive and withdraw monetary resources because of (anticipated) ﬁerce opposition from vested

spending interests. Thus, in popular spending domains the association between attention and

spending changes is expected to be driven mainly by increased attention followed by relatively

large spending increases, not by decreased attention followed by relatively large budget cuts.

Conversely, in spending domains in which a majority of the US public prefers less spending,

the relationship between congressional attention and public spending may be driven by decreased

congressional attention followed by increased spending investments advocated by special interest in

times of less outside visibility, not by congressional attention peaks paving the way for large pub-

licly supported spending cuts. Subsystem loyalties or ﬁerce opposition from groups or agencies with

intensive interests in a given unpopular spending program may not suﬃce to fully avoid cutbacks

during macro-political attention peaks. However, as opposed to popular spending programs,

unpopular programs are certainly not expected to beneﬁt from increased macro-political attention,

but to be built up in quieter times of decreasing macro-political attention (see Mortensen 2009).

Utilizing data from the US policy agendas project (see www.policyagendas.org) the empirical

analyses support this conditional model of the agenda-policy relationship. In popular spending

Policy agenda-setting studies

171

domains such as crime, environment, and health where most of the US public in every poll

since 1970 has expressed a preference for increased spending, increases in congressional attention

to these domains are followed by spending increases. In comparison, decreases in attention to

popular spending domains are not to a similar extent followed by spending cuts. The reverse

happens in unpopular spending domains where a majority of the US public in all polls since

1970 has expressed a preference for reduced spending (i.e. space exploration, foreign aid and

welfare). Here, decreases in congressional attention are followed by spending increases, whereas

attention increases do not show a spending eﬀect. In other words, well-organized spending

coalitions working within domains that are not very popular with the general public can better

pursue their spending interests in times of decreasing outside attention and may be able to block

large spending cuts when attention increases.

While these results are consistent with Redford’s (1969) classic conception of macro and

subsystem politics (see also Baumgartner and Jones 1993), the two studies oﬀer the ﬁrst large N

empirical illustration of these ideas and advocate a more explicit conﬂuence of the policy agenda-

setting perspective developed by Baumgartner and Jones and more mainstream perspectives on

representation and congressional policy-making. Furthermore, they demonstrate that policy chan-

ges also happen when issues are not politicized, but these changes can hardly be explained by

factors like party competition and public opinion.

More important in times of low attention is subsystem politics dominated by organized

interest groups, experts, advocacy coalitions, and bureaucrats with vested interests in the given

subject. A good example of policy change without a major surge in macro-political attention is

tobacco regulation, which in many West European countries has undergone substantial change

over the past few years without being subject to party competition or strong support in public

opinion. Western Europe has seen a surge in restrictions on smoking that is hard to explain by a

surge in macro-political attention (cf. Albæk et al. 2007). Thus, a change in macro-political atten-

tion is an important, but not a necessary condition for policy change, and it is not a satisfactory

stand-alone explanation if we want to understand directions of policy change.

One should note how these agenda-setting-inspired models of the attention-policy link diﬀer

from opinion-policy studies focusing on the broader concept of public saliency (see e.g. Page

and Shapiro 1983; Soroka 2003; Baumgartner and Jones 2004). While the latter tend to focus

on an issue’s saliency to the public, the former focus on the policy-making process and the

preferences of the relevant policy-makers. For instance, an issue might be salient to the general

public without being subject to macro-political intervention because party competition does

not form around that issue, just as macro-political intervention may anticipate, reduce, or per-

haps even prevent increased public saliency of an issue. Furthermore, some problems do not

have simple cures and will turn into major problems despite political action, but from a gov-

ernmental re-election point of view, it may be too late to respond at that time. In a similar vein,

as also implied by Redford’s (1969) distinction between subsystem and macro-political actors, all

policy-makers are most likely neither equally nor constantly concerned about potential voter

reactions. This explicit focus on the policy-makers and the policy-making process also suggests

that one considers not only the majority attitudes of the general public but also the demands

and interests of specialized policy advocates with more intensive interests in the given policy

programs. In other words, the ﬁndings reported in this section seem to call for a more general

model of public policy-making that transcends the classic division between models of electoral

competition and responsiveness to the mass public, on the one hand, and models that stress the

power of special interests and subsystem actors, on the other. In this way, Schattschneider’s (1960)

strong legacy is still very signiﬁcant in current developments in the literature on agenda-setting

and public policy.

Green-Pedersen and Mortensen

172

Conclusion

While early agenda-setting studies were somewhat narrowly concerned with describing diﬀer-

ent agendas and studying agenda-setting per se, the recent wave of agenda-setting studies is much

more inspired by and in dialogue with other major approaches to the study of public policy.

Examples include literature on party competition, for instance the “politics matters” approach,

literature on government coalitions, institutionalism, representation and the role of public

opinion, mediatization of politics, as well as more classical literature on public spending and

budgeting.

This, of course, implies that there is no narrow theoretical focus uniting the new scholars of

agenda-setting and public policy, for instance the mandate theory underlying the major research

program on party manifestoes (see e.g. Klingemann et al. 1994). What characterizes the group of

agenda-setting scholars is a common interest in the concept of attention as well as rather strictly

standardized approaches to how agenda-setting data are collected and coded across time and

countries.

Still, it is possible to draw some general insights about what we have recently learned from an

agenda-setting approach to public policy. First, shifts of party governments and elections are not

strongly related to shifts in government agendas and policy. As the review of studies shows, this

does not imply that political parties and their competition for electoral support are unimportant for

political agendas and thus public policies, but only that elections, even when they lead to a change

in government color, are not very central for understanding the dynamic of political agenda and

public policy. Instead, the competition ongoing in-between elections between government and

opposition parties has, for instance, been shown to be important for policy change.

Second, changes in attention have important policy consequences but this does not imply that

policy change only happens as a result of changes in attention, nor does it imply a linear rela-

tionship where changes in attention push public policies in certain directions. Rather, changes

in attention are important for public policy because they change the policy process. When an issue

suddenly reaches the macro-political agenda, factors like party competition and public opinion

become important to explain public policies. When issues receive limited macro-political attention,

public policies are better explained by the nature of the political subsystems related to the issue.

Third, attention to political issues and thus the public policy agenda shows less cross-national

variation than one would expect from an institutional approach. As the health care example showed,

the issues and related problems matter and they might be universal at least across rather closely

integrated (e.g. West European, countries). This does not mean that variation does not exist as the

examples about immigration in Denmark and Sweden and morality issues in Denmark and the

Netherlands showed. However, diﬀerent dynamics of party competition seem to be a more important

cause of variation than institutional diﬀerences.

Bibliography

Albæk, Erik, Christoﬀer Green-Pedersen, Lars Beer Nielsen. 2007. ‘Making tobacco consumption a political

issue in the United States and Denmark: The dynamics of issue expansion in comparative perspective’,

Journal of Comparative Policy Analysis: Research and Practice 9(1): 1–20.

Bachrach, Peter and Morton S. Baratz. 1962. ‘Two faces of power’, American Political Science Review 56:

947–52.

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago:

University of Chicago Press.

—— and Bryan D. Jones. 2004. ‘Representation and agenda-setting’, Policy Studies Journal 32(1): 1–24.

——, Christoﬀer Green-Pedersen and Bryan D. Jones. 2006. ‘Comparative studies of policy agendas’, Journal

of European Public Policy 13(7): 955–70.

Policy agenda-setting studies

173

Bennett, W. Lance. 1990. ‘Toward a theory of press-state relations in the United States’, Journal of

Communication 40(2): 103–25.

——, Regina C. Lawrence and Steven Livingston. 2007. When the Press Fails. Political Power and the News

Media From Iraq to Katrina. Chicago: University of Chicago Press.

Breeman, G., D. Lowery, C. Poppelaars, S. L. Resodihardjo, A. Timmermans and J. de Vries. 2009.

‘Political attention in a coalition system: Analyzing queen’s speeches in the Netherlands 1945–2007’,

Acta Politica 44(1): 1–27.

Cobb, Roger W. and Charles D. Elder. 1983. Participation in American politics. The Dynamics of Agenda-building.

Baltimore, MD: Johns Hopkins University Press.

Engeli, Isabella, Christoﬀer Green-Pedersen and Lars T. Larsen (eds). 2012. Morality Politics in Western

Europe: Parties, Agendas, and Policy Choices. London: Palgrave Macmillan.

Green-Pedersen, Christoﬀer. 2007. ‘The conﬂict of conﬂicts in comparative perspective: Euthanasia as a

political issue in Denmark, Belgium, and the Netherlands’, Comparative Politics 39(3): 273–91.

——and John Wilkerson. 2006. ‘How agenda-setting attributes shape politics: Problem attention, agenda

dynamics and comparative health policy developments in the US and Denmark’, Journal of European

Public Policy 13(7): 1039–52.

——and Jesper Krogstrup. 2009. ‘Immigration as a political issue in Denmark and Sweden’, European

Journal of Political Research 47(5): 610–34.

——and P.B. Mortensen. 2010. ‘Who sets the agenda and who responds to it in the Danish parliament? A

new model of issue competition and agenda-setting’, European Journal of Political Research 49(2): 257–81

Imbeau, Louis M., François Petry and Moktar Lamari. 2001. ‘Left-right party ideology and government

policies: A meta-analysis’, European Journal of Political Research 40(1): 1–29.

Jennings, W. and P. John. 2009. ‘The dynamics of political attention: public opinion and the queen’s

speech

in the United

Kingdom’, American Journal of Political Science 53(4): 838–54.

——, Shaun Bevan, Arco Timmermans, Laura Chaques, Gerard Breeman, Sylvain Brouard, Christoﬀer

Green-Pedersen, Peter John, Peter B. Mortensen and Anna Palau. 2011. ‘Eﬀects of the core functions of

government on the diversity of executive agendas.’ Comparative Political Studies 59(1): 74–98.

John, P. and W. Jennings. (2010. ‘Punctuations and turning points in British politics: The policy agenda of

the Queen’s speech, 1940-2005’, British Journal of Political Science 40(3): 561–86.

Kingdon, John W. 1995. Agendas, Alternatives, and Public Policies. New York: HarperCollins College Publishers.

Klingemann, Hans-Dieter, Richard Hoﬀerbert and Ian Budge. 1994. Parties, Policies, and Democracy.

Boulder, CO: Westview Press.

Mortensen, Peter B. 2009. ‘Political attention and public spending in the United States’, Policy Studies

Journal 37(3): 435–55.

—— 2010. ‘Political attention and public policy: a study of how agenda-setting matters’, Scandinavian

Political Studies 33(4): 356–80.

——, Cristoﬀer Green-Pedersen, Gerard Breeman, Laura Chaqués Bonafont, Will Jennings, Peter John,

Anna M. Palau Rogué, Arco Timmermans. 2011. ‘Comparing government agendas: Executive speeches

in the Netherlands, United Kingdom, and Denmark’, Comparative Political Studies 44(8): 973–1000.

Page, Benjamin I. and Robert Y. Shapiro. 1983. ‘Eﬀects of public opinion on policy’, American Political

Science Review 77(1): 175–90.

Redford, Emmette S. 1969. Democracy in the Administrative State. New York: Oxford University Press.

Schattschneider, Elmer E. 1960. The Semisovereign People. New York: Holt, Rinehart and Winston.

Schmidt, Manfred G. 1996. ‘When parties matter: A review of the possibilities and limits of partisan

inﬂuence

on public policy’, European Journal of

Political Research 30(2): 155–83.

Seeberg, Henrik. 2011. ‘Parties and politics: how policy changes when the opposition party sets the

agenda’, Aarhus University Department of Political Science.

Soroka, Stuart N. 2003. ‘Media, public opinion and foreign policy’, Harvard International Journal of Press and

Politics 8(1): 27–48.

Sulkin, Tracy. 2005. Issue Politics in Congress. Cambridge; Cambridge University Press.

Zahariadis, Nikolaos. 1995. Markets, States and Public Policy: Privatization in Britain and France. Ann Arbor,

MI: University of Michigan Press.

Green-Pedersen and Mortensen

174

Focusing events and policy windows

Thomas A. Birkland and Sarah E. DeYoung

Introduction: focusing events in the agenda-setting process

Policy scholars have long argued that the “stages” model of the policy process is not a useful

model for generating testable hypotheses about policy-making (Nakamura 1987; Sabatier 1988;

Sabatier 1991a, 1991b). But scholars still ﬁnd the stages as useful “sites” for the study of

important elements of the policy process, from problem recognition through implementation

(deLeon 1999). Among the most intensively studied stages is agenda-setting, which is the pro-

cess by which some issues gain and others lose attention among policy-makers and the public.

Agenda-setting is important to all theories of the policy process because “the deﬁnition of the

alternatives is the supreme instrument of power” (Schattschneider 1975: 66). Groups—or, more

speciﬁcally, advocacy coalitions, in Sabatier’s Advocacy Coalition Framework—engage in rheto-

rical battles, in many diﬀerent venues, to gain access to the agenda while attempting to deny

agenda access to other actors (Cobb and Ross 1997). Group competition, a major driving force

in agenda-setting, can be ﬁerce because of the limited capacity of any system or institution to

accommodate all issues and ideas (Walker 1977; Baumgartner and Jones 1993; Cobb and Elder

1983). This competition is over both which problems are most important, and over what causes

and solutions surround any one problem (Hilgartner and Bosk 1988; Lawrence and Birkland

2000, 2004; Birkland and Lawrence 2009). The agenda-setting process is therefore a system of

sifting issues, problems and ideas, and implicitly assigning priorities to these issues.

Focusing events as an aspect of agenda-setting

John Kingdon argues that agenda change is driven by two broad phenomena: changes in indicators

of underlying problems, which lead to debates over whether and to what extent a problem

exists and is worthy of action; and focusing events, or sudden shocks to policy systems that lead to

attention and potential policy change.

John Kingdon used the term “focusing event” within a general discussion of “Focusing Events,

Crises, and Symbols” (Kingdon 2003, 94–100). Kingdon calls focusing events a “little push”“like

a crisis or disaster that comes along to call attention to the problem, a powerful symbol that

catches on, or the personal experience of a policymaker.” Kingdon found that some policy

175

domains are less prone to sudden events than are low-visibility policy domains that are prone to

sudden shocks (Birkland 1997) which trigger a great deal of rapid attention, followed by a steep

drop-oﬀ in attention after the acute phase of a disaster or crisis (Downs 1972). Kingdon further

notes that focusing events gain their agenda-setting power by aggregating their harms in one

place and time. A plane crash that kills 200 people gets more attention than 200 single fatal car

wrecks; aviation is therefore a domain that is more prone to focusing events than is automobile

safety, where problems tend to be discovered through the accumulation of information. A

starker example in the United States would be the example of the 9/11 attacks, which were

socially, politically, and economically devastating, even though the death toll of about 3,000

people is less than 10 percent of annual road fatalities in the United States. Other problems—

cancer, heart disease, other forms of accidents—kill far more people than terrorist attacks, but

focusing events are much more sudden, and their harms more easily spoken of in terms of a place

(“Oklahoma City”) or a time (“9/11”).

Before Kingdon’s book was published, there was a broad acknowledgment that disasters and

crises were important in the policy process, but little systematic research on their agenda-setting

power. Cobb and Elder examine events they call “circumstantial reactors,” such as the 1969

Santa Barbara oil spill, “that led to a reconsideration of the whole question of oﬀshore drilling

regulations” (1983: 83). Broader ecological change, such as population growth, economic shifts,

and black migration, are also circumstantial reactors, but these changes are much more subtle

than natural disasters or technological accidents, and are therefore less likely to serve as a time—

and space—deﬁnite rallying point for group attention and mobilization. As Kingdon argues,

subtle changes are less likely to be viewed as “events” than are sudden, dramatic, and visible

problems. These elements of the policy process have a greater inﬂuence on the problem stream

and on the accumulation of ideas and evidence to justify policy change.

Traditionally, sudden events were thought to “simply bowl over everything standing in the

way of prominence on the agenda” (Kingdon 2003: 96). Kingdon portrays focusing events as

being more varied and subtle, and includes as focusing events personal experiences of policy-

makers with matters of personal interest, such as disease that aﬀect them or their families. Kingdon

also argues that “the emergence and diﬀusion of a powerful symbol” is a focusing event that

“acts … as reinforcement for something already taking place” (2003: 97). When symbols of

events propagate—the elderly woman in New Orleans sheltering herself from the rain with an

American ﬂag, the raising of the ﬂag at Ground Zero in a manner reminiscent of the raising of

the ﬂag at Iwo Jima, or the images of oil-soaked wildlife after oil spills—these symbols would not

have their power if it were not for the event itself. The propagation of the symbol ampliﬁes the

focusing power of the event, particularly if the symbol is particularly evocative (Birkland and

Lawrence 2002). But Kingdon seems to argue that the sudden uptake of a symbol or idea—

what scholars in the new media call a meme—is itself a focusing event.

Kingdon’s notion of a focusing event is important, but the term has often been uncritically

adopted

in the literature,

often as a term of art that is never precisely deﬁned. This may be because

Kingdon deﬁned and described focusing events in a way that is too broad to guide focused

empirical research. He conﬂates sudden crises and shocks with individual experience and symbol

propagation that are often reﬂective of what Peter May (1992) calls political learning about more

eﬀective policy arguments, which is less about the idea of focusing attention through an event, the

latter term suggesting a phenomenon that is easily placed in time.

Second, Kingdon compares sudden events with political events, such as protest activity, that

are purposefully caused, such as the 1963 March on Washington for Jobs and Freedom, a key

event in the history of the civil rights movement. Including these events in the deﬁnition of

focusing events confuses the role of political mobilization in the “politics” stream with the

Birkland and DeYoung

176

revelation and depiction of problems in the “problem” stream. “Shocking” and sudden focusing

events are not purposefully caused, and can be viewed as exogenous to a policy community or

domain, even if the policies and practices adopted by key members of that community make

such events more likely.

Third, Kingdon’sdeﬁ nition of a focusing event is retrodictive, which is unsurprising given

the nature of his research method. While this deﬁnition is suggestive of a way of empirically

studying the agenda-setting eﬀects of events, it was insuﬃcient to develop a testable model of

focusing event politics. Birkland (1997, 1998) focused his model of potential focusing events on

policy domains prone to crises more on the crisis and disaster aspect of this deﬁnition, a direc-

tion that has been broadly adopted, as we show here. Birkland deﬁnes a potential focusing event as

an event that is sudden, relatively rare, can be reasonably deﬁned as harmful or revealing the

possibility of potentially greater future harms, inﬂicts harms or suggests potential harms that are

or could be concentrated on a deﬁnable geographical area or community of interest, and that is

known to policy-makers and the public virtually simultaneously (1997: 22).

The term potential highlights that an event may inﬂuence the agenda, but one cannot say

with certainty precisely what, if any, that eﬀect will be, and that one can only study that event’s

eﬀects compared with a similar class of events. While we may “know” intuitively that an event—

the Deepwater Horizon Oil Spill, 9/11, or Hurricane Katrina—will gain a lot of attention, it is

diﬃcult to know in advance how focal that event will be. Will the event make a more sig-

niﬁcant and discernible diﬀerence on the agenda than competing events and issues? Will that

event have any discernible inﬂuence on policy changes? But we cannot collect a set of events from

one particular class of event and a priori decide that these events are equally and signiﬁcantly

inﬂuential. Only careful analysis can do so.

Foundations of focusing event theory

Focusing event theory is part of the broader scholarship on the agenda-setting process; it relies

heavily on the idea that agenda space is limited, and that groups and interests compete for

attention. One strand of this research focuses on the nature of the problems themselves including

their expansion to a larger more attentive audience (Cobb et al. 1976; Cobb and Elder 1983;

Cobb and Ross 1997) and how long that attention may last (Downs 1972). The nature of the

actors and the nature of the problems interact with each other to promote or impede issue

expansion. These theories are important because they help us to understand how groups seek to

ﬁnd or express the “meaning” of an event (Schattschneider 1975; Molotch and Lester 1974, as

cited in Best 2010; Nohrstedt 2010).

A focusing event, by deﬁnition, increases attention to a public issue or problem. Baumgartner

and Jones (1993) note that this attention is usually negative attention, and negative attention often

yields further attention, thereby moving issues closer to potential policy changes. However,

political elites do not always resist change, even if policy monopolies are reorganized. Focusing

events, particularly in policy domains characterized as “policies without publics,” (May 1990) yield

“internal mobilization” eﬀorts to promote the change that policy elites prefer (Cobb and Elder

1983). Indeed, Best (2010) cites Molotch and Lester’s (1975) claim that “actor-promoted events”

(APEs, in her term) are more likely to generate attention when the actors are elites with which

news media already have steady contacts, and whose actions are considered important by deﬁni-

tion. The messages these actors seek to convey are generally pro-status quo, or at least pro-elite,

to the extent that elites sometimes desire policy change, particularly in “policies without pub-

lics” (May 1990). Later we consider Best’s challenge to this proposition in her groundbreaking

work on purposive events and agenda-setting in the news media.

Focusing events and policy windows

177

A second strand is the use of language, stories, metaphors, and symbols, to advance or retard

issues on the agenda. Contained in this strand is the process of social construction, by which

societies collectively deﬁne and explain the nature and cause of problems. Students of the media

explain how issues gain the attention of journalists, how journalists and their sources use sym-

bols and stories to explain complex issues, and how news consumers respond to these issues and

symbols (Edelman 1967, 1988; Stone 1989, 2002; Majone 1989; Schneider and Ingram 1991).

Related to this is the literature on problem framing in government and the mass media (Entman

1993; Burnier 1994; Lawrence 2000), which has been proﬁtably applied to studies of focusing

events (Glascock 2004; Gunter 2005; Liu et al. 2011). Framing theory argues that that participants

in policy debate frame stories about problems to fulﬁll news-gathering routines designed to make

the story both eﬃcient and compelling (Bennett 2003), and to motivate action by its supporters,

inaction by its opponents, or both. These symbols are promoted by a policy entrepreneur, who

is an individualactive in the policy community because of their technical expertise in their ﬁeld,

political expertise, and ability to broker deals that lead to new programs and policies (Kingdon

2003; Mintrom 1997; Mintrom and Vergari 1998). Recent studies referencing focusing events

have further restated the roles of policy entrepreneurs (Gunter 2005; Wood 2006). Focusing

events often gain their power from the propagation of symbols, involvement of policy entre-

preneurs, as well as through group mobilization—all of which can overlap with one another in

order to cause momentum for agenda change and potential policy change.

The special role of group coalescence and mobilization

A third theoretical strand that is important for understanding focusing events within the agenda-

setting tradition is group coalescence and mobilization, “because of the impact of the nature of

the policy community on the policy process” (Baumgartner and Jones 1993: 43). Groups coalesce

to form advocacy coalitions based on mutual interests and values. Sabatier predicts that two to

four advocacy coalitions will form in most policy domains. However, some domains prone to

sudden disasters may be characterized as “policies without publics” (May 1990), in which policy-

making is often the domain of technical experts, with little public mobilization around particular

policy changes after a major crisis or disaster, (Birkland 1998) beyond the usual diﬀuse claims

that policies should work “better” in some way. These domains are characterized by one advocacy

coalition, which may be quite weak. In many policy domains characterized by focusing events,

the events drive an “internal mobilization” eﬀort among existing group members to resolve a

problem about which the public is broadly concerned, but is generally unable or unmotivated

to organize around. In other words, focusing events are not always triggers for mass mobilization or

sustained public attention.

However, policy domains that are part of the broader political and policy debate can see very

intense group activity. Birkland’s 1997 study found that the earthquake and hurricane policy

domains were very much characterized by an internal mobilization model of agenda-setting and

policy change, while in the oil spill and nuclear power policy domains, there were discernible

advocacy coalitions that were engaged in direct competition. Even here, though, there are impor-

tant diﬀerences. Birkland argued that oil spills have a much greater likelihood of mobilizing

groups to oppose oil shipping and exploration because of the very visible and deeply felt damage

done by these spills, as evidence by images of oily otters after Exxon Valdez, or idled ﬁshing

boats after the Deepwater Horizon spill. In this case, oil companies sought to “contain the scope

of conﬂict,” in Schattschneider’ s terms, while the highly polarized nuclear power policy domain

was less inﬂuenced by the dominant event, the Three Mile Island nuclear accident. The degree

of polarization therefore inﬂuences post-event politics (Walgrave and Verhulst 2009).

Birkland and DeYoung

178

Focusing events may trigger greater attention to problems and solutions because they increase

the likelihood of more inﬂuential and powerful actors entering the conﬂict on the side of policy

change (Schattschneider 1975; Baumgartner and Jones 1993), by way of greater claims of policy

failure and a more active search for solutions, leading to a greater likelihood of policy change

(Birkland 2006).

But one must not make too much of event-triggered group mobilization. Sometimes we

attribute greater power to events than they deserve. Kingdon notes that we cannot trace the dis-

covery or the propagation of a policy idea to its ﬁrst instance, lest we engage in “inﬁnite

regress.” In a similar manner, in some events we have to ask: Which came ﬁrst, the problem, or

the event? Among the public there is widespread belief that the Three Mile Island (TMI)

nuclear power plant accident in 1979 was the event that stalled the movement toward more

nuclear power plants for generating electricity. As Baumgartner and Jones note in their chapter

on nuclear power, the “policy image” of nuclear power had eroded before TMI—the major

milestones in that erosion were the creation of the Nuclear Regulatory Commission (NRC)

and the abolition of the Atomic Energy Commission and its congressional champion, the Joint

Committee on Atomic Energy (JCAE). Birkland (1997: Chapter 5) found that, at the time of

TMI, the nuclear power policy domain was highly polarized and the policy community was

already ﬁercely debating both the meaning of TMI (which was not, is still not, self-evident) and

the nature of nuclear power generally. The lack of one clear interpretation of the nature and

meaning of the TMI accident meant that all the participants in the debate could use the acci-

dent for their own rhetorical ends: as evidence either for the “defense in depth” notion of

nuclear safety, or for the idea that nuclear power contained unknown risks, that other incidents

were equally or more serious than TMI, and that heretofore poorly understood systems acci-

dents could yield catastrophic disasters (Perrow 1999). In the end, TMI was an important event,

and it was focal in its agenda-setting meaning, but its role in group mobilization and actual

policy-making is less clear.

Apart from group eﬀorts to expand issues, major events often reach the agenda without

group promotion through media propagation of news and symbols of the event. This coverage

of a sudden and shocking event, such as a particularly horriﬁc airline crash, or school shooting,

makes the public aware of these events without eﬀorts on the part of group leaders to induce

attention. This media propagation of symbols gives less powerful groups another advantage in

policy debates. Pro-change groups are relieved of the obligation to create and interpret powerful

images and symbols of the problem. Rather, groups only need to repeat the already existing

symbols that the media have seized upon as the most important in the current crisis. These obvious

symbols are likely to carry more emotional weight than industry or governmental assurances

that policy usually works well. Carpenter and Sin (2007) found that the images of fragile chil-

dren who had been poisoned by Sulfanilamide were inﬂuential in triggering a strong outcry that

led to more stringent regulation and licensing for medications in the Food and Drug Act of

1938. Thus, media-generated symbols of health, environmental, or other crises or catastrophes

are often used by groups as an important recruiting tool – thereby expanding the issue—and as

a form of evidence of the need for policy change.

To conclude, all these theoretical strands assume that agenda-setting is not a neutral, objec-

tive, or rational process. Rather, it is the result of a society, acting through its political and social

institutions, building a consensus on the meanings of problems and the range of acceptable

solutions. There are many possible constructions that compete with each other to tell the story

of why a problem is a problem, who beneﬁts or is harmed by the problem, whose fault it is,

and how it can be solved (Gusﬁeld 1981; Schneider and Ingram 1991; Schneider 2008; Stone

1989, 2002).

Focusing events and policy windows

179

Some empirical applications of focusing event theory

In this section we show how the application of the focusing event concept can reveal important

features of agenda-setting, group activity, and potential convergence or divergence in what a policy

community deems to be the “real ” problem revealed by a focusing event.

Ab initio, the inﬂuence of a particular event on the formal agenda of the news media and

political institutions should be detectable because focusing events are more sharply deﬁned in

time than are social, demographic, or ideological changes. Take, for example, both news media

and congressional attention to terrorism before and after the September 11 terrorist attacks (Figure

14.1). In this ﬁgure, data on the number of New York Times articles and the number of witnesses

appearing before congressional hearings on these issues are normalized to the same scale, for ease

of comparison. The key focusing event in this data is obvious, but the increase in attention is much

more acute in the news media than in Congress. The news media tend to focus to a much

greater extent on the seeming novelty of a particular event itself, while Congress tends to be

seized by the broader problem—in this case, terrorism—for a longer time than the media, whose

interest in the broader problem drops oﬀ rather rapidly.

The pattern shown in Figure 14.1 is not particularly surprising. But attention is not the same

as the substance of the agenda. Birkland found that the dominant topic of discussion in con-

gressional hearings can change considerably when testimony is oﬀered about a particular event,

compared with testimony about the general problem itself, as shown in Table 14.1.

Differences in the substance of discussion after focusing events

We know that focusing events gain increased attention, and we can plot that attention in terms

of media coverage and congressional activity. However, does the substance or content of the

discussion change in the presence of a recent event from the “typical” discussion?

Figure 14.1 Comparative attention to terrorism

Source: LexisNexis search for term terrorism in headline, lead, or key terms, by desk; testimony in

LexisNexis congressional testimony database, by term terrorism. Birkland 2006, Figure 2.1

Birkland and DeYoung

180

In most policy domains that are prone to sudden, dramatic events, there are two discernible domi-

nant topics: those that dominate inter-event periods, when focusing events are often dim mem-

ories, and periods in which an issue dominates the agenda as a direct consequence of a recent

event. Table 14.1 reﬂects this proposition. These categories were designed so that parallels could

be drawn among the four domains: the term “disaster relief,” for example, is roughly analogous

to the term “cleanup” in the industrial domains because they both involve the immediate response

to a disaster or accident. In all but the hurricane domains, the dominant issue on the agenda

changes when an event is speciﬁcally mentioned. In the other domains, the existence of an event

on the agenda inﬂuences the actual substance of discussion in the domain. The oil spill and

nuclear power policy domains have at two broad advocacy coalitions, generally a pro-industry

and pro-environment coalition. There appears to be one advocacy coalition in earthquake policy,

a professional community that works to educate Congress and to advocate for policy change intra-

disaster. Such a community is not present, or not called upon, by Congress, and most testimony

in the hurricane domain is therefore about the rapid and generous delivery of disaster relief.

This research led Birkland to suggest that focusing events led to a convergence of opinion

about the nature of problems on the agenda, and to assume that an event led to something like

convergence toward a policy solution. Indeed, in the oil spill domain, this supposition was

strongly conﬁrmed by the passage of the Oil Pollution Act of 1990 soon after the Exxon Valdez

oil spill; the oil spill broke a 14-year legislative logjam. This convergence proposition was also built

on the normative conviction that there should be some convergence on one or very few problem

deﬁnitions, which should yield a narrow range of policy options from which to choose.

Contrary to this supposition, Hilgartner and Bosk (1988) and Lawrence (2000) argued that

the elevation of an issue on the agenda can yield a competition over how to frame or describe the

problem, thereby creating a proliferation of “causal stories” (Stone 1989) about why some events

happened, and what should be done. Lawrence and Birkland found that in the case of a parti-

cular focusing event—the Littleton school shootings—considerable attention was paid to the

“school violence” problem, broadly conceived, but the causal stories vary considerably, as shown

in Figure 14.2.

Here, we see considerable divergence in the dominant frames or causal stories, where the news

media assigned more importance to guns, rank and ﬁle members of Congress to guns, pop culture,

and a potpourri of other causes, while the creation of additional school programs—or the increased

funding or enforcement e ﬀort devoted to them—was the modal frame in legislation. The expla-

nation is simple: Congress, not being directly responsible for the management of schools, politically

constrained on the gun issue, and constitutionally constrained to regulate speech and expression,

settled on where it can have the most visible inﬂuence. In this case, Congress, in seeking a policy

solution, largely turned to distributive spending to encourage greater eﬀorts by schools to engage

in a wide range of social service programs to prevent school violence (Birkland and Lawrence 2009).

Table 14.1 The substance of the agenda when events are or are not mentioned in testimony

Testimony on speciﬁc event All other testimony All testimony

Earthquake Disaster relief NEHRP Disaster relief

Hurricane Disaster relief Disaster relief Disaster relief

Oil spill Spill cleanup and costs Liability, compensation

and costs

Liability, compensation

and costs

Nuclear Power Cleanup Licensing Licensing

Source: Birkland (1998), Table 1

Focusing events and policy windows

181

Finally, it is possible to study the ideas contained in legislation in a disaster-prone domain and

consider whether legislative ideas converge to a focused few ideas, or expand to cover much

wider ground. The results of this analysis are provided in Table 14.2, which shows the main ideas

contained in legislation that followed major aviation security incidents (including the aftermath

of the 1996 ValuJet and TWA crashes, the latter of which was widely believed to have been the

result of a terrorist bombing, even though it was not).

This empirical work shows that thinking on focusing events has evolved, from the simple

idea that focusing events “bowl over” other issues, to the more sophisticated idea that focusing

events have variable inﬂuences on policy communities, on the nature of policy debate, and the

content of policy ideas in diﬀerent policy domains.

Broader applications of the focusing event idea

Since Birkland’s (1997) reﬁnement of Kingdon’s (2003) deﬁnition of focusing events, other scholars

have applied or mentioned focusing event theory as an important aspect of the policy process.

This work generally falls into three broad categories: research that cites Kingdon or Birkland

uncritically, accepting the deﬁnitions without any critical analysis, often simply to acknowledge

that events matter; work that cites Kingdon or Birkland and use the focusing event theory to

build, reinforce, or expand the deﬁnitions of focusing events and their causal mechanisms; and

work that seems to us to misapply the focusing event concept to other phenomena that better

explain the processes being described.

A great deal of work relies on Kingdon’s or Birkland’s discussion of focusing events by simply

acknowledging the importance of focusing events, without careful dissection of agenda-setting

theory. Works that cite Kingdon only (e.g. Oakley 2009; Prince 2010; Vaughan and Arsneault

Figure 14.2 Dominant problem frames in school violence, by venue

Source: Based on Table 1 in Lawrence and Birkland (2004)

Birkland and DeYoung

182

Table 14.2 Key issues addressed in aviation security legislation

Topic Aviation Security

Improvement Act

of 1990

Federal Aviation

Authorization Act

of 1996

Federal Aviation

Administration

Authorization Bill

Airport Security

Improvement Act

of 2000

Aviation and

Transportation

Security Act

Homeland

Security Act

of 2002

Airport access control  

Baggage matching  

Background checks of employees  

Cargo and mail security  

Cockpit security 

Employee ID systems 

Explosives and explosives detection  

Create or restore air marshals 

Modify existing organizations  

Create new organizations 

Passenger proﬁling 

Allow pilots to carry fatal weapons 

Allow pilots to carry nonfatal weapons 

Certiﬁcation of screening companies 

Require screening personnel be US citizens 

Require all airport personnel be screened

(including ﬂight crews)



Screeners—general issues  

Make screeners federal employees 

Provide security training to the aircrew 

Provide security training to pilots 

Total 7 8 1 1 17 5

Source: Birkland (2004).

2008) are works that contain overviews of major agenda-setting theory, with a focus on the

streams model, but that fall short of carefully diﬀerentiating focusing events from other drivers

of agenda change. Similarly, works that cite Birkland but not Kingdon (e.g. Ackleson 2005;

Chenoweth and Clarke 2010; Davis 2006; Einsiedel 2000; Melis and Pulwarty 2001; Verhulst

and Walgrave 2009) often simply recapitulate Birkland’sdeﬁnition of a focusing event testing

the deﬁnition or the underlying theory. The second body expands on theory by applying a case

and then building an argument for whether the case or event can be considered as a focusing

event, usually by using quantitative and qualitative data to reinforce or analyze how the case adheres

to the focusing event concept. These studies may also reconsider the deﬁnition of focusing events

in light of the new data and interpretation. For example, Busenberg (2000) makes an interesting

case for the second-order eﬀects of policy changes that can be engendered by focusing events.

He showed how the enactment of the Oil Pollution Act (OPA) of 1990 soon after the Exxon

Valdez oil spill created an atmosphere in which coalitions and oil companies were more willing

to collaborate to prevent and mitigate future spills, and that the OPA induced companies more

open to innovative strategies than they had been before the spill and the OPA.

A particularly promising study is Best’s (2010) analysis of how local newspapers in Denver

treated news events related to homelessness. Best coded story frames to consider, among other

things, whether stories were episodic or thematic, and whether the events being covered were

“actor promoted events” (APEs) or “non-actor promoted events” (NAPEs). Best draws on a wide

range of literature on problem framing to develop sophisticated models to ﬁnd that NAPEs

were highly likely to trigger thematic stories about the problem of homelessness, which appears

to support the idea, implicit in Birkland’s work and explicit in work by Lawrence (2000) and

Molotch and Lester (1975) that NAPEs will trigger greater thematic coverage. But, as we descri-

bed earlier, the idea that an event triggers a broader discussion does not mean that the potential

explanations in the media focus on broader systemic problems.

Best found that, in news coverage of high-proﬁle murders of homeless men in Denver, news-

papers were more prone to “problematize” (in Lawrence’s term) individual choices to remain

homeless than they were to cast light on systemic causes of homelessness. Contrary to her

expectations, APEs were found to be more likely to challenge the status quo, which Best claims

is contrary to Molotch and Lester’s theory that APEs promote the status quo. But focusing

event theory and agenda-setting theory could explain the pro-change orientation of elites as a

form of internal mobilization among a set of elite who perceive homelessness to be a problem

deserving attention and policy change. Best also notes that the articles that contained content

that promoted change and homelessness amelioration were most often written or promoted by

organizations that were dedicated to social change (2010: 87).

The third category of literature is characterized by a form of overreaching in which other

phenomena are characterized as focusing events. For example, Collins and Kapuçu (2008) discuss

the importance of warning systems for tornadoes, using “tornadoes as a focusing event.” But

their use of the term is synonymous with a presumably attention-grabbing event (the tornado

itself). Indeed, many uses of the term do not delve beyond this meaning.

Another common misuse of the concept comes when the surprising result of an election is

cited as a focusing event, as when Béland (2005) and Blankenau (2001) state or imply that the

1991 election of Harris Woﬀord to the US Senate was a focusing event that propelled health

care reform onto the agenda. In Kingdon’s streams metaphor, these surprising results are best

understood as a change in the political stream that can be harnessed, as was done by health care

reformers, to claim that the political atmosphere has changed such that a window of opportu-

nity for policy change is more likely. An election is not a focusing event because it is expected,

planned, and part of normal governmental routine. And it is very diﬃcult to discern whether

Birkland and DeYoung

184

Woﬀord’s election was more a reﬂection of growing attention to health care reform than a trigger

of it. In any case, the election is better accommodated in other aspects of Kingdon’s model, and

need not be force-ﬁtted into the focusing event framework.

Similarly, some researchers imply that indicators are synonymous with focusing events, and

they confuse an indicator change with a focusing event, as in the following inference: “Indicators

are inherently interconnected with focusing events and feedback in the sense that they reﬂect an

objective measure of the former and are prone to subjective constructs regarding the latter”

(Galland and McDaniels 2008: 519). This claim is problematic because indicators are as much sym-

bols as they are facts. But while the essential idea that indicators and focusing events interact is a good

insight, what distinguishes the agenda-setting power of a focusing event is its symbolic power in a

particular place and time, not in indicators of a problem. This is not to deny the idea that a focusing

event can trigger a search for more and better indicators of the problem revealed by the event.

Finally, we also found examples in which focusing events are mentioned without reference

to the literature at all, as when Green-Pederson (2007: 624) argues that “Focusing events … are

hard to predict or even include in variable-oriented explanations.” The latter claim may be true,

but without any theoretical warrant or appeal to the literature as the basis for this claim.

Most of the studies we found were not comparative studies of similar types of events within a

broader class of events, such as “disasters” or “natural disasters” or “industrial accidents.” Birk-

land’s 1997 study revealed considerable diﬀerences in the outcomes of event-driven debates. By

contrast, the studies we reviewed here made few such comparisons of policy community com-

parisons for “matching” focusing events with corresponding communities. Most research focuses

on policy changes over time in connection with one event in one community. (e.g. Carpenter

and Sin 2007) or comparing the policy diﬀerences across countries (Lowry 2006), rather than across

events within similar domains, such as Birkland’s (1997) comparison of hurricane and earthquake

coalition organization.

Fishman (1999) makes a valuable distinction between Type One (common) and Type Two

(novel) focusing events. This distinction is crucial in understanding the way in which the event

may cause a “window of opportunity” (Kingdon 2003) because the event type will have varying

contextual factors that will inﬂuence possible policy solutions, as well as on framing and symbol

propagation.

Conclusion and future directions: focusing events and policy change

A common feature of the focusing event literature is that it often enthusiastically applies the

focusing event concept to all manner of events. Many of these events are not rare, or sudden, or

harmful, in the sense that Birkland’sdeﬁnition suggests. For those policy domains in which

focusing events are important one can at least make a prima facie case that events can often be

precursors—or, if we might be so bold, to be triggers—of policy change. Birkland shows this

phenomenon in studying how the Exxon Valdez oil spill broke a 14-year-long legislative

deadlock over comprehensive oil spill liability legislation. The September 11 terrorist attacks are

widely understood to have triggered major policy change in the United States (in the form of

homeland security initiatives and the restructuring of the Federal Emergency Management

Agency (FEMA) into the Department of Homeland Security (DHS)), although the shape and

nature of that change is very broad and cannot be said to have resulted from some sort of

sudden consensus about the problem revealed by the event. Instead, the event inﬂuenced many

policy communities and policy regimes (May et al. 2006). But the mechanism by which agenda

change is turned into policy change is not yet well deﬁned; we can tell stories about these

changes, but the theory remains thin.

Focusing events and policy windows

185

Birkland (2006) outlined a theory of event-driven policy change and learning which is beyond

the scope of this chapter. To summarize, he argues that focusing events reveal policy failures —

which are the fodder for interest group debate, because these failures reveal the problems that

policy entrepreneurs join to “solutions” in the post-event window of opportunity. But, as we have

discussed, any event is a potential focusing event that has the potential to open a window, and, even

if the window opens, has the potential to yield policy change. Birkland’s model allows, therefore,

for an accumulation of knowledge driven by events, until, at some point, an event brings together

all other event-driven knowledge at a point where it is possible for the policy system to show

evidence of “learning” from the accumulation of events. This formulation is attractive, because

it ties to the literature on ideas and learning from the policy process and from organizational

theory. However, these bodies of literature are not fully reconciled in a way that would support

policy change. But recent research shows that the theoretical construction is promising and can

be accommodated in broader theoretical frameworks, such as the ACF (Albright 2011).

Among the works we reviewed for this chapter, there seems to be a sense—not explicitly stated,

however—that Birkland’sdeﬁnition of focusing event may be so restrictive that there would be

few events to study, thereby deﬁning too small a set of phenomena as focusing events. The varia-

bility of the implicit or explicit deﬁnitions of the term focusing event suggest that it may be

useful to attempt to create some categories or typologies of events, much as Birkland (1997:

Chapter 6) did. Neither Birkland nor other scholars have followed up on these tentative categories

with more careful research, but Fishman’s (1999) notion of Type One and Type two events is

promising. To the extent that future research is undertaken on focusing events, our goal should

be to improve the explanatory and predictive power of any theory of these events. Such a

theory should lend itself to replication and testing across multiple policy domains over many

years. A good foundation has been laid in the social sciences, and with continued sudden, dra-

matic, and unsettling events gaining attention, there is no paucity of cases or of policy domains

subject to these events, all of which provide fruitful material for research.

Bibliography

Ackleson, Jason. 2005. ‘Border security technologies: Local and regional implications’, Review of Policy

Research 22(2): 137–55.

Albright, Elizabeth A. 2011. ‘Policy change and learning in response to extreme ﬂood events in Hungary:

An advocacy coalition approach’, Policy Studies Journal 39(3): 485–511.

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago:

University of Chicago Press.

Béland, D. 2005. ‘Ideas and social policy: An institutionalist perspective’, Social Policy and Administration

39(1): 1–18.

Bennett, W. Lance. 2003. News, The Politics of Illusion. New York: Longman.

Best, R. 2010. ‘Situation or social problem: The inﬂuence of events on media coverage of homelessness’,

Social Problems 57(1): 74–91.

Birkland, Thomas A. 1997. After Disaster: Agenda Setting, Public Policy and Focusing Events. Washington, DC:

Georgetown University Press.

—— 1998. ‘Focusing events, mobilization and agenda setting’ , Journal of Public Policy 18(3): 53 – 74.

—— 2004. ‘Learning and policy improvement after disaster: The case of aviation security’, American

Behavioral Scientist 48(3): 341–64.

—— 2006. Lessons of Disaster. Washington, DC: Georgetown University Press.

—— and Regina G. Lawrence. 2002. ‘The social and political meaning of the Exxon Valdez oil spill’, Spill

Science and Technology Bulletin 7(1–2) (June): 17–22.

——and Regina G. Lawrence. 2009. ‘Media framing and policy change after Columbine’, American

Behavioral Scientist 52(10): 1405–25.

Blankenau,

Joe. 2001. ‘The fate

of national health insurance in Canada and the United States: A multiple

streams explanation’, Policy Studies Journal 29(1): 38–55.

Birkland and DeYoung

186

Burnier, DeLysa. 1994. ‘Constructing political reality: Language, symbols and meaning in politics: A

review essay’, Political Research Quarterly 47(1): 239–53.

Busenberg, George J. 2000. ‘Innovation, learning and policy evolution in hazardous systems’, American

Behavioral Scientist 44(4): 679–91.

Carpenter, Daniel and Gisela Sin. 2007. ‘Policy tragedy and the emergence of regulation: The Food, Drug

and Cosmetic Act of 1938: 1’, Studies in American Political Development 21(2): 149.

Chenoweth, E. and S.E. Clarke. 2010. ‘All terrorism is local: Resources, nested institutions and governance

for urban homeland security in the American federal system’, Political Research Quarterly 63(3): 495–507.

Cobb, Roger W. and Charles D. Elder. 1983. Participation in American Politics: The Dynamics of Agenda-

Building. Baltimore, MD: Johns Hopkins University Press.

——and Marc Howard Ross. 1997. Cultural Strategies of Agenda Denial: Avoidance, Attack, and Redeﬁnition.

Lawrence, KS: University Press of Kansas.

——, Jeannie-Keith Ross and Marc Howard Ross. 1976. ‘Agenda building as a comparative political

process’, American Political Science Review 70(1): 126–38.

Collins, Matthew, L. and Naim Kapuçu. 2008. ‘Early warning systems and disaster preparedness and

response in local government’, Disaster Prevention and Management 17(5): 587–600.

Davis, Charles. 2006. ‘Western wildﬁres: A policy change perspective’, Review of Policy Research 23(1): 115–27.

deLeon, Peter. 1999. ‘The stages approach to the policy process: What has it done? Where is it going’,in

Theories of the Policy Process. Boulder, CO: Westview Press.

Downs, Anthony. 1972. ‘Up and down with ecology: The issue attention cycle’, The Public Interest 28

(Summer): 38–50.

Edelman, Murray J. 1967. The Symbolic Uses of Politics. Urbana, IL: University of Illinois Press.

—— 1988. Constructing

the Political Spectacle.

Chicago: University of Chicago Press.

Einsiedel, Edna F. 2000. ‘Cloning and its discontents; A Canadian perspective’, Nature Biotechnology 18(9):

943–44.

Entman, Robert M. 1993. ‘Framing: Toward classiﬁcation of a fractured paradigm’, Journal of Communication

42(4): 51–9.

Fishman, Donald, A. 1999. ‘ValuJet Flight 592: Crisis communication theory blended and extended’,

Communication Quarterly 47(4): 345.

Galland, Daniel and Timothy McDaniels. 2008. ‘Are new industry policies precautionary? The case of

salmon aquaculture siting policy in British Columbia’, Environmental Science and Policy 11(6) (October):

517–32, doi:10.1016/j.envsci.2008.05.002.

Glascock, Jack. 2004. ‘The Jasper dragging death: Crisis communication and the community newspaper’ ,

Communication Studies 55(1): 29.

Green-Pedersen, Christoﬀer. 2007. ‘The growing importance of issue competition: The changing nature

of party competition in Western Europe’, Political Studies 55(3): 607–28.

Gunter, Valerie J. 2005. ‘News media and technological risks’, Sociological Quarterly 46(4): 671–98.

Gusﬁeld, Joseph. 1981. The Culture of Public Problems: Drinking Driving and the Symbolic Order. Chicago:

University of Chicago Press.

Hilgartner, James and Charles Bosk. 1988. ‘The rise and fall of social problems: A public arenas model’,

American Journal of Sociology 94(1): 53–78.

Kingdon, John W. 2003. Agendas, Alternatives, and Public Policies, 2nd edn, Longman Classics in Political

Science. New York: Longman.

Lawrence, Regina G. 2000. The Politics of Force: Media and the Construction of Police Brutality. Berkeley, CA:

University of California Press.

——and Thomas A. Birkland. 2000. ‘The politics of event driven problem deﬁnition: School violence,

media frames, and policy responses’,

paper presented at the annual meeting of the American Political

Science Association. Washington,

DC.

——and Thomas A. Birkland. 2004. ‘Guns, Hollywood, and criminal justice: Deﬁning the school shootings

problem across public arenas’, Social Science Quarterly 85(5): 1193–207.

Liu, X.S., E. Lindquist and A. Vedlitz. 2011. ‘Explaining media and congressional attention to global climate

change, 1969–2005: An empirical test of agenda-setting theory’, Political Research Quarterly 64(2): 405–19.

Lowry, William. 2006. ‘Potential focusing projects and policy change’, Policy Studies Journal 34(3): 313–35.

Majone, Giandomenico. 1989. Evidence, Argument and Persuasion in the Policy Process. New Haven, CT:

Yale University Press.

May, Peter J. 1990. ‘Reconsidering policy design: Policies and publics’, Journal of Public Policy 11(2): 187–206.

—— 1992. ‘Policy learning and failure’, Journal of Public Policy 12(4): 331–54.

Focusing events and policy windows

187

——, Joshua Sapotichne and Samuel Workman. 2006. ‘Policy coherence and policy domains’, Policy Studies

Journal 34(3): 381–403.

Melis, T.S and Pulwarty, R.S. 2001. ‘Climate extremes and adaptive management on the Colorado River:

Lessons from the 1997–1998 ENSO event’, Journal of Environmental Management 63(3): 307.

Mintrom, Michael. 1997. ‘Policy entrepreneurs and the diﬀusion of innovation’, American Journal of Political

Science 41(3) (July): 738–70.

——and Sandra Vergari. 1998. ‘Policy networks and innovation diﬀusion: The case of state education

reforms’, The Journal of Politics 60(1): 126–48.

Molotch, Harvey and Marilyn Lester. 1974. ‘News as purposive behavior’, American Sociological Review

39(February): 101–13.

——and Marilyn Lester. 1975. ‘Accidental news: The great oil spill as local occurrence and national event’,

American Journal of Sociology 81(2): 235 – 61.

Nakamura, Robert T. 1987. ‘The textbook policy process and implementation research’, Policy Studies

Journal 7(1): 142–54.

Nohrstedt, D. 2010. ‘Do advocacy coalitions matter? Crisis and change in Swedish nuclear energy policy’,

Journal of Public Administration Research and Theory 20(2): 309–33.

Oakley, M.R. 2009. ‘Agenda setting and state policy diﬀusion: The eﬀects of media attention, state court

decisions and policy learning on fetal killing policy’, Social Science Quarterly 90(1): 164–78.

Perrow, Charles. 1999. Normal Accidents Replacement: Living With High-Risk Technologies. Princeton NJ:

Princeton University Press.

Prince,

M.J. 2010. ‘Avoiding blame,

doing good, and claiming credit: Reforming Canadian income

security’, Canadian Public Administration—Administration Publique du Canada 53(3): 293–322.

Sabatier, Paul A. 1988. ‘An advocacy coalition framework of policy change and the role of policy-oriented

learning therein’, Policy Sciences 21: 129–68.

—— 1991a. ‘Political science and public policy and toward better theories of the policy PR’, PS: Political

Science and Politics 24 (June):144–56.

—— 1991b. ‘Toward better theories of the policy process’, PS: Political Science and Politics 24(2): 144–56.

Schattschneider, E.E. 1975. The Semisovereign People. Hinsdale, IL: The Dryden Press.

Schneider, Anne and Helen Ingram. 1991. ‘The social construction of target populations: Implications for

politics and policy’, American Political Science Review 87(2): 334–48.

Schneider, Saundra. 2008. ‘Who’s to blame? (Mis) perceptions of the intergovernmental response to disasters’,

Publius: The Journal of Federalism 38(4) (September 1): 715–38. doi:10.1093/publius/pjn019.

Stone, Deborah A. 1989. ‘Causal stories and the formation of policy agendas’, Political Science Quarterly 104

(2): 281–300.

—— 2002. Policy Paradox: The Art of Political Decision Making. New York: Norton.

Vaughan, Shannon K. and Shelly Arsneault. 2008. ‘Not-for-proﬁt advocacy: Challenging policy images

and pursuing policy change’, Review of Policy Research 25(5): 411–28.

Verhulst, J. and S. Walgrave. 2009. ‘The ﬁrst time is the hardest? A cross-national and cross-issue comparison

of ﬁrst-time protest participants’, Political

Behavior 31(3): 455–84.

Walgrave, S.

and J. Verhulst. 2009. ‘Government stance and internal diversity of protest: A comparative

study of protest against the war in Iraq in eight countries’, Social Forces 87(3): 1355–87.

Walker, Jack L. 1977. ‘Setting the agenda in the US senate: A theory of problem selection’, British Journal

of Political Science 7: 423–45.

Wood, Robert S. 2006. ‘Tobacco’s tipping point: The master settlement agreement as a focusing event’,

Policy Studies Journal 34(3): 419–36.

Birkland and DeYoung

188

Agenda-setting and

political discourse

Major analytical frameworks and

their application

David A. Rochefort and Kevin P. Donnelly

Three questions are central to the agenda-setting perspective in public policy analysis:

1 Among all competing concerns of individuals and groups, what determines which ones will

succeed in gaining the active attention of political decision-makers?

2 How do problems and solutions come to be matched together in public policy formulation?

3 Who are the key actors in agenda-setting, and by what means do they pursue their objectives?

At times, circumstances are such that it is neither puzzling nor questionable why oﬃcials would

be preoccupied with a given predicament. That is, the perception of threat is so reﬂexive, the

dimension of calamity so tangible and evident, no intermediation is necessary between social

awareness of a problem, on the one hand, and government engagement, on the other. Yet issues

of this type, which do not merely occupy space on the radar screen of government but command it,

are exceptional. The broader class of misfortunes and ills in society do not receive automatic attention

from public policy-makers. Rather, they must vie for standing within an appropriate institutional

arena—legislative, administrative, or judicial—already busy with matters current and pending.

Powerful interests outside and inside government can be inﬂuential in gaining consideration

for an issue. However, the more visible an issue, the more widespread its signiﬁcance, and the

more controversial its claims, the less likely will it be that some narrow group of actors, however

well positioned and resourceful, might control its fate. In the politics of public policy-making,

most issues are malleable entities subject to interpretation and prioritization in a contest over the

construction of meaning. Although factual information helps to document a problem’s existence,

rarely is there consensus about which facts are most relevant, how they should be assembled to

create a picture of the whole, or what the implications are concerning those things government

should, or should not, do in remedying a situation. Sorting out these possibilities may be

likened to an exercise in collective political cognition, one rooted in opinion, values, and beliefs

as much as objective analysis.

189

The “argumentative turn” in policy studies is a phrase that has come into usage as a way of

referring to the critical role of discourse within this disorderly process (Fischer and Forester

1993). Majone has written:

Argumentation is the key process through which citizens and policymakers arrive at moral

judgments and policy choices. Public discussion mobilizes the knowledge, experience, and

interest of many people, while focusing their attention on a limited range of issues. Each

participant is encouraged to adjust his view of reality, and even to change his values, as a

result of the process of reciprocal persuasion.

(1989: 2)

Language is more than just a serviceable instrument of con ﬂict and compromise in public debate,

however. Language can also be a vehicle for expressing identity, aﬃrming or challenging the

hierarchical status quo, and embedding discussion of issue speciﬁcs within larger categories of

social understanding (see, e.g., Edelman 1988). All are pertinent to a polity’s response to demands

for change through eﬀective policy choice.

The purpose of this chapter is to review three major veins of research on the role of political

discourse in public policy development with special focus on the identiﬁcation and typing of

issues within the agenda-setting phase. The three areas of scholarship are problem deﬁnition,

framing, and political narrative. Each possesses elements distinctive from the concerns and approa-

ches of the other two, although important points of overlap exist. All perspectives arise from

multiple disciplinary sources. Perhaps owing to this intellectual admixture, scholars situated

within particular literatures often have not adequately acknowledged the relevance of similar

work in other ﬁelds. Our aim, therefore, is to promote a synthesis toward greater interrelationship

and accumulation of knowledge going forward. Following an overview of main concepts

within these parallel bodies of work, we will demonstrate their applicability via a pair of case

studies of recent public policy controversies in Great Britain and the United States.

Problem deﬁnition

In a slim volume called The Semisovereign People published in 1960, E. E. Schattschneider out-

lined an understanding of the linkage among social conﬂict, issue creation, and government

response that continues to inspire research and writing in the public policy ﬁeld several decades

later. Schattschneider’s main interest was the “mobilization of bias” determining which concerns

get organized into and out of the “political game.” One critical asset in the competition for

political attention is large-scale public support, which can either be enhanced or reduced by the

way an issue is deﬁned. Indeed, for Schattschneider, “the deﬁnition of the alternatives is the

supreme instrument of power” (1960: 69).

Building on this work, Cobb and Elder (1972) elaborated a model of agenda-setting that

identiﬁes issue characteristics as a key factor in conﬂict expansion. In general, issues successfully

cast in terms of certain qualities—ambiguity and abstraction, social signiﬁcance, long-term rele-

vance, lack of technical complexity, and familiarity—possess greater potential for visibility and

salience. Thus, the description of issues, or problems, by “policy entrepreneurs” can be an impor-

tant strategy for gaining advantage in the process of political gate keeping that limits the policy

agenda (Kingdon 1995).

As noted by Cobb et al. (1976: 126), “Agenda building is a process ideally suited to com-

parative analysis” (see also Baumgartner et al. 2006). This is so because all political communities

face the same basic challenge of deciding which issues should be given priority in oﬃcial

Rochefort and Donnelly

190

decision-making. These scholars suggest a framework for cross-cultural investigation centering

on the major characteristics of issue careers, including the way public issues are initiated, the degree

to which issue demands are speciﬁc or general in form, the process of issue expansion, and an

issue’s method of entrance onto the formal agenda. Use of language is pervasive across these

dimensions of the policy process and can help to explain both issue-speciﬁc and regime-speciﬁc

tendencies.

Rochefort and Cobb (1993, 1994) examined more closely the categories of claims within the

discourse of problem deﬁnition. Providing an “anatomy of problem description,” they focused

on ﬁve areas of discussion that repeatedly arise in political debate over social issues: problem

causation; nature of the problem (severity, incidence, novelty, proximity, crisis status); character-

istics of the problem population; end-means orientation of problem deﬁner; and nature of the

solution. In eﬀect, this listing of categories provides a template for analyzing problem deﬁnition

as “a distinctive form of rhetoric made up of a habitual vocabulary” (1994: 15).

Each of these dimensions of description can be addressed in a way calculated to maximize a

problem’s perceived importance and handling as a “public issue” rather than as a “private trouble”

(Mills 1959). Problems attributed to social causes beyond individual control, that are widespread

in scope, have severe impacts, threaten the “average citizen,” have reached a dangerous crisis

point, involve sympathetic problem populations, and point to solutions viewed as eﬀective and

aﬀordable represent predictable items of attention inside the political arena. At the same time,

the polarity of these same dimensions can be reversed by problem deﬁners seeking to “contain”

public support for an issue and deem it unworthy of decision-makers’ precious time and resources.

Hirschman (1991) identiﬁed a counter-rhetorical strategy evident across various periods of history

to undermine reform proposals by portraying them as “perverse” (likely to backﬁre), “futile”

(beyond the eﬀective capability of government), and “jeopardizing” (prone to damage an existing

policy or system). Hirschman presents these arguments as the archetypal conservative response

to progressive proposals, although they are equally apt in patterning the discourse of objection

by liberals to policy innovations from the political right.

As Stone (2002) notes, the “language of counting” is prominent in problem deﬁnition. Num-

bers serve to document both the occurrence of a problem and its impacts. Numbers may lend

the appearance of precision to policy arguments, but, in fact, they are subject to manipulation,

open to diﬀering interpretations, and can acquire symbolic meanings just as happens with words

and images. It is not merely that most social problems are associated with “high” and “low”

est

imates.

Measurements can diﬀer widely as a result of time frame selected, categorization of people

and events, the reference points and comparisons used in establishing context, and other choices

supportive of, or detrimental to, the interests of a problem deﬁner.

In the eﬀort to persuade, practitioners of problem deﬁnition resort to language that is factual

and language that is ﬁgurative, sometimes combining the two for maximum impact. Figurative

language refers to words, phrases, and comparisons that convey meanings beyond their literal con-

tent. Analogies are statements that compare problems, or issues, by asserting a logical equiva-

lence, often leaving to inference the nature and extent of similarities between the two.

Analogies can be used to underscore the severity of a situation, suggest causal relationships, and

make predictions about the likely outcome of policy action or neglect. Metaphors are a closely

related form of comparison that link items through implied argument. British scholar Adrian

Beard (2000) observes: “Metaphor is deeply embedded in the way we construct the world

around us and the way that world is constructed for us by others” (p.21). Especially prominent

in the world of politics, Beard notes, are metaphors of sport and war.

One area of problem deﬁnition scholarship that draws directly from the discipline of sociol-

ogy, particularly the “social construction of reality” perspective, is the consideration of problem

Agenda-setting and political discourse

191

ownership (see, e.g., Gusﬁeld 1981). Sometimes what is at stake in deﬁnitional disputes is the

determination of which group in society will gain recognition as established experts for explaining

and resolving a given social problem. Status, inﬂuence, and resources all potentially hinge on

which professional, disciplinary, or moral point of view wins out.

Although problem deﬁnition is integral in issue creation, it also relates to the function of

policy design. Causation is a basic feature of description that can steer attention to a certain method,

level, or object of government intervention because it is in keeping with a problem’s attributed

origins. Sometimes problems achieve agenda standing when they are deﬁned as well suited for

the application of a policy approach already well established and widely supported. In this sense,

it is said, solutions give rise to problems, as well as the other way around. Characterization of

problem populations —the individuals and groups who are seen as aﬀected by, or the source of,

a social problem—conditions the types of instruments viewed by policy-makers as appropriate

forms of treatment due to judgments about blame, deservingness, deviancy, and potential for

change. Not just the substance of programmatic action, namely beneﬁts and burdens, is shaped

by these components of problem deﬁnition, but also the techniques of program implementation

and the rhetoric that imbues the chosen policy (Schneider and Ingram 1997; Rochefort 1986).

Framing

That reality is complex, multidimensional, inherently confusing, and overwhelming in its entirety

constitutes the fundamental premise of the “framing” perspective. A natural cognitive process is

to narrow the ﬁeld of perception to make reality more manageable for interpretation and for

use as a guide to action. As Noakes and Johnston (2005: 2) write: “In the simplest of terms,

framing functions in much the same way as a frame around a picture: attention gets focused on

what is relevant and important and away from extraneous items in the ﬁeld of view.”

Framing enables individuals and groups to make sense of the world, but because it is selective

with regard to information considered, many diﬀerent coherent meanings are possible. Rein

and Schön (1993: 147) discuss the creation of multiple social realities, which leads people facing

a common situation to “see diﬀerent things, make diﬀerent interpretations of the way things are,

and support diﬀerent courses of action concerning what is to be done, by whom, and how to do

it.” The framing approach echoes the problem deﬁnition perspective in underscoring the “malle-

ability” of social problems and, in particular, the divergent impressions that result when a causal

picture having multiple, simultaneous, and sequential inﬂuences is cast in partial terms. However,

the framing paradigm weaves this insight into a model broad enough to encompass thematic,

ideological, cultural, group identity, and motivational elements (Noakes and Johnston 2005).

The intellectual sources of framing analysis are multidisciplinary to an exceptional degree.

Anthropologist Gregory Bateson was one of the earliest scholars to make use of the concept of a

frame (Noakes and Johnston 2005). Subsequently, various applications of framing have appeared

in linguistics, social psychology, media studies, and artiﬁcial intelligence, in addition to political

science and policy analysis. The framing perspective has also emerged as one of the most

inﬂuential paradigms for theoretical and empirical research on social movements, an inherently

cross-disciplinary area with bearing on public policy initiation.

In tracing the impact of framing within political science, Druckman (2010) focuses on two

main types of application. First is the question of how an individual forms preferences, that is “a

rank ordering of a set of objects or alternative actions” (2010: 280). By representing the item

under consideration in diﬀerent ways or contexts, framing has been shown to produce diﬀerent

valuative outcomes. Second is the role of framing in political communication and a speaker’s

ability, at times, to evoke varying audience reactions depending on which facet or dimension of

Rochefort and Donnelly

192

a topic has been emphasized. Iyengar (1996) pursued a similar line of inquiry with respect to public

aﬀairs information in the media, ﬁnding, for example, that “episodic” treatment of social pro-

blems focusing on isolated instances versus more contextual “thematic” reports tends to result in

viewers favoring individualistic explanations for those problems. As Entman (1993) has noted,

to the extent that public opinion proves susceptible to manipulation, framing research conﬁrms

the possibility of elite control, thereby raising far-reaching implications for democratic theory.

Rein and Schön’s(1993)workon“frame analysis” represents a methodical, conceptually sophis-

ticated extension of these ideas into the policy realm (see also Schön and Rein 1994). These

authors examine policy controversies in terms of rival frames that are distinguished by alternative

factual interpretations, social meanings, and normative recommendations. The policy process is

a battle to set the agenda by means of one of these divergent models of reality. How this battle

turns out will serve the symbolic and material interests that shape, and are advanced by, a certain

complex of beliefs, values, and perspectives. None of this means frames are easy to identify,

however. On the contrary, frames exist in the world of (often unspoken) mental constructs. Fur-

ther, similar policy actions may align with diﬀerent frames, or the same frame can give rise to

multiple policy solutions. And yet, when a frame has been successfully decoded, it not only has

great explanatory power, it can also be the ﬁrst step in an emotional and cognitive transaction

by which starkly contending interests begin to ﬁnd common ground.

A frame sets the larger tableau within which political language is used. Statements of caus-

ality, metaphorical comparisons, claims of professional ownership, and larger story narratives all

may animate a policy debate, although the proposition of scholars in this ﬁeld is that relevance

and impact will be determined by compatibility with the dominant understanding of what is at

stake in an issue conﬂict. In their study of the US death penalty, Baumgartner et al. (2008)

found that media coverage shifted in line with the rise of the innocence frame, with elements

critical of capital punishment becoming much more common in news stories.

Often, language performs the crucial function of naming a policy terrain so as to accentuate

the overarching frame of interpretation (Rein and Schön 1993): female liberation, health care

rights, environmental justice, workforce competitiveness, and so on. Since frames consist of tone

as well as substance, there is also a relationship between framing and the expressive aspects of

policy discourse, as would be seen, for example, in harsh condemnation of “shiftless welfare cheats,”

sympathetic support for “victims of cancer,” and compassionate identiﬁcation with neighbors

“at the mercy of rising prices at the gas pump.” In a study of Druze Arabs in Israel, Krebs and

Jackson showed the importance of political rhetoric in framing a campaign for expanded citizenship

rights as the collective reward for military sacriﬁce:

In short, the Druze trapped Jewish leaders in a rhetorical cul-de-sac in which … they

maneuvered their Jewish opponents onto a rhetorical playing ﬁeld on which the Druze

could not lose, for no rebuttal would have been acceptable to key audiences, both domestic

and international.

(2007: 52).

In so far as framing contributes to an issue’s or a problem’s perceived degree of importance, it

has obvious repercussions for determination of agenda priorities. Framing is also relevant when

analyzing the content of disputes in the politics of agenda-setting, although researchers diﬀer on the

pattern. Baumgartner and Jones (1993) have stressed “noncontradictory argumentation” in which

rival frames are advanced without direct challenging of the alternatives, while other scholars have

focused on the clash between opposing frames and the discursive and other tactics used by

advocates to gain the upper hand within such competitions (see, e.g., Cobb and Ross 1997).

Agenda-setting and political discourse

193

Narrative

Many of the same observations made about problem deﬁnition and framing apply to the narrative

perspective on public policy-making. Scholars view narratives as devices that bring order and

meaning to a situation. Patterson and Monroe (1998: 315) state that narrative “refers to the ways in

which we construct disparate facts in our own worlds and weave them together cognitively in

order to make sense of our reality.” Further, Jones and McBeth (2010: 330) write that “narra-

tive cognition may be fundamental to a meaningful human existence.” Narratives make use of

metaphors, symbols, analogies, synecdoche, and other forms of ﬁgurative language in heightening

a message’s impact (Stone 2002). Narrative analysis also has its roots in disparate ﬁelds, most notably

anthropology, literary theory, history, marketing, neuroscience, psychology, and psychoanalysis.

What is distinctive about the study of narratives and public policy is the focus on stories as a

means for representing reality. A story is an account with a guiding arc structured around a

beginning, middle, and end. Most simply, there is a problem, an explanation for occurrence of

the problem, and a tension-resolving solution (Fischer 2003). In fulﬁlling the form of a story,

policy narratives contain predictable elements (Patterson and Monroe 1998). There are characters

or actors, with special prominence given to the roles of victims, villains, and heroes. There is a

point of view, which is the speaker’s perspective of what is important, true, worthy of comment,

and goes without saying. There is a sequential ordering of events to give coherence to a problem’s

origins and development. And there is the setting, which is the socio-cultural context in which a

speaker embeds the account.

The purpose of constructing such a story is neither neutral nor desultory. Rather, it is to

make a point, to register an objection, and to recommend a remedy or, as some put it, to propound

a moral. In other words, policy narratives have a normative element centrally concerned with

human intention. As Fischer (2003: 163) states, “the narrative is especially geared to the goals of

the actors and the way changing goals and intentions causally contribute to social change. It

seeks to comprehend and convey the direction and purpose of human aﬀairs.” A full-bodied,

well-articulated story can go far in dispelling the confusing facts and conﬂicting voices sur-

rounding a public policy issue, superimposing an interpretation of events that is “both lovely

and useful” (Kaplan 1993: 176).

Stepping back from the particulars, it is possible to recognize the repetition of several common

story lines within public policy discourse. Stone (2002) identiﬁes two main types, “stories of

decline” and “stories of control.” The former describes circumstances in which conditions are

deteriorating, disaster looms, and therefore swift action is needed to avert catastrophe. Stone

calls attention to subtypes of stories of decline: the “stymied progress story” (positive change in

an area is being undone), and the “change is only an illusion story ” (contrary to appearances,

reality is worse or better than appreciated). Stories of control focus on problems that had been

understood as beyond eﬀective social intervention, but in fact can now be meliorated by public

policy action. Diﬀerent versions of this story form develop themes of “conspiracy” (powerful

groups or entities have actually been controlling things behind the scenes for their own advantage),

or “blaming the victim” (people aﬀected

by problems have

brought these diﬃculties on them-

selves). With respect to the latter, Douglas Torgerson (1996), a Canadian policy specialist,

emphasizes the need for deconstructing stigmatized identities in social policy to uncover the

discursive opposition of a “privileged center” versus a “neglected margin.” Whatever the genre,

all policy narratives are likely to intermingle statistics, journalistic reporting, anecdotal portrayals,

and other factual and emotive supports for the sake of an explicit or implicit argument.

Ross examines competing narratives with regard to sense making of the attacks of September 11,

2011, from diﬀerent national perspectives. He bases his discussion on the succinct premise that:

Rochefort and Donnelly

194

Narratives are explanations for events (large and small) in the form of short, common-sense

accounts (stories) that often seem simple. However, the powerful images they contain and

the judgments they make about the motivations and actions of their own group, and

others, are emotionally signiﬁcant for groups and individuals.

(2011: 1)

A dominant narrative for Americans, which depicts the attack as evil behavior motivated by

religious and political extremism, construes 9/11 as an act of war. Here, the response that seems

appropriate is massive military intervention abroad to defend the United States proactively by

tracking down and eliminating the terrorists and their supporters. A diﬀerent narrative that is

heard in many parts of the Islamic world may or may not deem the attack as evil incarnate, but

relates it to objectionable activities by the United States in the Middle East over a period of

years. The American military response is predicted to extend these past injustices by harming

innocent Muslim populations while ignoring the precipitating factors and conditions that gave

rise to 9/11. Analyzing the contradictions between such narratives, Ross underscores the sig-

niﬁcance of divergent group beliefs and experiences, facts that are selectively remembered and

interpreted, and preexisting narratives within a culture that guide collective responses to stress.

Jones and McBeth (2010) have introduced a Narrative Policy Framework (NPF) for assessing

the impact of policy narratives by means of empirical quantitative methodology. Their goal, in

short, is to subject a postpositivist concept to positivist standards of evaluation. As outlined, NPF

takes into account the principal elements of narrative form (plot, setting, characters, and moral)

while categorizing generalized story content with respect to ideology (or another belief system).

The methodology then uses these inputs to generate hypotheses concerning the potential impact,

or utility, of narrative for individuals and groups within a policy dispute. Elaborating on this

approach, Shanahan et al. (2011) drew upon the NPF to formulate an additional set of hypotheses

incorporating independent and dependent variables of interest concerning policy outcomes,

policy learning, policy change, public opinion, and strategy from Sabatier and Jenkins-Smith’s

(1993) well known Advocacy Coalition Framework.

Two cases

During the summer of 2011 London experienced days of rioting marked by violence, social

unrest, and destruction on a massive scale. The turbulence generated a sweeping national exam-

ination of conscience in British society on topics ranging from income inequality to the decline

of moral virtue. The fall of 2011 witnessed an outpouring of citizen protests on Wall Street and

in dozens of cities around the United States. Few anticipated this “Occupy” movement and the

strong chord its message struck with large segments of the American people. A brief review of each

of these cases will serve to illustrate the potent role of language in political situations surrounding

the emergence (and reemergence) of divisive social issues.

Riots in the United Kingdom

Late in the evening of Saturday, August 6, 2011, riots broke out in the north London neigh-

borhood of Tottenham. Rioters set buildings on ﬁre, clashed with police, and smashed the win-

dows of local shops. In subsequent days, the violence and destruction spread to neighboring cities

and towns. According to the Guardian newspaper, by the third night “Buildings were torched,

shops ransacked, and oﬃcers attacked with makeshift missiles and petrol bombs as gangs of

hooded and masked youths laid waste to streets right across the city” (Dodd and Davies 2011).

Agenda-setting and political discourse

195

Calm did not return until the fourth night following Prime Minister David Cameron’s decision

to deploy an additional 10,000 police oﬃcers across the streets of London.

Both during and immediately following the riots, political commentary focused on a central

question of problem deﬁnition and narrative composition: Who or what was to blame for this

historic display of social unrest? The immediate precipitant was obvious. Days earlier, police had

shot and killed Mark Duggan when attempting to arrest him in an investigation of gun vio-

lence. A peaceful protest in Tottenham denouncing the police role in Duggan’s death, which

some suggested was racially motivated, ultimately got out of hand (Faiola 2011). Yet, according

to the New York Times, “the circumstances of Mr. Duggan’s death appeared to be remote from

the forces driving the riots, at least in the assessment of many of those who are most familiar

with the neighborhoods aﬀected.” Instead, this incident was simply the “original trigger for the

violence” (Burns 2011). While most seem to accept the truth of this observation, this is where

agreement ended in regard to underlying cause of the riots.

Drastic cuts to services, or “austerity measures,” surfaced as one prominent culprit noted in the

media. Consider the following account oﬀered by Nina Power of the Guardian, who suggested

a “context to London’s riots that can’t be ignored.”

Since the coalition came to power just over a year ago, the country has seen multiple student

protests, occupations of dozens of universities, several strikes, a half-a-million-strong trade

union marches and now unrest on the streets of the capital … Each of these events was

sparked by a diﬀerent cause, yet all take place against a backdrop of brutal cuts and enforced

austerity measures. The government knows very well that it is taking a gamble, and that its

policies run the risk of sparking mass unrest on a scale we haven’t seen since the early

1980s. With people taking to the streets of Tottenham, Edmonton, Brixton and elsewhere

over the past few nights, we could be about to see the government enter a sustained and

serious losing streak.

(8 August 2011)

Power also proposed basic income inequality as a critical driver of the riots. After ﬁrst establishing

the link between inequality and social issues, Power stated that “[d]ecades of individualism,

competition and state-encouraged selﬁshness—combined with a systematic crushing of unions and

the ever-increasing criminalisation of dissent—have made Britain one of the most unequal countries

in the developed world.” Ravi Somaiya of the New York Times echoed this view in stating that

“Economic malaise and cuts in spending and services instituted by the Conservative-led gov-

ernment have been recurring ﬂashpoints for months” (Somaiya 2011). Political commentators

were not the only ones delivering this message. In the aftermath of the riots, Labour Party leader

Ed Miliband called the cause of the riots “complex,” but frequently pointed to Britain’s “unequal

society” as a key contributor (Dunt 2011). The long-term pattern of deterioration within this

constructed narrative is consistent with Stone’s

(2002) “stories of decline.” By

assigning blame to

the Government as a guilty party in igniting the riots, this interpretation also pointed to a logical

solution—roll back the cuts and create a more equal society. Little in the way of speciﬁc policy

proposals has yet accompanied the linking of inequality to the rioting. However, in a major speech

delivered against the backdrop of his old secondary school, Miliband set the stage for possible future

action by urging the governing party to establish a formal “commission of inquiry” to investigate the

cause of the riots, and he tasked it, in particular, with examining “issues of inequality” (Miliband 2011).

Other commentary, however, tagged the rioters as reckless law-breakers skilled at exploiting

New Age social media outlets. Quickly dispensing with the notion that the riots were in some

way a valid political expression, The Daily Telegraph oﬀered the following:

Rochefort and Donnelly

196

What we have experienced in London and elsewhere since Saturday night is a wholly new

phenomenon: violent disorder whose sole intent is criminal. These are not protesters, as the

BBC stupidly insisted on calling them: they are looters and vandals and thieves. The unstruc-

tured nature of their violence has made it exceedingly diﬃcult for the police to gather any

worthwhile intelligence on them. Yet their sophisticated use of social media allows them to

muster at will. They have had the whip hand for four days and the police have constantly

had to play catch-up.

(Telegraph View 2011)

Government oﬃcials in London expressed parallel sentiments. During a press conference

just days following the initial night of rioting, deputy assistant commissioner to the Metropoli-

tan police, Steve Kavanagh, blamed the rioting on “really inﬂammatory, inaccurate” messages

on Twitter, adding that “[s]ocial media and other methods have been used to organise these

levels of greed and criminality” (Halliday 2011). Going further, Mike Butcher, a digital aide to

London’s mayor, told the BBC that “[m]obile phones have become weaponised in their cap-

ability of spreading information about where to target next.” Butcher remarked it was “unbe-

lievable” that BlackBerry Messenger had not been shut down during the riots (Minicucci 2011).

Ultimately, government authorities chose not to interrupt service. However, this criminal,

rather than political, framing of the riots impacted solution development. London police

authorities launched an oﬃcial investigation into the use of BlackBerrys as an organizing tool by

the rioters.

A broader theme in discourse about the riots was the issue of moral decay. Prime Minister

Cameron pointed to the UK’s “slow-motion moral collapse” as overriding cause of the riots:

These riots were not about race: the perpetrators and the victims were white, black and

Asian. These riots were not about government cuts: they were directed at high street stores,

not Parliament. And these riots were not about poverty: that insults the millions of people

who, whatever the hardship, would never dream of making others suﬀer like this.

No, this was about behaviour. People showing indiﬀerence to right and wrong. People

with a twisted moral code. People with a complete absence of self-restraint.

(15 August 2011)

Cameron announced that his plan to ﬁxa“broken” Britain would be back on the “top of [his]

political agenda.” This policy vision, which centered on the goal of reversing general social

decay, originated during the previous 2010 election, but it was dusted oﬀ as the ideal response

to the current violence.

In a Wall Street Journal op-ed, Britain’s chief rabbi, Jonathan Sacks, wrote that no one should

have been surprised by the London riots. In his view, “Britain is the latest country to pay the

price for what happened half a century ago in one of the most radical transformations in the

history of the West. In virtually every Western society in the 1960s there was a moral revolu-

tion, an abandonment of its entire traditional ethic of self-restraint” (Sacks 2011). Not to be

outdone, another religious leader, the Archbishop of Canterbury, Rowan Williams, sermonized

on moral failings by criticizing an “educational philosophy at every level [that] has been more

and more dominated by an instrumentalist model; less and less concerned with a building of

virtue, character and citizenship.” The solution, according to Dr. Williams, could be found in

“rebuilding the skills of parenting in some of our communities, [and] in rebuilding education

itself” (Walton 2011). These, of course, are not new ideas in socially conservative circles. In

eﬀect, the riots provided a context, not to mention a pretext,

for attaching existing solutions to a

Agenda-setting and political discourse

197

new problem, which is a well-known sequence in problem deﬁnition (Kingdon 1995). The

dynamic of religious ﬁgures asserting the relevance of their special brand of expertise also represents

a striking illustration of the struggle for “problem ownership” that frequently occurs in public

policy debate.

Occupy Wall Street

On September 17, 2011, a protest movement emerged on the streets of Lower Manhattan, the

likes of which had not been seen in the US for decades. On that day, a small group of protesters

gathered to “occupy” parts of Wall Street and camp out in nearby Zuccotti Park in a dramatic

attempt to draw attention to growing inequalities in American wealth and income. Although

this initial action turned out to be fairly modest in scope, within weeks similar demonstrations

cropped up in dozens of cities across the nation and the world. Increasingly larger groups of pro-

testers claiming solidarity with the Occupy movement set up encampments in city parks and

other public spaces from Boston to Oakland, Nashville to Reno. Proudly “leaderless” and vague

in its goals, the Occupy movement nonetheless captured the attention of law-makers and

dramatically shifted the focus of American political discourse.

In the early days of the Occupy protests, political commentary focused on explaining the

movement itself. An array of competing narratives resulted. Tying the movement to the “story

of our economic woes,” Paul Krugman (2011), a liberal columnist for the New York Times,

outlined the following “three acts”:

In the ﬁrst act, bankers took advantage of deregulation to run wild (and pay themselves

princely sums), inﬂating huge bubbles through reckless lending. In the second act, the

bubbles burst—but bankers were bailed out by taxpayers, with remarkably few strings

attached, even as ordinary workers continued to suﬀer the consequences of the bankers’

sins. And, in the third act, bankers showed their gratitude by turning on the people who

had saved them, throwing their support—and the wealth they still possessed thanks to the

bailouts—behind politicians who promised to keep their taxes low and dismantle the mild

regulations erected in the aftermath of the crisis. Given this history, how can you not

applaud the protesters for ﬁnally taking a stand?

Starkly diﬀerent was the take of conservative columnist for the Washington Post, Charles Krau-

thammer, who linked the Occupy movement to President Obama’s recent calls for the rich to

“pay their fair share”:

To the villainy-of-the-rich theme emanating from Washington, a child is born: Occupy

Wall Street. Starbucks-sipping, Levi’s-clad, iPhone-clutching protesters denounce corpo-

rate America even as they weep for Steve Jobs, corporate titan, billionaire eight times over.

These indignant indolents saddled with their $50,000 student loans and English degrees

have decided that their lack of gainful employment is rooted in the malice of the million-

aires on whose homes they are now marching—to the applause of Democrats suﬀering

acute Tea Party envy and now salivating at the energy these big-government anarchists will

presumably give their cause.

(13 October 2011)

Aside from reﬂecting the ideological predispositions of these two pundits, such accounts spoke

to a broader eﬀort on the part of media ﬁgures to contextualize the ongoing social unrest.

Rochefort and Donnelly

198

Certain political commentators drew analogies to other grassroots movements. Nicholas Kristof,

an international news reporter for the New York Times, drew a parallel between the Occupy

movement and the pro-democracy protesters in Cairo’s Tahrir Square:

[T]here is the same cohort of alienated young people, and the same savvy use of Twitter

and other social media to recruit more participants. Most of all, there’s a similar tide of

youthful frustration with a political and economic system that protesters regard as broken,

corrupt, unresponsive and unaccountable.

(Kristof 2011)

Use of this particular analogy sent a powerful message by framing the Occupy movement as part of

a global phenomenon in which a virtuous younger generation had stood up to an ineﬀective

and illegitimate power structure. Other observers, including President Obama, compared the

Occupy and Tea Party movements, the latter coalescing two years earlier around the goals of

shrinking government and an originalist interpretation of the Constitution. During an interview

Obama suggested:

In some ways, they’re not that diﬀerent from some of the protests that we saw coming

from the Tea Party. Both on the left and the right, I think people feel separated from their

government. They feel that their institutions aren’t looking out for them.

(Dwyer 2011).

This comparison both anointed (prematurely) the Occupy movement as a potent political force

and presented the idea that a counterweight to the Tea Party movement was now emerging.

The formulation of comparisons to movements as diverse as the Arab Spring and the Tea Party

is testament to the amorphous nature of the Occupy protests. However, it also illustrates the

extent to which analogies portray political events in dramatically diﬀerent lights in the creative

application of political discourse.

Despite its ambiguity, the Occupy movement articulated an agenda clear and motivating

enough to draw thousands of Americans into the streets. This may be due, in part, to the fact

that protesters proposed a new framework for debating US economic policy. Unlike during the

previous two years, in which the national debt, government spending, and draconian cost-cutting

measures dominated the discourse, the Occupy movement spotlighted economic inequality as a

transcendent issue. A New York Times editorial chastised the “chattering classes [who] keep com-

plaining that the marchers lack a clear message and speciﬁc policy prescriptions,” and oﬀered the

following defense:

At this point, protest is the message: income inequality is grinding down that middle class,

increasing the ranks of the poor, and threatening to create a permanent underclass of able,

willing but jobless people. On one level, the protesters, most of them young, are giving

voice to a generation of lost opportunity.

(New York Times Editorial 2011)

Along with their physical presence, the protesters supported their inequality frame with pow-

erful sloganizing. Put simply, “We are the 99%.” This phrase eﬀectively hit home the move-

ment’s essential complaint—that wealth is highly concentrated in the top 1 percent of the

population while the status quo serves the ﬁnancial elite at the expense of the many. Similarly,

by featuring a “raised ﬁst”—the universal symbol of solidarity and resistance—on the Occupy

Agenda-setting and political discourse

199

Wall Street website and at rallies across the nation, the group reinforced the scope of their

solution: “A revolution of the mind as well as the body politic” (OccupyWallSt.org).

Critics of the Occupy movement portrayed its participants in a negative light to discredit

their message. In her appearance on the Fox News “Hannity” program, conservative author and

commentator Ann Coulter assured the viewing audience that “none of these kids at ‘Occupy

Wall Street’ are coming from places like Yale, they are coming from the most bush league schools

which is why none of them can even say why they are there.” Further, Coulter expressed con-

ﬁdence that “Americans are reacting the way they should, which is with hilarity and revulsion”

(Coulter 2011). Even more contemptuously, talk radio show host Rush Limbaugh told his

audience:

[W]hen I was ten years old … I was more self-suﬃcient than this parade of human debris

calling itself Occupy Wall Street … If you look at the minutes and read their website, these

people are announcing that they’re parasites, that they are not self-suﬃcient and that they

are totally dependent on the very things and people they ’re protesting.

(Limbaugh 2011)

Here was a classic tactic of conﬂict containment via rhetoric: defuse the issue and block its expansion

by means of a direct verbal attack on a protest group and its leaders (Cobb and Elder 1972).

Other more mainstream critics elected to challenge the inequality frame head on. Repre-

sentative Paul Ryan, in a prominent speech before the Heritage Foundation, asked the audience

to consider “[w]hether we are a nation that still believes in equality of opportunity, or whether

we are moving away from that, and toward an insistence on equality of outcome.” The latter,

Ryan suggested, constitutes “the moral basis of class warfare—a false morality that confuses fairness

with redistribution, and promotes class envy instead of social mobility” (Ryan 2011). Mitt Romney,

front-runner for the 2012 Republican presidential nomination, responded to a question about

the Occupy protests by saying: “I think it’sdangerous,this[is]classwarfare” (Boxer 2011). The contest

of economic frames was classic: inequality versus competition, fairness versus winner-take-all.

The fact that prominent Republicans saw the need to defend, not merely promote, their

preferred economic frame illustrates the surprising strength of the Occupy movement. What

remains unclear at the time of this writing, however, is the long-term impact of such rhetorical

point-counterpoint on US politics and public policy.

Discussion and conclusion

The two cases we have discussed may be notable for their timeliness and for the signiﬁcance of

issues being raised, but they are not unique from a discursive standpoint. The patterns of argu-

mentation, descriptive construction, and storytelling are commonplace, making it easy to demon-

strate the applicability of the problem deﬁnition, framing, and narrative perspectives across the

political landscape. This should not be surprising if it is true, as has sometimes been claimed,

that “all language is political” and “all politics is language.” Yet the very pervasiveness of poli-

tical discourse inclines the subject to certain ambiguities that should be noted at the close of this

review.

First, although much work in the ﬁeld of discourse analysis has been devoted to typing the

rhetorical strategies that recur in politics, the raw material in any given case study is always unique,

with its own combination of themes, accents, and emotions. This situation makes it likely that

no two analysts will interpret a case precisely the same way, either in identifying critical ele-

ments of language or in explicating their role in the outcome of a conﬂict. The focus on politics

Rochefort and Donnelly

200

as a “text” has, in short, opened the door to all the creativity and all the inconsistency that one

ﬁnds in other ﬁelds of textual analysis, such as literary studies. Thus, even though it is possible to

recognize a close connection among the problem deﬁnition, framing, and narrative approaches,

these frameworks have oddly served as alternative lenses at times, that is, as intellectual spheres

with their own reference points, observational criteria, and conceptual intentions. Whether this

pluralism has proved boon or bane to the broad community of scholars and students interested

in the impact of political language is questionable.

Second, it is understandable that the ﬂuidity of political discourse analysis has now given rise

to attempts to pin down crucial variables through quantitative measurement. The Narrative

Policy Framework of Jones and McBeth (2010) has already been cited as an illustration of this

quantifying impulse, one that aims to gain a place for language variables within the policy stu-

dies literature with respect to such basic processes as policy change and policy learning. Simi-

larly, the speciﬁcation, tabulation, and interrelating of basic numerical indicators has been highly

useful in documenting the rise and fall of rival policy frames for those working within this

perspective (see, e.g., Baumgartner et al. 2008; Baumgartner and Jones 1993). Various media

and survey studies have also conﬁrmed hypothesized relationships between language construc-

tions and political attitudes of diﬀerent kinds. So far, however, the ambitiousness of this quan-

titative vein of work, worthwhile though it may be, continues to outstrip its accomplishments,

nor has it done much to integrate the divergent traditions of scholarship within this area.

Third, a persistent barrier to positivist analysis in searching for regularities in the expressions

of political language, on the one hand, and their impact on agenda setting, on the other, is

the shifting meaning of similar language and narratives depending on context. Within the

American national healthcare debate, it has been intriguing to note the contradictory uses of

what might be called the “rhetoric of foreign standards. ” In the hands of diﬀerent polemicists,

the same examples of national health systems elsewhere have been used either to shame the

United States for being a “laggard” in providing universal coverage and cost control, or to praise

the US health care system as the “best in the world.” Even aside from such blatant paradoxes,

important nuances of discursive meaning are likely whenever attention shifts from one policy

domain to another. Thus, Rochefort et al. (1998) found substantial diﬀerences in rhetorical uses

of the concept of “community” within policy debates concerning mental health, criminal jus-

tice, and social service programs in the US and Canadian province of Quebec. Underlying

similar sounding language were both divergent value assumptions and alternative pathways for

policy design.

A concern with the potent use of language in politics is not new. It dates back at least as far

as Plato’s academy in ancient Greece and the ﬁrst formal analyses of rhetoric written by Aris-

totle, former student and faculty member there. Today, however, researchers from many dif-

ferent disciplines who share an interest in the politics of public policymaking are injecting new

energy and concepts into this ﬁeld of study. This academic focus is being matched and doubtless

heightened by a dramatic expansion in the techniques and forms of media available to the practi-

tioners of political language, inside and outside government. The purpose of this chapter has

been to signal promising trends in the major scholarly approaches surrounding this development

while encouraging greater awareness of their intellectual common ground.

Bibliography

Bardach, Eugene. 2009. A Practical Guide for Policy Analysis. Washington, DC: CQ Press.

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago:

University of Chicago Press.

Agenda-setting and political discourse

201

——, Green-Pedersen, and Bryan D. Jones. 2006. ‘Comparative studies of policy agendas ’, Journal of European

Public Policy 13(7): 959–74.

——, Suzanna L. De Boef and Amber E. Boydstun. 2008. The Decline of the Death Penalty and the Discovery

of Innocence. Cambridge: Cambridge University Press.

Beard, Adrian. 2000. The Language of Politics. London: Routledge.

Boxer, Sarah B. 2011. ‘Romney: Wall Street protests “Class Warfare’, National Journal, 5 October. Available at:

http://nationaljournal.com/2012-presidential-campaign/romney-wall-street-protests-class-warfare–20111004.

Burns, John F. 2011. ‘Cameron deploys 10,000 more oﬃcers to riots’, New York Times, 10 August, p. A4.

Cameron, David. 2011. Speech on the ﬁght-back after the riots, Witney, Great Britain, 15 August.

Available at: http://www.newstatesman.com/politics/2011/08/society-ﬁght-work-rights.

Cobb, Roger W. and Charles D. Elder. 1972. Participation in American Politics: The Dynamics of Agenda-

Building. Boston: Allyn & Bacon.

——and Marc H. Ross (eds). 1997. Cultural Strategies of Agenda Denial: Avoidance, Attack, and Redeﬁnition.

Lawrence, KS: University Press of Kansas.

——, Jennie Keith-Ross and Marc Howard Ross. 1976. ‘Agenda building as a comparative political process’,

American Political Science Review 70(1): 126–38.

Coulter, Ann. 2011. ‘Remarks on Fox News’“Hannity”,’ 14 October. Available at: http://www.foxnews.

com/on-air/hannity/2011/10/17/ann-coulter-sounds-brainless-wall-street-protesters.

Dodd, Vikram and Caroline Davies. 2011. ‘London riots escalate as police battle for control’, The Guardian,8

August. Available at: http://www.guardian.co.uk/uk/2011/aug/08/london-riots-escalate-police-battle.

Druckman, James N. 2010. ‘What’s it all about? Framing in political science’, in Gideon Keren (ed.)

Perspectives on Framing. New York: Psychology Press, 279–302

Dunt, Ian. 2011. ‘Miliband: if Cameron doesn’t

hold an inquiry

I will’, Politics.co.uk, 12 August. Available

at: http://www.politics.co.uk/news/2011/08/12/miliband-pushes-for-national-inquiry-into-eng.

Dwyer, Devin. 2011. ‘Obama: Occupy Wall Street “not that diﬀerent” from Tea Party protests’, ABC

News, 18 October. Available at: http://abcnews.go.com/blogs/politics/2011/10/obama-occupy-wall-

street-not-that-diﬀerent-from-tea-party-protests/.

Edelman, Murray. 1988. Constructing the Political Spectacle. Chicago: University of Chicago Press.

Entman, Robert M. 1993. ‘Framing: Toward clariﬁcation of a fractured paradigm’, Journal of Communication

43(4): 51–8.

Faiola, Anthony. 2011. ‘Riots breakout in London neighborhood’ , Washington Post, 7 August. Available at:

http://www.washingtonpost.com/world/riots-break-out-in-london-neighborhood/2011/08/07/gIQAn0

sA0I_story.html.

Fischer, Frank. 2003. Reframing Public Policy: Discursive Politics and Deliberative Practices. Oxford: Oxford

University Press.

——and John Forester (eds). 1993. The Argumentative Turn in Policy Analysis and Planning. Durham, NC:

Duke University Press.

Gusﬁeld, Joseph R. 1981. The Culture of Public Problems. Chicago: University of Chicago Press.

Halliday, Josh. 2011. ‘London riots: BlackBerry to help police probe Messenger looting “role”’, Guardian,

8 August. Available at: http://www.guardian.co.uk/uk/2011/aug/08/london-riots-blackberry-messenger-

looting.

Hirschman, Albert. 1991. The Rhetoric of Reaction: Perversity, Futility, Jeopardy. Cambridge, MA: Belknap

Press.

Iyengar, Shanto. 1996. ‘ Framing responsibility for political issues’, Annals of the American Academy of Political

and Social Science 546(1): 59–70.

Jones, Michael D. and Mark K. McBeth. 2010. ‘A narrative policy framework: Clear enough to be

wrong?’ Policy Studies Journal 38(2): 329–53.

Kaplan, Thomas J. 1993. ‘Reading policy narratives: beginnings, middles, and ends’, 167–85, in Frank

Fischer and John Forester (eds) The Argumentative Turn in Policy Analysis and Planning. Durham, NC:

Duke University Press.

Kingdon,

John. 1995. Agendas, Alternatives,

and Public Policies. Boston: Little, Brown.

Krauthammer, Charles. 2011. ‘The scapegoat strategy’, Washington Post, 13 October. Available at: http://

www.washingtonpost.com/opinions/the-scapegoat-strategy/2011/10/13/gIQArNWViL_story.html.

Krebs, Ronald R. and Patrick Thaddeus Jackson. 2007. ‘Twisting tongues and twisting arms: The power

of political rhetoric’, European Journal of International Relations 13(1): 35–66.

Kristof, Nicholas. 2011. ‘The bankers and the revolutionaries ’, New York Times, 1 October, SR11.

Krugman, Paul. 2011. ‘Confronting the malefactors’, New York Times, 6 October, A27.

Rochefort and Donnelly

202

Limbaugh, Rush. 2011. ‘Remarks on the “Rush Limbaugh Show”’, 5 October. Available at: http://www.

rushlimbaugh.com/daily/2011/10/05/anti_capitalist_protestors_face_dilemma_should_they_buy_stuﬀ.

Majone, Giandomenico. 1989. Evidence, Argument, and Persuasion in the Policy Process. New Haven, CT:

Yale University Press.

Miliband, Ed. 2011. Speech on the London riots, Haverstock School in Chalk Farm, Great Britain, 15

August. Available at: http://www.newstatesman.com/politics/2011/08/society-young-heard-riots.

Mills, C. Wright. 1959. The Sociological Imagination. Oxford: Oxford University Press.

Minicucci, Daniela. 2011. ‘The role of digital and social media in the London riots’, Global News, 9 August.

Available at: http://www.globalnews.ca/feature/6442461111/story.html.

New York Times editorial. 2011. ‘Protesters against Wall Street: It’s obvious what they want’, New York

Times, 9 October, p. SR10

Noakes, John A. and Hank Johnston. 2005. ‘Frames of protest: A roadmap to a perspective’,1–29, in

Hank Johnston and John A. Noakes (eds) Frames of Protest: Social Movements and the Framing Perspective.

Oxford: Rowman and Littleﬁeld.

OccupyWallSt.org. 2011. ‘A modest call to action on this September 17th’, 17 September. Available at:

http://occupywallst.org/article/September_Revolution/.

Patterson, Molly and Kristen Renwick Monroe. 1998. ‘Narrative in political science’, Annual Review of

Political Science 1: 315–31.

Power, Nina. 2011. ‘There is a context to London’s riots that can’t be ignored’, Guardian, 8 August.

Available at: http://www.guardian.co.uk/commentisfree/2011/aug/08/context-london-riots.

Rein, Martin and Donald Schön. 1993. ‘Reframing policy discourse’, 145–66, in Frank Fischer and John

Forester (eds) The Argumentative Turn in Policy Analysis and Planning. Durham, NC: Duke University Press.

Rochefort, David A. 1986. American Social Welfare Policy: Dynamics of Formulation and Change. Boulder,

CO: Westview Press.

——and Roger W. Cobb. 1993. ‘Problem deﬁnition, agenda access, and policy choice’, Policy Studies

Journal 21(1):

56–71.

——and Roger W.

Cobb. 1994. ‘Problem deﬁnition: An emerging perspective’ in David A. Rochefort

and Roger W. Cobb (eds) The Politics of Problem Deﬁnition: Shaping the Policy Agenda. Lawrence, KS:

University Press of Kansas, 1–31.

——, Michael Rosenberg and Deena White. 1998. ‘Community as a policy instrument: A comparative

analysis’, Policy Studies Journal 26(3): 548–68.

Ross, Marc H. 2011. ‘The political psychology of competing narratives: September 11 and beyond’, Social

Science Research Council. Available at: http://essays.ssrc.org/sept11/essays/ross.htm.

Ryan, Paul. 2011. Remarks before the Heritage Foundation, Washington, DC, 26 October. Available at:

http://blog.heritage.org/2011/10/26/video-rep-paul-ryan-on-saving-the-american-idea/.

Sabatier, Paul A. and Hank Jenkins-Smith. 1993. Policy Change and Learning: An Advocacy Coalition

Approach. Boulder, CO: Westview Press.

Sacks, Jonathan. 2011. ‘Reversing the decay of London undone’, Wall Street Journal, 20 August. Available

at: http://online.wsj.com/article/SB10001424053111903639404576516252066723110.html.

Schattschneider, E.E. 1960. The Semisovereign People: A Realist’s View of Democracy in America. Hinsdale, IL:

Dryden Press.

Schneider, Anne and Helen Ingram. 1997. Policy Design for Democracy. Lawrence, KS: University of Kansas

Press.

Schön, Donald A. and Martin Rein. 1994. Frame Reﬂection: Toward the Resolution of Intractable Policy Controversies.

New York: Basic Books.

Shanahan, Elizabeth A., Michael D. Jones and Mark K. McBeth. 2011. ‘Policy narratives and policy processes’,

Policy Studies Journal 39(3): 535–61.

Somaiya, Ravi. 2011. ‘London sees twin perils converging to fuel riot’, New York Times, 8 August, A4.

Stone, Deborah. 2002. Policy Paradox: The Art of Political Decision Making. New York: Norton.

Telegraph View editorial. 2011. ‘The criminals who shame our nation’, The Telegraph, 9 August. Available

at: http://www.telegraph.co.uk/comment/telegraph-view/8691352/The-criminals-who-shame-our-nat

ion.html.

Torgerson, Douglas. 1996. ‘Power and insight in policy discourse: Post-positivism and problem deﬁnition’ ,

Laurent Dobuzinskis, Michael

Howlett, and David Laycock (eds) Policy Studies in Canada: The State of

the Art. Toronto: University of Toronto Press, 266–98.

Walton, Jeﬀ. 2011. ‘Church of England bishops: Lack of virtue at fault for London riots’, The Institute on

Religion and Democracy, 22 August. Available at: http://www.theird.org/page.aspx?pid=2012.

Agenda-setting and political discourse

203

Mass media and policy-making

Stuart Soroka, Stephen Farnsworth, Andrea Lawlor and Lori Young

Mass media can, and often do, play a critical role in policy-making. The typical view of media

is that they matter in the early stages of the policy process—that media can help to set an

agenda, which is then adopted and dealt with by politicians, policy-makers, and other actors.

The impact of media is rarely so constrained, however. Our argument here, in short, is that

media matter, not just at the beginning but throughout the policy process.

Many of the standard accounts of policy-making have a much too narrow view of the timing

of media eﬀects. That said, the ways in which mass media can matter are relatively well under-

stood. Existing work tells us that media can draw and sustain public attention to particular issues.

They can change the discourse around a policy debate by framing or deﬁning an issue using

dialogue or rhetoric to persuade or dissuade the public. Media can establish the nature, sources,

and consequences of policy issues in ways that fundamentally change not just the attention paid

to those issues, but the diﬀerent types of policy solutions sought. Media can draw attention to the

players involved in the policy process and can aid, abet or hinder their cause by highlighting

their role in policy-making. Media can also act as a critical conduit between governments and

publics, informing publics about government actions and policies, and helping to convey public

attitudes to government oﬃcials.

Allowing for the possibility that any and all of these eﬀects can be evident not just in the

early stages but throughout the policy process makes clear the potentially powerful impact we

believe that media can have on policy. Indeed, mass media are in the unique position of having

a regular, marked impact on policy, but from outside the formal political sphere, often without

even being recognized as a policy player.

This chapter reviews the state of the literature on media and policy-making. It reviews two

of the most prominent theories in the study of media and policy-making: agenda-setting and

issue framing. It then considers some of the normative implications of the regular impact of

media on policy-making. Is the fact that media matter to policy-making a good thing? There

are beneﬁ ts, to be sure, but also costs, and we consider below the costs associated with the well-

known event-driven, sensationalist tendencies in media content. We then ﬁnish with a brief

example from Canadian environmental news coverage; an example which illustrates some of

the problematic tendencies in media content, and highlights some of the issues with media as a

policy actor.

204

Agenda-setting and issue attentiveness

The policy agenda-setting literature has its roots in early work in political behavior focused on

how media coverage of political events impacts electoral outcomes. Berelson et al.’s (1954) seminal

study on voting, for instance, notes that the media persuades individuals by prioritizing parti-

cular stories over others, or by airing a greater volume of stories related to some policy domains,

but not others. McCombs and Shaw’s (1972) Chapel Hill study, which spawned a vast literature

on public agenda-setting, examines the media’s role in focusing public attention on particular

issues, concluding that the media can eﬀectively “set” the public agenda by consistently and pro-

minently featuring issues in their news coverage. Cobb and Elder’s (1972) early work is the policy-

oriented equivalent, and these authors were followed by a growing body of literature focused

on the sources of the policy agenda, that is, the “general set of issues that are communicated in a

hierarchy of importance at a point in time” (Cobb and Elder [1972]1983: 14).

Though by no means the only source of the policy agenda, existing work suggests that media

can be an important one. Consider work by Flickinger (1983) and Mayer (1991), for instance, on

the role of media in the rise of consumer protection as a policy issue, or Pritchard ’ s (1986) work

on the impact of media coverage on the decision to prosecute murderers (also see Pritchard’s

(1992) review of the role of the news media in public policy). All of this work points toward an

impact of media on policy-makers that is very similar to what Cohen famously observed about the

public, namely, that the mass media “may not be successful much of the time in telling people

what to think, but it is stunningly successful in telling its readers what to think about” (Cohen

1963: 13). The policy equivalent—mass media may not deﬁne the nature or direction of policy

change, but can certainly steer attention towards certain policy domains over others—is of no small

consequence. Recent work by both Kingdon (1995) and Baumgartner and Jones (1993) has

made this especially clear. For each, issue attentiveness is a critical precursor of policy change.

The same was true in Anthony Downs’ (1972) work on the “issue attention cycle,” the cano-

nical model of public issue attentiveness. Downs suggested that (policy) issues move cyclically in

and out of the public consciousness. The life cycle of an issue moves incrementally from periods

of low to high salience, ultimately retreating to the background after the public has moved onto

to other issues. The notion itself is not earth-shattering, but Downs’ description of the process is

valuable. It also points to the role of media in aﬀecting change in issue salience.

Note too that while most of the preceding work focuses on the role of media in aﬀecting

some kind of aggregate policy agenda, the role of media in setting the policy agenda can also be

seen at the individual level, that is, impacting individual political and policy actors directly.

Politicians are aﬀected by media in the same way as ordinary citizens (Eilders 2001; Dearing and

Rogers 1996). They rely on media cues to prioritize information and to disseminate public opi-

nion (Walgrave and Van Aelst 2006: 100; Cook et al. 1983). Politicians can be even more sus-

ceptible to media content depending on whether they are subject to electoral punishment on a

set of issues (Kingdon 1984). However, legislators, like voters, cannot pay attention to all issues

at a given time; their attention is ﬁnite, and therefore they tend to focus on key issues that are

beneﬁcial to their constituents, country and indeed, their own career. Unsurprisingly, there are often

high levels of congruence among these issues and the issues that are of concern to the public

(Baumgartner et al. 1997: 350).

Of all this said, media quite clearly do not matter to all policy issues all the all the time, and there

is a growing body of work exploring the ways in which media inﬂuence varies systematically

across issues. Some work has compared the impact of media across issues directly. Soroka (2002),

for instance, argues for three diﬀerent issue types, where “sensational” issues—characterized by low

complexity and the possibility for dramatic events—are those for which media are most likely to

Mass media and policy-making

205

play a leading role. (Also see Walgrave et al. 2007.) Indeed, there are a number of studies suggesting

that the complexity of issues seriously constrains the potential for media eﬀects, on both the public

and policy-makers (e.g. Zucker 1978; Yagade and Dozier 1990), and a series of studies that focus

on the role of often sudden and unexpected “focusing events” in attracting media, and then policy,

attention to issues (e.g. Birkland 1998; Kingdon 1994). The impact of media on policy also appears

to be contingent on the source of the news: reliable and respected news outlets have more impact

than marginal and questionable news sources (Bartels 1996). Additionally, the possibility that

media have a marked impact on the political agenda increases when there is heightened coverage

of the same issue, at the same time, by diﬀerent media outlets (Eilders 2000, 2001).

Even with these caveats, the accumulated literature suggests an important (and often indepen-

dent) role for media in determining which issues are important, and when—for the public, and

for policy-makers as well. Setting the agenda is just one way in which media may matter to

policy, however. We explore a second below.

Issue framing

As with agenda-setting, the framing literature spans analyses of both policy-making and public

opinion. Druckman oﬀers a cogent deﬁnition:

a framing eﬀect is said to occur when, in the of describing an issue or event, a speaker’s

emphasis on a subset of potentially relevant considerations causes individuals to focus on

these considerations when constructing their opinions.

(2001: 1042; drawing on work by Rabin 1998, among others)

The extent to which this kind of framing is seen as separate from agenda-setting varies somewhat—

some authors choose to see framing mainly as shifts in attentiveness to sub-issues (see, e.g.

McCombs et al. 1997; McCombs 2004). Others argue that while agenda-setting looks at story

selection as determinants of public perceptions of issue importance, framing looks at the way

those issues are presented (see, e.g. Price and Tewksbury 1997: 184). We take the view here, in

line with a vast body of work on policy-making, that it is worth considering framing as distinct

from agenda-setting.

Framing theory is based on the belief that how an issue is characterized to an audience will

inﬂuence how it is understood (see, e.g. Scheufele and Tewksbury 2007). Put diﬀerently, issue

framing refers to the selective exposure of information to an audience with the intent of shaping

their understanding of an issue; it is the “selection of—and emphasis upon—particular attributes

for the news media agenda when talking about an object” (McCombs 2004: 87). Media can

apply frames to issues in order to organize a storyline around a series of events (e.g. Gamson and

Modigliani 1987), or to instruct or persuade the audience in how to evaluate the information

being given to them in a predictable way (e.g. McQuail 1994). Framing issues presuppose that

the information presented to the public will change the way that they view an issue. In other

words, framing an issue involves “select[ing] some aspects of a perceived reality … in such a

way as to promote a particular problem deﬁnition, causal interpretation, moral evaluation, and/

or treatment recommendation” (Entman 1993: 52). To frame a story is to often withhold some

information or prioritize some facts over others. In fact, most frames are, according to Entman

(1993), as deﬁned by what they omit or obscure as what they include. (And note that motives

for inclusion or exclusion of information may be conscious or unconscious, see Gamson 1989.)

Frames often inﬂuence the direction of policy by pulling values or emotion into the discus-

sion. Stone argues that problem identiﬁcation often focuses on framing a story in a way that

Soroka et al.

206

attributes cause and assigns blame (1989: 282). According to Iyengar (1991), thematic frames (those

that give information about general trends such as poverty or social welfare) tend to promote a

sense of social or institutional responsibility, while episodic frames (those that reference indivi-

duals or personal experiences or and stories) place responsibility on the individual. Frame types

prime social values diﬀerently, and simultaneously establish the salience of the subject and promote

a particular policy direction (Sniderman et al. 1991).

Framing is not an activity for media alone, of course. Policymakers are not simply aﬀected by

issue framing in media, but they actively engage in policy framing. Indeed, the policy literature

on framing is much more focused on the ways in which policy re-framing by politicians and/or

bureaucrats can shift attentiveness to or attitudes towards an issue. Consider Edelman’s (1997)

work on the role of language and symbols in politics, for instance, connected to a broader body

of constructionist accounts of the policy process (e.g. Lasswell 1949; Spector and Kitsuse 1977;

Rosenau 1993). Consider also Fischer’s (2003) discussion of politicians’ use of rhetorical devices

to frame an issue and sway public opinion. Using the example of hiring policy, Gamson and

Modigliani (1987) show that formal terms in thematic, policy-oriented frames (i.e. “aﬃrmative

action”) may lend the frame more credibility than vernacular or emotionally charged language

(i.e. “preferential treatment”), thereby impacting how audiences receive the frame and accept

policy direction from elites (see also Gamson 1992).

Media as a policy actor: challenges

Clearly, media matter to policy, throughout the policy process, and in many diﬀerent ways. Is

this a good thing? It is an inevitable thing, surely—it is nearly impossible to imagine modern

politics and policy-making without some kind of media involvement, after all. Even so, we should

consider some of the issues related to media’s role in the policy process. Here, we review the

considerable body of work that criticizes the nature and tone of media content on policy issues.

As we shall see, this work raises serious questions about the potential for and diﬃculties with the

role of mass media in policy-making.

Note ﬁrst that work in political communication routinely oﬀers largely critical assessments of the

ability of news content to contribute to informed public debate, particularly with respect to

television news. Biases in reporter coverage and the use of active conﬂict frames—key journalistic

tools in other areas of media coverage—are highlighted as important constraints. Studies fault

reporters for focusing on the trivial, for being too closely tied to oﬃcial sources, for not providing

their viewers with enough context to understand contentious policy options, for their bias and

for a lack of technical proﬁciency in the matters about which they write (Entman 2004;

Farnsworth and Lichter 2006, 2011; Herman and Chomsky 1988; Iyengar 1991; Iyengar and

Kinder 1987; Larson 2001; McChesney 1999; Patterson 1994).

Research also questions the possibility that journalistic norms serve to create misleading news

stories. Coverage of environmental issues stands as one example of this phenomenon. Even with

major issues that have been repeatedly featured in the media, such as the global warming debate,

scholars note that attempting to provide roughly equal treatment of both sides of a story can

“distort” the reality of widespread scientiﬁc agreement regarding climate change (Boykoﬀ 2005).

Rather than focus on the preponderance of scientiﬁc evidence that supports the global warming

hypothesis, scholars argue that US news reports have tended to give roughly equal weight to

skeptics with little peer-reviewed evidence (Boykoﬀ 2005; Mooney 2004).

Other journalistic norms may also reduce the likelihood of eﬀective/informative policy discussion

in the mass media. Reporters often prefer conﬂict frames to increase news consumer interest,

but the news reports that emerge often lack suﬃcient context (Iyengar 1991; Iyengar and Kinder

Mass media and policy-making

207

1987). Emphasizing the conﬂict frame of environmental debates decreases public awareness of

the scientiﬁc consensus regarding the existence of human-triggered climate change, for instance

(Corbett and Durfee 2008; Nisbet and Myers 2007). Content analyses of climate change news

in the New York Times, the Washington Post, the Los Angeles Times and the Wall Street Journal from

1998 through 2002 suggest that the journalistic attempts to be even-handed sowed far greater

public doubts about global warming than exist within the scientiﬁc community (Boykoﬀ and

Boykoﬀ 2007). Studies of television news have found that these attempts to provide balance in

stories made scientiﬁc ﬁndings on climate change appear to the public as far more tentative than

they actually were (Boykoﬀ 2007a).

Additional problems with media coverage of complicated policy domains have emerged in

other studies. First, scientiﬁc uncertainty and other technical matters tend to be papered over by

reporters who do not mention that correlation is not causation, or that preliminary ﬁndings are

tentative (Murray et al. 2001). Complexities, in other words, that too often are ignored in favor

of a more compelling and deﬁnitive, if less accurate, narrative. In cases where policy specialists oﬀer

an interpretation that is too nuanced or too technical (others might say “accurate”), reporters

are tempted to rely on environmental activists who are quicker with a pithy quote, even though

they may not possess the credentials of the less-quotable scientiﬁc experts (Lichter and Rothman

1999). Part of the problem may be the relatively limited policy expertise possessed by many

reporters. Work suggests, for instance, that a lack of expertise has been important in journalists’

susceptibility to marginal claims of potential health hazards, such as the relative dangers posed by

pesticides on apples or Bisphenol A (BPA) in water bottles when compared, for example, to

cigarette smoking or obesity (Murray et al. 2001; Lichter 2009).

Second, when elected oﬃcials weigh in on policy, political issues tend to become increasingly

prominent in the public discussion. That is, the interaction of politics and policy in the media

tends to reduce attention to substantive policy matters and refocus concern on estimations of

politically viable policy options (Miller et al. 1990; Wilkins and Patterson 1991).

Third, and perhaps most relevant, one problem in policy coverage by the media is that the

very long-term nature of some policy domains works against the traditional newsroom norms of

timeliness and novelty (McCright and Dunlap 2003; Trumbo 1995). While news coverage can

focus intensely on scientiﬁc issues when a hurricane makes landfall or when a severe drought

decimates crop yields, media attention can evaporate as quickly as it emerges (Mazur 2009; Mazur

and Lee 1993). For instance, content analysis of climate change reports in the New York Times

and Washington Post from 1980 to 1995 shows an attention cycle of media interest in global

warming, where coverage increases in the early stages of discussion but erodes over time

(McComas and Shanahan 1999). Early coverage was anchored by dire projections from scientists,

while a middle phase of coverage focused on disagreements among scientists to maintain interest,

and a later phase of reduced coverage concentrated on the economic costs and political debates

over potential remedies. The rapid turnover of issues in mainstream news works against gradual

long-term stories like climate change, particularly if the dire early predictions do not appear to

come to pass shortly after they are made (Stevens 1993).

An expository analysis: environmental coverage in Canada

The preceding section has drawn on work focused on environmental policy in particular for good

reason: there is a growing body of work dealing with the problems with media coverage of envir-

onmental issues. A recurring theme is the tendency for a strong connection between environ-

mental coverage and major events—or, more precisely, a lack of environmental coverage in the

absence of major weather or climate events. For instance, a comparison of global warming news

Soroka et al.

208

in top circulation newspapers in the US and the UK found a much greater volume of coverage

in the UK, though the number of news reports in both nations increased notably (and tem-

porarily) around key environmental events, such as new expert reports on greenhouse gas

emissions caused by air travel (a key issue of the G8 summit in June 2005), and the release of Al

Gore’s ﬁlm, An Inconvenient Truth, a year later (Boykoﬀ 2007a).

The biggest challenge in media coverage of environmental politics may be, then, that despite

the long-term nature of many environmental issues, environmental news is really not all that

diﬀerent from other types of news: as with coverage of a wide range of policy issues, events with

immediate impacts are both easier and more attractive to cover than continuous monitoring of a

known issue. And in the absence of such events, regardless of the actual state of the environ-

ment, environmental issues will disappear from the media agenda (and quite possibly the policy

agenda as well).

This event-driven tendency in media is relatively easily illustrated. We focus here on a body

of content-analytic data of all news stories from the nightly newscast for CTV in Canada (the

nightly newscast with the largest audience shares in each country) from 1999 to 2009. Data were

gathered using full-text indices in Nexis, and relying on the Nexis topic ﬁeld to identify stories

for which a “major theme” (based on the Nexis subject-coding scheme) was the environment.

In total, the database includes 1,789 news stories.

We begin by subject-coding stories based on an iterative automated process. We ﬁrst look at

frequencies for all words and phrases in the stories (analyzed together), using WordStat. We

then take the most common words and phrases—those relating to the environment—and build

a dictionary of commonly used terms. We add these terms, gradually, to the topic dictionary—the

product of a combination of data processing and common sense. A story is then coded as a given

topic if at least two keywords for that category were present in the text. The resulting codes thus

are indicators of the prevalence of speciﬁc language relating to various environmental issues.

Our focus here is on two topics in particular: pollution/climate change, and weather and

natural disasters.

Figure 16.1 shows trends in coverage for both topics over time. What is most

important to draw from the ﬁgure is (a) the tendency for weather and natural disasters to pro-

duce more news than pollution and climate change and (b) the lack of any upward trend in

coverage of environmental issues in spite of increasing environmental diﬃculties worldwide.

Media coverage of pollution and climate change does not, at least in terms of the volume of

coverage, appear to reﬂect trends in environmental indicators.

The more striking ﬁ nding from these data is shown in Table 16.1, however. The table pre-

sents results from a relatively simple Granger causality test—a statistical test of the temporal

relationship between the number of articles relating to disasters/weather and the number of

articles dealing with pollution/climate change.

Analyses rely on weekly data, and Granger tests proceed as follows. Total current coverage of

disasters/weather is regressed on the last week’s coverage of disasters/weather, as well as the last

week’s coverage of pollution/climate change. Results show, controlling for past coverage of

disasters/weather, whether pollution/climate change coverage systematically leads to disasters/

weather coverage. The same model is also estimated in the opposite direction: pollution/climate

change coverage is regressed on last week’s coverage of pollution/climate change, and last week’s

coverage of disasters/weather. Drawing on both models, we have a good sense of the extent to

which disasters/weather leads pollution/climate change, and vice versa.

Both models are estimated simultaneously using OLS vector autoregression; estimated coef-

ﬁcients are shown in Table 16.1. In the ﬁrst column, we see that current coverage of disasters/

weather is related to the previous week’s coverage of disasters/weather. (The coeﬃcient is .32,

and is statistically signiﬁcant.) The same is not true for the previous week’s coverage of

Mass media and policy-making

209

Figure 16.1 Coverage of natural disasters/weather and pollution/climate change by year

Source: Stuart Soroka, Stephen Farnsworth, Lori Young and Andrea Lawlor (2009) ‘Environment

and energy policy: Comparing reports from US and Canadian television news’. Paper presented

at the Annual Meeting of the American Political Science Association, Toronto.

Soroka et al.

210

pollution/climate change. That is, there is no relationship between current coverage of weather

or disasters, and the preceding week’s reports on pollution or climate change.

The second column includes the model for current coverage of pollution or climate change.

Here, we see that current coverage of pollution and climate change is related to last week’s coverage

of climate change (a coeﬃcient of .26), but also to last week’s coverage of disasters and events (a

coeﬃcient of .06). There is, then, evidence here of a unidirectional causal eﬀect: coverage of

pollution and climate change is systematically (and positively) led by coverage of weather and

disasters. (Put diﬀerently, coverage of disasters and events “Granger-causes” coverage of pollution

and climate change.)

In sum, these data suggest that substantive coverage of environmental themes such as air pollu-

tion, global warming and climate change increases after major weather-related disasters and events.

Indeed, one interpretation—drawing on the literature discussed above—is that coverage of pollu-

tion and climate change is dependent on weather-related disasters and events. As we might expect,

reporters apparently ﬁnd it hard to write about climate change without reference to high tem-

peratures and weather-related disasters. A sustained conversation on the environment is unlikely,

our ﬁndings suggest, to exclude a steady stream of ﬂoods, hurricanes, ice storms, power blackouts,

and other climate mayhem.

Conclusions

There is little doubt that mass media play an important role in policy-making. A growing body of

literature highlights this role, both in terms of issue attentiveness and policy framing. We have argued

here, however, that it is important to note, ﬁrst, that media matter not just at the beginning but

throughout the policy process, and second, that media’s involvement in the policy process poses

some real diﬃculties. In short, the complexities of policy-making are likely not to be well served

by well-known tendencies in media coverage. That media matter to policy-making seems beyond

doubt. Whether their contribution tends to be positive on balance is, however, up for discussion.

Notes

1 This section draws on analyses in Soroka et al. 2009.

2 The ﬁrst is captured using the following keywords: acid rain, air, carbon, climate change, emission*, global

warming, greenhouse gas*, pollution, pollute*, toxic, toxin*, ozone; the second is based on: ﬂood*, forest

ﬁre*, wildﬁre*, wild ﬁre*, hurricane*, storm*, drought*, rain*, wind*, ice, weather, tornado*.

Table 16.1 Coverage of natural disasters/weather and pollution/climate change

DV:

disasters/

weather

DV: pollution/climate Change

Disasters/weather

t-1

.315* .061

(.109) (.035)

Pollution/climate Change

t-1

.019 .259*

(.051) (.043)

Constant 1.114* .776*

(.109) (.090)

R-sq .101 .079

Notes: N=520. * p < .05; a p < .10. Estimates rely on weekly data. Cells contain coefﬁcients from an OLS vector

regression, with standard errors in parentheses.

Source: Author

Mass media and policy-making

211

Bibliography

Bartels, L.M. 1996, September. ‘Politicians and the press: Who leads, who follows? in APSA. San

Francisco, CA.

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago:

University of Chicago Press.

——, Bryan D. Jones and Beth L. Leech. 1997. ‘Media attention and Congressional agendas’, in Shanto

Iyengar and Richard Reeves (eds) Do the Media Govern? Politicians, Voters, and Reporters in America

Thousand Oaks, CA: Sage.

Berelson, Bernard R., Paul F. Lazarsfeld and William N. McPhee. 1954. Voting: A Study of Opinion Formation

in a Presidential Campaign. Chicago: University of Chicago Press.

Berger, B. 2001. ‘Private issues and public policy: Locating the corporate agenda in agenda-setting theory’,

Journal of Public Relations Research 13(2): 91–126.

Birkland, Thomas A. 1998. ‘Focusing events, mobilization, and agenda setting’, Journal of Public Policy

18(1): 53–74.

—— 2007. ‘Agenda setting in public policy’ , in Frank Fischer, Gerald J. Miller and Mara S. Sidney (eds)

Handbook of Public Policy Analysis. London: Taylor & Francis.

Boykoff, Maxwell. 2005. ‘The disconnect of news reporting from scientiﬁc evidence’, Nieman Reports,

Harvard University, winter: 86.

—— 2007a. ‘From convergence to contention: United States mass media representations of anthropogenic

climate change science’, Transactions of the Institute of British Geographers 32: 477–89.

—— 2007b. ‘Flogging a dead norm? Newspaper coverage of anthropogenic climate change in the United

States and the United Kingdom from 2003–2006’, Area 39: 470–81.

——and Jules M. Boykoﬀ. 2007. ‘Climate change and journalistic norms: A case study of US mass media

coverage’, Geoforum 38: 1190–204.

Cobb, R.W. and C.D. Elder. [1972]1983. Participation in American Politics: The Dynamics of Agenda-Building.

Boston: Allyn & Bacon.

Cohen, Bernard C. 1963. The Press and Foreign Policy.

Princeton, NJ: Princeton

University Press.

Cook, Fay Lomax, Tom R. Tyler, Edward G. Goetz, Margaret T. Gordon, David Protess, Donna R. Leﬀ,

and Harvey L. Molotch. 1983. ‘Media and agenda setting: Eﬀects on the public, interest group leaders,

policy makers, and policy’, The Public Opinion Quarterly 47(1): 16– 35.

Corbett, Julia B. and Jessica Durfee. 2008. ‘Testing public (un)certainty of science: Media representations

of global warming’, Science Communication 26: 129–51.

Dearing, James W. and Everett M. Rogers. 1996. Agenda-Setting. Thousand Oaks, CA: Sage.

Downs, Anthony. 1972. ‘Up and down with ecology: The “issue attention cycle”’, The Public Interest 28: 35–50.

Druckman, James N. 2001. ‘On the limits of framing eﬀects: Who can frame?’ The Journal of Politics 63:

1041–1066.

——and Michael Parkin. 2005. ‘The impact of media bias: How editorial slant aﬀects voters’, The Journal of

Politics 67(4): 1030–49.

Edelman, M. 1964. The Symbolic Uses of Politics. Chicago: University of Illinois Press.

—— 1977. Political Language. New York: Academic Press.

Eilders, C. 2000. ‘Media as political actors? Issue focusing and selective emphasis in the German quality

press’, German Politics 9(3): 181–206.

—— 2001. Conﬂict and Consonance in Media Opinion: Political Positions of Five German Quality Papers. Berlin:

WZB Discussion Papers.

Entman, Robert. 1993. ‘Framing: Toward clariﬁcation of a fractured paradigm’, Journal of Communication

43(4): 51–8.

—— 2004. Projections

of Power: Framing

News, Public Opinion and Foreign Policy. Chicago: University of

Chicago Press.

Farnsworth, Stephen J. and S. Robert Lichter. 2006. The Mediated Presidency: Television News and Presidential

Governance. Lanham, MD: Rowman & Littleﬁeld.

—— and S. Robert Lichter. 2011. The Nightly News Nightmare: Media Coverage of US Presidential Elections,

1988–2008. Lanham, MD: Rowman and Littleﬁeld. Third Edition.

Fischer, Frank. 2003. Reframing Public Policy: Discursive Politics and Deliberative Practices. Oxford: Oxford

University Press.

Flickinger, Richard. 1983. ‘The comparative politics of agenda-setting: The emergence of consumer

protection as a public policy issue in Britain and the United States’, Policy Studies Review 2(3): 429–44.

Soroka et al.

212

Gamson, W.A. 1989. ‘News as framing: Comments on Graber’, American Behavioral Scientist 33: 157–66.

—— 1992. Talking Politics, New York, NY: Cambridge University Press.

——and A. Modigliani. 1987. ‘The changing culture of aﬃrmative action’, in R.G. Braungart and M.M.

Braungart (eds) Research in Political Sociology. Greenwich, CT: JAI Press.

Gandy, O.H., Jr. 1982. Beyond Agenda Setting: Information Subsidies and Public Policy. Norwood, NJ: Ablex.

Herman, Edward S. and Noam Chomsky. 1988. Manufacturing Consent: The Political Economy of the Mass

Media. New York: Panetheon.

Hilgartner, Stephen and Charles L. Bosk. 1988. ‘The rise and fall of social problems: A public arenas

model’, American Journal of Sociology 94(1): 53 – 78.

Iyengar, Shanto. 1991. Is Anyone Responsible? How Television Frames Political Issues. Chicago: University of

Chicago Press.

—— 1993. ‘Agenda-setting and beyond: Television news and the strength of political issues’, in William

Riker (ed.) Agenda Formation. Ann Arbor, MI: University of Michigan Press.

——and Donald R. Kinder. 1987. News That Matters. Chicago: University of Chicago Press.

Jensen, Carsten. 2009. ‘Policy punctuations in mature welfare states’, Journal of Public Policy 29(3): 287–303.

Kingdon, John. 1984. Agenda-Setting, Alternatives and Public Policies. New York: Harper Collins.

—— 1995. Agendas, Alternatives, and Public Policies. New York: HarperCollins College Publishers.

Lang, Kurt and Gladys Engel Lang. 1966. ‘The mass media and voting’, in Bernard Berelson and Morris

Janowitz (eds) Reader in Public Opinion and Communication. New York: Free Press.

Larson, Stephanie. 2001. ‘Poll coverage of the 2000 presidential campaign on network news’, paper

delivered at the 2001 Annual Meeting of the American Political Science Association, San Francisco.

Lasswell, Harold D. 1949. ‘Style in the language of politics’, in Harold Lasswell and N. Leites (eds) The

Language of Politics: Studies in Quantitative Semantics. New York: George Stewart, 20–39.

Lichter, S. Robert. 2009. ‘Are chemicals killing us? Experts speak out’,

STATS (21 May report). Available

at: http://stats.org/stories/2009/are_chemicals_killing_us.html (accessed

2 August, 2009).

Lichter, S. Robert, and Stanley Rothman. 1999. Environmental Cancer: A Political Disease? New Haven,

CT: Yale University Press.

McChesney, Robert. 1999. Rich Media, Poor Democracy: Communications Politics in Dubious Times. New

York: Times Press.

McComas, Katherine and James Shanahan. 1999. ‘Telling stories about global climate change: Measuring

the impact of narratives on issue cycles’, Communication Research 26: 30–57.

McCombs, M. 2004. Setting the Agenda: The Mass Media and Public Opinion. Malden, MA: Blackwell

Publishing Inc.

——and Donald L. Shaw. 1972. ‘The agenda-setting function of mass media’, Public Opinion Quarterly

36(2): 176–87.

——and J.H. Zhu. 1995. ‘Capacity, diversity, and volatility of the public agenda: Trends from 1954 to

1994’, Public Opinion Quarterly 59: 495–525.

——, D.L. Shaw, and D. Weaver. 1997. Communication and Democracy: Exploring the Intellectual Frontiers in

Agenda-Setting Theory. Mahwah: NJ: Lawrence Erlbaum.

McCright, Aaron M. and Riley E. Dunlap. 2003. ‘Defeating Kyoto: The conservative movement’s impact

on US climate change policy’, Social Problems 50: 348–73.

McQuail, D. 1994. Mass Communication Theory: An Introduction, 3rd edn. Thousand Oaks, CA: Sage.

Mayer, Robert N. 1991. ‘Gone yesterday, here today: Consumer issues in the agenda-setting process’,

Journal of Social Issues 47(1): 21–39.

Mazur, Allan. 2009. ‘American generation of environmental warnings: Avian inﬂuenza and global warming’,

Human Ecology Review 16: 17–26.

——and Linling Lee. 2003. ‘Sounding the global alarm: Environmental issues in the US national news’,

Social Studies of Science 23: 681–720.

Miller, S. M., A. Evers and H. Winterberger. 1990. ‘The evolving welfare state mixes’,inShifts in

the

Welfare Mix: Their

Impact on Work, Social Services and Welfare Policies. Frankfurt: Campus Verlag,

371–88.

Mooney, Chris. 2004. ‘Blinded by science: How ‘balanced’ coverage lets the scientiﬁc fringe hijack reality’,

Columbia Journalism Review, November/December.

Murray, David, Joel Schwartz and S. Robert Lichter. 2001. It Ain’t Necessarily So: How Media Make and

Unmake the Scientiﬁc Picture of Reality. Lanham, MD: Rowman & Littleﬁeld.

Nelson, Barbara. 1984. Making an Issue of Child Abuse: Political Agenda-Setting for Social Problems. Chicago:

University of Chicago Press.

Mass media and policy-making

213

Nisbet, Matthew C. and Teresa Myers. 2007. ‘Twenty years of public opinion about global warming’,

Public Opinion Quarterly 71: 444–70.

Patterson, Thomas E. 1994. Out of Order. New York: Vintage.

Peters, B. Guy and Brian W. Hogwood. 1985. The Pathology of Public Policy. New York: Oxford University

Press.

Price, V. and D. Tewksbury. 1997. ‘News values and public opinion: A theoretical account of media

priming and framing’, in G.A. Barett and F.J. Boster (eds) Progress in Communication Sciences: Advances in

Persuasion. Greenwich, CT: Ablex.

Pritchard, David. 1986. ‘Homicide and bargained justice: The agenda-setting eﬀect of crime news on

prosecutors’, Public Opinion Quarterly 50(2): 143–59.

—— 1992. ‘The news media and public policy agendas’, in. D. Kennamer (ed.) Public Opinion, the Press

and Public Policy. Westport, CT: Praeger.

Rabin, Matthew. 1998. ‘Psychology and Economics’, Journal of Economic Literature 36: 11–46.

Rosenau, Pauline V. 1993. ‘Anticipating a post-modern policy current?’ Policy Currents 3: 1–4.

Schattschneider, E.E. 1960. The Semisovereign People. New York: Rineheart and Wilson.

Scheufele, Dietram A. and David Tewksbury. 2007. ‘Framing, agenda setting, and priming: The evolution

of three media eﬀects models’, Journal of Communication 57(1): 9–20.

Sniderman, P.M., R.A. Brody and P.E. Tetlock. 1991. Reasoning and Choice: Explorations in Political Psychology .

Cambridge: Cambridge University Press.

Soroka, Stuart. 2002. Agenda-Setting Dynamics in Canada. Vancouver: University of British Columbia Press.

——, Stephen Farnsworth, Lori Young and Andrea Lawlor. 2009. ‘Environment and energy policy:

Comparing reports from US and Canadian television news’, paper presented at the Annual Meeting of

the American Political Science Association, Toronto.

Spector, Malcolm and John I. Kitsuse. 1977. Constructing Social Problems. Menlo Park, CA: Cummings.

Stevens, W.K. 1993. ‘Scientists confront renewed backlash on global warming’, New York Times,1

4Se

ptember.

Stone, Deborah. 1989. ‘Causal stories and the formation of policy agendas’, Political Science Quarterly 104(2):

281–300.

Trumbo, Craig. 1995. ‘Longitudinal modeling of public issues: An application of the agenda-setting process

to the issue of global warming’, Journalism and Mass Communication Monographs 152(August): 1–57.

—— 1996. ‘Constructing climate change: Claims and frames in US news coverage of an environmental

issue’, Public Understanding of Science 5: 269–83.

Tuchman, G. 1978. Making News: A Study in the Construction of Reality. New York: Free Press.

Walgrave, Stefaan and Peter Van Aelst. 2006. ‘The contingency of the mass media’s political agenda setting

power: toward a preliminary theory’, Journal of Communication 56(1): 88–109.

——, Stuart Soroka and Michiel Nuytemans. 2007. ‘The mass media’s political agenda-setting power: A

longitudinal analysis of media, parliament, and government in Belgium (1993 to 2000) ’, Comparative

Political Studies 41(6): 814–36.

Wilkins, Lee and Philip Patterson. 1991. ‘Science as symbol: The media chills the greenhouse eﬀect’,in

Lee Wilkins and Philip Patterson (eds) Risky Business: Communication Issues of Science, Risk and Public

Policy. Westport, CT: Greenwood, 159–76.

Yagade, A. and Dozier D.M. 1990. ‘The media agenda-setting eﬀect of concrete versus abstract issues’,

Journalism Quarterly 67: 3–10.

Zhu, J.H. 1992. ‘Issue competition and attention distraction: A zero-sum theory of agenda setting’, Journalism

Quarterly 69: 825–36.

Zucker, H.G. 1978. ‘The variable nature of news media inﬂuence’,

in B.D. Ruben (ed.) Communication

Yearbook, vol 2. New

Brunswick, NJ: Transaction Books.

Soroka et al.

214

Part V

Understanding the

formulation process

Policy design and transfer

Anne Schneider

Policy design, whether conceptualized as a verb referring to the process of formulating public

policy, or as a noun describing the content of public policy remains largely uncharted territory.

This chapter reviews what we know about policy designs and develops ideas about how the

study of policy design can contribute to an understanding of policy transfer and diﬀusion. Policy

scholars have long been interested in understanding how policies are copied, stolen, or diﬀuse

across jurisdictions, but much of the attention has been on the politics and geographic proxi-

mity of the “innovating” states—that is, the ones that were early adopters (Walker 1969). The

focus in this chapter is on what types of policy designs are most likely to transfer and which

ones may be subject to truncated transfer or to not move at all (Schneider and Ingram 1988).

Theories of policy invention and expert decision-making suggest that individuals search through

large amounts of relevant information stored in memory, reason by analogies, make comparisons,

and either copy or simulate patterns of information to produce desired consequences. These

consequences may focus on “good” public policy that resolves important collective problems,

or on gaining political capital, or a broad array of more narrow-gauged goals. On the other hand,

social psychologists have documented that even experts are subject to a wide range of cognitive

biases, and these biases ﬁnd their way into public policy designs (Kahneman et al. 1982). Rather

than simply a benign process of eﬃcient decision-making, cognitive biases may operate at the

expense of policy that would serve democratic ends. In the remaining parts of this chapter, I

will ﬁrst explore policy designing as it commonly occurs with an emphasis on the role of decision

heuristics and cognitive biases, and then develop a framework for analyzing policy design—that

is, policy content. From the analysis of the elements of design found in almost all public poli-

cies, it is possible to identify a number of commonly copied designs and their implications for

democracy.

Background

Two very diﬀerent approaches to the study of policy design have emerged in the literature

(Schneider and Ingram 1988, 1993, 1997, 2005; Howlett and Lejano 2011; Linder and Peters 1985;

Bobrow and Dryzek (1987). What these approaches have in common is a desire to improve the

consequences of public policy—the outputs or outcomes of policy. They diﬀer considerably,

217

however, in the focus of attention, deﬁne the term diﬀerently, and also come at the problem of

producing what might be considered “good” public policy from very diﬀerent perspectives.

The rationalist approach deﬁnes policy design as a rational and systematic eﬀort to achieve

eﬀective and eﬃcient policy. Howlett and Lejano (2013), for example, deﬁne policy design as

the eﬀort to more or less systematically develop eﬃcient and eﬀective policies through the

application of knowledge about policy means gained from experience, and reason, the devel-

opment and adoption of courses of action that are likely to succeed in attaining their desired

goals or aim.

(forthcoming: 3)

Design, in this approach, is mainly used as a verb and is a particular type of process emphasizing

rational, scientiﬁc, and logical thinking toward identifying the most eﬀective and eﬃcient means

to a given end.

The second tradition, best developed by Schneider and Ingram (1997), but tracing its roots

also to Dahl and Lindblom (1953), Forester (1989), Schőn (1994); Stone (1988) begins with

policy design as policy content—broadly understood—and not just those aspects of the policy

text and practices that are rationalistic, grounded in science or logic, or even systematically put

into place. Policy design, according to Schneider and Ingram (1997: 2)

refers to the content or substance of public policy—the blueprints, architecture, discourses,

and aesthetics of policy in both its instrumental and symbolic form.

Policy designing is the process through which this policy content is produced, and (again) not

just the systematic and rational in terms of clearly stated public policy goals, but also the cognitive

biases embedded in the policy as well as attributes damaging to the linkage between public

policy and democratic principles, such as the intentionally manipulative, deceptive, illogical,

mean-spirited, and unscientiﬁc factors that inﬂuenced the choice of design elements.

This chapter begins with the presumption that policy design is more a process of copying and

adaptation than of invention, at least at the policy-maker level. Model legislation and guide-

lines commonly are drafted by policy entrepreneurs who attempt to spread a particular policy

idea throughout the nation. And, policy-makers pay close attention to the policy designs being

used elsewhere, as has been well-documented in the studies of policy diﬀusion (Karch 2007).

Second, the process of designing is not necessarily intended to produce the policy goals that are

touted as the probable result, but are ﬁlled with compromises, attempts to build suﬃcient

support for passage, rationales that may have little to do with the actual motivations, and out-

right deception. Policy-makers have both political and policy goals, as well as personal goals

that inﬂuence their choice of design elements. Designs contain embedded social construc-

tions, images, and symbols that send messages to target populations and the broader public

about who is deserving of what, and why (Schneider and Ingram 1993, 1997, 2005). Designs

embed, as well, the cognitive biases and decision heuristics of those who crafted the content of

the policy. For these reasons, it is essential that analysts develop a framework for systematically

analyzing and comparing policy designs (content). It is no longer suﬃcient just to say that it is

“education policy,” or “three strikes and you’re out,” or an “anti-immigration” bill. More

information is needed about the actual array of elements in the design. The study of policy transfer,

innovation, and diﬀusion typically has not paid much attention to actually what was transferred

or copied across jurisdictional lines, other than the very broad popularized title for the policy—

such as an immigration law “as

tough at Arizona’s.” Thus,

a more careful and comprehensive

Schneider

218

approach to the study of policy design can contribute to a number of areas of policy and

political research.

To limit the study of policy design entirely to those aspects that are rational and systematic is

not necessary and makes a serious mistake in understanding the reasons certain designs are

chosen and the consequences of those designs for democracy. Policy designs can fruitfully be

understood with an analogy to the design of a city. A city probably has some elements that

were designed intentionally to be eﬃcient and eﬀective—such as the system of streets and roads,

the placement of light rail and traﬃc lights, the water and power delivery systems. But some

aspects—many in fact—of a design were not put in place by any overall intentional or sys-

tematic or scientiﬁc process. The used car parked on the sidewalk is an aspect of the design of a

city at a particular moment, as is the green grass or desert landscaping at a new housing devel-

opment. At any given moment, every element in the city is a part of its design. And, the design

of a city is constantly changing as new houses are built, new roads put in place, changes in

traﬃc patterns, landscaping, and so on. Yet, it is possible to stop the scene at any moment and

capture the exact design of that city, at that moment.

This chapter builds on an understanding of how policy designs are crafted and chosen. First,

ample research points out that designs often are copied, borrowed, or stolen from similar poli-

cies elsewhere (Schneider and Ingram 1988). With the exception of some very early studies of

policy diﬀusion (Gray 1973), most of the research has largely ignored the characteristics of the

policies themselves, and focused on which kinds of cities, states, or nations are most “innova-

tive”—deﬁned as “early adopters.” Thus, a huge uncharted area of research is why some kinds

of policy designs are copied and spread whereas others are truncated and others never even begin

to move across jurisdictional lines (see Karch 2007). Second, the reasoning process through

which people arrive at decisions relies heavily on decision heuristics—deviations from rational

decision-making—that introduce biases into policy design. And equally important, knowledge

about how people reason has become a fundamental tool of political leaders who manipulate

media and citizens by using policy arguments and designs as a technique to gain public support

for policy that actually will not produce the consequences touted for it and may work to the

actual disadvantage of those that the rhetoric implies will gain from it. Third, this chapter draws

on Schneider and Ingram’s theory of degenerative policy-making and policy design to identify

characteristics of policy that often are copied and appropriated from other jurisdictions

(Schneider and Ingram 1993, 1997).

Policy design as a verb: cognitive biases and decision heuristics

This discussion begins with the insight from sociology and social psychology that people have a

tendency to identify with an “in” group and embellish its virtues as well as ﬁnd “the other”

from which they not only distance themselves, but socially construct as dangerous, undeserving,

lazy, stupid, or other undesirable epitaphs. This apparently fundamental trait of modern human

beings plays a central role in the ways that decision heuristics and cognitive biases aﬀect politics

and public policy.

The “in group” preference along with the theory of decision heuristics and cognitive biases

form the micro-level theory of this approach to policy design. Kahneman et al. (1982) and

others have identiﬁed a number of common heuristics (rules of thumb) that people—even

experts—rely on to make decisions. One of the most common in the process of policy design is

availability. Availability refers to the ease by which a particular instance or occurrence or causal

event is brought to mind. Those that are most easily brought to mind include those that are

more recent, more dynamic or colorful, more consistent with biases and prejudices. Why, for

Policy design and transfer

219

example, have the American states altered their policies since the early 1970s to produce the highest

incarceration rate in the world? One could contend that it is simply to gain political capital. But

that brings up the next question—why does it work to produce political capital? And a pro-

mising explanation is that most people ﬁnd it easy to imagine that if you threaten enough, and

punish severely enough, and catch people when they commit a crime; they will stop committing

crimes. Why has “tough on immigration” legislation spread like wildﬁre across the United States

in the past several years, but the “Utah Compact” that advocates a much more benign approach

even though thoroughly grounded in the principles of conservatives has gained almost no traction?

Saying that “tough on immigration” produces political capital, again, is not enough: why does it

“work” to produce such presumably large political payoﬀs? Decision theory suggests it is easy

for people to buy into the notion that people who are in this country illegally are deceptive,

criminal, lazy, dangerous and other negative social constructions. Decision theory suggests that

this image is considerably easier to imagine than one in which “illegal aliens” as they are com-

monly called, are good, honest, hard-working people who value their families. Why did policy-

makers leading up to the 2012 elections tie every conceivable policy proposal to “job creation”?

Decision theory suggests that it is easier for people to imagine that if businesses both large and

small had more money (through tax cuts), they would create more jobs than to believe that if

the government spent more money, it would create more jobs. The highly negative social con-

struction of “government,” compared with the far more positive construction of “business” in the

ﬁrst decade of the twenty-ﬁrst century is a prime reason. As Stimson (1991) has shown, “national

“moods” exist and exert a powerful inﬂuence over what it is that people ﬁnd “easy to believe.”

Social psychologists have also documented an “availability cascade” in which an assertion is

repeated often enough that the simple repetition adds to the perception of its accuracy.

The combination of policy-makers themselves relying on cognitive biases when they propose

policy and the ease of convincing the public of their contentions has produced an alarming lack

of rationality in the designing process. As David Brooks (2011) said, the American public is

suﬀering from intellectual laziness of massive proportions. He speaks speciﬁcally of the problems

introduced by contribution bias. Research has shown consistently that people sort through all of

the purported facts and evidence that comes to their attention, but they primarily focus only on

the facts, studies, and anecdotal incidents that conﬁrm what they already know and they easily

dismiss even the most prestigious studies if those contradict their current knowledge. Challenges

to the methodology of global warming studies, for example, come almost exclusively from those

who already believe that either global warming is not occurring or that humans have no hand

in it. In spite of massive scientiﬁc consensus, leading candidates for president such as Rick Perry

of Texas in 2012 repeatedly paid attention only to the “facts” that already supported his beliefs.

Others play into this same tendency toward intellectual laziness when they use speciﬁc incidents,

such as Hurricane Katrina, as “proof” that global warming is occurring.

Closely related to “availability,” illusory

correlations are another

example of cognitive bias

that inﬂuences policies designs. Illusory correlations are those that are easy to imagine but have

no scientiﬁc basis. For example, it is easy to believe that if a business has more money, it will

expand and create more jobs. If America sends its military forces into a small, weak country like

Vietnam, Iraq, or Afghanistan, the war will be won quickly and easily. And if not, it is easy to

justify the mistaken correlation on the grounds, for example, that enough troops were not sent.

Anchoring is another heuristic important in understanding policy designing. The idea here is

that people start from a particular point and then make small adjustments in various directions but

commonly do not make adequate adjustments. Change the anchor point, and a diﬀerent policy

design will emerge. Another heuristic is overconﬁdence in the prediction of eﬀects—even

experts become over time more conﬁdent that a highly unlikely event will actually occur. For

Schneider

220

example, scientists involved in nuclear energy research consistently overestimate the likelihood

of nuclear accidents (Kahneman et al. 1982).

Decision-makers have become considerably more savvy about the persuasive techniques that

may be used to gain public support (or opposition) and the arguments surrounding policy

choices have become particularly disingenuous. The ﬁndings of the social psychologists are used

not to correct for cognitive biases, but to rely on them as ways to shift public opinion and

opinion leaders toward a particular point of view. The social construction of “government” has

become so negative that everyone wants to run as an outsider. The social construction of the word

“socialist” is viewed as so antithetical to American values that simply labeling a policy proposal

as “socialist” is commonly used as a way to kill the idea even though the United States has

dozens, even hundreds, of programs that by almost any deﬁnition are “ socialist” ranging from

public schools to the ownership of the Green Bay Packers!

An example of how public oﬃcials use the insights from cognitive and decision research to

reinforce certain biases is found in Senator John McCain’s response when he was asked what he

thought about the cause of Arizona’s most destructive forest ﬁre in history—the Wallow ﬁre of

2011. He had just been briefed about the Wallow ﬁre, and it had been compared to the second

most destructive ﬁre in Arizona history, the Rodeo-Chedeski ﬁre, both of which occurred in

the northeast part of Arizona. McCain was asked:

How do you explain two massive record wildﬁres, all being said, human caused, but how

do you explain such huge ﬁres and how do you ﬁx this from happening?

(Arizona Republic, B1)

His response:

Well ﬁrst of all we are concerned about particular areas down on the border where there is

substantial evidence that some of these ﬁres are caused by people who have crossed our

border illegally. They have set ﬁres because they want to signal others, they have set ﬁres to

keep warm, and they have set ﬁres in order to divert law enforcement agents and agencies

from them. So the answer to that part of the problem is, get a secure border.

Liberals quickly jumped all over this comment as there was no evidence that immigrants had

started the ﬁre and both of these ﬁres were more than 300 miles away from the Mexican border.

McCain retaliated, pointing out that he did not say the Wallow ﬁre was caused by an illegal

immigrant—and it was factually correct that he had not made that statement. (Two white male

campers had been arrested for letting their campﬁre get out of control and starting this ﬁre that

destroyed more than 500,000 acres). His oﬃce retaliated with this statement:

The facts are clear. For years, federal, state and local oﬃcials

have stated that smugglers and

illegal immigrants have

caused ﬁres on our southern border.

This statement went on to say that McCain’soﬃce had asked for a study and that it would be

out in about a month. And on July 8, 2011, Jim Pena, a US Forest Service associate deputy

chief, testi ﬁed before the House Subcommittee on National Parks, Forests and Public Lands

that between 2002 and 2011, 457 ﬁres were determined to be human-caused in the southwest

border area of the Coronado National Forest. Pena added, “Forest Service investigators have

been able to identify the individuals responsible in 31 of those ﬁres. Of those 31 ﬁres, it was

determined that undocumented aliens were responsible for starting ﬁve.”

Policy design and transfer

221

The Arizona Republic’s “fact check” upheld McCain’s position that he had not speciﬁcally said

that illegal immigrants had caused the ﬁre. Yet, there was a clear association being made between

illegals and ﬁre—an illusory correlation that played directly from the “availability heuristic” to

make it “easier to believe” that illegal immigrants are dangerous and the border must be secured

to protect Arizonans from wildﬁres and other dangers that illegal immigrants bring into the state.

Furthermore, although it is true that the statement from McCain’soﬃce only said that “facts are

clear” and that “illegal immigrants have caused ﬁres on the southern border,” the statement

deceives the reader by not going on to say that ﬁve of 431 ﬁres were known to have been

started by illegal immigrants.

Framing eﬀects are well known in the public policy literature and policy-makers have taken

advantage of these to draw together a number of diﬀerent cognitive biases as a way to make a

persuasive argument. Policy analysts (and media “fact checkers”) may contribute signiﬁcantly to

improved policy design by paying careful attention not only to “fact checking” but also to the

types of cognitive biases built into the rationales and the ways that political leaders are using the

“intellectual laziness” of the public to gain support for legislation. It is not enough to just look

for the rational, instrumental logic of policy but to unmask the persuasive devices being used to

thwart rational, logical thinking and reliance on well-established facts.

Policy design as a noun (policy content)

Public policy has a common-sense understanding of being everything that government, or gov-

ernment agencies or street-level case workers do. This understanding means that public policy

includes documents such as constitutions, statutes, and court cases as well as agency guidelines,

state and local laws, and actual practices such as the ways that street-level case workers apply the

policy to a particular person or organization. Each of these contains a policy design, but at dif-

ferent levels of analysis. At whatever level the focus might be, an adequate empirical description

of policy design is essential. Development of particular types of designs—with elements that tend

to cluster together—is essential for linking politics to policy design and policy design to policy

consequences. Particular types and styles of political dynamics do not produce policy outcomes

directly, as is sometimes implied. Policy design is not a black box that has no bearing on con-

sequences. Rather, political dynamics produce policy designs and these, in turn, “feed forward”

to produce not only policy results but also to inﬂuence the subsequent “politics” of the society.

In fact, it will be argued below that certain kinds of political dynamics are associated with cer-

tain types of policy designs and that these designs, in turn, have consequences for achievement

(or lack thereof) of the purported policy goals, and also systematic consequences for social identity,

political participation, justice, political gain or loss, and democracy in the broad sense of the word.

Policy designs embed aspects of the politics that created them, and the designs tend to reproduce

the same kinds of politics from which they emerged (Schneider and Ingram 1993, 2005).

This approach to policy design places its study within the empirical and normative tradition

of Lasswell’s “who gets what, why, and how” (Lasswell 1936). Thus, the study of policy design—

as policy content—is a broad, coherent, and essential part of political science. This approach has

led to a new very important area of study called “policy centric” orientation; and it has led to

new ways of thinking about the empirical and normative aspects of the policy-making process

itself (see Mettler and Soss 2004).

From a normative perspective, public policy in a democracy is charged with the task of creating

democratic institutions; encouraging and promoting an active, engaged, and informed citizenry;

promoting justice as fairness for all people; and solving collective problems in ways that are

eﬃcient and eﬀective.

Schneider

222

From an empirical perspective, all public policies contain a set of central elements that can be

described, studied and compared. The argument made below is that these tend to cluster in

identiﬁable patterns that can be identiﬁed and classi ﬁed as certain types of policy designs. The

question is then, what kinds of policies, under what kinds of conditions, are most likely to be

copied, stolen, or borrowed by what kinds of political entities?

The fundamental elements of designs include goals or problems to be solved, target popula-

tions, beneﬁts or burdens to distribute, and linkage mechanisms of tools, rationales, underlying

assumptions, an implementation structure, and social constructions (of targets, goals, and other

elements) (Schneider and Ingram 1997). These elements are found in all levels of policy—from

constitutions, to statutes, to agency guidelines, to state and local laws, and even to the discretion

that street-level agents of government use when they actually put a policy to work.

Goals of a policy at times may seem obvious, since written policy often contains a statement

of the problem to be solved or goals to be achieved. Goals or problems to be solved refer (on

the one hand) to the intentional aspects of designs and indicate what the purported consequences

are desired to be. On the other hand, actual goals may be masked and the policy itself strate-

gically attached to some widely acknowledged public problem that the policy has no realistic

chance of impacting with its true goals being largely hidden from sight (Stone 1988; Kingdon

1984). For example, in the run-up to the 2012 election, “job creation” was the purported number

one priority of almost all candidates from every party; yet policy proposals directly conﬂicting

with one another attached themselves to this goal with utter disregard of the logic in the con-

nection. Perhaps the most memorable example of policies conveniently attaching themselves to

a problem came in the summer and fall before the 2000 presidential election when candidate

George Bush was arguing that tax cuts were needed because the government had a substantial

surplus. As the economy deteriorated in late summer and early fall, he switched his argument saying

that tax cuts were needed to stimulate the economy and ward oﬀ a recession.

Arizona’s Senate Bill 1070, widely characterized as the strictest law again undocumented

people in the United States, carries the beneﬁcent title, “Support Our Law Enforcement and

Safe Neighborhood Acts.” The stated intent of the law is to make “attrition through enforce-

ment the public policy of all of Arizona.” The purpose, in other words, is to rid the state of

undocumented immigrants. The governor’s rationale for signing the bill, however, focused on

crime, violence, Mexican drug cartels, and the failure of the federal government in keeping

undocumented people out of the country. Opponents argued that Arizona’s economy would

be severely damaged by the legislation both because of a likely boycott and because ridding the

state of workers and consumers is not the way to “grow the economy.” The political consequences

of the Hispanic vote were largely disregarded also even though almost one-third of the Arizona

adult population are of Hispanic origin. No evidence was presented or needed that undocu-

mented people damage the quality of life in Arizona—illusory correlations, “ availability” and the

apparent desire to identify and punish the “other” were suﬃcient. An early poll showed that 70

percent of the population agreed with the legislation, although that number dropped substantially

over several months.

Target populations are the people, groups, and organizations that are impacted by the policy.

Target populations are critical to policy eﬀectiveness, since the target groups must coproduce or

behave in ways intended by the policy if the policy is to achieve its purported goals. In the

example above regarding Arizona’s Senate Bill 1070, one of the reported purposes is to rid

the state of undocumented people. These people must either “get the message” and leave, or

the provisions of the law that enable local law enforcement to identify the undocumented and

refer them to federal authorities for deportation must be operative. Further, the federal oﬃcials

are the only ones that can actually deport someone. Target populations, however, may not be

Policy design and transfer

223

chosen on the basis of goal achievement, at all. If the goal is to reduce crime, then shifting law

enforcement responsibilities from crime prevention and detection to ﬁnding undocumented

people who have not committed any crime, may not be an eﬀective or eﬃcient method of crime

prevention. Target groups may be selected on the basis of their direct linkage to goal achieve-

ment, but they also are selected on the basis of political power, wealth, images (particularly

whether they are “deserving” or “undeserving,” or other principles that send messages about

the values of the society. Furthermore, people’s experiences with government as targets shapes

their view of government and their understanding of their role in government. As Schneider

and Ingram argued:

the way targets are treated by policy is central to justice, citizenship, support for democratic

institutions, and democratic problem solving. Policy can either reinforce or undermine

government legitimacy and sense of civic duty.

(1997: 85)

Policy-makers have a choice among diﬀerent target groups that can be involved or excluded

from policy. To create jobs, for example, policy can develop public works programs that actu-

ally create job programs that hire people and put them to work on road construction, bridges,

libraries, tutoring for school children, and aides for child protective services. Alternatively,

policy can cut corporate taxes with the contention that cutting corporate taxes will enable

corporations to expand thereby hiring people. Or policy can create jobs by providing tax credits

for new hires, and so on.

Policies allocate values—beneﬁts and burdens—to target groups and this, too, oﬀers policy-

makers a wide range of choices for solving any given problem. An entire chain of target groups

may be involved as intermediate linkages between policy design and goal achievement.

Some may be in line for beneﬁts, others for burdens. Target groups are sent messages by policy

design of what they deserve or do not deserve. Persons without jobs during the virtual collapse

of the US economy were told, on the one hand, that they deserved jobless beneﬁts and

that the amount of time they would be eligible would be extended, and on the other hand, that

paying jobless beneﬁts encouraged people to not look for jobs and in fact became a factor that

prolonged unemployment.

Rules refer to who is to do what, when, with what resources and limitations. Rules include

the eligibility criteria for receiving beneﬁts, as well as the behaviors that would lead to arrest and

punishment. Tools are the incentive and disincentives built into the design intended to insure

that agencies and target groups and any other players in the system take the policy-preferred

actions. Tools have been categorized in terms of the underlying behavioral assumptions: authority

(just doing what one is told to do); positive incentives, negative incentives or punishment, actual

force, symbolic and hortatory tools (persuasive devices), learning tools, and capacity-building

tools are examples. Tools are quite interchangeable and policy designers have a wide range of

choices regarding how to get target groups to follow the rules.

Implementation structures are often built into the policy design. These may be top-down com-

mand and control types of implementation or follow the ideas of adaptive management and

leave considerable discretion to the local implementers.

Rationales are the reasons, justiﬁcations, or legitimations for the design itself. These may be

straightforward or deceptive, logically consistent or inconsistent, and based on a wide range of dif-

ferent values ranging from justice-oriented rationales to economic, scientiﬁc or religious. Ratio-

nales may reﬂect the actual reasons, or the real reasons may be carefully masked as policy-makers

attempt to tie the policy to a popular problem for the purpose of gaining political capital and

Schneider

224

support. Underlying assumptions typically are not laid out in the policy design itself but are brought

to the analysis by the policy analyst. There are three types of assumptions: technical assumptions

that are scientiﬁc and can be tested with scientiﬁc research (including social sciences); normative

assumptions that the policy design is fair, just, compassionate, treats people as they ought to be

treated, consistent with prevailing values; and behavioral assumptions about what people need

to be motivated to act in policy-preferred ways.

Policy designs contain—in addition to these empirical elements—other more judgmental

characteristics such as the type of message (positive, negative) being sent to target populations;

the extent of deception in the design and the rationales used to justify it; and the strength of the

logic holding the elements together. These, too, are important aspects of design for the policy

analyst to unmask.

Commonly copied policy designs

One of the least developed aspects of the study of policy design is in the development of design

patterns – clusters of elements that commonly go together. The approach that will be presented

here builds from my work and that of my colleague Helen Ingram to classify design types in terms

of the power and the social construction of the target populations and whether beneﬁts or burdens

are being distributed. We hypothesized that certain design elements tend to be used, depending

on this classiﬁcation. Furthermore, subsequent work has contended that some types of policy

design will be more subject to path dependency than other types, and in the remaining part of

this chapter, the point will be made that some types are more likely to transfer than others.

Schneider and Ingram have argued that a great deal can be understood about the elements of

policy design by taking into account the political power of target populations, their social con-

struction (such as “deserving” or “undeserving”) and whether beneﬁts or burdens are being

distributed. Using this two-dimensional framework, target populations can be clustered generally

into one of four types: advantaged groups (politically powerful and positively constructed as deser-

ving, intelligent, hard-working, and other positive values of the society); contenders who are

politically powerful but carry a negative image as “greedy,”“dishonest,”“undeserving of what they

have”; dependents (politically weak but constructed as “good but weak or helpless” people); and

deviants (politically weak and carrying the image of dangerous, violent, mean, selﬁsh). Most of

the research using this framework has focused on the allocation of beneﬁts and burdens consistent

with the theory and on the consequences of policy design for political attitudes and participation.

Far less has been done with the ideas that some types of designs are more apt to transfer or diﬀuse

across jurisdictions than others, or are more (or less) likely to converge over time. Schneider

(2006) has shown that a particular type of design—policies that inﬂict harsh punishment on

those who break the law have been far more path-dependent, over time, than policies that

utilize rehabilitation, probation, house arrest, ﬁnes, or other less draconian policies. This study

based on state-level incarceration data from 1927 through 2003 showed far longer periods of

continuous increases in incarceration rates across almost all states than periods of decreases. The

research also found that state-level incarceration rates tend to change in the same direction, at the

same time, as if some kind of national “mood” were inﬂuencing all of them at once, or that policy

designs were diﬀusing rapidly throughout the nation even though decisions about incarceration

are made at the state and local level and in response to the state and local laws.

The most likely designs to transfer are those that ﬁt the dominant patterns identiﬁed by

Schneider and Ingram. The ﬁrst of these is providing beneﬁts to advantaged groups. The design

elements are expected generally to be quite similar across jurisdictions: strong statutes; tight

implementation structures but very little oversight or assessment of results; “popular” rationales

Policy design and transfer

225

that ﬁt into prevailing values, capacity building and learning tools for target groups; and uni-

versalistic rules of eligibility. These designs are politically popular regardless of whether one is

liberal or conservative since the target populations are considered deserving of what they receive

and the rationales are structured carefully to ﬁt prevailing values. Also, it is diﬃcult to generate

opposition except on rational/instrumental or scientiﬁc grounds that the means will not pro-

duce the desired ends—but since most do not pay much attention to such logic, opposition is

diﬃcult to generate. Strong statutes provide suﬃcient detail to ensure that the beneﬁts are

actually delivered and not sidetracked in some ways. Once a program area, however, becomes

“oversubscribed”—whereby so many beneﬁts are being provided that it becomes politically

risky and diﬃcult to sustain the rationales or create new ones—implementing agencies can be

expected to be given discretion to determine eligibility. Implementation structures are tight in

that there are clear lines of authority and responsibility for carrying out the statutory intent. If

policy is not clear on the linkages between funding and desired outcomes, however, learning

and capacity-building tools may be used to help to guarantee that the programs work. Over-

sight, however, is weak or nonexistent or largely illusory except to prevent fraud that would

endanger the program. Rationales will depend on the prevailing “ideas in good currency.” If

jobs are needed and valued, the beneﬁts will be touted as contributing to job creation. If national

security is involved, the rationale will be based on national security. If fairness or justice is involved

and if that is a value currently in vogue, then it would be used. Policy tools will be largely

benign—learning tools providing adequate discretion even to target populations, capacity build-

ing, and direct handouts. Rules will tend toward the universalistic without means testing and

without other tight restrictions. Beneﬁts will tend to be provided now, not well into the future,

and signiﬁcant sums of money or tax breaks can be expected. Far less likely to transfer are designs

that inﬂict burdens on advantaged populations, but when regulations are transferred, design

elements will tend to have weak statutes (providing multiple loopholes for advantaged groups to

escape), highly particularistic (only some of the advantaged are subject to the rules), and regulatory

activities can expect to be underfunded.

A second commonly copied type of design is expected to be providing punishment and dis-

cipline to deviant groups. These designs are expected also to involve universalistic rules (everyone

who breaks a particular law is supposed to be treated the same way as others); the rationales

focus on negative social constructions of the target groups as “dangerous,”“untrustworthy,”

“already had their chance,”“violent,” and so on. The logic connecting design elements to desired

results is largely unimportant provided that the target groups can continue to be constructed as

undeserving. Even though there is considerable political gain and ideological commitment to

punishing the undeserving, the political leaders do not want to spend much money on this task

if it can be avoided. Evaluation to discern program eﬀectiveness is seldom used when punishment

is provided to “undeserving” people.

A third commonly copied design involves the provision of services or the discipline of various

kinds of dependent populations (children, women, mothers, disabled, etc.). These designs can

generally be expected to involve very particularistic rather than universalistic rules, typically means-

tested, far more “talk” than “funding,” especially from the federal level; tools to induce desired

behavior can be expected to rely on authority, persuasion, and expectations that the clients will

establish their own eligibility—no one will help them to do that. Rules contained in these policies

especially when beneﬁts are being distributed are expected to be strict, with complex eligibility

requirements. In contrast to the types of design discussed above where beneﬁts are provided to

advantaged groups or punishment to deviant groups, provision of beneﬁts to dependents typically

will require evaluation studies to “prove ” the

worthiness of the

policy. Rationales will focus on

justice, fairness, and when discipline or punishment is involved—it is “for their own good.”

Schneider

226

A fourth type of commonly copied designs pertains to attempts to provide beneﬁts to contenders

(such as Wall Street, corporations, “big labor”, “big banks”) at the same time that the public

stance is to discipline and restrict these groups. Whether the design elements that are likely to

transfer will be the same across jurisdictions simply is not known but the theory strongly sug-

gests that beneﬁcial policy for contenders will transfer and will be the most deceptive of all the

types of designs.

Conclusions

Given the fact that public policies have increasingly been used for political and ideological purposes

rather than to ﬁnd the best solution to a collective problem, policy analysts are well advised to

study carefully the design elements and analyze them not simply in terms of their rationality,

but also in terms of who gets what, who wins and who loses, whose image is emboldened and

who is denigrated, and what are the messages being sent to the target populations and the broader

public. What are the impacts on political participation and attitudes? Is it the case that policy

designs embed the social constructions and assumptions and then reproduce that same pattern of

politics? Is policy itself partly responsible for the politics that ensue, as Lowi so famously argued

many years ago? It is not suﬃcient to simply analyze policy designs in terms of their technical

logic and whether they achieve their purported goals? The full array of elements along with the

images, hidden messages, underlying assumptions, deception, and illogic need to be probed by

the policy analyst.

Bibliography

Arizona Republic, ‘Sen. John McCain ties ﬁres to illegal border crossings’, June 20, 2011, B1.

—— ‘The issue: McCain’s comments on migrants, Wallow Fire’, by Ginger Rough, August 27, 2011.

Bobrow, Davis B. and John Dryzek. 1987. Policy Analysis by Design. Pittsburgh, PA: University of Pittsburgh

Press.

Brooks, David. 2011. The Social Animal: The Hidden Sources of Love, Character, and Achievement. New York:

Random House Digital, Inc.

Dahl, Robert and Charles Lindblom. 1953. Politics, Economics, and Welfare. New York: Harper.

Forester, John. 1989. Planning in the Face of Power. Cambridge, MA: MIT Press.

Gray, Virginia. 1973. ‘Innovation in the states: A diﬀusion study’, American Political Science Review 67(4):

1174–85.

Howlett, Michael and Raul Lejano. 2013. ‘Tales from the crypt: The rise and fall (and rebirth?) of policy

design’, Administration and Society (forthcoming)

Ingram, Helen and Anne L. Schneider. 1990. ‘Improving implementation through framing smarter statutes’,

Journal of Public Policy 10(1): 66–87.

——and Anne L. Schneider. 1991. ‘The choice of target populations’, Administration and Society 23(3):

333–56.

James, Oliver and Martin Lodge. 2003. ‘The limitations of “policy transfer” and “lesson drawing” for

public policy research’, Political Studies Review 1: 179–93.

Kahneman, Daniel, Paul Slovic and Amos Tversky. 1982. Judgment under Uncertainty: Herusitics and Biases.

Cambridge: Cambridge University Press.

Karch, Andrew. 2007. Democratic Laboratories: Policy Diﬀusion among the American States. Lansing, MI: University

of Michigan Press.

Kingdon, John. 1984. Agendas, Alternatives, and Public Policies. New York: Longman.

Lasswell, Harold. 1936. Who Gets What, When, and How? New York: McGraw Hill.

Lieberman, Robert, Helen Ingram and Anne Schneider. 1995. ‘Social

constructions’, American Political

Science Review 89(2):

437–44.

Linder, Stephen H. and Guy Peters. 1985. ‘From social theory to policy design’, Journal of Public Policy 4(3):

237–59.

Policy design and transfer

227

Mettler, Suzanne and Joe Soss. 2004. ‘ The consequences of public policy for democratic citizenship:

bridging policy studies and mass politics’, Perspectives on Politics 2(1): 55–73.

Schneider, Anne Larason. 2006. ‘Patterns of change in the use of imprisonment in the American states: An

integration of path dependence, punctuated equilibrium and policy design approaches’, Political Research

Quarterly 59(3): 457–70.

——and Helen Ingram. 1988. ‘Systematically pinching ideas: A comparative approach to policy design’,

Journal of Public Policy 8(1): 61–80.

——and Helen Ingram. 1990. ‘Behavioral assumptions of policy tools’, Journal of Politics 52(2): 510–29.

——and Helen Ingram. 1993. ‘The social construction of target populations’, The American Political Science

Review 87(2): 334–46.

——and Helen Ingram. 1997. Policy Design for Democracy. Lawrence, KS: University of Kansas Press.

——and Helen Ingram (eds). 2005. Deserving and Entitled: Social Construction and Public Policy. Albany, NY:

SUNY Press.

——and Mara Sidney. 2009. ‘What’s next for policy design/social construction theory?’, Policy Studies

Journal 37(1): 103–19.

Schön, Donald and Martin Rein. 1994. Frame Reﬂection: Toward the Resolution of Intractable Controversies.

New York: Basic Books.

Stone, D.A. 1988. Policy Paradox and Political Reason. Glenview, IL: Scott, Foresman.

Stimson, James. 1991. Public Opinion in America . Boulder, CO: Westview Press.

Utah Compact. 2011. http://www.deseretnews.com/article/700080758/Oﬃcial-text-of-Utah-Compact-

declaration-on-immigration-reform.html (Accessed 10 September 2011).

Walker, Jack. 1969. ‘The diﬀusion of innovations among the American states’, American Political Science

Review 63(3): 880–99.

Schneider

228

Epistemic communities

Claire A. Dunlop

Introduction

Peter M. Haas formulated the epistemic communities framework as a means of exploring the

inﬂuence of knowledge-based experts in international policy-making. Speciﬁcally, the approach was

designed to address decision-making instances characterized by technical complexity and uncer-

tainty. Control over the production of knowledge enables epistemic communities to articulate

cause-and-eﬀect relationships and so frame issues for collective debate and export their policy pro-

jects globally. Despite the framework being two decades old, there are still relatively few studies

which explicitly test or develop the concept theoretically, thus making it diﬃcult to assess what we

have collectively and cumulatively learned about this topic. This chapter attempts to systematize

key developments in the literature. The ﬁrst section outlines the concept locating it in the politics

of ideas literature. The second explores the state of the art in the empirical studies deploying the

epistemic communities framework. The third section considers the theoretical challenges that

researchers face when attempting to study epistemic communities. Discussion here proposes ﬁve

possible causal pathways through which we can explain how epistemic communities help decision-

makers learn. The chapter closes with a brief sketch of potential future research frontiers for

scholars interested in the power of knowledge and epistemic communities in public policy.

Epistemic communities and the politics of ideas

Epistemic communities are groups of professionals, often from a variety of diﬀerent disciplines,

which produce policy-relevant knowledge about complex technical issues (Haas 1992a: 16). Such

communities embody a belief system around an issue which contains four knowledge elements:

(1) a shared set of normative and principled beliefs, which provide a value-based rationale

for the social action of community members; (2) shared causal beliefs, which are derived

from their analysis of practices leading or contributing to a central set of problems in their

domain and which then serve as the basis for elucidating the multiple linkages between possible

policy actions and desired outcomes; (3) shared notions of validity – that is, intersubjective,

internally deﬁned criteria for weighing and validating knowledge in the domain of their

229

expertise; and (4) a common policy enterprise – that is, a set of common practices associated

with a set of problems to which their professional competence is directed, presumably out of

the conviction that human welfare will be enhanced as a consequence.

(1992a: 3)

The concept was the subject of an International Organization (IO) special edition in 1992 edited by

Haas. In this, the full framework was articulated (1992a) and empirically explored in environmental,

security, trade and international political economy cases studies. This volume still represents the

keystone of the epistemic communities literature. Emphasizing experts’ inﬂuence over decision-

maker learning as a potentially central mechanism eﬀecting policy development and change, the

epistemic communities framework aims to make sense of policy-making in conditions of

uncertainty and technical complexity. In these settings, decision-makers’ preferences will be less

clear to them transforming policy-making into an exercise in learning rather than bargaining.

As the main vehicles for authoritative consensual knowledge, epistemic communities have the

ability to help formulate policy in three main ways: elucidating cause-and-eﬀect relationships

and providing advice on the likely results of various courses of action; shedding light on the

complex interlinkages between issues, and helping to deﬁne the self-interest of states (Haas 1992a).

The ‘value added’ of the approach is that it highlights the importance of actors that are able to

deﬁne complex problems, particularly in the early policy design stages of the policy cycle where the

uncertainty of novel policy problems is at its peak. The framework is intended to complement

existing theories of policy-making that focus upon interests and the calculation of costs and

beneﬁts (rational choice), identities and socialization (constructivism), policy legacies (historical

institutionalism) and the use of words and discourse (poststructuralism) (1992a: 6). The core analy-

tical point here is that in the absence of epistemic communities to frame complex issues and proﬀer

new ideas, policy-making would follow more conventional, unreﬂective paths (Haas 2011).

The advent of the epistemic communities framework should be seen as part of the wider

‘renaissance of knowledge’ (Radaelli 1995). In political science in the 1990s, scholars were

reminded that, along with the traditional interests and institutions, there was a third ‘ i ’ . Ideas also

matter in explaining political decision-making. Although material power, identities and policy

legacies have not disappeared from analysis, this ideational turn emphasizes decision-makers as

‘sentient’ agents (Schmidt 2010). As such, they are sensitive to new ideas or new representations

of existing ones. The politics of ideas agenda has been followed enthusiastically and resulted in

empirical analysis that sheds light on how public policy emerges from new ways of thinking,

beliefs, rhetoric and discourse. The contribution of the epistemic communities framework is to

remind us that ‘ideas would be sterile without carriers’ (Haas 1992a: 27). This is an anthro-

pomorphic conceptualization of knowledge (Radaelli 1997: 169) where those experts who

create the knowledge on which decision-makers depend are central to political analysis. Thus,

to identify an epistemic community is to identify a set of actors with the professional and social

stature to make authoritative claims on politically pertinent and socially relevant issues of the

day. Their success is dependent, not simply on their epistemic resources but also on their political

acumen. Epistemic communities must persuade decision-makers, and successfully navigate the

machinery of government by insinuating themselves into bureaucratic positions, if their consensual

knowledge is to inform policy choices.

The empirical analysis of epistemic communities

The ﬁrst step in mapping the ﬁeld of epistemic communities’ scholarship involves exploring the

empirical analysis deploying the term. Bibliographic analysis provides the broad canvas on which

Dunlop

230

we can sketch out some of the signiﬁcant details. Using a citation search, we identiﬁed some

638 articles, book chapters and books citing Haas’s 1992 introductory article to the IO special

edition.

Results conﬁrm that the idea of epistemic communities is ﬁrmly embedded in the social

sciences lexicon; in the past two decades, the concept has travelled across a wide range of dis-

ciplines – 54 in fact (see Figure 18.1

). Such wide appeal ﬁts with the framework’s aim to explain

the role of experts in the technically complex policy problems which dominate the policy

process in contemporary society. As we would expect Government and Law, Public Adminis-

tration and International Relations (IR) lead the way in citations, though interestingly IR lags a

little way behind the ﬁrst two political science sub-ﬁelds (see Figure 18.2).

Despite epistemic communities’ obvious resonance across social scientiﬁc disciplines, a review

of the literature reveals superﬁcial engagement. Speciﬁcally, the actual identiﬁcation of epistemic

communities using the fourfold belief system that deﬁnes them remains rare. Rather, the term is

more frequently used metaphorically to describe any group of experts giving policy advice. A

succinct explanation for this state of aﬀairs is oﬀered by Wright: ‘[A]ctually identifying these

communities … can be a diﬃcult process’ (1997: 11). Practical obstacles, such as identifying,

locating and gaining access to those believed to be members of an epistemic community, may

frustrate attempts to engage with the approach as an analytical tool.

Identiﬁcation of an epistemic community’s belief system presents a further empirical challenge.

In most cases, authors delineate beliefs using the qualitative ‘soaking and poking’ (Fenno 1986)

Figure 18.1 Bibliographic analysis of 638 articles, book chapters and books citing Haas’s 1992

introductory article to the IO special edition

Source: Author

Epistemic communities

231

strategies advocated by Haas. This is especially appropriate when attempting to uncover what

can be sensitive information – particularly experts’ normative and policy beliefs about an issue.

However, in addition to analysing scientiﬁc documents and conference proceedings, and conducting

interviews, some scholars have used quantitative surveys (Radaelli and O’Connor 2009; Wright

1997) to gather this kind of information. Criticisms have, however, been voiced about the neglect

of how epistemic communities’ beliefs about an issue are mediated by experts’ own deeper socio-

political beliefs and psychological motivations (Finlayson 2004; Zelikow, 1994). This question

of social conditioning is explored by Mitchell et al. (2007) in relation to nuclear scientists in the

United States (US) and European Union (EU) member states. Using survey data, they conclude that

values (expressed on the traditional left-right spectrum) and national context matter for epistemic

communities’ policy preferences.

The challenges of using epistemic communities as an analytical framework are not only prac-

tical. Getting to grips with the structure and power dynamics that may exist within an epistemic

community is important if we are to understand belief system formation and message framing.

For example, in their study of an epistemic community of legal scholars, van Waarden and Drahos

(2002) uncover a hierarchical structure where some actors are more equal than others: ‘it mat-

ters not only what some member says, but who says it’ (2002: 930). In a related vein, Drake and

Nicolaïdis’s (1992) and Dunlop’s (2009) case studies illustrate the impact that a division of labour

between disciplines within these groups can have on belief system formation and communication.

Yet, in the politics literature, these examples are the exception rather than the rule. In most

Figure 18.2 Bibliographic analysis of 638 articles, book chapters and books citing Haas’s 1992

introductory article to the IO special edition (subject areas receiving 11 citations or more)

Source: Author

Dunlop

232

cases, where scholars do map out the beliefs which epistemic communities embody, the com-

munity itself is rarely the centre of analytical attention – usually losing out to interest groups and

institutions. Disciplinary preferences provide a plausible explanation for this. Political scientists’

attention is naturally drawn toward political institutions rather than the world of professionals

and experts. Here, the increasing attention being paid by sociologists, well-schooled in the

power of professions (Abbott 1988), will be valuable to the epistemic communities approach.

For

example, Roth’s (2008a, 2008b) analysis of epistemic communities of embryologists and zebra-

ﬁsh experts, respectively, forms the basis for a discussion on the internal structure and politics of

these communities. Lorenz-Meyer’s (2010) recent work on the membership conﬁgurations

within communities, and experience of actually being a member, is similarly promising. By

exploring the internal workings of epistemic communities, analysts can uncover how experts

expand their political power and ability to inﬂuence.

The epistemic communities framework was originally developed to explain the role of transna-

tional experts in fostering international policy coordination – something reﬂected in the pro-

nounced number of studies focussing on the EU (for example, Kaelberer 2003; van Waarden and

Drahos 2002; Zito 2001).

The bibliographic review also reveals, however, that these communities

are not restricted to inter-, trans- or supranational levels. Epistemic communities can be located

at any level of government. Indeed, they are commonly national entities, and policy transfer

concepts (see Chapter 17 of this volume) have usefully been deployed to explore what happens

to ideas as they move from an epistemic community in one jurisdiction to another (Albert and

Laberge 2007; Prince 2010; Melo 2004). Studies have also highlighted the more antagonistic

relationship that can develop between epistemic communities in diﬀerent places. In her studies of

hormone growth promoters and the biotech milk yield enhancer bovine somatotrophin (rbST),

Dunlop illustrates how epistemic communities in the EU and US, oﬀering contrasting interpreta-

tions of the same scientiﬁc evidence, were used to justify international policy divergence (2007,

2009, 2010). Harrison’s cross-national study of dioxins in pulp mill eﬄuents and paper products

reports a similar role for experts in cross-national policy divergence (2002). This underlines the

fact that the concept’s analytical leverage is founded on the potentially powerful role experts and

knowledge can play in overcoming collective action dilemmas in uncertain contexts regardless of

jurisdiction and government level. Empirical studies have also successfully applied epistemic

communities to more local settings – for example exploring interagency cooperation in California

(Thomas 1997); forest policy change in British Columbia (Kamieniecki 2000); medical ethics in

the Czech Republic (Simek et al. 2010) and transport policy in Belgium (Albrechts, 2001).

The empirical literature also highlights research design issues associated with locating epistemic

communities in dynamic policy environments. Two main challenges are evident: how to explore

epistemic communities’ interactions with other actors in policy arenas, and designing research

that is able to capture the temporal dimension of epistemic communities’ inﬂuence.

To understand how epistemic communities interact with other policy actors, we ﬁrst need some

background on what distinguishes epistemic communities from others in the policy process.

Their blend of beliefs and highly specialized expertise distinguish epistemic communities from

interest groups and policy networks (Haas 1992a: 22). This enables them to make legitimate claims

to being the main producers of knowledge in an issue area (Dunlop 2009). This speciﬁcation of

diﬀerence does not imply that Haas attempts to advance epistemic communities as morally

superior to other actors in the policy process (see the debate between Toke 1999 and Dunlop

2000). Rather, the aim is to record the view that expert groups’ main interactions are expected

to be with decision-makers and that they are not a brand of specialized interest groups with

non-material aims (Haas 1992a: 19). For Haas, the crucial claim to distinctiveness rests on ele-

ments (2) and (3) of their belief system which result in consensual knowledge. Haas consistently

Epistemic communities

233

implies the non-negotiable nature of these two elements – without them, the epistemic commu-

nity ceases to function as an authoritative voice of advice. Accordingly, in the event of a serious

challenge to the causal ‘world view’ of the community (2) which could not be settled internally

(3), the community would withhold policy advice: ‘Unlike an interest group, if confronted

with anomalous data, they would retract their advice or suspend judgement’ (Haas 1990: 55).

But what of the real world of politics – where constellations of political actors interact, argue

and bargain? Being ‘apolitical’ (yet ‘politically empowered’) seems unfeasible. The emphasis on

consensual knowledge may overstate the inﬂuence these expert enclaves alone can have. Interest

groups and rival scientiﬁc groups interact with epistemic communities learning from and arguing

with them. The need to explore epistemic communities’ propensity to build alliances was a

theme taken up by James Sebenius in the 1992 IO special edition. Using insights from nego-

tiation analysis, Sebenius argues that an epistemic community’sinﬂuence may emanate from

bargaining with other actors in an attempt to convert their ‘natural coalition’ of believers into a

‘winning coalition’, pushing forward a shared policy enterprise (characteristic (4) of an epistemic

community) (Sebenius 1992: 325). Therefore, epistemic communities have to be politically proac-

tive players to convey their message, interacting with a multiplicity of other actors where it is to

be expected that inﬂuence is variable and contingent as wider strategic games are played out. In

such a scenario, the full potential of consensual knowledge may only be realized through the

involvement of other, more politically astute, groups.

In the literature, this interaction is often captured by subsuming epistemic communities within

interest-based approaches. Most commonly, empirical studies which explore alliances between

epistemic communities and interest groups use the advocacy coalition framework (for example,

Dudley and Richardson 1996; Elliott and Schlaepfer 2001; Meijerink 2005). However, with the

interest-based approaches doing the analytical work, epistemic communities’ conceptual devel-

opment is stunted. Alliances with interest groups raise key issues that are still to be addressed. What

happens to the epistemic community itself as alliances stabilize? What aspect(s) of an epistemic

community’s belief system are most vulnerable to change, if any? Unlike Sabatier’s advocacy

coalition framework approach, the epistemic communities framework does not explicitly explore

the weight of the beliefs relative to one another. Exploring the susceptibility of a belief system

to change would tell analysts a good deal about a community’s propensity towards coalition

building in a given policy area. Exploring the extent to which one element could dominate

another, and how this is inﬂuenced by the membership of a community, could also help build

bridges with the emerging research on interest groups. For example, by delineating the relative

importance of normative and policy-oriented beliefs, synergies may be possible between epistemic

communities and advocacy networks (Keck and Sikkink 1998) and epistemic communities and

Communities of Practice (Wenger 1999).

While we have empirical studies of cooperation, there is little evidence that epistemic com-

munities are willing to engage in conﬂict with interest groups. Rather, conﬂict appears reserved

for other expert groups. For example, the epistemic battles between rival groups of experts,

anticipated by Haas (1990: 57, 1992b: 44), has been explored empirically by Youde (2007). In

his study of health policy in South Africa, Youde identiﬁes the emergence of both an epistemic

community and a ‘counter-epistemic community’ oﬀering fundamentally diﬀerent understandings

of AIDS and radically diﬀerent policy prescriptions.

This intervention is extremely valuable. In the two decades since the epistemic communities

concept was ﬁrst

introduced, communication technology has changed beyond all recognition

with the advent

of the Internet. Simultaneously, the technical policy dilemmas that decision-

makers encounter have increased in quantity and complexity – most notably climate change.

The democratization of information (if not knowledge) and increased urgency of complex

Dunlop

234

policy-making has resulted in a huge expansion in the number and types of groups that classify

themselves as experts. These trends are reﬂected in the increased willingness of scholars to talk

in terms of multiple epistemic communities (Stephens et al. 2011) that engage in ‘contests of

credibility’ (Epstein 1996: 3). The prescience of the approach is also reﬂected by the empirical

cases studies. The framework has been used to explore policy problems as diverse as agricultural

policy (for example, Coleman and Skogstad 1995); ﬁnancial policy (Rethel 2010); economics and

banking (for example, Ikenberry 1992); forestry (Shankland and Hasenclever 2011); defence and

security (Howorth 2004); health policy (for example, Ogden et al. 2003), and transport (for

example, Schot and Schipper 2011). Reﬂecting the approach’s empirical roots, environmental

case studies are by far the most common – something which is set to continue (for example,

Haas 1990, 1992b; Peterson 1992). Nine of the 118 articles citing Haas’s 1992a introduction in

2010 and 2011 were directly related to climate change and sustainable development (for

example, Lovell and MacKenzie 2011).

While epistemic communities are discrete actors, they are located in dynamic policy envir-

onments. This dynamism requires not only that research is designed in ways which illuminate

how epistemic communities relate to other policy actors; we must also attend to the temporal

dimension of policy-making. A focus on the ‘here and now’ of policy-making means that

analysis may miss epistemic communities’ inﬂuence. There are three reasons for taking the long

view; the ﬁrst concerns the internal operation of an epistemic community. While we might expect

experts’ inﬂuence to be biased towards the early stages of the policy process, the back story of

how a group of experts came together, and the process of knowledge creation, still matters.

Scientiﬁc consensus takes time to construct; knowledge is often incomplete and judgements of

cause and eﬀect partial. If we focus only on what has occurred when the consensus has been

reached and the community has entered the political arena, we may miss earlier episodes of

political behaviour where the expert enclave has, for example, reﬁned its position in order to

gain access to decision-makers (Dunlop 2007).

In the same vein, empirical studies are required to illuminate what happens to epistemic com-

munities over time, as they interact with decision-makers and other actors in the policy process.

This issue of time is again something which Haas acknowledges in pointing out that the

knowledge consensus represents a ‘temporally bounded notion of truth’ (1992a: 23, emphasis added).

Over time, this particular ‘truth’ will evolve, be challenged and altered, or debunked and replaced,

as new discoveries dictate. We need to empirically explore what happens to these communities

during this evolutionary process. Of speciﬁc interest is the impact that political exposure has on

an epistemic community’s belief system. Adler and Haas state that: ‘[A]s epistemic communities

consolidate and expand their political and bureaucratic inﬂuence internationally, additional ideas

may be incorporated into the core community beliefs’ (1992a: 374). Successful examination of

the stability and strength of the diﬀ erent elements of an epistemic community’s belief system

requires that the four elements’ characteristics are problematized clearly in empirical studies and

their importance relative to each other elucidated (Dunlop 2000).

second reason for

analysing epistemic communities over time concerns when epistemic

communities are located in the policy cycle. It is well established that experts are active in the

early agenda-setting and issue framing stages of policy-making and less visible as decisions are

being set and policy implemented. However, if we only front-load our attention to these early

stages, analysis may miss a reappearance or reincarnation of an epistemic community in the

policy evaluation and feedback stages.

Finally, the causal mechanism associated with the approach also requires that we take the

long view. Learning processes have their own temporal dimension – with ‘policy-oriented

learning’ and ‘enlightenment’ happening over protracted periods of time (Sabatier 1988; Weiss

Epistemic communities

235

1979). And so, we must look at the ‘moving picture’ rather than the ‘snapshot’ (Pierson 1996).

Long after an epistemic community has left the political arena, its ideas may be inﬂuencing

policy-making.

Theoretical challenges

While the epistemic communities framework is empirically well travelled, the analytical hor-

izons of the approach have not been broadened to the same extent. Despite being the con-

ceptual name to drop in studies of technical issues in political science and beyond, very little

work has been done to interrogate and develop the concept theoretically. The tendency in the

discipline to commit substantial time and resources to reﬁning and reforming such conceptual

frameworks – for example the last three decades’ reﬁnement of policy networks (for example

Thatcher 1998; see Chapter 12 of this volume) and the advocacy coalition framework (most

recently Weible et al. 2011; see Chapter 10 of this volume) – has not been repeated in the case

of epistemic communities. Given the empirical, practical and disciplinary barriers outlined already,

such underdevelopment is perhaps unsurprising. This is not to say that the analytical challenges

associated with the approach are unknown, however. On the contrary, many of the articles

citing Haas’s work oﬀer conceptual criticisms. But these are made in passing and have not been

related back to the framework itself. This section outlines some of the analytical gauntlets that

have been thrown down and proposes ways that these might be picked up by future scholars

focused on explanatory analysis of epistemic communities.

To understand how epistemic communities inﬂuence decision-makers’ learning we must ﬁrst

clarify our dependent variable. Asking what is actually being explained in epistemic community

analysis is perhaps not as self-evident a question as it may seem. In his case studies, Haas uncovers

epistemic communities as the catalysts for international policy coordination (1990, 1992a). This

policy outcome should not however be the standard against which an epistemic community’s

inﬂuence is judged. These cases show the maximum inﬂuence experts can have; they are extreme

examples. If we look only for international policy coordination, the results are liable to disappoint.

Epistemic communities are rarely ‘live’ actors at the decision-making or implementation stages

of the policy cycle. So, we must lower our expectations and give epistemic communities their

proper analytical place. What we are interested in is how epistemic communities help decision-

makers learn. Policy outputs should not be the primary focus as epistemic communities’ inﬂuence

over those will usually be indirect.

So, what do decision-makers learn from epistemic communities? In his exposition of the epis-

temic communities’ framework, Haas provides a clear deﬁnition of learning. Learning is a pro-

cess of informing decision-makers’ beliefs about the four key components of complex technical

issues embodied by epistemic communities with particular attention drawn to epistemic com-

munities’ inﬂuence on the ‘substantive nature of … policy arrangements’ and their more overtly

political role as the ‘nonsystemic origins of state interests’ (1992a: 4). Experts’ potential to stimulate

learning is assured by the control they enjoy over the production of knowledge relating to an

issue (ibid.: 2) and the inﬂuence they exert is a function of decision-makers’ uncertainty.

Despite this clarity and speciﬁcity, we still know relatively little about the possible forms of

learning that may arise between epistemic communities and decision-makers. The framework

emphasizes a form of learning that depicts control over knowledge as something which epis-

temic communities have and decision-makers, whose bounded rationality initiates their call for

advice, do not have (1992a: 14–16). This deﬁnition of uncertainty is problematic, however.

Epistemic communities’ entry into the policy arena is a function of decision-makers’ technical

uncertainty, however by conﬂating

this uncertainty with comprehensively bounded rationality

Dunlop

236

we are left with a single learning category where decision-makers experience extreme epistemic

deﬁciencies which need to be ﬁlled. And so, despite disaggregating belief systems into individual

components in the deﬁnition of what an epistemic community actually is, the epistemic resources

experts pass on to decision-makers are portrayed as a uniﬁed good.

This ‘deﬁcit model’ of learning (Dunlop 2009), where epistemic communities are required to

ﬁll gaps in decision-makers’ knowledge, reﬂects the empirical case which informed the original

development of the epistemic communities approach. Haas’s major study Saving the Mediterranean

(1990) illustrates the extreme level of inﬂuence that epistemic communities can have. In this

case, the epistemic community had the power to make decision-makers cooperate and develop

policies in ways they would not otherwise have. Such ideal typical cases yield important

examples of what is possible, however they also highlight extreme phenomena. If it is to explain

the role of epistemic communities in ‘who learns what, when, to whose beneﬁt and why’ (Adler

and Haas 1992: 370), the analytical framework must be able to account for the variety of these

actors’ learning interactions. While epistemic communities’ central role in information production

gives them the authoritative status to occupy this role, we cannot assume that they will be able

to exert control over every aspect of what is known about an issue or that the control they do

enjoy will be uniform across time and space. To understand the diﬀerent roles epistemic com-

munities occupy in the policy process, we must have a clear, coherent and systematic way of

describing variation in this explanandum.

One promising way of characterizing the learning that epistemic communities can eﬀect is to

focus upon the ways in which decision-makers’ preferences mediate the learning processes in

which they engage. This is something that Haas himself acknowledges, describing a scenario

where decision-makers rely on epistemic communities for only certain aspects of policy knowl-

edge (1992a: 15). Where decision-makers’ preferences are unclear or controversial, uncertainty

is generated and gaps appear into which epistemic communities can insinuate themselves. Learning

may also be diﬀerentiated across time and space as well as knowledge components; epistemic

communities can be ‘called in’ to provide one type of input and ﬁnd that the scope and intensity of

their inﬂuence may increase, decrease or change in emphasis (ibid.: 16).

In recent research, a typology has been developed to capture the variety of ways in which

decision-makers actually learn from epistemic communities (Dunlop 2009). Using a ﬂexible deﬁ-

nition of uncertainty for example, Dunlop (ibid.) suggests that variety is best captured by dif-

ferentiating between the control enjoyed by decision-makers and epistemic communities over

the production of substantive knowledge – or means – that informs policy on the one hand and

the policy objectives – or ends – to which that knowledge is directed on the other. The four

components identiﬁed by Haas can be reclassiﬁed

quite simply into

the foci of means and ends

with shared normative beliefs (component 1) cause and eﬀect postulates (component 2) and

intersubjective understandings of validity (component 3) making up the substantive means pro-

duced around an issue and the most overtly political component of epistemic communities the

common policy enterprise (component 4) equating to the end objectives. The implications of

this distinction for the types of epistemic community-decision-maker learning exchanges that may

prevail are elaborated using a typology of adult learning from the education literature which

delineates the four possible learning situations (see Dunlop 2009 for an exposition of the typology).

Dunlop’s empirical exploration of the typology provides grounds for optimism suggesting that

each of the types are readily distinguishable from one another and provide a parsimonious account

of learning exchanges and basis for ordering empirical ﬁndings across cases and comparison within

them. Of course, further empirical assurances are required that the typology does not contain self-

contradictory ideas. Suﬃce it to say that research needs to unpack the nature of uncertainty and

what it is in epistemic terms that decision-makers are open to learning from epistemic communities.

Epistemic communities

237

With regard to the independent variable, we can break down the explanatory challenges posed

by epistemic communities. Following Coleman (1986), explanation of epistemic communities’

inﬂuence in the policy process consists of unpacking three elements in turn: (a) the microfounda-

tions of the macro-level independent variable (i.e. epistemic communities and the learning they

eﬀect); (b) the micro-micro causal logics that characterize the learning interactions between

epistemic communities and decision-makers (and other policy actors), and (c) the mechanisms that

turn learning at the micro level into macro outputs of the dependent variable (i.e. learning in

government and, ultimately, policy outputs). Against Coleman’s ‘bathtub’ ideal of rigorous research,

the epistemic communities’ literature reveals the various conceptual challenges to be met.

a. Microfoundations of epistemic communities and learning: To uncover the microfoundations of

epistemic communities’ inﬂuence in the policy process, we must focus on the conditions of

their emergence. Empirical studies reveal at least two distinct epistemic community types –

each distinguished by the extent to which decision-makers exert control over their existence

(Dunlop, 2010). Epistemic communities were originally conceived as ‘evolutionary’ entities;

self-regulating enclaves of experts that existed ‘out there’ in the academic and research world

that entered the political arena by responding to decision-makers’ call for advice (Haas 1989,

1992a). In such ‘ideal typical’ scenarios, decision-makers’ uncertainty is such that epistemic

communities are able to oﬀer both substantive policy advice (means) and policy proposals (ends).

However, another type of expert enclave is apparent in the literature. Some epistemic

communities are what can be described as ‘governmental’–where members have been

deliberately selected by decision-makers to justify a predetermined policy decision (Dunlop

2010) or depoliticize an emotive issue (Peterson 2001). For example, in Ikenberry’s study of

Anglo-American post-war economic settlement (1992), an epistemic community was assembled

by decision-makers to provide technical and normative guidance to facilitate a move away

from policies based on unregulated free trade. A similar instance of decision-makers deli-

miting the policy ends is provided by Verdun in her study of the Delors’ Committee that was

appointed to provide the substantive epistemic means to supranational elites aiming to deli-

ver the policy end of Economic and Monetary Union (EMU) in the EU (1999). It is only

by identifying epistemic communities’ microfoundations, that the scope of decision-makers’

epistemic needs, and the likely range of epistemic communities’ inﬂuence, can be discerned.

b. Explaining learning interactions between epistemic communities and policy elites: On the micro-micro

interactions between epistemic communities and decision-makers, the causal mechanism

associated with the framework is obviously learning. However, learning as an explanation

takes many forms. Indeed, a closer look at the epistemic communities literature suggests that

learning is a ‘master mechanism’ (see Hedström and Swedberg 1998 for a deeper discussion

of causal mechanisms) which can be unpacked into a variety of speciﬁc causal logics. The

ﬁve causal pathways with most analytical potential are suggested and discussed in turn.

i Learning as instrumental: Haas’s epistemic communities framework, as it stands, implies two

causal rationales. The ﬁrst of these treats decision-maker learning as underpinned by the

Lasswellian desire to make ‘better public policy’.

Here, interactions between epistemic

commu

nities and decision-makers are guided by a functional logic where beliefs are updated

in a Bayesian manner (Radaelli 2009). Epistemic communities’ supply the policy-relevant

information required by decision-makers to update their beliefs. Public policy, in this view,

is a matter of trial-and-error and experts’ role is more overtly technical than political.

ii Learning as persuasion and socialization: Haas also emphasizes that the framework is under-

pinned by ‘limited constructivism’ (1992a: 23) stressing the importance of persuasion

and reﬂection in the redeﬁnition of decision-makers’ interests and identities (Checkel

Dunlop

238

2001; Jacobson 2000: 160–1). In Haas’s Mediterranean study (1990) for example, the

epistemic community which grew from the United Nations Environment Programme

(UNEP) persuaded decision-makers across nations to internalize a new set of environmental

standards and common understanding of the problem. The international diﬀusion of

new environmental norms resulted in eﬀective policy coordination.

iii Learning as calculation: Moving beyond the causal logics originally envisaged for epistemic

communities, more overtly political learning styles may exist where interaction is underpinned

by a ‘logic of consequence’ and behaviour a function of calculation. In such interest-driven

accounts, epistemic communities and decision-makers act in ways that maximize their utility

where their engagement in a learning relationship is a product of the sanctions and rewards

they perceive such behaviour will bring. The beneﬁt of this logic is that it enables a move

away from simply treating learning as something which is value driven, or based on tech-

nical rationale. Rather, decision-makers can control a policy domain or gain political advan-

tage by using and, perhaps sourcing, specialist knowledge in a strategic manner (Chwieroth

2007; Dunlop and James 2007; Kohler-Koch 2000; Niemann 1998). For example, principal-

agent modelling has been used to explore the role of epistemic communities is delivering

pre-determined policy choices in an eﬃcient and credible manner (Dunlop 2010). We

should be clear, it is not simply decision-makers that can be treated as rational actors.

Epistemic communities’ belief systems suggest that while they do not have direct pecuniary

incentives, they are still interested parties. Because of their socio-political beliefs these are

experts who want rather than need to be in the policy arena. They are self-selecting policy

actors driven by normative and policy beliefs. Moreover, as creators of knowledge, they

are essentially ‘residual claimants’ (Alchian and Demsetz 1972); long-term shareholders

in a product who may compete with other ‘claim makers’ to control an epistemic ﬁeld.

iv Learning as legitimacy: The sociologies of science, knowledge and professions have obvious,

but rarely used, insights for epistemic communities. Speciﬁcally, the logic of practice

thesis (Bourdieu 1990) can be used to illuminate the interaction between experts and

decision-makers as a game where action is informed by the cultural dispositions of the

two sets of actors involved. In relation to advisory games, epistemic communities as

symbols of authority and legitimacy has been criticized in the Science and Technology

Studies (STS) literature (Jasanoﬀ 1996; Lidskog and Sundqvist 2002; Walker 2001) and

this ﬁeld oﬀers important analytical resources for the framework. Epistemic communities’

practices as authority claiming actors can be explored using Gieryn’s (1983) notions of

boundary work. In contrast to the logic of appropriateness, epistemic communities in

this view draw a ﬁrm line, a boundary, between themselves and the decision-makers they

advise. By emphasizing the diﬀerences in their professional identities, epistemic com-

munities monopolize epistemic resources and aim to impose their view of the world by

dint of their epistemic authority. Decision-maker learning in this scenario is not about

uncovering ‘what works’, but rather is symbolic and legitimacy seeking.

v Learning as unreﬂective: The idea of learning as unreﬂective is more than a little contra-

dictory. But, we can imagine scenarios where learning from an epistemic community is

an automatic response from decision-makers. Following Hopf’s thesis on the ‘logic of

habit’ in international relations (2010), habitual behaviour ﬂourishes in environments

where uncertainty is absent. Where an epistemic community has become institutiona-

lized, and has successfully attenuated uncertainty and advisors’ trustworthiness is assured

(ibid.: 16), decision-makers may be content to unquestioningly adopt its ideas thereafter.

The proposition that epistemic communities’ interpretations

could be adopted

whole-

sale through routinized behaviour of decision-makers is perhaps not as controversial as it

Epistemic communities

239

may at ﬁrst appear. Epistemic communities that have been institutionalized in bureaucracies

may command considerable resources – institutional as well as epistemic (Finnemore

1993). This is especially plausible in technically complex policy domains where the

political stakes are relatively low or remain hidden.

Of course, if considered in isolation, each of these conceptual lenses may obscure as much

as they reveal. Certainly, mono-causal explanation will not take analysts very far in the

complex world of policy learning. Rather, these logics are better used in combination. For

example, in her study on hormone growth promoters, Dunlop (2010) found that explana-

tions emphasizing a calculative logic were important at the start of the relationship between

European Commission oﬃcials and their epistemic community but gave way to learning

underpinned by socialization further down the line.

vi Relating the micro world of learning to the macro world of policy outputs: Finally, research must relate

the micro world of epistemic community-decision-makers’ interactions to the macro-level

world of government learning and policy outputs. The challenges posed by such aggrega-

tion are considerable. Analysts must capture how decision-maker learning becomes trans-

formed into learning at a systemic level. They must also pinpoint the source of those

lessons as being the epistemic community. Methodological choices may hold the key here.

The painstaking qualitative process-tracing originally advocated by Haas (1992a) remains

the most suitable method to uncover whether socially accepted ideas can actually be traced

back to a group of experts. Counterfactual analysis, also suggested by Haas, is a valuable yet

underused tool which can help to construct a rigorous account of learning that would not

have occurred in the absence of an expert enclave creating new policy relevant ideas.

Conclusions and future research agenda

The epistemic communities approach speaks directly to the complex challenges faced by decision-

makers in uncertain policy environments. Accordingly, the approach has proved valuable in

explaining the role of experts in a variety of political settings and has remarkable disciplinary

reach beyond its roots in IR. Yet arguably, we are still some way oﬀ the ‘reﬂective research

program’ envisaged two decades ago (Adler and Haas 1992).

Empirical research on epistemic communities has ensured that the term is now part of the

social science lexicon. To deepen our understanding, research is required that explores how epis-

temic communities emerge in the ﬁrst place and what happens to them as they interact with

decision-makers and other policy actors over time. Moving from exploratory to explanatory mode,

future research interrogating the causal mechanism of learning would be welcome. In particular,

following major analytical themes which have emerged in political science in the past few years,

the role of professional practice and habit in how decision-makers learn from epistemic communities

would represent important additions to the research agenda. Finally, with regard to methodol-

ogy, again epistemic communities should take inspiration from where analytical energies are being

deployed in the social sciences. Speciﬁcally, it would be useful for scholars to invest research

time in two areas: the role of temporal variables in epistemic communities’ inﬂuence and the

transformation of learning at the micro level to learning at the systemic level.

Acknowledgements

Claire A. Dunlop gratefully acknowledges the support of the Economic and Social Research Council

(grant R00429034387) and European Research Council, grant on Analysis of Learning in Regulatory

Governance, ALREG, available at: http://centres.exeter.ac.uk/ceg/research/ALREG/index.php.

Dunlop

240

Notes

1 The citation search involved two steps. The ﬁrst was to perform a citation search using the ISI social

sciences database – this yielded 583 articles. The second was a manual search for works not cited in the

ISI—a further 55 articles, books and book chapters citing Haas (1992a) were uncovered at this stage.

2 The ego-network analysis of the citation data presented in Figures 18.1 and 18.2 were conducted using

UCINET network analysis software (Borgatti et al. 1999).

3 There have been 45 research articles citing Haas’s 1992 IO introduction in journals classiﬁ ed under

Sociology by the Web of Science. Just over half of these have been published since 2006 (Web of

Science citation search 22 July 2011). Perhaps surprisingly, only 17 of these are published in journals

which identify as including Science and Technology Studies (STS).

4 It is instructive to note that the leading journal publishing work citing epistemic communities is the

Journal of European Public Policy (which recorded 40 of the 583 articles listed in the ISI).

Bibliography

Abbott, A. 1998. The System of Professions. Chicago: University of Chicago Press.

Adler, E. and P.M. Haas. 1992. ‘Conclusion: Epistemic communities, world order and the creation of a

reﬂective research program’, International Organisation 46(1): 367–90.

Albert, M. and S. Laberge. 2007. ‘The legitimation and dissemination processes of the innovation system

approach: The case of the Canadian and Quebec science and technology policy’, Science Technology and

Human Values 32(2): 221–49.

Albrechts, L. 2001. ‘How to proceed from image and discourse to action: As applied to the Flemish diamond’,

Urban Studies 38(4): 733–45.

Alchian, A.A. and H. Demsetz. 1972. ‘Production, information costs and economic organization’, American

Economic Review 62, 777–95.

Borgatti, S.P., M.G. Everett and L.C. Freeman. 1999. UCINET 6.0 Version 1.00. Natick, MA: Analytic

Technologies.

Bourdieu, P. 1990. Logic of Practice. Stanford, CA: Stanford University Press.

Checkel, J. 2001. ‘Why comply? Social learning and European identity change’, International Organization

55(3): 553–88.

Chwieroth, J. 2007. ‘Neoliberal economists and capital account liberalization in emerging markets’, International

Organization 61(2): 443–63.

Coleman, J.S. 1986. ‘Social theory, social research, and a theory of action’, American Journal of Sociology 91:

1309–35.

——and G. Skogstad. 1995. ‘Neo-liberalism, policy networks, and policy change: Agricultural policy

reform in Australia and Canada’, Australian Journal of Political Science 30(2): 242–63.

Drake, W.J. and K. Nicolaïdis. 1992. ‘Ideas, interests and institutionalisation: “Trade in services” and the

Uruguay Round’, International

Organization 46: 1, 37–100.

Dudley,

G. and J.J. Richardson. 1996. ‘Why does policy change over time? Adversarial policy communities,

alternative policy arenas and British trunk roads policy, 1945– 1995’, Journal of European Public Policy 3(1):

63–83.

Dunlop, C.A. 2000. ‘Epistemic communities: A reply to toke’, Politics 20(3): 137 – 44.

—— 2007. ‘Up and down the pecking order, what matters and when in issue deﬁnition: the case of rbST

in the EU’, Journal of European Public Policy 14(1): 39–58.

—— 2009. ‘Policy transfer as learning: Capturing variation in what decision–makers learn from epistemic

communities’, Policy Studies 30(3): 291–313.

——. 2010. ‘Epistemic communities and two goals of delegation ’, Science and Public Policy 37(3): 205–17.

——and O. James. 2007. ‘Principal-agent modelling and learning: The European commission, experts and

agricultural hormone growth promoters ’, Public Policy and Administration 22(4): 403–22.

Elliott, C. and R. Schlaepfer. 2001. ‘Understanding forest certiﬁcation using the advocacy coalition framework’,

Forest Policy and Economics 2(3–4): 257–66.

Epstein, S. 1996. Impure Science: AIDS, Activism, and the Politics of Knowledge. Berkeley, CA: University of

California Press.

Fenno, R.F. 1986. ‘Observation, context, and sequence in the study of politics’, American Political Science

Review 80(1): 3–15.

Finlayson, A. 2004. ‘Political

science, political ideas

and rhetoric’, Economy and Society 33(4): 528–49.

Epistemic communities

241

Finnemore, M. 1993. ‘International organizations as teachers of norms: The United-Nations Educational

Scientiﬁc and Cultural Organization and Science Policy ’, International Organization 47(4): 565–97.

Gieryn, T.F. 1983. ‘Boundary-work and the demarcation of science from non-science: Strains and interests

in professional ideologies of scientists’, American Sociological Review 48(6): 781–95.

Haas, P.M. 1989. ‘Do regimes matter? Epistemic communities and Mediterranean pollution control’,

International Organization 43(3): 377–403.

—— 1990. Saving the Mediterranean – The Politics of International Environmental Co-operation. New York:

Columbia University Press.

—— 1992a. ‘Introduction: Epistemic communities and international policy co-ordination’, International

Organization 46(1): 1–36.

—— 1992b. ‘Obtaining international environmental protection through epistemic consensus’, in I.H.

Rowlands and M. Greene (eds) Global Environmental Change and International Relations. Basingstoke:

Palgrave Macmillan.

—— 2011. ‘Epistemic communities’,inIPSA Encyclopedia of Political Science, New York: Sage.

Harrison, K.A. 2002. ‘Ideas and environmental standard-setting: A comparative study of regulation of the

pulp and paper industry’, Governance 15(1): 65–96.

Hedström, P. and R. Swedberg (eds). 1998. Social Mechanisms. Cambridge: Cambridge University Press.

Hopf, T. 2010. ‘The logic of habit in international relations’, European Journal of International Relations

16(4): 539–61.

Howorth, J. 2004. ‘Discourse, ideas, and epistemic communities in European security and defence policy’,

West European Politics 27(2): 211–34.

Ikenberry, G.J. 1992. ‘A world economy restored: Expert consensus and the Anglo-American postwar

settlement’ in International

Organization 46: 1, 289–322.

Jacobson,

H.K. 2000. ‘International institutions and system transformation’, Annual Review of Political Science

3: 149–66.

Jasanoﬀ, S. 1996. ‘Science and norms in global environmental regimes’, in F. Hampson Osler and J. Reppy

(eds). Earthly Goods: Environmental Change and Social Justice. Ithaca, NY: Cornell University Press.

Kaelberer, M. 2003. ‘Knowledge, power and monetary bargaining: Central bankers and the creation of

monetary union in Europe’, Journal of European Public Policy 10(3): 365 – 79.

Kamieniecki, S. 2000. ‘Testing alternative theories of agenda setting: Forest policy change in British

Columbia, Canada’, Policy Studies Journal 28(1): 176–89.

Keck, M. and K. Sikkink. 1998. Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca,

NY: Cornell University Press.

Kohler-Koch, B. 2000. ‘Framing; the bottleneck of constructing legitimate institutions ’, Journal of European

Public Policy 7(4): 513–31.

Lidskog, R. and G. Sundqvist. 2002. ‘The role of science in environmental regimes: The case of LRTAP’,

European Journal of International Relations 8(1): 77–101.

Litﬁn, K.T. 1994. Ozone Discourses: Science and Politics in Global Environmental Cooperation. New York:

Columbia University Press.

Lorenz, D. 2010. ‘Possibilities of enacting and researching epistemic communities’, Sociological Research

Online 15(2).

Lorenz-Meyer, Dagmar. 2010. ‘Possibilities of enacting and researching epistemic communities’, Sociological

Research Online 15(2). Available at: http://www.socresonline.org.uk/15/2/13.html.

Lovell, H. and D. MacKenzie. 2011. ‘Accounting for carbon: The role of accounting professional organisations

in governing climate change’, Antipode 43(3): 704–30.

Meijerink, S. 2005. ‘Understanding

policy stability and

change. The interplay of advocacy coalitions and

epistemic communities, windows of opportunity, and Dutch coastal ﬂooding policy 1945–2003’, Journal

of European Public Policy 2(6): 1060–77.

Melo, M.A. 2004. ‘Institutional choice and the diﬀusion of policy paradigms: Brazil and the second wave

of pension reform’, International Political Science Review 25(3): 320–41.

Mitchell, N.J., H. Jenkins-Smith, K. Herron and G. Whitten. 2007. ‘Elite beliefs, epistemic communities

and the Atlantic divide: Scientists’ nuclear policy preferences in the United States and European Union’,

British Journal of Political Science 37: 753–64.

Niemann, A. 1998. ‘The PHARE programme and the concept of spillover: Neofunctionalism in the

making’, Journal of European Public Policy 5(3): 428–46.

Ogden, J., G. Walt and L. Lush. 2003. ‘The politics of “branding” in policy transfer: The case of DOTS

for tuberculosis control’, Social Science and Medicine 57(1): 179–88.

Dunlop

242

Peterson, J. 2001. ‘The choice for EU theorists: Establishing a common framework for analysis’, European

Journal of Political Theory 39(3): 289–318.

Peterson, M.J. 1992. ‘Whalers, cetologists, environmentalists and the international management of whaling’,

International Organization 46(1): 147–86.

Pierson, P. 1996. ‘The path to European integration: A historical institutionalist analysis’, Comparative

Political Studies 29(2): 123–63.

Prince, R. 2010. ‘Globalizing the creative industries concept: Travelling policy and transnational policy

communities’, Journal of Arts Management Law and Society 40(2): 119–39.

Radaelli, C.M. 1995. ‘The role of knowledge in the policy process’, Journal of European Public Policy 2(2): 159–83.

—— 1997. The Politics of Corporate Taxation in the EU: Knowledge and International Policy Agendas. London:

Routledge.

—— 2009. ‘Measuring policy learning: Regulatory impact assessment in Europe ’, Journal of European Public

Policy 16(8): 1145–64.

—— and K. O’Connor. 2009. ‘How bureaucratic elites imagine Europe: Towards convergence of governance

beliefs?’, Journal of European Public Policy 16(7): 971–89.

Rethel, L. 2010. ‘The new ﬁnancial development paradigm and Asian bond markets’, New Political Economy

15(4): 493–517.

Roth, C. 2008a. ‘Epistemic networks: Formalizing distributed cognition’, Sociologie Du Travail 50(3): 353–71.

—— 2008b. ‘Co-evolution of authors and concepts in epistemic networks the “zebraﬁsh” community’,

Revue

Francaise de Sociologie 49(3).

Sabatier,

P.A. 1988. ‘An advocacy coalition framework of policy change and the role of policy oriented

learning therein’, Policy Sciences 21: 129–68.

Schmidt, V.A. 2010. ‘Taking ideas and discourse seriously: Explaining change through discourse institutionalism

as the fourth “new institutionalism”’, European Political Science Review 2(1): 1–25.

Schot, J. and F. Schipper. 2011. ‘Experts and European transport integration, 1945–1958’, Journal of European

Public Policy 18(2): 274–93.

Sebenius, J.K. 1992. ‘Challenging conventional explanations of international co-operation: Negotiation

analysis and the case of epistemic communities’ in International Organisation 46(1): 323–65.

Shankland, A. and L. Hasenclever. 2011. ‘Indigenous peoples and the regulation of REDD plus in Brazil:

Beyond the war of the worlds?’, International Development Bulletin 42(3): 80–8.

Simek, J., L. Zamykalova and M. Mesanyova. 2010. ‘Ethics committee or community? Examining the

identity of Czech ethics committees in the period of transition’, Journal of Medical Ethics 36(9): 548–52.

Stephens, J.C., A. Hansson, Y. Liu, H. de Coninck and S. Vajjhala. 2011. ‘Characterizing the international

carbon capture and storage community’, Global Environmental Change-Human and Policy Dimensions 21(2):

379–90.

Thatcher, M. 1998. ‘The development of policy network analyses’, Journal of Theoretical Politics 10(4): 389–416.

Thomas, C.W. 1997. ‘Public management as interagency co-operation: Testing the epistemic community

theory at the domestic level’, Journal of Public Administration Research and Theory 7(2): 221–46.

Toke,

D. 1999. ‘Epistemic communities

and environmental group’,inPolitics 19(2): 97–102.

van Waarden, F. and M. Drahos. 2002. ‘Courts and (epistemic) communities in the convergence of

competition policies’, Journal of European Public Policy 9(6): 913–34.

Verdun, A. 1999. ‘The role of the Delors Committee and EMU’, Journal of European Public Policy 6(2): 308–29.

Walker, K.J. 2001. ‘Uncertainty, epistemic communities and public policy’, in J. Handmer, T. Norton and

S. Dovers. (eds) Ecology, Uncertainty and Policy. Englewood Cliﬀs, NJ: Prentice Hall.

Weible, C.M., P.A. Sabatier, H.C. Jenkins-Smith, D. Nohrstedt, A.D. Henry and P. deLeon. 2011. ‘A

quarter century of the advocacy coalition framework’, Policy Studies Journal 39(3): 349–60.

Weiss, C. 1979. ‘The many meanings of research utilization’, Public Administration 39: 426–31.

Wenger, E. 1999. Communities of Practice. New York: Cambridge University Press.

Wright, K. 1997. ‘Knowledge and expertise in European conventional arms control negotiations: an

epistemic community?’, The European Policy Process Occasional Papers, No. 41 Department of Government,

University of Essex.

Youde, J. 2007. AIDS, South Africa and the Politics of Knowledge. Aldershot: Ashgate.

Zelikow, P. 1994. ‘Foreign policy engineering: From theory to practice and back again’, International

Security 18(4): 143–71.

Zito, A.R. 2001. ‘Epistemic communities, European Union governance and the public voice’, Science and

Public Policy 28(6): 465–76.

Epistemic communities

243

Policy appraisal

John Turnpenny, Camilla Adelle and Andrew Jorda n

Introduction

For anyone interested in the role of knowledge in policy-making, these are fascinating times.

Terms like policy analysis, policy formulation, impact assessment and policy appraisal are fre-

quently used, and not just by academics: the media, government, non-governmental organisations,

interest groups and research funders are all visibly concerned with how evidence is collected,

marshalled, communicated, digested and used. One might be forgiven for using such terms inter-

changeably, or at least rather loosely, but they are quite distinct, and distinctive. This chapter is

speciﬁcally about the role of policy appraisal in knowledge collection, review and utilisation

processes.

There are several important reasons why policy appraisal should interest public policy scho-

lars. First, it belongs to a family of other, perhaps, more familiar concepts. We argue policy

appraisal is a very speciﬁc type of policy analysis, a term which covers the use of analytical methods –

formal or informal – in any part of policy-making from agenda-setting to implementation.

So-called ex ante policy appraisal – the focus of this chapter – speciﬁcally relates to policy for-

mulation: ‘how policy options are formulated within government’ (Howlett 2011: 18). While

part of the process of formulation involves a choice among instruments for implementing policy

objectives, like regulation, taxes and voluntary agreements (Bardach 2005; Howlett and Ramesh

2003), the process of formulation itself involves activities such as appraising knowledge or evi-

dence, engaging in dialogue about the nature of policy problems and solutions, and identifying

and assessing the impacts of diﬀerent policy options (Howlett 2011).

Second, policy appraisal is a symptom of broader changes in policy-making which began in

the latter part of the twentieth century. Many jurisdictions have since embarked on reform

programmes intended to better structure and manage their regulatory processes (Allio 2007).

This interest in ‘Better Regulation’ was ignited in part by the need to cope with the increased

competition created by more globalised markets, as well as the international diﬀusion of the

Organisation for Economic Co-operation and Development’s (OECD) core principles of regulatory

reform (ibid.). Radaelli (2007: 191) describes Better Regulation as a type of meta-governance

because of ‘its emphasis on standards and rules which, instead of governing speciﬁc sectors or

economic actors, steer the process of rule formulation, adoption, enforcement, and evaluation’.

244

As we will see below, policy appraisal has become an important enabler of Better Regulation

among other things, spreading rapidly throughout the OECD and beyond.

Third, having now secured this role, appraisal processes have themselves become important

new sites of political behaviour (Turnpenny et al. 2009). Thus, investigating how policy appraisal

systems are structured and operate in practice provides another way to study governance via its

policy instruments (see Hood 2007 and Lascoumes and Le Galès 2007 among others). While

the idea of policy appraisal owes a theoretical debt to the founding fathers of policy analysis

(e.g. Lasswell 1947; Simon 1976), paradoxically, contemporary policy analysts have been slow

to explore appraisal’s policy and political consequences. Appraisals are contested sources of

legitimacy, accountability and normative justiﬁcation for public action. Accordingly, they inform

us of the theoretical presuppositions and legitimacy standards embedded in public policy (Lascoumes

and Le Galès 2007). Investigating the use of appraisal can also improve our understanding of the

administrative capacity and eﬀectiveness of public policy (Chaker et al. 2006; Hahn and Tetlock

2008). There is great potential to inform political science discussion on capacity and imple-

mentation. Finally, as a highly visible form of evidence use, appraisal oﬀers a fresh angle on some

rather old debates in public policy such as decision-making (Russel and Jordan 2009), policy

coordination (Schout and Jordan 2007), and the development and (non)-utilisation of evidence

in politics (Dowie 1996; Pawson 2002; Sanderson 2006). In fact, in this chapter we argue that

policy appraisal is a recent and more institutionalised manifestation of policy formulation activities

which have always been carried out in various forms.

This chapter provides an introduction to the principles and dynamic practices of policy apprai-

sal. We ﬁrst examine diﬀering ideas about what appraisal is, or could be, starting from the most

basic deﬁnitions and how these have evolved over time as ever more ambitious expectations

have emerged as to what it can deliver in practice. We then discuss how policy appraisal has been

working in practice in diﬀerent contexts – particularly how it has become institutionalised in

many jurisdictions since the mid-1990s. These observations illustrate well the politicised nature

of knowledge use in policy-making and the obstacles to its institutionalisation. We then show

how several distinct types of research on these subjects have developed, sometimes in parallel

with practices, sometimes apart from them. The literatures on appraisal are advancing rapidly, but

we show that they often take us back to many fundamental debates in policy analysis, namely between

proponents of fairly linear models of knowledge transfer between experts and policy-makers,

and those adopting more ‘post-positivist’ positions. In the ﬁnal section, we assess the relationship

between research and practice across diﬀerent types of policy appraisal research, and question

some of the basic assumptions made about appraisal in the light of theory, research and practice.

Finally, we speculate on the possible future direction of policy appraisal research and practice.

What is policy appraisal?

Deﬁnitions of policy appraisal abound. One of the most widely cited is: ‘[that] family of ex ante

techniques and procedures. … that seek to inform decision makers by predicting and evaluating

the consequences of various activities according to certain conventions’ (Owens et al. 2004:

1944). Another, less well-known deﬁnition, is ‘a test or judgment of some policy, with the aim

to inform the decision makers on the suitability, desirability, eﬀectiveness or eﬃciency of it” (de

Ridder 2006: 21). Regardless of the deﬁnition, we are arguably living in “an age of assessment”

(Rayner 2003: 163). The assessment or ‘appraisal’ of projects (e.g. through Environmental Impact

Assessment—EIA) has been routinely undertaken in many countries since the 1970s (Jay et al.

2007). This was followed by attempts to institute appraisal at the more strategic level of plans

and programmes (e.g. Strategic Environmental Assessment—SEA) (Bina 2007).

Policy appraisal

245

Put simply, policy appraisal is appraisal conducted at the policy level. It attempts to formalise

the decision-making process in a series of steps to be undertaken when developing a policy.

The end product is a report which describes the results of each step. The steps vary depending

on the jurisdiction applying the system (see the section ‘How is policy appraisal practised?’), but

at a general level they usually include: identifying the problem to be addressed by the proposed

policy; deﬁning the objectives of the proposed policy; identifying the diﬀerent policy options to

pursue these objectives; analysing the potential impacts of each option; comparing the options

by weighing up the negative and positive impacts for each; and setting out plans for monitoring

and evaluating the policy once it is implemented.

These are very similar to the standard steps in EIA and SEA (e.g. Barrow 1997). However,

the motives for appraising have varied enormously across jurisdictions and policy ﬁelds, from

environmental protection through to reducing regulatory burdens and promoting a neo-liberal

economic agenda (Kirkpatrick and Parker 2007; Jacob et al. 2007; Hertin et al. 2009b; OECD

2008). Diﬀerent subtypes have emerged such as Regulatory Impact Assessment (RIA) (e.g. Radaelli

2004b) and Sustainability Impact Assessment (SIA) (e.g. Kirkpatrick and Lee 2001), each with a

slightly diﬀerent conception of the basic steps outlined above. Appraisal systems in turn harness

a wide range of policy appraisal tools such as cost-beneﬁt analysis, scenario analysis and com-

puter modelling (Carley 1980; de Ridder et al. 2007; Nilsson et al. 2008). Until now, we have

mostly been concerned with textbook deﬁnitions and understanding of appraisal. In the next

section we turn to look at the very varied ways that policy appraisal has actually been practised

over the past 20 years or so.

How is policy appraisal practised?

Precursory forms of policy appraisal, which focused mainly on the assessment of economic and

administrative impacts of regulation, were introduced as early as 1966 (Denmark) and 1971

(USA) (OECD 2009). However, the practice spread only slowly at ﬁrst, with Finland and

Canada following in the 1970s and Australia, the UK, the Netherlands and Germany adopting

some form of policy appraisal in the mid-1980s. There was, however, a rapid rise in adoption in

the second half of the 1990s, following OECD recommendations on regulatory reform (OECD

1995), and again in the mid-2000s following the launch of the European Commission’s Impact

Assessment system in 2003 (European Commission 2002). ‘Better Regulation’ has ranked high

on the European Union’s (EU) agenda for well over a decade now (Allio 2007), and the EU,

and especially the European Commission, was a highly visible adopter of policy appraisal, as part

of a number of measures designed to improve and streamline its decision-making processes in

order ‘to improve the quality and coherence of the policy development process’ (European

Commission 2002: 2).

Many of the early policy appraisal systems were subsequently revised to include a wider range

of impacts. For example, the UK system introduced in 1986 as a Compliance Cost Assessment was

replaced in 1996 by a broader system of regulatory appraisal, but it did not appear in anything

near its current form of RIA until 1998. All 31 OCED countries had by 2008 either adopted,

or were in the process of adopting, policy appraisal (OECD 2009). A recent survey of 17 European

countries found that all of them had adopted some form of policy appraisal, although some coun-

tries were still in the early stages of implementation (Adelle et al. 2010). The most recent countries

to adopt policy appraisal have tended to be EU member states in Central and Eastern Europe

(CEE) such as Estonia and Lithuania (ibid.; De Francesco 2010), although there are some geo-

graphical exceptions such as Ireland which only introduced its system in 2005 (Adelle et al.

2010). There are also early CEE adopters like Hungary (1994) (see Staronova 2010). Recent

Turnpenny et al.

246

adoption appears to be one of the predictors of weak implementation of policy appraisal, with

many of the more sophisticated and robust systems being found in the older EU member states

(Adelle et al. 2010). In addition, policy appraisal is beginning to spread beyond the OECD and

EU (De Francesco 2010) and become a ‘global norm’ (Jacobs 2006). Interest in policy appraisal

in middle- and lower-income countries is increasing, albeit from a relatively low base (Kirkpatrick

and Parker 2004; Kirkpatrick et al. 2004).

This widespread diﬀusion of policy appraisal systems has not, however, led to a standard appraisal

approach. Radaelli (2005: 924) argues that it is an idea which has ‘travelled lightly’ around the

world producing ‘diﬀusion without convergence’. He claims that this has led to the introduc-

tion of a policy appraisal ‘bottle’ which contains many diﬀerent ‘wines’. Essentially most policy

appraisal systems (the ‘bottle’) draw on certain common elements: they are often – but not always –

supported by a legislative act making their application mandatory; they consist of similar pro-

cedural steps set out in oﬃcial ‘guidance’ documents (see above); they are undertaken by the

oﬃcial responsible for policy development; and they result in a written document, which may

or may not be made public via a government website.

However, these common features disguise the many diﬀerent ways policy appraisal is imple-

mented in practice (the ‘wine’). For example, in some cases, policy appraisal only exists on paper as

a ‘tick-box’ exercise (Radaelli 2005). Radaelli attempted to account for this variety by exam-

ining how institutional and political context matters in the process of diﬀusion. A number of

comparative surveys of policy appraisal practices around Europe have indeed revealed vast diversity

in the practice of policy appraisal. They have uncovered diﬀ erent institutional frameworks,

objectives, uses of policy appraisal tools, quality of reports, as well as the eﬀects of cultures and

the place of appraisal within policy processes (e.g. Jacob et al. 2008; Adelle et al. 2010; Hertin et al.

2009a). These and other insights into the everyday practices of policy appraisal have emerged

from a rapidly developing academic literature, which we summarise in the next section.

Policy appraisal research

Policy appraisal research falls into diﬀerent broad types. Some is strongly (although often implicitly)

premised on the positivist belief that more ‘rational’ policy-making can be achieved by applying

analytical tools. The idea that appraisal exists to bring in evidence to counter interest-based policy-

making, to integrate cross-cutting issues, and to increase cooperation between diﬀerent depart-

ments is widespread, not least in the guideline documents prepared for government oﬃcials. It

is based on a ‘ rational model’ of linear knowledge transfer between experts and policy-makers.

Other research draws on a post-positivist epistemology, and emphasises the importance of

revealing, understanding and working with normative values in studying the relationship between

evidence and policy-making. Such an approach fundamentally challenges the underlying

assumption that policy appraisal is purely about informing (and thus ‘improving’) policy, and

hence challenges the straightforward deﬁnition of what policy appraisal is – is it a tool, a method,

an instrument of public policy (e.g. Radaelli 2008: 6) or an administrative coordination mechan-

ism (Jordan and Schout 2006) for example? Policy appraisal can also, with varying degrees of

explicitness, be used to delay or dilute unwanted regulation, or to exert political control on

government departments and agencies.

Positivist approaches

Turnpenny et al. (2009) tried to make sense of these advances by subdividing the literature into

four main types. The ﬁrst two may be thought of as broadly positivist, while the third and

Policy appraisal

247

fourth are more post-positivist in their orienation. The ﬁrst type concerns the design of appraisal

systems themselves. This often highly technical literature is rooted in the environmental man-

agement, economics and policy analysis communities. It addresses the design of speciﬁc tools

and methods for policy appraisal such as Cost Beneﬁt Analysis, the so-called Standard Cost

Model, and more general tools such as computer models and scenarios (e.g. Quade 1989; Dunn

2004). It also includes attempts to inform the design of appraisal systems (e.g. Lee 2006; OECD

2008), and often takes the form of ‘how to’ textbooks (e.g. Barrow 1997) and handbooks (e.g.

European Commission 2009), for which there is a steady demand amongst the practitioners

who perform appraisals. For useful overall reviews of this type of literature, see Eales et al. 2005;

Ness et al. 2007; de Ridder et al. 2007 and Tamborra 2002.

The second type of research concerns the assessment of the operation of policy appraisal designs ‘in

practice’, in individual jurisdictions and internationally (e.g. Lee and Kirkpatrick 2004; Renda

2006; Jacob et al. 2007; Hertin et al. 2009b; EEAC 2006). Initial attempts (e.g. Harrington et al.

2000) measured quality by comparing the contents of appraisal reports with oﬃcial guidance.

Subsequent research extended this approach somewhat to emphasise aspects of the process of

appraisal using a more in-depth case study approach, often including interviews with those

involved (e.g. TEP 2007). This research has generated a fairly consistent picture of the empirical

‘reality’ of appraisal, namely that: there is a gap between the aims of appraisal and its imple-

mentation; the economic aspects of policy all too easily crowd out other (e.g. social and environ-

mental) aspects; appraisals tend to be performed at a relatively late stage in the policy process

and consequently have little or no inﬂuence over the ﬁnal decisions made; consultation is often

limited to the ‘usual suspects’ who have participated before or who have large resources; and

formal appraisal tools such as computer modelling are patchily used despite the strong political

invocation to use them. Most studies identify recommendations to policy-makers on how to

improve the quality of their appraisal systems. Many such recommendations focus on micro-

level constraints, such as calls for more training and guidance (EEAC 2006; DBR 2004; Jacobs

2006; NAO 2007; TEP 2007; Wilkinson et al. 2004). Another common recommendation is to

start the policy appraisal earlier in the policy process, when more options are likely to be open

(Renda 2006; TEP 2007; Wilkinson et al. 2004). Researchers such as Turnpenny et al. (2008) also

emphasised the need to address higher-level constraints through, for example: stronger political

leadership (Jacob et al. 2008; Russel and Jordan 2007); the creation of oversight and quality assur-

ance mechanisms (DBR 2004; TEP 2007; Torriti 2007; Wilkinson et al. 2004); and a greater

understanding, acceptance and use of appraisal tools (Jacob et al. 2008; Nilsson et al. 2008; de Ridder

et al. 2007; Turnpenny et al. 2008).

Post-positivist approaches

Research of the third and fourth types adopts a more post-positivist view. In the third type,

analysts have tried to go beyond calling for micro-level changes such as better training and political

support, and explored the wider political and institutional context in which appraisal systems

operate. Policy appraisal is, as all evidence use in policy-making is, a political activity (see, for

example, Hall (1989) on evidence generally, Richardson (2005) on EIA, and Turnpenny et al.

(2008) on policy appraisal). Such research searches for evidence that appraisal has led to policy

change via processes of learning, and attempts to understand what is aﬀecting such processes,

such as the constraining eﬀect of the institutional context in which appraisal takes place (Jacob et al.

2008; Russel and Jordan 2009; Thiel 2008; Turnpenny et al. 2008). For example, Thiel (2008)

revealed how appraisal in the European Commission is shaped by departmental cultures and

prior experiences of appraisal. There are a relatively small number of studies employing this type

Turnpenny et al.

248

of research (e.g. Cashmore et al. 2008; Hertin et al. 2009a; Hertin et al. 2009b; Jacob et al. 2008;

Nilsson 2006; Thiel and König 2008; Torriti 2007; Turnpenny et al. 2008). A distinction is

often made between single-loop (or instrumental) learning where ‘knowledge directly informs

concrete decisions by providing speciﬁc information on the design of policies’ (Hertin et al.

2009a: 1187), and double-loop (or conceptual) learning where ‘knowledge “enlightens” policy-

makers by slowly feeding new information, ideas and perspectives into the policy system’ (ibid.).

Clearly the former is more consistent with a positivist epistemology, whereas the latter is more

consistent with a post-positivist one. Much of this literature explicitly endorses more deliberative

approaches (e.g. Bond and Morrison-Saunders 2011; Cashmore et al. 2008; Owens et al. 2004).

However, a positivist epistemology continues to provide the foundation for most methodolo-

gical development (Bond and Morrison-Saunders 2011; Hertin et al. 2009a, 2009b). As a con-

sequence, while policy appraisal could potentially provide a new venue for conceptual learning,

it appears that this rarely occurs in practice (Hertin et al. 2009a: 1196).

The fourth type of research investigates the underlying motivations for appraising in the ﬁrst

place, such as to facilitate political control of bureaucracy (e.g. Radaelli and Meuwese 2010;

Radaelli 2008), following perceived ‘best practice’ in other states (e.g. Radaelli 2005), and

attempts to ‘modernise’ the state (Hood and Peters 2004; Radaelli and Meuwese 2010). Studies

of the diﬀusion of policy appraisal practices across and within jurisdictions have emphasised how

diﬀerent actors shape appraisal structures and practices to suit their preferences (Radaelli 2004a,

2004b, 2005). Research has also focused on the intended and unintended consequences of policy

appraisal. It treats appraisal as ‘a good lens on the changing nature of the regulatory state in the EU

and its member states’ (Radaelli and Meuwese 2009: 651). In such accounts, the claim is that

politically contested questions (such as who is in charge of the law-making process?) are transferred

into more technocratic venues such as policy appraisal processes, where they can be more easily

resolved (Radaelli and Meuwese 2010: 143). Appraisal thus becomes a way to unite otherwise

incompatible political visions; in many respects, a form of politics by more technocratic means.

Policy appraisal research-practice relationships

There have been developments in both the research on, and practice of, policy appraisal, but to what

extent have research and practice actually informed each other? Susan Owens and her colleagues

(2004) expected the practices of, and the research on, appraisal to co-evolve in a steadily more

reﬂexive, or mutually informing, direction. But while there are signs that policy appraisal research is

moving away from its positivist roots, both research and practice remain heavily informed by it.

According to Adelle et al. (2012), research on the design of appraisal systems continues apace.

Consequently, there are now myriad policy appraisal designs and tools that can be consulted.

However, the practices of policy appraisal appear to be informing such research only in a rather

general sense, namely through demands from policy-makers for policy appraisal tools, methods

and process designs. Consequently, there remains a gap between ‘sophisticated tool knowledge

and pragmatic policy-making’ (de Ridder et al. 2007: 436). Thus, while both research and practice

still appear to be heavily informed by the positivist model, it is questionable if innovation in

both is proceeding in a mutually informative manner.

Regarding research on the assessment of the operation of policy appraisal, while at ﬁrst there was little

evidence that such research was producing changes in policy appraisal practices (e.g. Cecot et al.

2007; Renda 2006), the picture is now changing (e.g. European Court of Auditors 2010). There is

some evidence that as a direct consequence of research, policy appraisal systems in some jur-

isdictions have been revised to improve the impact of appraisals on political decision-making,

for example by setting up dedicated bodies to review the quality of appraisal and/or moving

Policy appraisal

249

responsibility for appraisal from specialised departments to the heart of government (Jacob et al.

2008: 11). However, such changes have not (yet) amounted to a serious reconsideration of the

theoretical underpinnings of appraisal practices or research. In mainly pursuing change at the

micro and occasionally meso levels, they essentially remain rooted in a positivist epistemology. At

present practitioners seem content to commission yet more research on the quality of the policy

appraisal systems in their jurisdictions (e.g. European Court of Auditors 2010; NAO 2009),

without really investigating some of the underlying contextual issues.

There is still relatively little post-positivist work. Perhaps this is not surprising considering the

diﬃculties associated with studying learning processes over long time periods (Owens et al. 2004;

Radaelli 2009). Nonetheless, if research and practices were proceeding in parallel we would

expect to ﬁnd more evidence of policy appraisal practices being informed by post-positivistic

recommendations, such as more deliberative approaches. There is currently little evidence of

such ‘innovative’ approaches within policy appraisal practice or research.

The role of appraisal in policy analysis

Another way to understand the current place of policy appraisal is to view it against the back-

drop of policy analysis more generally. Doing this reveals that appraisal raises some very old

questions. Modern policy analysis is usually held to date from the late 1940s, particularly forged

by the work of Lasswell, Simon and Lindblom. But each of these had very diﬀerent ideas about the

nature and goals of policy, and hence disagreed about the form and function of policy analysis.

It is possible to divide the history of policy analysis into three broad ‘waves’. The ﬁrst wave,

from roughly the late 1940s to the late 1970s, corresponded to widespread use, and implicit

faith in, rationally based appraisal tools like cost-beneﬁt analysis, operations research and assessment

(Rossi et al. 2004; Wollmann 2007). But the limits to such tools soon became evident, and the

promise of analytically based answers to every policy question soon faded.

The ‘second wave’ (Wollmann 2007) began as limits on the ability to predict every even-

tuality became clearer, and the politics surrounding use of analysis became more intense (Weiss

1979; Palumbo 1987). During the early 1980s, an ‘anti-analytical’ mood of government coupled

with budgetary retrenchment pressure, particularly in the USA and the UK, resulted in a scaling

back of the policy units who once espoused analytical methods. But since the mid-1990s, there

has been a revival in the use of policy analytical tools in practical policy-making, including policy

appraisal. This shift corresponded with a drive to more ‘evidence-based’ approach to policy-

making in an era of New Public Management (e.g. Radaelli 1995; Dowie 1996; Pawson 2002;

Sanderson 2006; Nutley et al. 2007). Critically – and somewhat in tension with this ‘third wave’ of

use – there has been a blossoming of more sophisticated research on the discursive, argumentative

and political aspects of policy analysis (e.g. Majone 1989; Fischer 1995; John 1998). The tone of

this research has changed from ‘discover and predict’ to a recognition of the ‘power of language,

persuasion and argument’ ( John 1988: 154) – the importance of revealing, understanding and

working with normative values expressed in a ‘post-positivist epistemology’ (Fischer 1995). How-

ever, as we have seen, this ‘post-positivist turn’ has made little impact so far on policy appraisal

research or practice. Policy appraisal is thus both a symptom of these long-term trends in policy

analysis, and a site of everyday politics around the use of knowledge in the policy process.

Conclusion

In this chapter we have shown that policy appraisal is an important aspect of policy analysis

which has spread rapidly and secured a high proﬁle. It now appears in many diﬀerent guises,

Turnpenny et al.

250

each with diﬀerent aims. It is variously: a method or procedure for analysis; a policy instrument;

a site of political behaviour; and a technique for policy harmonisation across states. As it enjoys

its current and relatively rapid phase of proliferation, research on policy appraisal is moving

towards the mainstream of public policy research, on several diﬀerent fronts. The broadly

‘positivist’ and broadly ‘post-positivist’ approaches to understanding policy appraisal reﬂect wider

debates within policy analysis and evidence use. But while policy appraisal has emerged as a

signiﬁcant part of the policy-making process in the last ten years or so, research and practice are

only weakly informed by post-positivist research developments elsewhere (Adelle et al. 2012).

This is partly because of the lack of a ready audience for such work. Many policy appraisal practi-

tioners have made a signiﬁ cant political and/or resource commitment to forms of policy appraisal

informed by the ‘rational’ model which they are reluctant to surrender. In part it is also due to

the nature of the post-positivist research, which questions the very purpose of doing appraisal.

What might this mean for the future of policy appraisal? In the second decade of the twenty-

ﬁrst century, perhaps another policy analysis ‘wave’ is breaking. The tension between the broadly

positivist ‘evidence-based policy’ underpinning of policy appraisal in practice, and the post-

positivist challenges to it, echo the transition from the ﬁrst to the second wave in the mid-late

1970s. Challenges to the evidence-based policy-making discourses of the early 2000s in the light

of economic crises and ideological retrenchment may have signiﬁcant implications for the future

direction of policy appraisal. Its continued rise, coupled with more political pressure, oﬀers many

opportunities for research which examines underlying motivations for appraisal, and methods

for deducing this. Research could lead to a fuller understanding of when, why and how appraisal

practices depart from the positivist model, depending on the speciﬁc context(s) in which

appraisal is deployed. Work which seeks to mediate between the positivist and the post-positivist

approaches will also play a more signiﬁcant role. Policy appraisal researchers are well placed to

shape new developments. These are indeed fascinating times.

Bibliography

Adelle, C., A. Jordan, J. Turnpenny, S. Bartke, T. Bournaris, K. Jacob, H. Kuittinen, K. Mäkinen,

C. Moulogianni, T. Nõmmann, L.E. Larsen, K. Peterson, M. Perez Soba, L. Roupioz, B. Sánchez,

D. Wascher, and S. Weiland. 2010. ‘A summary of user needs and expectations with regards to impact

assessment’. Unpublished LIAISE project report prepared under contract from the European Commission.

Contract No. 243826.

——, A. Jordan and J.R. Turnpenny. 2012. ‘Proceeding in parallel or drifting apart? A systematic review

of policy appraisal research and practices’, Environment and Planning C 30(3): 400–414.

Allio, L. 2007. ‘Better regulation and impact assessment in the European Commission’, in C. Kirkpatrick

and D. Parker (eds) Regulatory Impact Assessment: Towards Better Regulation? Cheltenham: Edward Elgar.

Bardach, E. 2005. A Practical Guide for Policy Analysis: The Eightfold Path to More Eﬀective Problem-Solving.

Washington, DC: CQ Press.

Barrow, C.J. 1997. Environmental and Social Impact Assessment. London: Arnold.

Bina, O. 2007. ‘A critical review of the dominant lines of argumentation on the need for strategic environmental

assessment’, EIA Review 27: 585–606.

Bond, A. and A. Morrison-Saunders. 2011. ‘Re-evaluating sustainability assessment: Aligning the vision

and the practice’, EIA Review 31(1): 1–7.

Carley, M. 1980. Rational Techniques in Policy Analysis. London: Heinemann.

Cashmore, M. 2004. ‘The role of science in Environmental Impact Assessment: Process and procedure

versus purpose in the development of theory’, EIA Review 24: 403 –26.

——, A. Bond and D. Cobb. 2008. ‘The role and functioning of environmental assessment: Theoretical

reﬂections upon an empirical investigation of causation’, Journal of Environmental Management 88:

1233–48.

Cecot, C., R.W. Hahn and A. Renda. 2007. ‘A statistical analysis of the quality of impact assessment in the

European Union’, AEI- Brookings Joint Center for Regulatory Studies. Working Paper 07-09.

Policy appraisal

251

Chaker, A., K. El-Fadl, L. Chamas and B. Hatjian. 2006. ‘ A review of strategic environmental assessment

in 12 selected countries’, EIA Review 26(1): 15–56.

DBR (EU Directors of Better Regulation Group). 2004. ‘A comparative analysis of regulatory impact assessment

in ten EU countries ’, report prepared for the EU DBR Group, Dublin, May 2004 (accessed 15 August

2011). Available at: http://www.betterregulation.ie/eng/Publications/A_COMPARATIVE_ANALYSI

S_OF_REGULATORY_IMPACT_ASSESSMENT_IN_TEN_EU_COUNTRIES.html (accessed 15

August 2011).

De Francesco, F. 2010. ‘The diﬀusion process of regulatory impact analysis in EU and OECD member

states’. PhD thesis, University of Exeter.

de Ridder, W. 2006. Tool Use in Integrated Assessment: Integration and Synthesis Report. Bilthoven: Netherlands

Environmental Assessment Agency.

——, J. Turnpenny, M. Nilsson and A. von Raggamby. 2007. ‘A framework for tool selection and use in

integrated assessment for sustainable development’, Journal of Environmental Assessment Policy and Management

9(4): 423–41.

Dowie, J. 1996. ‘“Evidence-based”, “cost-eﬀective” and “preference-driven” medicine: decision analysis based

medical decision making is the pre-requisite’, Journal of Health Services Research and Policy 1(2): 104–13.

Dunn, W.N. 2004. Public Policy Analysis – An Introduction, 3rd edn. Mahwah, NJ: Pearson Prentice Hall.

Eales, R., S. Smith, C. Twigger-Ross, W. Sheate, E. Özdemiroglu, C. Fry, P. Tomlinson and C. Foan.

2005. ‘Emerging approaches to integrated appraisal in the UK’, Impact Assessment and Project Appraisal

23(2): 113–23.

EEAC (European Environment and Sustainable Development Advisory Councils). 2006. ‘Impact assessment

of European Commission policies: Achievements and prospects’, statement of the EEAC Working Group

on Governance, April.

European Commission. 2002. ‘Impact assessment’ (COM (2002) 276), Commission of the European

Communities, Brussels.

—— 2009. ‘Impact assessment guidance’ (SEC(2009)92), Commission of the European Communities, Brussels.

European Court of Auditors. 2010. ‘Impact assessments in the EU institutions: Do they support decision-making?’

Special Report No. 3, European Court of Auditors, Luxembourg.

Fischer,

F. 1995 Evaluating Public

Policy. Boulder, CO: Wadsworth.

Hahn, R.W. and P.C. Tetlock. 2008. ‘Has economic analysis improved regulatory decisions?’, Journal of

Economic Perspectives 22(1): 67–84.

Hall, P.A. (ed.). 1989. The Political Power of Economic Ideas. Princeton, NJ: Princeton University Press.

Harrington, W., R.D. Morgenstern and P. Nelson. 2000. ‘On the accuracy of regulatory cost estimates’,

Journal of Policy Analysis and Management 19: 297–322.

Hertin, J., J. Turnpenny, A. Jordan, M. Nilsson, D. Russel and B. Nykvist. 2009a. ‘Rationalising the

policy mess? Ex ante assessment and the utilisation of knowledge in the policy process’, Environment and

Planning A 41: 1185–1200

——, K. Jacob, U. Pesch and C. Pacchi. 2009b. ‘The production and use of knowledge in regulatory

impact assessment: An empirical analysis’, Forest Policy and Economics 11(5–6): 413–21.

Hood, C. 2007. ‘Intellectual obsolescence and intellectual makeovers: Reﬂections on the tools of government

after two decades’, Governance 20(1): 127–44.

——and B.G. Peters. 2004. ‘The middle aging of New Public Management: Into the age of paradox?’,

Journal of Public Administration Research and Theory 14: 267–82

Howlett, M. 2011. Designing Public Policies: Principles and Instruments. London: Routledge.

——and M. Ramesh. 2003. Studying Public Policy: Policy Cycles and Policy Subsystems. Oxford: Oxford

University Press.

Jacob, K., J. Hertin and A. Volkery. 2007. ‘Considering environmental aspects in integrated Impact

Assessment: Lessons learned and challenges ahead’, in C. George and C. Kirkpatrick (eds) Impact

Assessment for a New Europe and Beyond. Cheltenham: Edward Elgar.

——, J. Hertin, P. Hjerp, C. Radaelli, A. Meuwese, O. Wolf, C. Pacchi, and K. Rennings. 2008. ‘Improving

the practice of impact assessment’, EVIA (Evaluating Integrated Impact Assessments) project report.

Project No. 028889, European Commission Sixth Framework Programme.

Jacobs, S.H. 2006. ‘Regulatory impact analysis in regulatory process, method, and co-operation: Lessons

for Canada from international trends’.

Policy Horizons Canada Working paper 026. Government of

Canada, Ottawa.

Jay, S.,

C. Jones, P. Slinn and C. Wood. 2007. ‘Environmental impact assessment: Retrospect and prospect’,

EIA Review 27: 287–300.

Turnpenny et al.

252

John, P. 1998. Analysing Public Policy. London: Continuum.

Jordan, A. and A. Schout. 2006. The Coordination of the European Union. Oxford: Oxford University Press.

Kirkpatrick, C. and N. Lee. 2001. ‘Methodologies for sustainability impact assessments of proposals for

new trade agreements’, Journal of Environmental Assessment, Policy and Management 3(3): 395–412.

——and D. Parker (eds). 2007. Regulatory Impact Assessment: Towards Better Regulation? Cheltenham:

Edward Elgar Publishing.

——and D. Parker. 2004. ‘Regulatory impact assessment and regulatory governance in developing countries’,

Public Administration and Development 24(4): 333–44.

——D. Parker and Y.-F. Zhang. 2004. ‘Regulatory impact assessment in developing and transition

economies: A survey of current practice’, Public Money and Management 24(5): 291–6.

Lascoumes, P. and P. Le Galès. 2007. ‘Introduction: Understanding public policy through its instruments:

From the nature of instruments to the sociology of public policy instrumentation’, Governance 20(1): 1–21.

Lasswell, H.D. 1947. The Analysis of Political Behavior: An Empirical Approach. London: Routledge.

Lee, N. 2006. ‘ Bridging the gap between theory and practice in integrated assessment’, EIA Review 26: 57–78.

——and C. Kirkpatrick. 2004. ‘A pilot study on the quality of European Commission extended impact

assessment’. Draft For Consultation. Impact Assessment Research Centre, Institute for Development

Policy and Management, University of Manchester, 21 June.

Majone, G. 1989. Evidence, Argument and Persuasion in the Policy Process. New Haven, CT: Yale University

Press.

NAO (National Audit Oﬃce). 2007. Evaluation of Regulatory Impact Assessments 2006–07. London: The

Stationery Oﬃce.

—— 2009. Delivering High Quality Impact Assessments. London: The Stationery Oﬃce.

Ness, B., E. Urbel-Piirsalu, S. Anderberg and L. Olsson. 2007. ‘Categorising tools for sustainability assessment’,

Ecological Economics 60: 498–508.

Nilsson, M. 2006. ‘The

role of assessments

and institutions for policy learning: A study on Swedish climate

and nuclear policy formation’, Policy Sciences 38: 225–49.

——, A. Jordan, J. Turnpenny, J. Hertin, B. Nykvist and D. Russel. 2008. ‘The use and non-use of policy

appraisal in public policymaking: An analysis of three European countries and the European Union’,

Policy Sciences 41(4): 335–55.

Nutley, S.M., I. Walter and H.T.O. Davies. 2007. Using Evidence: How Research Can Inform Public Services.

Bristol: The Policy Press.

OECD (Organisation for Economic Co-operation and Development). 1995. Recommendation of the Council

of the OECD on Improving the Quality of Government Regulation. Paris: OECD.

—— 2008. Building an Institutional Framework for Regulatory Impact Assessment: Guidance for Policy-makers.

Paris: OECD Publications.

—— 2009. Indicators of Regulatory Management Systems. Paris: OECD Publications.

Owens, S., T. Rayner and O. Bina. 2004. ‘New agendas for appraisal: Reﬂections on theory, practice and

research’, Environment and Planning A 36: 1943–59.

Palumbo, D.J. 1987. ‘Politics and evaluation’, in D.J. Palumbo (ed.) The Politics of Program Evaluation.

Newbury Park, CA: Sage, 12–46.

Pawson, R. 2002. ‘Evidence-based policy: In search of a method’, Evaluation 8(2): 157–81.

Quade, E.S. 1989. Analysis for Public Decisions , 3rd edn. Mahwah, NJ: Prentice Hall.

Radaelli, C. 1995. ‘The role of knowledge in the policy process’, Journal of European Public Policy 2: 159–83

—— 2004a. ‘Getting to grips with quality in the diﬀusion of regulatory impact assessment in Europe’,

Public Money and Management 5: 271–6.

—— 2004b. ‘The diﬀusion of regulatory impact analysis: Best practice or lesson-drawing’, European Journal

Political Research 43: 723–47.

—— 2005. ‘Diﬀusion

without convergence: How political context shapes the adoption of regulatory

impact assessment’, Journal of European Public Policy 12: 924–43.

—— 2007. ‘Whither better regulation for the Lisbon agenda?’, Journal of European Public Policy 14(2): 190–207.

—— 2008. ‘Evidence-based policy and political control: What does Regulatory Impact Assessment tell us?’,

paper for the European Consortium for Political Research Joint Sessions, workshop on ‘The politics of

evidence-based policy-making’, Rennes, 11–16 April.

—— 2009. ‘Measuring policy learning: Regulatory impact assessment in Europe ’, Journal of European Public

Policy 16: 1145–64.

——and A.C.M. Meuwese. 2009. ‘Better regulation in Europe: Between public management and regulatory

reform’, Public Administration 87: 639–54.

Policy appraisal

253

——and A.C.M. Meuwese. 2010. ‘Hard questions, hard solutions: Proceduralisation through impact

assessment in the EU’, West European Politics 33: 136–53.

Rayner, S. 2003. ‘Democracy in the age of assessment: Reﬂections on the roles of expertise and democracy in

the public sector decision making’, Science and Public Policy 3: 163–70.

Renda, A. 2006. Impact Assessment in the EU: The State of the Art and the Art of the State. Brussels: Centre for

Policy Studies.

Richardson, T. 2005. ‘Environmental assessment and planning theory: Four short stories about power,

multiple rationality and ethics’, EIA Review 25: 341–65.

Rossi, P.H., M.W, Lipsey and H.E. Freeman. 2004. Evaluation: A Systematic Approach, 7th edn. Thousand

Oaks, CA: Sage.

Russel, D. and A. Jordan. 2007. ‘Gearing-up governance for sustainable development: Patterns of policy

appraisal in UK central government’, Journal of Environmental Planning and Management 50: 1–21.

——and A.J. Jordan. 2009. ‘Joining up or pulling apart? The use of appraisal to coordinate policy-making

for sustainable development’, Environment and Planning A 41(5): 1201–16.

Sanderson, I. 2006. ‘Complexity, “practical rationality” and evidence-based policy making’, Policy and

Politics 34(1): 115–32.

Schout, A. and A. Jordan. 2007. ‘What do “new” modes of governance demand of public administrations?’,

paper presented at the Connex Workshop on Governing the European Union: Policy instruments in a

multi-level polity, Paris, 21 and 22 June. Sciences Po—Centre d’études européennes.

Simon, H.A. 1976. Administrative Behavior: A Study of Decision-making Processes in Administrative Organization,

3rd edn. London: Collier Macmillan.

Staronova, K. 2010. ‘Regulatory impact assessment: Formal institutionalization and practice’, Journal of

Public Policy 30(1):

117–36.

Tamborra, M.

2002. Tools for Sustainability Impact Assessment of EU Policies: The Contribution of the EU Socio-economic

Research. Brussels: European Commission.

TEP (The Evaluation Partnership). 2007. ‘Evaluation of the Commission’s Impact Assessment system’.

Final Report Submitted to Secretariat-General of the European Commission (Contract Number

SG-02/2006).

Thiel, A. 2008. ‘Changing administrative culture and political debate? A review of impact assessment

practices in the European Commission’, Zeitschrift für Umweltpolitik and Umweltrecht 31: 5–27.

——and B. König. 2008. ‘An institutional analysis of the use of modelling tools within the European

Commission’, in K. Helming and and H. Wiggering (eds) Sustainability Impact Assessment of Land Use

Policies. Heidelberg: Springer.

Torriti, J. 2007. ‘Impact assessment in the EU: A tool for better regulation, less regulation, or less bad

regulation?’ Journal of Risk Research 10(2): 239 – 76.

Turnpenny, J., M. Nilsson, D. Russel, A. Jordan, J. Hertin and B. Nykvist. 2008. ‘Why is integrating

policy assessment so hard? A comparative analysis of the institutional capacities and constraints’ , Journal of

Environmental Planning and Management 51(6): 759–75.

——, C. Radaelli, A. Jordan and K. Jacob. 2009. ‘The policy and politics of policy appraisal: Emerging

trends and new directions’, Journal of European Public Policy 16: 640–53.

Weiss, C. 1979. ‘The many meanings of research utilization’, Public Administration Review 39(5): 426–31.

Wilkinson, D., M. Fergusson, C. Bowyer, J. Brown, A. Ladefoged, C. Monkhouse and A. Zdanowicz.

2004. Sustainable Development in the European Commission’s Integrated Impact Assessments for 2003: Final

Report. London: Institute for European Environmental Policy.

Wollmann, H. 2007. ‘Policy evaluation and evaluation research’, in F. Fischer, G.J. Miller and M.S Sidney

(eds) Handbook of Public Policy Analysis: Theory, Politics and Methods. London: CRC Press, 393 – 402.

Turnpenny et al.

254

Policy analytical styles

Igor S. Mayer, C. Els van Daalen and Pieter W.G. Bots

Introduction: perspectives on policy analysis: a framework for

understanding and design

Policy analysis is a multifaceted ﬁeld in which a variety of diﬀerent activities and ambitions have

found a place. Some policy analysts conduct quantitative or qualitative research while others recon-

struct and analyse political discourse or set up citizen forums. Some policy analysts are independent

researchers, some are process facilitators, while others act as political advisers (Dror 1967; Jenkins-

Smith 1982; Durning and Osuna 1994). The debate on the discipline – for example, on its foun-

dations, underlying values and methods – is conducted in a fragmented way (Dunn [1981] 1994;

Brewer and DeLeon 1983; Hogwood and Gunn 1984; Bobrow and Dryzek 1987; Wildavsky

1987; DeLeon 1988; MacRae and Whittington 1997; Hawkesworth 1988; House and Shull 1991;

Weimer and Vining 1992; Fischer and Forester 1993; White 1994; Radin 1997; Mayer 1997;

Hoppe 1998; Shulock 1999; Lynn 1999).

This is a pity, because it tends to sideline a reﬂection on the relationship between applied research,

the use and development of methods in relation to policy advice and policy processes.

The variety and multifaceted nature of policy analysis makes it clear that there is no single, let

alone ‘best’, way of conducting policy analyses. The discipline consists of many diﬀerent schools,

approaches, roles and methods. The observed diversity of policy analysis does give rise to numerous

questions. If we are unable to construct cohesion and unity behind this great diversity, we cannot

speak of a discipline. What relationship exists between the diﬀerent schools and activities of policy

analysis? Do they exclude each other or are there – in practice – numerous hybrids and com-

binations? What conceptual framework do we have at our disposal if we need to demarcate the

discipline, design new methods and approaches, or evaluate projects? Can we enrich the metho-

dological toolbox by adding new methods? What is the relationship between policy analysis methods

and new insights from the policy sciences, such as interactive policy development and process man-

agement (Edelenbos 1999; de Bruijn et al. 2002)? These are important questions that we obviously

cannot answer in full and all at once, but for which we modestly hope to provide a framework.

Untangling and explaining

The great diversity of views, schools and methods easily causes confusion and gives rise to the

need for insight into the discipline for insiders and outsiders alike (Radin 1997; Lynn 1999).

255

Various attempts have been made to untangle and explain policy analysis as a methodical discipline.

Some well-known examples of models in which activities and methods are systematically rela-

ted can be found in Dunn (1994), Brewer and DeLeon (1983), Hogwood and Gunn (1984),

Bobrow and Dryzek (1987), Miser and Quade (1985), Patton and Sawicki (1986), Weimer and

Vining (1992) and Mayer (1997).

It is precisely because of the varied developments in policy analysis and the diﬀuse image that

they create of the ﬁeld, that this chapter seeks to make the ﬁeld transparent and to structure it

with the help of a framework or conceptual model. Structuring will not take place by choosing

a speciﬁc author, perspective or school, but rather by displaying the variety of views of policy

analysis. It is not our intention to adopt a normative standpoint on what the most preferable form

or style of policy analysis should be. This chapter provides a framework for positioning the diﬀer-

ent perspectives and for highlighting the implications of choosing a perspective when designing

or evaluating a policy analysis project.

The presented conceptual model therefore has three functions. First, structuring the ﬁeld into

activities and styles provides a greater insight into and overview of the diversity of policy analysis.

The model is a means to demarcate and understand the ﬁeld as a whole. Second, when designing a

particular policy analysis project, the analysts will select methods and tools they consider to be

appropriate. The model can support choosing existing methods and designing new methods. Third,

we believe that the quality of a policy analysis project can be judged from diﬀerent perspectives.

The model helps to formulate the values pertaining to a perspective, values from which criteria

for the evaluation of a policy analysis project can be derived.

Research approach

In our attempt to (re)structure the diﬀerent styles and construct a conceptual model, we reviewed

the authoritative literature on the development of policy analysis and policy analysis styles. This

review led to an important observation. All characterisations of policy analysis are inclusive of a

limited number of preferred styles but are also exclusive of other styles, either because these are

not considered at all or because they are criticised as not being (eﬀective) policy analysis (e.g.

Lawlor 1996). From the present literature, a preliminary classiﬁcation of policy analysis activities,

roles and values was constructed.

A set of interacting activities

Our strategy in developing the model has been to ﬁrst address the question: ‘What general

activities do policy analysts perform when it comes to supporting policy and policy processes?’

We identiﬁed six major clusters of activities. They are:

1 Research and analyse: this cluster of activities matches with a perspective on policy analysis as

knowledge generation. Knowledge institutions, such as statistical agencies, semi-scientiﬁc

research institutions, and research agencies, gather and analyse, on request and on their own

initiative, knowledge and information for policy purposes. It is possible that the political

agenda inﬂuences their research priorities, but the results of their autonomous research activities

may also inﬂuence the political agenda. Translation of the results of their research into a policy

design or recommendation is not a primary part of their task or mission. It is up to the political

system to identify consequences and draw conclusions from the best available knowledge.

2 Design and recommend: when suﬃcient data and information have been gathered in earlier

research, a policy analysis will focus on translating the available knowledge into new

Mayer et al.

256

policy, either by making recommendations or by making a complete policy design. Recom-

mendations will typically be the result of comparing the eﬀects of diﬀerent policy alter-

natives and weighing the options based on various criteria. Policy analysts in this way

are supportive to the policy process by translating available knowledge into new policy

either by advising or by making (partial) policy designs in terms of ‘actions-means-ends’.A

complete policy design typically involves generating a set of alternative strategies that

each consists of several tactics aimed at achieving particular objectives or sub-goals (see

Walker 1988).

3 Clarify arguments and values: there will always be implicit normative and ethical

questions and opinions behind public policy. Prolonged conﬂicts and social issues that

turn into stalemates often come about through fundamental normative and argu-

mentative diﬀerences (van Eeten, 2001; Fischer and Forester 1993). Abortion, euthanasia

and drilling for natural gas in protected areas are examples of such issues. Policy analysis

may not only make instrumental recommendations for policy-making; it may also

analyse the values and argumentation systems that underpin social and political debate.

Moreover, policy analysis seeks to improve the quality of debate by identifying the one-

sided or limited nature of arguments or showing where blind spots exist in the debate

(Hoppe 1998).

4 Provide strategic advice: policy analysis will often be a strategic, client-oriented activity.

The substantive or procedural advice will be made dependent on the analysis of the ﬁeld

of forces that exist, namely the environment in which the client and his problem are

located. The policy analyst will advise the client on the most eﬀective strategy for achieving

certain goals given a certain political constellation, namely the nature of the environment

in which the client operates, the likely countersteps of opponents, and so on.

5 Democratise: in the ‘democratise’ cluster of activities, policy analysis does not have a value-

free orientation, but a normative and ethical objective: it should further equal access to,

and inﬂuence on, the policy process for all stakeholders (DeLeon 1988, 1994; Lerner and

Lasswell 1951). Experts and elites are more likely to be involved and carry greater weight

than ordinary citizens and laymen (Fischer 1990). Policy analysis can try to correct this

inequality by calling attention to views and opinions typically overlooked in policy-making

and decision-making (Fischer 2000).

6 Mediate: policy analysts can play a role in enhancing the knowledge actors have about

their own position, about the actors’ room for manoeuvring, and in looking for possible

compromises and win-win options. In addition, they can be involved in designing the rules

and procedures for negotiating in a policy-making or decision-making process and mana-

ging the interaction and progress of that process. The mediation cluster comprises diﬀerent

types of activities, with a focus on analysing contextual factors (e.g. dependency analysis,

transaction analysis), and designing, and possibly also facilitating, meetings in which dif-

ferent stakeholders and decision-makers consult and negotiate. The policy analyst can be

involved during the design of the negotiation process as well as its execution.

In real-life cases and projects, a policy analyst will combine one or more activities, albeit not

all at the same time. When more activities are combined, a policy analysis project will become

richer and more comprehensive, but also more complex.

The hexagon in Figure 20.1 is a diagrammatic representation of these six activities. The

theoretical foundation will be discussed later in this chapter, when we show the policy analysis

styles and values on which the clustering of activities has been based. In this section, we focus

on the six activities and illustrate these with the help of examples based on policy analyses.

Policy analytical styles

257

At the end of this section, we will look at the relations between the various activities in

more detail.

Policy analysis styles

It is the objective of our model to clarify and understand the discipline of policy analysis. Numerous

schools of thought, paradigms and models can be found in the policy analysis literature (Bobrow

and Dryzek 1987; DeLeon 1988; Hawkesworth 1988; House and Shull 1991; Mayer 1997). In this

chapter, we will refer to styles of policy analysis rather than to a paradigm, model or school. Based

on the schools discussed in the literature and the framework of our model, we have identiﬁed

six policy analysis styles. They are:

1 a rational style;

2 an argumentative style;

3 a client advice style;

4 a participatory style;

5 a process style;

6 an interactive style.

Figure 20.2 shows how these styles relate to the activities discussed above. Below, we will brieﬂy

discuss the styles in an archetypical manner. We will focus on the arguments that are used by

proponents of these styles.

The rational style

The rational style is shaped to a large degree by assumptions about knowledge and reality, and

by a relatively large distance between the object and subject of study: it is assumed that the

world is to a large extent empirically knowable and often measurable. Knowledge used for policy

must be capable of withstanding scientiﬁc scrutiny. The role of knowledge in policy is a positive

Figure 20.1 Policy analysis tasks

Source: Mayer et al.. 2004. ‘Perspectives on policy analysis: A framework for understanding and

design’, International Journal of Technology, Policy and Management 4(1): 169–91

Mayer et al.

258

one, namely a greater insight into causes, eﬀects, nature and scale produces better policy (Weiss

and Bucuvalas 1980). Policy should come about – preferably – in neat phases, from preparation

to execution, with support through research in each phase. An example of this policy analysis

approach is the systems analysis method developed by the RAND Corporation (Quade 1989;

Miser and Quade 1985). The advice on policy regarding the Eastern Scheldt storm surge barrier

in the Netherlands was obtained using this method (Goeller et al. 1977). This style is discussed

in many general textbooks on methods of policy analysis (Patton and Sawicki 1986; MacRae

and Whittington 1997; House and Shull 1991).

Argumentative style

This style assumes that policy is made, defended and criticised through the medium of language.

The basic assumption of the argumentative style is therefore that when analysing policy, it is

important to devote attention to aspects related to the language game that takes place around a

policy problem or issue. Attention will shift to the debate and the place in the debate of arguments,

rhetoric, symbolism and stories (Fischer and Forester 1993; Fischer 1995; van Eeten 2001). Argu-

ments aim to have an eﬀect on the public. The positions of parties and the argumentations in a

policy discourse are not always clear and unambiguous, however. Therefore, policy analysis will

make policy easier to understand by illustrating the argumentations and the quality thereof schema-

tically and making a judgement based on criteria such as justiﬁcation, logic and richness (Dunn 1982,

1994). But the ambition of argumentative policy analysis is to use such an analysis to produce

recommendations and improvements in situations where parties have been talking at cross-purposes

for many years: a dialogue between the deaf (van Eeten 2001). The argumentative style assumes that

it can make the structure and progress of the discourse transparent by means of interpretive and

qualitative methods and techniques, and can also bring about improvements by identifying caveats

in the debate or searching for arguments and standpoints that can bridge the gap between oppo-

nents. This style of policy analysis centres on discourse and argumentation analyses so as to frame

the diﬀerent standpoints of clusters of parties and, if possible, change and inﬂuence them.

Figure 20.2 Policy analysis styles

Source: Mayer et al. (2004) ‘Perspectives on policy analysis: A framework for understanding and

design’, International Journal of Technology, Policy and Management 4(1): 169–91

Policy analytical styles

259

Client advice style

In a number of respects, the client advice style is based on assumptions that policy-making

occurs in a complex and rather chaotic arena. There are numerous players, with diﬀerent interests

and strategies (de Bruijn and ten Heuvelhof 2000; de Bruijn et al. 2002). Therefore, it is wise to

gain insight into the various objectives, means and interests of the actors involved. For that

reason, the analysis of this complex environment is important and can be undertaken analyti-

cally and systematically by such means as stakeholder analyses, although intuition and soft

information deﬁnitely play a role. Besides knowledge and insights gained through research,

policy analysis is largely a question of politico-strategic insight and skills including client-

analyst communication. In addition to being a skill, methodical and explicit, policy analysis is

also an art in which tacit knowledge plays an important role (Wildavsky 1987). Depending on

orientation, the client advice style involves a more design-oriented approach or a strategic,

process-driven approach.

Participatory style

Participatory policy analysis views the relationship between research and advice on the one

hand and policy and politics on the other by looking at society critically (Fischer 1990, 2000).

Here it is assumed that not all sections of the population have ready access to policy systems.

Researchers, economic elites, institutionalised non-governmental organisations and politicians

dominate policy discussions and decisions about major social issues (Jasanoﬀ 1990). Researchers,

stakeholders and policy-makers will even change roles and positions within one and the same

system. Certain subjects and also certain groups of actors are often excluded from the social

debate. This is referred to as the technocratic criticism of policy analysis (Fischer 1990). Participatory

policy analysis assumes that citizens can have a voice and be or become interested enough to

deliberate on substantive and politically diﬃcult questions (Dryzek 1990; Fishkin 1991; Durning

1993; DeLeon 1994; Mayer 1997; Fischer 2000). The policy analyst can take on a facilitating role

in such a debate by promoting equality and openness in the design and by giving ordinary citizens

and laymen a role alongside others (Mayer 1997).

Process style

Just as in a game of chess, the parties that participate in a policy-making process will exhibit

strategic behaviour in the pursuit of their own objectives and achievement of the best possible

positions, even if such action runs counter to the public interest formulated in policy (de Bruijn

et al. 2002). It is perfectly understandable that, in controversial and complex issues, opponents

will underpin their case with conﬂicting research reports. Impartial experts do not exist and a

solution by way of new reports and studies can aggravate the problem in a certain sense. In fact,

knowledge is (not much more than) negotiated knowledge. It is better to negotiate and reach

agreements about the use of the results of a study or jointly contracting research (de Bruijn et al.

2002). The process style of policy analysis is based on the assumption that substantive aspects of

a policy problem are, in fact, coordinate or perhaps even subordinate to the procedural aspects

of a policy problem. The analyst or process manager creates ‘loose coupling’ of procedural

aspects and substantive aspects of a problem. Procedural aspects are understood to be the orga-

nisation of decision-making or the way in which parties jointly arrive at solutions to a problem.

To that end, agreements can be reached through ‘mediation and negotiation’. If the procedural

sides of a policy-making or decision-making process have been thought through properly, it

Mayer et al.

260

will greatly increase the likelihood of substantive problems being resolved. Substantive problems

can be made part of a process design, for example, by placing the diﬀerent substantive aspects

on the agenda.

Interactive style

The interactive style of policy analysis assumes that individuals – experts, analysts, clients, sta-

keholders and target groups – have or may have diﬀering views of the ‘same’ policy problem.

An insight relevant to policy can be obtained by bringing about a confrontation and interaction

of diﬀerent views. The interactive style has a strong socio-constructive foundation. Diﬀerent

views of reality can be valid simultaneously. Through continuous interaction and interpretation –

the ‘hermeneutic circle’–it is possible to gain an ‘insight’ (Guba and Lincoln 1989).

In an interactive style of policy analysis, target groups and stakeholders are usually invited to

structure problems or devise solutions in structured working meetings at which policy analysis

techniques may be used (Mason and Mitroﬀ 1981). This brings about a multiple interaction

whereby the views and insights of the analyst, the client and also the participants are enriched

(Edelenbos 1999). In other words, participants learn about their own views in relation to

those of others, and have an opportunity to reﬁne those views. The selection of views is

obviously crucial. Political considerations – the power to obstruct – and enrichment argu-

ments – what citizens really think – may be interwoven. What matters is the quality of the

obtained insights in combination with the heterogeneity of opinions and interests. If policy

analysis concerns the redevelopment of a city square, for example, stakeholders such as local

residents and business people can be consulted by means of workshops about the problems they

experience with the present arrangement of the square and their wishes with regard to the new

plans. The interactive style assumes that a process like this is informative for decision-makers and

planners, is more likely to lead to acceptance and fulﬁlment of the plans, and can bring about all

kinds of positive eﬀects among the participants (learning about each other and about policy

processes) (ibid.).

Deﬁnition of archetypal styles

Figure 20.3 shows the policy analysis styles placed in an ‘archetypal’ way between the diﬀerent

activities. This rightly suggests that a style balances on two important activities. This balance

does not necessarily need to be in equilibrium. Participatory policy analysis balances between

democratisation and clariﬁcation of values and arguments. The emphasis may be more on one

activity than on the other: citizens can be directly involved in discussions about genetic technology,

or the analyst may be mainly interested in the value systems, arguments and opinions of citizens

about the technology and may want to systematise them for the purpose of policy advice.

The argumentative style balances between research and analysis and clariﬁcation of values and

arguments. Some argumentative policy analysts attempt to improve the quality of policy by

testing the policy design as thoroughly as possible, or by building on consistency, validity and so

on, of the underlying arguments (Dunn 1994). This is based on the principle that ‘claims’ must

be substantiated up by facts (‘backings’). The ‘formal logic’ is dominant in this setting. Others

reconstruct arguments, not in relation to scientiﬁc quality, but according to their variety and

richness. This allows greater scope for normative systems, religion and intuitive arguments

(Fischer 1995).

In a similar way, the rational style balances between researched analysis and advisory design;

the interactive style between democratisation ambitions and mediation activities; the client advice

Policy analytical styles

261

style between substantive design and strategic advice; and the process style between strategic advice

and mediation. The styles of policy analysis may thus have diﬀerent manifestations and emphases.

A focus on a certain activity may result in a style leaning more towards one activity than to another.

Combining activities

In the preceding sections of this chapter, we have diﬀerentiated between the policy analysis

styles by showing that they balance between two activities. It is also possible to let go of the

balance and to make combinations of activities that are not adjacent to one another. In other

words, a policy analytic arrangement will be made whereby two or more activities that are oppo-

site, rather than adjacent, to each other in the hexagon of Figure 20.1 can be combined. This

kind of combination or arrangement, symbolised by the dashed diagonals in the hexagon, is

achievable in two ways:

1 The activities can be carried out sequentially or separately, either in various parts of

one policy analysis project or in diﬀerent complementary or competing projects; i.e. a

form of methodological triangulation of activities. As part of a policy analysis project

focusing on climate change, for example, ﬁrst research can be conducted by experts using

climate models (activity: research) and subsequently the perceptions and arguments of

ordinary citizens and laymen regarding climate change can be mapped out (activity: clarify

arguments).

2 The various activities can be integrated into one design or method. As part of a project

focusing on climate change, for example, climate models can be used to get various groups

Figure 20.3 The underlying values and criteria of policy analysis

Source: Mayer et al. (2004) ‘Perspectives on policy analysis: A framework for understanding and

design’, International Journal of Technology, Policy and Management 4(1): 169–91

Mayer et al.

262

of stakeholders, experts, politicians and so on to jointly generate and test policy proposals,

while obtaining feedback from representative citizen panels. Such a design would integrate

various activities: research, design, democratise and mediate.

Underlying values

Evaluation criteria

In addition to demarcating and understanding the ﬁeld of policy analysis and designing a policy

analysis project, our model has a third function: evaluation of policy analysis projects and

methods (Twaalfhoven 1999). The various activities and styles are based on underlying values

and orientations. The values determine in what way a policy analyst or others will view the

quality of the policy analysis study and the criteria that will be applied to examine it. These

criteria can be made explicit by addressing the following questions:

 Rational style: what is good knowledge?

 Argumentative style: what is good for the debate?

 Client advice style: what is good for the client/problem owner?

 Participatory style: what is good for democratic society?

 Process style: what is good for the process?

 Interactive style: what is good for mutual understanding?

Figure 20.3 shows that the activities in the top half of the hexagon are primarily object-oriented

activities: a system, a policy design, an argumentative analysis. The activities at the bottom are

subject-oriented activities. They focus primarily on the interaction between citizens, stakeholders,

the analyst and the client. Whereas the top-half activities are usually captured in a product – e.g.

a report, a design, a computer model – the eﬀects of the bottom-half activities are usually cap-

tured in the quality of the process itself: increased support base, mutual understanding, citizen-

ship, learning. The distinction ‘object–subject’ translates into the types of evaluation criteria to

be applied. Object-oriented policy analysts will judge the quality of a policy analysis by its sci-

entiﬁc rigor or the substantive insights it has yielded. Subject-oriented policy analysts will base

their judgement on the contribution of the orchestrated interaction between stakeholders to the

decision-making process. The turning point between object and subject-oriented activities lies

with ‘clariﬁcation of values and arguments’ and ‘provision of strategic advice ’. These can be either

object-oriented and/or subject-oriented.

Figure 20.3 also shows that the activities on the left-hand side are judged by idealistic and

generic criteria for good policy analysis, such as validity, reliability, consistency, fairness, equality

or openness. The activities on the right-hand side of the hexagon are judged by pragmatic and

particular criteria, such as workability, usability, opportunity, feasibility or acceptance.

These criteria for evaluating the quality of a policy analysis project or method are summarised

in Box 20.1 and appear in the corners of the hexagon in Figure 20.3.

The role of the policy analyst

As the presented model is based on activities, styles and their associated values, it also generates

and organises the positive and negative images, the metaphors, of the policy analyst (Dror 1967;

Jenkins-Smith 1982; Durning and Osuna 1994). Some policy analysts allow themselves to be

guided mainly by their wish to conduct objective scientiﬁc research; these are the objective

Policy analytical styles

263

technicians. In contrast, others seek interaction with their client; these are the client advisers or

counsellors. Some advocate a clear standpoint such as a more stringent environmental policy;

these are the issue activists. How the role of a policy analyst is perceived depends on one’s own

values and position in a policy process. A skilful strategic advisor, for example, may be highly

appreciated by his/her client, but portrayed as a hired gun by his/her client’s opponents. In

Table 20.1, positive and negative images of the role of the policy analysts are depicted for each

activity.

Box 20.1 Translation of values into quality criteria

Research and analyse

Policy analysis will be judged by substantive (scientiﬁc) quality criteria, such as validity and

reliability, the use and integration of state-of-the-art knowledge, the quality of data gathering

and the formal argumentation and validation of conclusions.

Design and recommend

Policy analysis will be judged by instrumental criteria of policy relevance, such as usability

and accessibility for policy-makers, action orientation and utilisation, presentation and

communication of advice, weighing up of alternatives, clear choices and so on.

Clarify values and arguments

Policy analysis will be judged by quality of argumentation and debate criteria such as formal

logic (consistency), informal logic (rhetoric and sophism), and quality of the debate in terms

such as richness, layering and openness of arguments.

Advise strategically

Policy analysis will be judged by pragmatic and political effectiveness criteria, such as the

‘workability’ of advice, political cleverness and proactive thinking, greater insight (for the

client) in the complex environments (political and strategic dynamics, forces and powers),

targeting and achievement of goals.

Democratise

Policy analysis will be judged by democratic legitimacy criteria, such as openness and

transparency of the policy-making process, representation and equality of participants and

interests, absence of manipulation and so on.

Mediate

Policy analysis will be judged by external acceptance and learning criteria, such as the agree-

ment that mutually independent actors reach on the process and/or content, support for

and commitment to the negotiating process and solutions, learning about other problem

perceptions and solutions.

Mayer et al.

264

Perspectives on the ﬁeld of policy analysis

Figure 20.4 presents the complete conceptual model in which policy analysis activities are related

to underlying styles and values and the policy analyst’s roles. The ﬁgure enables us to demarcate

all manifestations and varieties of policy analysis and also to develop new approaches and methods.

Methods developed mainly within one style of policy analysis can be combined with insights

from another style and adapted to new activities. Below, we will brieﬂy recapitulate the functions

of the model, namely demarcate, design and evaluate.

Reﬂection on policy analysis

Policy analysis is characterised by ambitions but also by ambivalences. Various approaches criti-

cise each other and it is very diﬃcult to deﬁne and describe what policy analysis is. The added

value of our model is that it shows why policy analysis is ambivalent and elusive, namely because

the proponents and opponents reason from diﬀerent points of departure, about what they are

doing, why they are doing it, and the limitations or conversely the richness of the discipline. It

is not our intention in this chapter to adopt a position on our preferred form of policy analysis,

even if we were to have one. Depending on one’s own position, one may accept the wide picture

of policy analysis as depicted in our model, but it is likely that many will argue that certain styles

or activities are not (proper) policy analysis (e.g. Lawlor 1996): for those critics, the hexagon

may turn into a straight line, a triangle or a square. The problem, of course, is that there will be

no disciplinary consensus on what activities and styles to cut from the hexagon and on what grounds.

For every policy analytic style there are both proponents and critics. Given the actual and desirable

development of the various deﬁnitions of policy analysis, we are of the opinion that it is better

to deﬁne the discipline too widely than too narrowly. The integrated conceptual model oﬀers

full scope without losing the unity of the policy analysis and causing the disintegration of the ﬁeld.

The model oﬀers the possibility to examine policy analyses already performed and to relate these to

each other. The model seeks to provide a foothold, or a framework, for demarcating the wide

ﬁeld of work.

Table 20.1 Positive and negative images of the policy analyst

Activity Positive role image Negative role image

Research and analyse Independent scientist;

objective researcher.

A-moral researcher;

technocrat.

Design and recommend Independent expert;

engineer; impartial adviser.

Desk expert;

‘back seat driver’.

Clarify values and arguments Logician or ethicist;

narrator.

Linguistic purist;

‘journalist’.

Advise strategically Involved client adviser;

client counsellor; policy

entrepreneur

‘Hired gun’

Democratise Democratic (issue) advocate. Missionary;

utopian.

Mediate Facilitator; mediator;

process manager.

Manipulator;

‘relativist’.

Source: Mayer et al. (2004). ‘Perspectives on policy analysis: A framework for understanding and design’, International

Journal of Technology, Policy and Management 4(1): 169–91

Policy analytical styles

265

Design of a policy analysis

The model provides an overview of the wealth of possibilities of policy analysis studies and the

interrelationships between them, and can be of help in reﬂecting consciously and creatively on

the design of a policy analysis. As a rule, policy analysis projects require a customised design. It is

possible, however, to fall back on standard methods of policy analysis, although the choice and

combination of methods will depend on the problem under examination. The model deﬁnitely

does not seek to prescribe instrumentally how a policy analysis should be designed. The oppo-

site is the case, because we advocate creativity and innovation in designing approaches, actions

and methods. Innovative combinations of researching, designing, recommending, mediating, argu-

mentation and democratisation can be made. If desired, a rational style of policy analysis may be

combined, for example, with a process style. This would ‘interweave’ analytical or scientiﬁc study

in mediation processes between parties (de Bruijn et al. 2002).

We consider the design of policy analysis to include the development of new methods of

policy analysis so as to allow a good integration of sub-activities. In point of fact, the history of

policy analysis is characterised by the repeated application of creative and intelligent combinations

of methods; methods that originated in one domain are translated into applications for other

domains. The, by now, classical Delphi and scenario methods came about as methods for studying

the future, but are currently used for strategic advice, mediation and even democratisation in

policy Delphis, interactive scenario methods and scenario workshops (Mayer 1997). Cross-impact

techniques and stakeholder analysis techniques, which came about as methods for advising clients,

now have principally interactive applications and are used for mediation. Consensus conferences

Figure 20.4 Overview of the complete hexagon model of policy analysis

Source: Mayer et al. (2004). ‘Perspectives on policy analysis: A framework for understanding and

design’, International Journal of Technology, Policy and Management 4(1): 169–91

Mayer et al.

266

that came about as a method for study and mediation between leading experts in medical sci-

entiﬁc controversies have been transformed into methods for democratising and for public

participation (Mayer 1997; Fischer 2000).

Evaluation of a policy analysis

Each policy analysis activity is based (implicitly) on values concerning the quality and purposes

of the policy analysis. Therefore, policy analysis projects can be examined from diﬀerent per-

spectives. This may lead to diﬀerent opinions about success or failure, quality or shortcomings.

A substantively thorough and valid study can be unusable for a client. A brilliant and workable

compromise that breaks a stalemate may be based on negotiated nonsense (van de Riet 2003) or

may violate or manipulate the interests of legitimate participants. Conﬂicts like these are almost

inherent in every evaluation of sizeable policy analysis projects. In the design and evaluation,

the policy analyst attempts to cope as well as possible with these tensions and dilemmas, either

by making choices or by ﬁnding new routes.

Conclusion

In this chapter we have presented a model for policy analysis based on six archetypal policy

analysis activities. This subdivision makes it possible to relate various policy analysis styles found

in the literature to each other and to analyse the characteristics of and diﬀerences between the

styles. Additionally, the activities provide pointers for evaluating policy analyses. By explicitly

identifying which activities are being pursued with the policy analysis, it is possible to use that

information as a basis for identifying success criteria for the policy analysis. The developed fra-

mework seeks to map out transparently the enormous variety of diﬀerent types of policy ana-

lyses and to allow them to be viewed in relation to each other. The model can also be used to

design policy analysis studies. By making explicit which activities are relevant in a particular policy

analysis, a conscious choice can be made for a certain policy analysis style and the policy analysis

methods can be selected in a well-founded way for the contribution made by the method or

technique to the activities that must be carried out.

While the developed framework provides pointers for reﬂection, design and evaluation, it is

not intended to be a rigid, prescriptive model. Rather, the intention is for the policy analyst to

be consciously working on the goal of the analysis in relation to the policy-making process and

to produce his own policy analysis design and evaluation.

Note

1 This chapter is a condensed and slightly adapted version of a journal article previously published as: I.S.

Mayer; P.W.G. Bots and C.E. van Daalen. (2004). ‘Perspectives on policy analysis: A framework for

understanding and design’, International Journal of Technology, Policy and Management 4(2): 169–91. The

material has been reused with permission of the copyright owner.

Bibliography

Bobrow, D. and J. Dryzek. 1987. Policy Analysis by Design . Pittsburgh, PA: University of Pittsburgh Press.

Brewer, G. and P. DeLeon. 1983. The Foundations of Policy Analysis. Homewood, IL: Dorsey.

de Bruijn, J. and E. ten Heuvelhof. 2000. Networks and Decision-Making. Utrecht: Lemma.

——, E. ten Heuvelhof and R. in ‘t Veld. 2002. Process Management, Why Project Management Fails in

Complex Decision-Making Processes. Dordrecht: Kluwer.

Policy analytical styles

267

DeLeon, P. 1988. Advice and Consent: The Development of the Policy Sciences. New York: Russell Sage

Foundation.

—— 1994. ‘Democracy and the policy sciences: Aspirations and operations’, Policy Studies Journal 22: 200–12.

Dror, Y. 1967. ‘Policy analysts: A new professional role in government service’, Public Administration

Review 27(3): 198.

Dryzek, J. 1990. Discursive Democracy; Politics, Policy and Political Science. Cambridge: Cambridge University

Press.

Dunn, W. [1981]1994. Public Policy Analysis: An Introduction, 1st and 2nd edns. Englewood Cliﬀs, NJ:

Prentice Hall.

—— 1982. ‘Reforms as arguments’, Knowledge, Creation, Diﬀusion, Utilization 3(3): 293–326.

Durning, D. 1993. ‘Participatory policy analysis in a social service agency: A case study’, Journal of Policy

Analysis and Management 12(2): 297–322.

——and W. Osuna. 1994. ‘Policy analysts, roles and value orientations: An empirical investigation using

Q methodology’, Journal of Policy Analysis and Management 13(44): 629–57.

Edelenbos, J. 1999. ‘Design and management of participatory public policy making’, Public Management

1(4): 569–78.

—— 2000. Process in Shape. PhD thesis. Lemma: Utrecht.

Fischer, F. 1990. Technocracy and the Politics of Expertise. Newbury Park, CA: Sage.

—— 1995. Evaluating Public Policy. Chicago: Nelson-Hall.

—— 2000. Citizens, Experts and the Environment. The Politics of Local Knowledge. Durham, NC: Duke

University Press.

——and J. Forester (eds). 1993. The Argumentative Turn in Policy Analysis and Planning. Durham, NC: Duke

University Press.

Fishkin, J. 1991. Democracy and Deliberation: New Directions for Democratic Reform. New Haven, CT: Yale

University Press.

Goeller, B., A. Abrahamse, B.H. Bigelow, J.G. Bolten, D.M. De Ferranti, J.C. DeHaven, T.F. Kirkwood,

R. Petruschell. 1977. Protecting an Estuary from Floods: A Policy Analysis of the Oosterscheld,

vol. 1. Santa

Monica,

CA: Summary Report, prepared for the Netherlands Rijkswaterstaat.

Guba, E. and Y. Lincoln. 1989. Fourth Generation Evaluation. Newbury Park, CA: Sage.

Hawkesworth, M. 1988. Theoretical Issues in Policy Analysis. Albany, NY: State University of New York

Press.

Hogwood, B. and L. Gunn. 1984. Policy Analysis for the Real World. Oxford: Oxford University Press.

Hoppe, R. 1998. ‘Policy analysis, science, and politics: From “speaking truth to power” to “making sense

together”’, Science and Public Policy 26(3): 201 – 10.

House, P. and R. Shull. 1991. The Practice of Policy Analysis; Forty Years of Art and Technology. Washington,

DC: Compass.

Jasanoﬀ, S. 1990. The Fifth Branch. Science Advisors as Policymakers. Cambridge, MA: Harvard University

Press.

Jenkins-Smith, H. 1982. ‘Professional roles of policy analysts’, Journal of Policy Analysis and Management 2(1):

88–100.

Lawlor, E. 1996. Book review of The Argumentative Turn in Policy Analysis and Planning, F. Fischer and

J. Forester. (eds) Journal of Policy Analysis and Management 15(1): 110–46.

Lerner, D. and Lasswell, H. (eds). 1951. The Policy Sciences, Recent Developments in Scope and Methods.

Stanford, CA: Stanford University Press.

Lynn, L. 1999. ‘A place at the table: Policy analysis, its postpositive critics, and the future of practice’,

Journal of Policy Analysis and Management 18(3): 411–24.

MacRae, D. and D. Whittington. 1997. Expert Advice for Policy Choice: Analysis and Discourse. Washington,

DC: Georgetown University Press.

Mason, R. and I. Mitroﬀ. 1981. Challenging Strategic Planning Assumptions. Theory, Cases and Techniques.

New York: John Wiley and Sons.

Mayer, I. 1997. Debating Technologies: A Methodological Contribution to the Design and Evaluation of Participatory

Policy Analysis. Tilburg: Tilburg University Press.

——, C. Els van Daalen and Pieter W.G. Bots. 2004. ‘Perspectives on policy analysis: A framework for

understanding and design’, International Journal of Technology, Policy and Management 4(1): 169–91.

Miser, Hugh J. and E.S Quade. 1985. Handbook of Systems Analysis. New York: North-Holland.

——and E. Quade. 1988. Handbook of Systems Analysis: Craft Issues and Procedural Choices. Chichester: John

Wiley

& Sons.

Mayer et al.

268

Patton, C. and D. Sawicki. 1986. Basic Methods of Policy Analysis and Planning. Englewood Cliﬀs, NJ: Prentice

Hall.

Quade, E. 1975. Analysis for Public Decisions. New York: Elsevier.

Radin, B. 1997. ‘Presidential address: The evolution of the policy analysis ﬁeld: \from conversation to

conversations’, Journal of Policy Analysis and Management 16(2): 204–18.

Shulock, N. 1999. ‘The paradox of policy analysis: If it is not used, why do we produce so much of it?’,

Journal of Policy Analysis and Management 18(2): 226–44.

Twaalfhoven, P. 1999. The Success of Policy Analysis Studies: An Actor Perspective. Delft: Eburon.

van De Riet, O. 2003. Policy analysis in Multi-Actor Settings: Navigating between Negotiated Nonsense and

Superﬂuous Knowledge. Delft: Eburon.

van Eeten, M. 2001. ‘Recasting intractable policy issues: The wider implications of the Netherlands civil

aviation controversy’, Journal of Policy Analysis and Management 20(3): 391–414.

Walker, W. 1988. ‘Generating and screening alternatives’, in H. Miser and E. Quade (eds) Handbook of

Systems Analysis: Craft Issues and Procedural Choices. Chichester: John Wiley & Sons.

Weimer, D. and A. Vining, A. 1992. Policy Analysis: Concepts and Practice. Englewood Cliﬀs, NJ: Prentice Hall.

Weiss, C. and M. Bucuvalas. 1980. Social Science Research and Decision-Making. New York: Columbia

University Press.

White, L. 1994. ‘Policy analysis as discourse’, Journal of Policy Analysis and Management 13(3): 506–25.

Wildavsky, A. 1987. Speaking Truth to Power: The Art and Craft of Policy Analysis. New Brunswick, NJ:

Transaction Books.

Policy analytical styles

269

Part VI

Understanding the

decision-making process

Bounded rationality and

public policy decision-making

Bryan D. Jones and H.F. Thomas III

Underlying most approaches to the study of policy processes is a model of human choice.

This model

rests upon critical aspects of the cognitive and emotional architectures of decision-makers. The parti-

cular aspects of human psychology that informs the major approaches in policy studies diﬀer, but all

share the common focus of requiring some assumptions about human nature. The Advocacy Coalition

Framework is based on an analysis of the attitudes, beliefs and values of coalition participants. Punc-

tuated Equilibrium Theory is based on shifts in the focus of collective attention, which requires an

individual choice model based on attention and short-term memory. The Institutional Analysis and

Development Framework centers on rational action by strategic actors, but with an understanding

of the limits of full rationality and a focus on the abilities of citizens in smaller associations to focus

on collective rather than individual beneﬁts. In each case, some elements of cognitive processing

beyond the maximization of goals assumed in the model of instrumental rationality, which is used

in economics and in public choice and neo-institutional analyses in political science, are present.

It is not much of a stretch to argue that the model underlying virtually all studies of policy

processes is bounded rationality. In this model, humans pursue goals, but they often fall prey to

mistakes in doing so. These mistakes are not simple, but complex results of human cognitive

architectures, and because the environments human decision-makers face is complex. Because

they are fundamentally related to human psychology, they are liable to be repeated. This approach

has the beneﬁt of keeping the traditional role of strategic action in the picture, but allows for

major deviations from optimum strategies.

In this chapter we review the genesis and development of bounded rationality and show how it has

inﬂuenced the study of policy processes. Further, this chapter reviews the general research on bounded

rationality relevant to the policy-making process, focusing speciﬁcally on bounded rationality as a

microfoundational underpinning for the study of public policy. We show how bounded rationality is

a superior microfoundation for most aspects of public policy, most particularly the prioritization of

policy problems, and show how the theory of preferences can be reconciled with bounded rationality.

Genesis and development of bounded rationality

Most scholars studying politics and government care little about the ﬁne details of the speciﬁcs of

human cognition; we are quite content to leave that to biologists, psychologists, and cognitive

273

scientists. What we cannot escape, however, is the need for some ﬁrm foundation capable of

linking human behavior to policy processes. That foundation must fulﬁll three criteria: ﬁrst, it

must do no harm (it should not mislead); second, it must allow us to move between individual-

level processes and organizational processes in a more or less seamless manner; and third, it

should be eﬃcient, in that it does not drag in speciﬁcs of human behavior that are not needed

to understand the policymaking process.

As we show below, the model of bounded rationality, as initially articulated by Herbert A.

Simon and expanded by Simon, organizational theorists such as James A. March, and cognitive

scientists, especially Allen Newell, does that, and that foundation has been available since 1958.

Simon made the ﬁrst bold step toward the development of a model of decision-making

capable of aiding the understanding of collective choice in organizations with the publication of

Administrative Behavior (Simon 1945). Simon admits that the model articulated there consisted

largely of “residual categories, and the positive characterization of the process of choice is very

incomplete” (Simon 1977: xxix). By 1958, however, all the elements for producing an orga-

nizational and policy science based on a positive model of choice were in place. The basic elements

of that model have been conﬁrmed and reconﬁrmed by cognitive scientists in the laboratory

and by students of organizations and policy processes in the ﬁeld. Yet the approach, rather than

serving as the undisputed decisional foundation for modern social science, has engendered much

confusion and controversy. In political science, much time was spent ﬁghting Simon’s approach

as “too scientiﬁc” for the humanistic study of politics; later the public choice approach simply

ignored the model. Economists “went on counting angels on the heads of neoclassical pins”

(Simon 1999: 113).

It is no accident that the behavioral model of choice came more or less directly from the beha-

vioral discipline of political science. Simon gives great credit to his participation as a student in

Charles Merriam’s behavioral revolution at the University of Chicago’s Department of Political

Science in the 1930s (Simon 1996a: Chapter 4). The scientiﬁc tenets of political behavioralism

were strong on observation and quantiﬁcation and not so strong on theory; the movement had

a clearly inductive ﬂavor. It demanded real-world observation—Merriam wanted to make a

diﬀerence in the conduct of public policy as well as in the conduct of scientiﬁc inquiry.

The tenets of bounded rationality

Simon developed bounded rationality as a critique of fully rational decision-making; what Simon

later termed the “behavioral theory of choice” was an attempt to state the positive aspects of a

theory of human choice based on scientiﬁc principles of observation and experiment rather than

the postulation and deduction characteristic of theoretical economics. We ﬁ rst review the princi-

ples of bounded rationality, and then turn to the modern conceptions of the behavioral theory

of choice. The tenets of bounded rationality can be summarized in four straightforward prin-

ciples. While research in many ﬁelds of social science has advanced our understanding over the

years, the basic formulation occurred in Administrative Behavior (Simon 1945).

Principle of intended rationality. Simon’s model is enshrined in the crucial principle of intended

rationality. That is, it starts with the notion that people are goal-oriented, but often fail to accom-

plish this intention because of the interaction between aspects of their cognitive architectures

and the fundamental complexity of the environment they face (Simon 1977: xxvii; March

1994). Intellectually, this notion did not begin with Simon; it probably began with Vilfredo

Pareto. In Mind and Society he distinguished logical, illogical, and non-logical behavior (Pareto

1935).

Logical behavior is rational choice; it is ends-means reasoning where means are appro-

priate to goals. Illogical behavior is behavior not rooted in ends-means thinking; Pareto thought

Jones and Thomas

274

little human behavior (at least of interest to a social scientist) was illogical. Non-logical thought

involved “ sentiments and residues” that could interfere with logical thinking. In a way, then,

we might term the principle of intended rationality the “Pareto-Simon Principle”.

The principle of intended rationality implies that we look at the goal-directed behavior of

people, and investigate the manner in which their cognitive and emotional constitutions con-

comitantly promote and interfere with goal-directed behavior. It implies, of course, that “Ration-

ality does not determine behavior… Instead, behavior is determined by the irrational and

nonrational elements that bound the area of rationality” (Simon 1945: 241).

Bounded rationality is not simply a lack of calculational ability (see Lupia et al. 2000: 9).

Simon, March, and Newell all stressed that calculations were a minimal part of the diﬃculty,

easily solved via notepads, calculating machines, or a bureau of accountants. Simon does write

extensively, however, about attention, emotion, habit, and memory, exploring the functionality

and dysfunctionality of these aspects of the architecture of human cognition. And it is true that

a prime component of the behavioral model of choice is diﬃculty in planning and executing

long behavior sequences in complex environments (Jones 2001: 61). But this aspect of the

model should not be confused with calculational diﬃculties.

Principle of adaptation. The principle of adaptation stems most directly from the studies of Allen

Newell and Herbert Simon in human problem-solving, and is best stated in Simon’s The Sciences

of the Artiﬁcial (1996b). The claim is that most human behavior is explained by the nature of the

“task environment. ” Given enough time, human thought takes on the shape of the tasks facing

it—that is, human thought is adaptive and basically rational. Simon put it this way: “There are

only a few ‘intrinsic’ characteristics of the inner environment of thinking beings that limit the

adaptation of thought to the shape of the problem environment. All else in thinking and problem-

solving behavior … is learned and is subject to improvement” (Simon 1996b: 54). From this

principle comes the inference that, in general, the more time one spends on a problem, the

more likely the decision-maker’s understanding of the problem will approximate the actual task

environment, and the limitations of human cognitive architecture fades (Newell 1990).

Psychologists stress the distinction between central and peripheral mental processing (Fiske

and Taylor 1991: 475–80). Kuklinski and Quirk put it this way:

In central processing, used when attention and motivation are high, people employ more

mental resources, think more systematically, and allow data to shape inferences. In peripheral

processing, used when attention and motivation are low, they employ fewer resources, rely

on simple heuristics, and use top-down, stereotypic inferences.

(2000: 163)

Principle of uncertainty: One of the major contributions of decision theory has been to under-

stand uncertainty in light of the calculus of probabilities. We speak of “expected utility” and

think of outcomes as following a probability distribution. Unfortunately this does not tell the

full story of uncertainty in human decision-making. Students of human choice in real-world or

in laboratory situations again and again ﬁnd that people have great diﬃculties in working with

probabilities, assessing risk, and making inferences where uncertainty is involved. Indeed, a whole

ﬁeld of endeavor has emerged that studies the factors responsible for perceptions of risk; clearly

these perceptions are not just rooted in “nature” but also involve human psychology.

An underlying tenet of bounded rationality from its early years centered on how human cog-

nitive architecture interacted with an uncertain world; bounded rationalists saw uncertainty as

far more fundamental to choice than the probability calculus implied (March 1994). If one’s

understanding of the causal factors involved in a problem is hazy or ambiguous, then the

Bounded rationality

275

uncertainty is not contained, but reverberates through the entire thought process. If one is uncertain

about how to specify outcomes, then one must also be uncertain about how to specify one’s

utility function. Simon termed this diﬃculty “the design problem” to denote that the fundamental

nature of specifying a problem-space within which to solve problems.

Principle of trade-oﬀs. People have a very diﬃcult time with trading oﬀ one goal against another

in a choice (Slovak 1990; Tetlock 2000). The classical economic model depicts trade-oﬀsas

smooth indiﬀerence curves, and modern rational choice theory oﬀers little new in the theore-

tical study of trade-oﬀs. The ﬁrst behavioral tool for understanding trade-oﬀs was Simon’s notion

of “satisﬁcing.” His idea that “administrative man”—an individual in an organization—chooses

alternatives that are “good enough” has led critics to claim that the notion is just a poverty-

stricken version of maximization (Lupia et al. 2000: 9). However, if one adds information and

decision-making cost constraints to choice, this will not cause bounded rationality to dissolve

into maximizing behavior. Satisﬁcing describes the cognitive diﬃculties people have with trade-

oﬀs. Because of the operation of limited attention spans, people work on goals sequentially. As a

consequence, trade-oﬀs among goals is very diﬃcult. The response, argues Simon, is for people

to set “aspiration levels” on the goals they wish to achieve. If a choice is “good enough” (that

is, if it exceeds aspiration levels) on all goals, then it is chosen.

Other models of choice among multiple goals have been developed, including the lexicographic

strategy (choose the strategy that maximizes gain on the most salient goal and ignore the rest) and

elimination by aspects (use a lexicographic strategy unless there is a tie among alternatives; then

and only then use a second goal to break the tie). In an important particular, people have consider-

able diﬃculty in trading oﬀ beneﬁts against losses, something that standard utility maximization

theory treats as straightforward (Kahneman and Tversky 1979).

The behavioral theory of choice

While great strides have been made in recent years by psychologists and behavioral economists

studying choices in the controlled arrangements of the laboratory, only serious ﬁeld study can

indicate how real choices are made in the “structured yet dynamic” environments of real choice

situations. Bounded rationality and the behavioral theory of choice came from organization theory;

indeed, March (1994) once noted that breakthroughs in the study of human cognition were

likely to come from a study of organizations. Laboratory ﬁ ndings of systematic violations of

principles of behavior based on expected utility calculation are dramatic and widespread (Camerer

1998; Camerer and Thaler 1995; Thaler 1991, 1992), yet this says little about choice in the ﬁeld

where behavior can be adaptive and responsive to multiple feedback streams (Laitin 1999). As a

consequence of the single-minded focus on experimental design and rejecting various tenets of

expected utility, behavioral economists have termed their own literature “the anomalies litera-

ture” (Thaler 1988, 1992). Several political scientists have criticized experimental psychology

and the behavioral economics literature for its seeming ad hoc nature, building ﬁndings experi-

mental eﬀect by experimental eﬀect. David Laitin (1999) notes that in everyday actions people

are adaptive, avoiding many of the traps set for them in experiments.

Much recent debate has centered on the viability of expected utility theory given the laboratory

ﬁndings. This debate is to a large extent misplaced, because since the late 1950s a full model of

choice, the behavioral model of choice, has been available. It avoids the “anomalies” problem,

is parsimonious, and yields more accurate predictions on aggregate choices such as we observe

in studying policy processes than does comprehensive rationality.

Bounded rationality points to the limits of rational adaptation; behavioral choice theory

provides a body of literature that shows how human choice works. As we noted above, bounded

Jones and Thomas

276

rationality and the associated behavioral theory of choice are open-ended; we do not know

everything about human choice and we learn more every year. But we know a lot, and we

know enough to specify the outlines of what aspects of human cognition must be incorporated

to formulate a general theory of human choice. We would cite the following:

1 Long-term memory: humans learn by encoding experience (direct or secondary) into rules

that specify action to be taken in response to categories of stimuli.

2 Short-term memory: human cognitive capacities include a “front end” that extracts features

from the world around them, categorizes them as relevant or irrelevant (in the former

case, they become “stimuli”) and prioritizes them.

3 Emotions set priorities: in an initial encounter with a choice situation, the major mechanism

for weighting the relevance of stimuli is emotion.

4 Central versus peripheral processing: when attention and emotion are aroused, information

processing shifts toward problem analysis and search. When they are not, the decision-maker

relies on prepackaged solutions.

5 The preparation-search trade-oﬀ: if the front-end system indicates a need for action, humans can

take two paths: draw upon previously prepared and stored rules specifying how to respond

to the category that the stimulus has been placed in, or search for new responses.

6 Identiﬁcation: People identify emotionally with the previously prepared solutions that they

have encoded in memory. They tend to become emotionally attached to their current reper-

toire of encoded solutions, even as the problems they face evolve. As a consequence, reliance

on prepared solutions dominates search.

Heuristics

Heuristic decision-making plays a key role in most of the basic theories of policy processes. The

notion was a critical component of the problem-solving studies of Newell and Simon (1972).

Heuristics, or shortcuts, are strategies that people employ to cope with the bounds of their

cognitive architectures. Jon Bendor (2010: 1) notes that political science harbors two orienta-

tions toward heuristics. On the one hand some political scientists report that “decision-makers

often manage to do ‘reasonably well’—even in complex tasks—despite their cognitive limitations.”

He puts Lindblom’s theory of “muddling through” (1959) and Wildavsky’s (1964) budgetary

studies in this category. On the other hand, in the work of other political scientists, “the emphasis

is on how people make mistakes even in simple tasks,” an orientation common in much of

psychology (where the terms “heuristics and biases” often go together), and in studies of voting

behavior in political science. Newell and Simon saw that both facets were operative at once. For

example, in their experiments, when people found that a problem-solving strategy would not

work, they found another through trial and error search. In contrast, when presented with a

problem logically similar to the one they just addressed, many went back to the failed ﬁrst-choice

strategy of the earlier problem.

A critical component of the problem-solving approach of Newell and Simon is the separation

of problem from solution spaces (Newell 1990). Most decision-making occurs without much

analysis of the problem-space. Decision-makers assume that they know the dimensions of a

problem and apply a known solution from memory. Basically, decision-makers can either search

memory for a prepackaged solution, or they can engage in search for new solutions. Because it

is less costly to use a prepackaged solution from memory, this is the route most typically taken.

The separation of problem-space and solution-space is a fundamental component of the

policy process ﬁeld. The garbage can theory of organizational decision-making rests on this

Bounded rationality

277

distinction (Cohen et al. 1972), as does the work of Kingdon (1984) on agenda-setting and

public policy. In these studies, solutions and problems are not linked causally, but are uniﬁed by

collective attention. When a problem emerges, solutions are likely to be attached without prior

analysis of the problem-space.

Fernandes and Simon (1999) applied the process-tracing methodology initially developed by

Newell and Simon (1972) in their problem-solving experiments to the complex and ill-structured

problems characteristic of policy issues. Fernandes and Simon wondered if the professional iden-

tiﬁcations among doctors led to diﬀerent problem-solving strategies. They found a heavy reli-

ance on what they termed a Know ! Recommend strategy among many participants that

hindered their use of information in problem-solving.

Ill-structured problems lend themselves to the application of prepackaged solution sets that

participants bring to the problem-solving enterprise. These prepackaged “solution sets” (Jones

and Bachelor 1993) can derive from ideology, or professional identiﬁcation, or current organi-

zational practices. Whatever its source, it suggests a major limitation in one of the fundamental

principles of bounded rationality: what we termed the principle of adaptation above. But it oper-

ates away from comprehensive rationality, and it cannot be construed as some sort of “heuristic”

shortcut for limiting search costs. It may be in moderation, but current studies suggest that it is

overused to the point of interfering with adaptation. This is an important ﬁnding capable of

being linked to collective decision-making (see Brown 2001).

The Know ! Recommend strategy has powerful implications for public policy. In many cases

in which attention and emotion are aroused, people may insist on rigidly reaching for old solutions

to problems, regardless of their applicability to the current issue. Hinich and Munger (1994)

note that ideologies organize the political (and policy) world, so it is not surprising that solutions

more amenable to a particular ideological world view are viewed as appropriate, even when

evidence shows them to have failed in the past. Jones and Williams (2008), who studied tax

policies, and Schrad (2010), who studied alcohol control, show in two very diﬀerent policy-

making situations the role of bad policy ideas in leading to policy failures. Bad policy ideas are

deﬁned as those resulting in a policy that “isn’t working, hasn’t ever worked, and runs the risk

of causing great damage” (Schrad 2010: 208). Yet, the committed activists in both cases rushed

to claim eﬀects for policies that had already been discredited.

This commitment to prior solutions, based on cognitive and emotional identiﬁcation, pro-

vide an important limit to the principle of adaptation. In such cases, much central processing

eﬀort is expended, but most of it goes to justifying previously employed solutions —a condition

known as motivated reasoning (Kunda 1990; Lodge and Taber 2000).

Organizations

The six aspects of human cognition, and their inﬂuences on heuristic decision-making, obviously

do not deal with every aspect of human behavior that could inﬂuence the policy process, but

they cover much ground, and lay the basis for a general behavioral theory of choice in organizations

and institutions. Most importantly, they form a basis for “scaling up” the behavior of humans to

organizations and policy processes. One of the continuing challenges for those scholars working

in the bounded rationality tradition is the movement from models of human decision-making

to those characterizing organizations and interactions of organizations in policy-making systems.

While organizations clearly free humans by extending their capacities to achieve goals, being

human inventions they also fall prey to aspects of human cognitive architecture in predictable

ways. Major aspects of the behavioral theory of organization parallel major facets of the behavioral

theory of human choice:

Jones and Thomas

278

1 Organizational memory: organizations encode experience into rules, routines, and standard

operating procedures that specify action to be taken in response to categories of stimuli.

2 Agenda-setting: organizational capacities include a “front end” that extracts features from the

world, categorizes them as relevant or irrelevant (in the former case, they become “stimuli”)

and prioritizes them. Agenda-setting in organizations parallels the short-term and attention

“bottleneck” (Simon’s term) aﬄicting human cognition.

3 Parallel processing: a major way that organizations expand human capacities is the ability to

process information in parallel. By decentralizing and delegating, organizations can process

multiple streams of input simultaneously (Simon 1983; Jones 1994). This organizational

strategy presupposes considerable “peripheral processing” relying on preprogrammed solutions.

4 Serial processing: the search for new solutions is activated only when previously prepared

solutions encoded in organizational routines are judged to be inadequate. Then organizations

move from peripheral to central processing (or from parallel processing to serial processing).

5 Emotional contagion: in policy-making, emotional commitment and contagion are crucial

elements in mobilizing for major initiatives. Moving from parallel to serial processing is

invariably accompanied by emotional arousal by participants (Jones 1994).

6 Identiﬁcation: people identify emotionally as well as cognitively with the organizations they

participate in, or even parts of an organization, which Simon termed “sub-goal identiﬁcation.”

Organizational identiﬁcation is a great resource for leaders. Patriotism or religious zeal or

even pride in performing their jobs can push people to actions that would be unthinkable

in a calm cost-beneﬁt analysis. But it also can make it diﬃcult for leaders to shift strategies

when they ﬁnd it necessary to do so.

The relationships between organizational decision-making and individual decision-making

are not metaphorical; they are causal (Jones 2001). One cannot really understand how organi-

zations operate without a strong sense of how individuals process information and make decisions.

As a consequence, a ﬁrm scientiﬁc foundation for policy studies must be rooted in a behavioral

approach to organizations (see Green and Thompson 2001).

Information-processing

A political science relying on behavioral choice theory will invariably be drawn toward the

study of information-processing and problem-solving. It will be less focused on questions of

preferences and equilibria. The nature of the behavioral assumptions inﬂuences the choice of

topics for study. It is not that preferences and equilibria are not interesting subjects, well worthy

of study. It is that they have been attended to far out of proportion to their explanatory power.

The focus on equilibrium processes has as its underpinning a narrow conception of choice,

based on the notion of ﬁxed preferences that lead actors to choose strategies that maximize

those preferences. It is often found in modern studies of legislatures and public agencies. Early

students of public administration were concerned with how organizations and democracy were

intermeshed, whether a pluralism of interests generated by Roosevelt’s “alphabet soup” of reg-

ulatory agencies could ﬁt with a single “overhead control” model of democracy. However, much

recent scholarship has relied heavily on the principal agent approach, in which the issue for a

principal, such as collectively acting legislature, involves controlling the actions of a delegate,

such as a bureaucratic agency, which receives delegated authority. This has resulted in a public

administration that has overly focused on the single issue of control—and control solely through

formal incentives (primarily punishment). The rich insights of behavioral choice have devolved

into a discredited Skinnerian psychology—the incentive controls the behavior. It is ironic that

Bounded rationality

279

Skinner’s study of pigeons and mice and the comprehensive rationality of economics both lead

to the same impoverished model of human choice: only formal incentives matter.

The adoption of an overly limited model of human behavior has led to inadequate devel-

opment of other aspects of bureaucratic behavior. Models based on principal agency have pro-

blems in conﬁrmation (see, for example, Brehm and Gates 1997; Balla 1998; Carpenter 1996;

Balla and Wright 2001). Brehm and Gates (1997: 578) comment that “the primary contribution

is … our ﬁnding of the overwhelming importance of attributes of the organizational culture in

determining subordinates’ levels of compliance.” Miller (1992) has shown the critical importance

of organizational culture as well.

In addition, public bureaucracies are important in the process of providing and interpreting

information for the legislative branch. Because the environment of policy-making is both

uncertain and ambiguous, delegating information processing to public agencies is common

and necessary in legislatures. One cannot control what one does not understand (Workman

et al. 2009).

Attention-driven choice in public policy

The basis for the behavioral model of choice is the processing of information. Information is not

predeﬁned or packaged; rather it is often vague, ambiguous, and most importantly, is generated

from multiple sources. The receiver of the information is as important as the sender (Jones 2001:

Chapter 4). In modern complex environments, neither individuals nor organizations respond

simply to stimuli. They must attend, prioritize, and select an appropriate response. As a consequence,

there is no clear, one-to-one mapping between potential stimuli or events and actions.

A major key to understanding information-processing in people and in organizations is the

allocation of attention. In his study of municipal budgeting, John Crecine (1969) noted that city

agencies developed attention rules that indicated what aspects of the environment ought to be

monitored for indicators of change that could need addressing. These rules did not tell the

agency what to do; only what to attend to. Similarly Jones (1980) found in a study of Chicago

building code enforcement that informal norms generally supplanted the complexities of the code,

but that supervisors occasionally sent out signals to ﬁeld inspectors that all violations were to be

recorded in potentially “hot” cases. Diﬀerential code enforcement resulted as a consequence of

these attention rules. Research by Armstrong et al. (2003) indicates that media attention to disease

is not simply related to mortality and morbidity of disease, and this attention seems to mediate

output indicators, such as investment in cures and related science.

Attention is diﬀerent from any other resource type variable because one cannot allocate it

proportionally to one’s priorities at any one instance. Attention is selective; select one aspect of

an environment for study and inattention must be paid to the rest of the environment. Attention is

partially under the control of a decision-maker, but cognitively we possess no comprehensive

system for monitoring when enough attention has been devoted to a topic. As a consequence,

shifts in attentiveness are in large part hostage to emotional arousal. Because attention shifts are

governed by emotion, they are unavoidably disjointed. Since, in many cases, the devotion of other

resources follows attention, past decisions are a residue of past allocations of attention. They

may or may not be consistent—great inconsistencies in choices are a result of attention.

The allocation of attention strongly aﬀects the manner in which individuals or organizations

prioritize problems for action. Because attention is disjointed and episodic, so is problem-

prioritization. Because problems are prioritized in policy-making systems through attention, and

attention is disjointed, policy-making is disjointed and episodic, not necessarily related to the

severity of indicators. Similarly attention to solutions given that a problem has been prioritized

Jones and Thomas

280

is governed by attention—oftentimes directed at solutions that have been used before, even

when they have been unsuccessful in the past (Jones and Williams 2008; Schrad 2010).

Disproportionate information-processing and institutional friction

So far, we have addressed the scaling up of the model of bounded rationality to the level of the

single organization. Now we will need to address full policy-making systems—those composed

of interacting organizations. The key to this scaling up is the notion of disproportionate information-

processing (Jones 2001). A policy-making system may be considered fully adaptive to the extent

that it responds appropriately to the information coming in from its environment. Yet systems

comprised of boundedly rational decision-makers are not capable of responding proportionately—

a direct implication of attention-driven choice and other mechanisms rooted in bounded

rationality (Jones and Baumgartner 2005). The result is episodic policy-making as policy-makers

shift from addressing one set of problems to a diﬀerent set.

Research focusing on explaining the episodic nature of policy change over time cites two

mechanisms: institutional friction and positive feedback. Disproportionate information-processing

in policy-making systems stems from a continual interaction between the conservative forces of

policy stasis governed by negative feedback processes and the cascades of policy action governed

by a collective sense of urgency among policy-makers. Jones and Baumgartner (2005) examine

how government attends to, interprets, and prioritizes diverse incoming information streams

within a boundedly rational framework. In their approach, human cognitive processing interacts

with policy-making institutions to determine responses to incoming information.

In the model, institutional friction results from the interaction of two kinds of costs: decision

costs, which are those imposed by the operation of institutional rules (such as those required by

the constitutional system of divided powers in American federal government) and cognitive costs

associated with boundedly rational decision-makers operating in a complex and changing

environment (Jones et al. 2003; Jones and Baumgartner 2005). Baumgartner et al. further deﬁne

the mechanism:

Institutional friction in the form of sunk costs, long-term budgetary commitments, identi-

ﬁcation with means rather than ends, and bureaucratic inertia makes it hard for govern-

ments to reduce attention to issues that are improving just as it inhibits them from paying

attention to problems that are just emerging.

(2009: 608)

Empirical results in the US and elsewhere show that increased friction from these sources lead

to more pronounced patterns of overreaction and underattention in policy-making. If friction is

understood as strong resistance to change over time (and not as the static concept of ‘gridlock’),

major shifts in policy can occur when it is overcome (Jones and Baumgartner 2005). This generally

occurs when “errors accumulate”—that is, when the system is “far enough” out of balance with

its environment that it lurches forward.

Disproportionate information-processing and positive feedback

The friction approach to incorporating disproportionate information processing into a model of

policy change is essentially passive—pressures build up and cause attention to shift from one

problem to another, which is then addressed. This resistance is in part a function of cognitive

limitations and heuristic decision-making, so it incorporates elements of bounded rationality. A

Bounded rationality

281

second aspect of disproportionality, based on positive feedback, is active. Positive feedback

operates to amplify a process rather than to damp it down (as is the case for negative feedback

processes). Positive feedback mechanisms can lead to contagion eﬀects and the breakdowns of

control mechanisms, and are well known in both natural and social sciences.

Well-ordered and equilibrium processes always involve negative feedback systems. The “wisdom

of crowds” argument (Surowiecki 2004) claims that decisions or judgments made through the

aggregation of individual choices often leads to accurate and rational conclusions. So it seems

that rational decisions in systems stem from averaging out the individual errors. The theory is

based on independent decision-making, each actor making his or her own judgment. Unfor-

tunately, communication and interaction among individuals can easily lead to the rapid break-

down of the theory, as mimicking can bias outcomes. Experimental evidence in sociology conﬁrms

this argument, suggesting that the breakdown of the “wisdom of crowds” eﬀect can be undermined

through social inﬂuence among actors (Lorenz et al. 2010).

A weight-guessing game, perhaps involving a carnival crowd oﬀering suggestions on a

particular hog’s size, oﬀers an illustrative example. Here, the independence of individuals and a

lack of communication would likely yield a close prediction of the hog’s actual size at the

average (Galton 1907). Yet if carnival attendees are allowed to communicate and discuss

their potential guesses while they examine the hog and place their bets, a bounded ration-

ality microfoundation would predict considerable deviation from the actual weight of the

hog as cue-taking and information cascades ensue. A propensity to mimic neighbors and

authority ﬁgures, take cues to reduce decisions costs, and the cascade of information (say from

one side of the hog to the other) would generate a set of guesses subject to ﬂuctuation, bias, and

inaccuracy.

This ampli ﬁcation mechanism is found in political bandwagons, in fads of fashion, and in

ﬁnancial markets, bubbles and crashes (Mackay 1841). Shiller (2005) discusses the psychology of

herding, or irrational exuberance, among investors whereby price increases beget further price

increases in a consecutive loop of speculative conﬁdence. Sornette (2004) and Youssefmir et al.

(1998) focus on how unchecked feedback can generate asset-pricing bubbles and cause their

subsequent collapse. Beyond ﬁnance, examples of positive feedback and cascades abound, and

include: city growth (Gabaix 1999), the superstardom of pop musicians (Chung and Cox 1994),

fashion trends and cultural change (Bikhchandani et al. 1992), website links (Barbási 2002),

restaurant popularity (Becker 1991), standing ovations (Miller and Page 2004), and credit for

scientiﬁc ﬁndings (Merton 1968), among others.

Political scientists address feedback mechanisms in research spanning the discipline (see Baum-

gartner and Jones 2002 for a review). For example, scholars of the news media (Boydstun 2008),

interest group behavior (Baumgartner and Leech 2001; Leech et al. 2005; Halpin 2011), agenda-

setting (Baumgartner and Jones 1993), state-level diﬀusion of innovations (Boushey 2011),

decision-making in Congress (Kingdon 1977; Matthews and Stimson 1975), presidential primaries

(Bartels 1988), collective action (Lohmann 1994) and social movements (Chong 1991) all discuss

the amplifying role of positive feedback in terms of cue-taking, bandwagons, cascades, contagion,

and momentum related to various political phenomena.

Linking these two mechanisms of positive feedback and institutional friction, Jones and

Baumgartner (2005) argue that cascades may be critically important to our understanding of

policy outcomes. They describe a gap between a distribution of real-world budget outlays and

another predicted by their friction model, and conclude that friction alone does not fully

explain budget distributions. They see positive feedback as a likely candidate for the diﬀerence,

and suggest that “it may turn out that cascades are the key to understanding how friction is

overcome in policy systems” (ibid. 139).

Jones and Thomas

282

Bounded rationality and stochastic processes

In rational models of action, people act proportionately to the information they receive. They

attend to everything simultaneously and make trade-oﬀseﬀortlessly. They avoid punctuated

outputs except where inputs are disjointed and episodic, or where rules do not allow for the

easy adjustment of outputs to inputs. Organizations composed of rational participants respond

directly and eﬃciently to information discounted by costs. Organizations composed of bound-

edly rational participants cannot avoid punctuated outputs, because they cannot adjust their

behaviors to incoming information of any degree of complexity.

Bounded rationality and behavioral choice lead to predictions about policy outcomes that

imply that organizational outputs will be disjointed and episodic regardless of ﬁrst, the input

stream and second, the cost structure of the organization. It is clear that decision costs in the policy-

making process can cause disjoint outputs. For example, in the American system of separated

powers, considerable changes in the preferences of policy-makers can occur without producing

policy change because of the need to assemble majorities in all responsible branches (Hammond

and Miller 1987). If we were able to discount these “decision costs” bounded rationality implies

that disjointed and episodic behavior would still occur. Examination of a number of diﬀerent

distributions of outcomes from political institutions and the policy-making process in the US,

ﬁnds that, regardless of the institutional cost structure, outputs are punctuated (Jones et al. 2003;

Jones and Baumgartner 2005).

Because of the disproportionate information processing described above, the traditional

approaches to analysis based on point prediction and regression analysis can be misleading. It is

too diﬃcult (and perhaps meaningless) to try to tie a particular event of ﬂow of information to a

particular outcome. Students of organizational processes have begun to make use of stochastic

process approaches, in which eﬀorts are made to understand the processes underlying an entire

distribution of outputs rather than a particular response to a policy innovation or other change

(Padgett 1980, 1981; Jones et al. 2003). Disproportionate information processing implies lepto-

kurtic (“fat-tailed”) distributions of output changes (such as annual changes in a government’s

budget) regardless of the input distribution. Because of the phenomenon of disproportionate

information-processing, decision-makers are forever underreacting to stimuli that require action,

and overreacting with cascading behavior when the problem is fully recognized (Jones 2001;

Jones et al. 2003). The underreacting results in the slender peak of the leptokurtic distribution,

while the fat tails indicate the rapid changes that occur in the overreacting stage.

Concluding comments

Psychologists, neuroscientists, behavioral economists, and students of political behavior have

accumulated a great amount of information since the studies of Newell and Simon 40 years ago. Not

only have these studies shown where failures in human rationality exist, they have also found

adaptive capacities in human action, even in complex modern environments (Gigerenzer and

Goldstein 1996; Gigerenzer 2001). In general, while the tenets of cognitive rationality have been

repeatedly undermined by these studies, those of bounded rationality have fared much better.

Policymaking systems are comprised of multiple organizations, each of which incorporates

numerous human actors. In order to serve as a microfoundation for understanding policy pro-

cess, we need to avoid getting overly consumed by the details of human cognition (see Bendor

2010: 44–7 for a discussion). What are the important elements of a microfoundation for the

study of policy processes, one rooted in the scientiﬁc and empirical ﬁndings on human cognation

and information-processing?

Bounded rationality

283

In this chapter we have discussed those elements of human cognition that we see as critical

for policy process scholars to understand, and shown how those elements may be scaled up to

the levels of the organization and the policy-making system. While models of strategic inter-

action among rational actors are (relatively) easy to scale up, the same cannot be said for a more

realistic empirically based cognitive foundation. As a consequence, less eﬀort has been expended

in achieving this scaling up than we think desirable.

Notes

1 This chapter draws heavily on Jones (2003).

2 Our thanks to Fred Thompson for drawing this link.

3 See Schelling 1978 for an early overview as well as Miller and Page 2007 for a recent review of positive

feedback in complex adaptive systems.

4 See also Pierson (2000) for a discussion of ‘increasing returns’ in political processes.

Bibliography

Alt, James E., Margaret Levi and Elinor Ostrom. 1999. Competition and Cooperation: Conversations With

Nobelists About Economics and Political Science. New York, Russell Sage Foundation.

Armstrong, Elizabeth M., Dan Carpenter and Marie E. Hojnacki. 2003. ‘Whose deaths matter? Attention to

disease in the public arena’, paper presented at the Conference on the Politics of Biomedical Research,

Woodrow Wilson School of Public and International Aﬀairs. Princeton, NJ: Princeton University, March.

Balla, Steven J. 1998. ‘Administrative procedures and political control of the bureaucracy’, American Political

Science Review 92: 663–73.

——and John R. Wright. 2001. ‘Interest groups, advisory committees, and congressional oversight’,

American Journal of Political Science 45: 799–81.

Barabási, Albert-László. 2002. Linked. New York: Penguin.

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago:

University of Chicago Press.

——and Bryan D. Jones, eds. 2002. Policy Dynamics. Chicago: University Press of Chicago.

——and Beth L. Leech. 2001. ‘Issue niches and policy bandwagons: Patterns of interest group involvement

in national politics’, Journal of Politics 63: 1191 –213.

——, Christian Breunig, Christoﬀer Green-Pedersen, Bryan D. Jones, Peter B. Mortensen, Michiel

Nuytemans and Stefaan Walgrave. 2009. ‘Punctuated equilibrium in comparative perspective’, American

Journal of Political Science 53(3): 603–20.

Becker, Gary S. 1991. ‘A note on restaurant pricing and other examples of social inﬂuence on price’,

Journal of Political Economy 99: 1109–16.

Bendor, Jonathan. 2010. Bounded Rationality and Politics. Berkeley, CA: University of California Press.

Bikhchandani, Sushil, David Hirshleifer and Ivo Welch. 1992. ‘A theory of fads, fashion, custom, and

cultural change as informational cascades’, Journal of Political Economy 100: 992–1026.

Boushey, Graeme. 2011. Policy Diﬀusion Dynamics. Cambridge: Cambridge University Press.

Boydstun, Amber E. 2008. ‘How policy issues become front-page news’,

PhD dissertation, Penn State University.

Brehm, John and

Scott Gates. 1993. ‘Donut shops and speed traps: Evaluating models of supervision on

police behavior’, American Journal of Political Science 37: 555–81.

—— 1997. Working, Shirking, and Sabotage. Ann Arbor, MI: University of Michigan Press.

Brown, Steven R. 2001. ‘Structural and functional information: A cautionary note to Fernandes and Simon’,

paper presented at the annual meeting of the Society for the Policy Sciences New Haven, Connecticut.

Camerer, Colin F. 1998. ‘Behavioral economics and nonrational organizational decision making’ in Jennifer

J. Halpern and Robert N. Stern (eds) Debating Rationality. Ithaca, NY: Cornell University Press.

——and Richard F. Thaler. 1995. ‘Ultimat ums, dictators and manners’, Journal of Economic Perspectives 9: 209–19.

Carpenter, Daniel. 1996. ‘Adaptive signal processing, hierarchy, and budgetary control in federal regulation’,

American Political Science Review 90: 283–302.

Chong, Dennis. 1991. Collective Action and the Civil Rights Movement. Chicago: University of Chicago Press.

Chung, Kee H. and Raymond A.K. Cox. 1994. ‘A stochastic model of superstardom: An application of

the Yule distribution’, Review of Economics and Statistics 76(4): 771–5.

Jones and Thomas

284

Cohen, Michael D., James G. March and Johann P. Olsen. 1972. ‘A garbage can model of organizational

choice’, Administrative Science Quarterly 17: 1–25.

Crecine, John. 1969. Government Problem-Solving. Chicago: Rand-McNally.

Fernandes, Ronald and Herbert A. Simon. 1999. ‘A study of how individuals solve complex and ill-structured

problems’, Policy Sciences 32: 225–45.

Fiske, Susan and Shelly Taylor. 1991. Social Cognition. New York: McGraw-Hill.

Galton, Francis. 1907. ‘Vox populi’, Nature 75: 450–1.

Gabaix, X. 1999. ‘Zipf’s Law and the growth of cities’, American Economic Review 89(2): 129–32.

Gigerenzer, Gerd. 2001. The Adaptive Toolbox. Cambridge, MA: MIT Press.

——and Daniel G. Goldstein. 1996. ‘Reasoning the fast and frugal way: Models of bounded rationality’,

Psychological Review 103: 650–69.

Granovetter, M.S. 1973. ‘The strength of weak ties’, American Journal of Sociology 78(6): 1360–80.

—— 1978. ‘Threshold models of collective behavior’, American Journal of Sociology 83(6): 1420.

Green, Mart T. and Fred Thompson. 2001. ‘Organizational process models of budgeting’, in John Bartle

(ed.) Research in Public Administration: Evolving Theories of Budgeting. San Francisco: JAI Press.

Halpin. D. 2011. ‘Organized interests and cascades of attention: Unpacking policy bandwagon

dynamics’, Governance, 24(2); 205–30.

Hammond, Thomas and Gary J. Miller. 1987. ‘The core of the constitution’, American Political Science

Review 81: 1155–74.

Hinich,

Melvin and Michael

Munger. 1994. Ideology and the Theory of Political Choice. Ann Arbor, MI:

University of Michigan Press.

Jones, Bryan D. 1980. Service Delivery in the City: Citizen Demand and Bureaucratic Response. New York: Longman.

—— 1994. Reconceiving Decision-Making in Democratic Politics. Chicago: University of Chicago Press.

—— 2001. Politics and the Architecture of Choice. Chicago: University of Chicago.

—— 2003. ‘Bounded rationality and political science: Lessons from public administration and public

policy’, Journal of Public Administration and Theory 13(4): 395–412.

——and Lynn W. Bachelor. 1993. The Sustaining Hand: Community Leadership and Corporate Power. Lawrence,

KS: University Press of Kansas.

——and Frank R. Baumgartner. 2005. The Politics of Attention: How Government Prioritizes Problems. Chicago:

University of Chicago Press.

——and Walter Williams. 2008. The Politics of Bad Ideas. New York: Pearson Longman.

——, Tracy Sulkin and Heather Larsen. 2003. ‘Punctuations in political institutions’, American Political

Science Review 97: 151–69.

——, Heather Larsen-Price and John Wilkerson. 2009. ‘Representation and American governing institutions’,

Journal of Politics 71: 277–90.

Kahneman, Daniel and Amos Tversky. 1979. ‘Prospect theory: An analysis of decision-making under risk’,

Econometrica 47: 263–91.

Kingdon, John W. 1977. ‘Models of legislative voting’, The Journal of Politics, 39(3): 563–95.

—— 1984. Agendas, Alternatives, and Public Policies. Boston: Little, Brown.

Kuklinski, James and Paul Quirk. 2000. ‘Reconsidering the rational public: Cognition, heuristics and mass

opinion’, in Arthur Lupia, Mathew D. McCubbins and Samuel L. Popkin (eds) Elements of Reason.

Cambridge: Cambridge University Press.

Kunda, Ziva. 1990. ‘The case for motivated reasoning’, Psychological Bulletin 108: 480–98.

Laitin,

David. 1999. ‘Identity choice

under conditions of uncertainty: reﬂections on Selten’s dualist meth-

odology’, in James Alt, Margaret Levi and Elinor Ostrom (eds) Competition and Cooperation: Conversations with

Nobelists about Economics and Political Science. New York: Russell Sage.

Leech, Beth L., Frank R. Baumgartner, Timothy La Pira and Nicolas A. Semanko. 2005. ‘Drawing lob-

byists to Washington: Government activity and interest-group mobilization’, Political Research Quarterly

58(1) (March): 19–30.

Lindblom, Charles E. 1959. ‘The science of “muddling through”’, Public Administration Review 19(2): 79–88.

Lodge, Milton and Charles Taber. 2000. ‘Three steps toward a theory of motivated political reasoning’,in

Arthur Lupia, Mathew D. McCubbins and Samuel L. Popkin (eds) Elements of Reason. Cambridge:

Cambridge University Press, 183–213.

Lohmann, Susanne. 1994. ‘The dynamics of informational cascades: The Monday demonstrations in

Leipzig, East Germany, 1989–1991’, World Politics 47: 42–101

Lorenz, Jan, Hekio Rauhut, Frank Schweitzer, and Dirk Helbing. 2010. ‘How social movements can

undermine the wisdom of crowd eﬀect’, PNAS 108(22): 9020–9025.

Bounded rationality

285

Lupia, Arthur, Mathew D. McCubbins and Samuel L. Popkin. 2000. ‘Beyond rationality: Reason and the

study of politics’, in Arthur Lupia, Mathew D. McCubbins and Samuel L. Popkin (eds) Elements of

Reason. Cambridge: Cambridge University Press.

Mackay, Charles. 1841. Extraordinary Popular Delusions and the Madness of Crowds . New York: Broadway.

March, James 1994. A Primer on Decision-Making. New York: The Free Press.

——and Herbert A. Simon. 1958. Organizations. New York: John Wiley & Sons.

Matthews, D.R. and James A. Stimson. 1975. ‘Yeas and nays: Normal decision-making in the US House

of Representatives’, New York: John Wiley & Sons.

Merton, Robert K. 1968. ‘The Matthew eﬀect in science’, Science 159: 56–63.

Miller, G.J. 1992. Managerian Dilemmas: The Political Economy of Hierarchy. New York: Cambridge University

Press, 216–333.

Miller, John H. and Scott E. Page. 2004. ‘The standing ovation problem’, Complexity 9: 8–16.

——and Scott E. Page. 2007. Complex Adaptive Systems. Princeton, NJ: Princeton University Press.

Mladenka, Kenneth. 1978. ‘Rules, service equity and distributional decisions’, Social Science Quarterly

59: 192–202.

Newell, Allen.1990. Uniﬁed Theories of Cognition. Cambridge, MA: Harvard University Press.

——and Herbert A. Simon. 1972. Human Problem Solving. Englewood Cliﬀs, NJ: Prentice Hall.

Padgett, John F. 1980. ‘Bounded rationality in budgetary research’, American Political Science Review 74: 354–72.

—— 1981. ‘Hierarchy and ecological control in federal budgetary decision making’, American Journal of

Sociology 87: 75–128.

Pareto, Vilfredo. 1935. Mind and Society. New York: Kessinger Publishing.

Pierson, Paul. 2000. ‘Increasing returns, path dependence, and the study of politics’, The

American Political

Science Review 94(2): 251–67.

Robinson,

Scott. 2003. ‘Punctuated equilibrium, bureaucratization, and budgetary changes in schools’,

paper presented at the Midwest Political Science Association, Chicago, April 4–6.

Schelling, Thomas. C. 1978. Micromotives and Macrobehaviour, New York: Norton.

Schrad, Mark Lawrence. 2010. The Political Power of Bad Ideas. Oxford: Oxford University Press.

Shiller, R.J. 2005. Irrational Exuberance. Princeton, NJ: Princeton University Press.

Simon, Herbert A. 1945. Administrative Behavior. New York: Macmillan.

—— 1977. ‘The logic of heuristic decision-making’, in R.S. Cohen and M.W. Wartofsky (eds) Models of

Discovery. Boston: D. Reidel.

—— 1983. Reason in Human Aﬀairs. Stanford, CA: Stanford University Press.

—— 1996a. Models of My Life, MIT edn. Cambridge, MA: MIT Press.

—— 1996b. The Sciences of the Artiﬁcial, 3rd edn. Cambridge, MA: MIT Press.

—— 1999. ‘The potlatch between political science and economics’, in James Alt, Margaret Levi and

Elinor Ostrom (eds) Conﬂict and Cooperation: Conversations with Nobelists about Economics and Political Science.

New York: Russel Sage.

Slovak, Paul. 1990. ‘Choice’, in Daniel N. Osherson and Edward E. Smith (eds) Thinking: An Invitation to

Cognitive Science, vol. 3. Cambridge, MA: MIT Press.

Sornette, D. 2004. Why Stock Markets Crash. Princeton, NJ: Princeton University Press.

Surowiecki, James. 2004. The Wisdom of Crowds. New York: Doubleday.

Thaler, Richard H. 1991. Quasi Rational Economics. New York: Russell Sage.

—— 1988. ‘Anomalies: the ultimatum game’, Journal of Economic Perspectives 2: 195–206.

—— 1992. The Winner’s Curse: Paradoxes and Anomalies of Economic Life. Princeton, NJ: Princeton University Press.

Tetlock, Philip. 2000. ‘Coping with trade-oﬀs: Psychological constraints and political implications’,i

Arthur Lupia, Mathew

D. McCubbins and Samuel L. Popkin (eds) Elements of Reason. Cambridge:

Cambridge University Press.

True, James L. 2000. ‘Avalanches and incrementalism: Making policy and budgets in the United States’,

American Review of Public Administration 30: 3–18.

——, Bryan D. Jones and Frank R. Baumgartner. 1999. ‘Punctuated equilibrium theory’, in Paul Sabatier

(ed.) Theories of the Policy Process. Boulder, CO: Westview Press.

Tversky, Amos. 1972. ‘Elimination by aspects: A theory of choice’, Psychological Review 79: 281–99.

Wildavsky, Aaron. 1964. The Politics of the Budgetary Process. Boston: Little, Brown.

Workman, Samuel, Bryan D. Jones and Ashley E. Jochim. 2009. ‘Information processing and policy

dynamics’, Policy Studies Journal 37(1): 75–92.

Youssefmir, Michael, Bernadro A. Huberman, and Tad Hogg. 1998. ‘Bubbles and market crashes’, Computational

Economics 12(2): 97–114.

Jones and Thomas

286

Incrementalism

Michael Hayes

Incrementalism is a pluralistic process of policy-making involving mutual adjustment among multi-

ple actors who typically disagree on objectives and start oﬀ from very diﬀerent conceptions of the

problem at hand. Incomplete knowledge and time constraints limit attention to alternatives diﬀer-

ing only marginally (incrementally) from previous policies. The necessity for bargaining and com-

promise virtually assures incremental outcomes. Major policy change will occur gradually, if at all, as

experience with policies generates demands for modiﬁcation or expansion in subsequent policy cycles.

Charles Lindblom advanced incrementalism as both an explanatory and a prescriptive model

of the policy process (Dahl and Lindblom 1953; Lindblom 1959, 1965, 1979; Braybrooke and

Lindblom 1963). In his view, incrementalism provides the best explanation of the policy process

because it is the best way to make policy, allowing policy-makers to proceed when the rational-

comprehensive ideal has broken down.

The ﬁrst section of this chapter will review how rational decision-making breaks down in

most cases and how incrementalism departs from the rational method on all points. The second

section will identify some circumstances under which nonincremental policy change may occur.

The third section will explore whether incrementalism really is the best way to make public

policies, as Lindblom asserts.

Incrementalism as an alternative to the rational ideal

Ideally, policies would result from a rational analysis culminating in a value-maximizing choice

after a thorough examination of all relevant alternatives. Unfortunately, this almost never happens,

for a variety of reasons.

The breakdown of rationality: For decisions to be made rationally, public problems must be

perceived and deﬁned accurately. In reality, there is no guarantee that problems will be perceived

at all and almost no likelihood that they will be deﬁned the same way by all participants. Items

will reach the agenda because they have attracted political support from organized interests or

the mass public, not because they have been rationally or scientiﬁcally identiﬁed.

Moreover, policy-makers must agree on objectives. This means more than just sharing the

same broad goals. All involved must also share the same priorities, ranking goals in the same way

and agreeing on exactly how to allocate ﬁnite resources. This condition is almost never satisﬁed.

287

Third, policy-makers must estimate accurately the consequences of all alternatives. Policy-makers

are almost never able to do this. On the contrary, the question of what consequences will follow

from various alternatives is almost always vigorously contested.

Clearly, rational decision-making is unattainable most of the time. Where there is no agreement

on objectives, it is impossible to say which values should be maximized. Where the knowledge

base is inadequate, there is no way to say which alternatives have the best consequences.

Incrementalism rejects the very idea of policies as the product of decisions. According to

Lindblom, policies are not the products of rational choice but rather the political resultants of par-

tisan mutual adjustment among various actors possessing diﬀerent information, adhering to diﬀerent

values, and driven by diﬀerent individual or group interests (Lindblom 1965).

The advantages of incrementalism: by contrast, incrementalism requires neither comprehensive infor-

mation nor agreement among policy-makers on objectives. It permits action where the rational

ideal oﬀers no guidance to policy-makers.

For example, policy-makers almost never face a “given” problem (Lindblom 1980: 24). Rather,

problems are brought to government through a process Lindblom terms the social fragmentation of

analysis (Braybrooke and Lindblom 1963: 104–6). No single actor possesses comprehensive infor-

mation on the problem. Rather, each brings to the table some portion of the knowledge that is

required to analyze the problem. The dispersal of essential knowledge throughout the system

makes some form of social fragmentation of analysis inevitable. No policy-maker could accu-

rately specify the varying value preferences of diﬀerent individuals or assess the impact of policy

proposals on diﬀerent groups without some input from these groups. Because disagreements can

be accommodated through bargaining, problems may be, and often are, acted on without ever

being fully deﬁned.

Incrementalism avoids the need to secure agreement on objectives by focusing on concrete

problems to be alleviated (for example, unemployment) rather than on abstract ideals to be

attained (a self-actualized citizenry). The process moves away from problems rather than toward

ideals, which can never be speciﬁed with suﬃcient precision to permit rational analysis. Lindblom

terms this element of the model remediality (Braybrooke and Lindblom 1963, 102–4).

Ends and means are typically considered simultaneously because diﬀerent alternatives embody

diﬀerent trade-oﬀs among values. When policy-makers choose one of several alternatives, they

are also choosing which trade-oﬀs should be made among values. In practice, ends and means

cannot be separated as called for by the rational model, with a decision on values preceding any

decision on alternatives. Any decision among alternatives is inherently and necessarily a decision

about values. Lindblom terms this element of the model the adjustment of objectives to policies

(Braybrooke and Lindblom 1963: 93–8).

Time constraints preclude a comprehensive examination of all alternative solutions. Policy-

makers must somehow limit their attention to a manageable number of options. In practice,

they accomplish this by limiting their focus to incremental alternatives (Braybrooke and Lindblom

1963: 88–90). Major policy departures are unlikely to be enacted. Established policies acquire

powerful supporters opposed to any signiﬁcant changes in policy. Thus prolonged consideration

of nonincremental policy proposals is likely to be a waste of time. Moreover, policy-makers build

up a reservoir of knowledge through experience with existing policies that they will be reluc-

tant to throw away by moving in an entirely diﬀerent direction. It makes more sense to build

on past policy than to begin again from the ground up.

Policy-makers further reduce the costs of analysis by focusing on the increments by which

various proposals diﬀer from one another and from past policies. Lindblom terms this process

margin-dependent choice (Braybrooke and Lindblom 1963: 83–8). In the budgetary process, for

example, legislators normally make no attempt to evaluate federal policies comprehensively.

Hayes

288

Rather, they focus on the increments by which spending will go up or down for various pro-

grams under diﬀerent proposals. Focusing on the margins permits an intelligent comparison of

proposals where a comprehensive analysis of the various competing policies would be infeasible

(Wildavsky 1992: 82–3).

Finally, the policy process is serial, or repetitive (Braybrooke and Lindblom 1963: 99–102).

Because policies can be modiﬁed in subsequent policy cycles, unanticipated consequences do

not pose a serious problem for incrementalism. They merely generate new problems to be dealt

with at a later date. Problems are considered unsolved as long as some publics continue to express

dissatisfaction with existing policies. By the same token, a policy may be regarded as “solved”

when it ﬁnally disappears from the agenda, crowded oﬀ by other problems now considered

more pressing (Wildavsky 1974: 60).

Preconditions for nonincremental change

While Lindblom is correct that policy-making will be incremental most of the time, research

suggests that nonincremental policy change may occur in response to crises, an aroused mass public

opinion, or the attainment of the conditions for rational decision-making. These three favorable

circumstances will be covered in turn below.

Crises as a source of nonincremental change: Lindblom identiﬁed conditions for nonincremental

change in a volume he coauthored with David Braybrooke (Braybrooke and Lindblom 1963:

66–79). He characterized problems as falling within four quadrants deﬁned by two continuous

dimensions. The ﬁrst dimension measured the policy-makers’ understanding of the problem at

hand; the second measured the degree of change a proposal would make from previous policy.

The four quadrants are shown below (see Table 22.1):

Within this scheme, rational decision-making is conﬁned to a relatively small number of

cases, primarily technical decisions made by professionals in the middle levels of the

bureaucracy—e.g. high understanding of problems is limited to policy alternatives that are

incremental in scope. There will be, in Lindblom’s view, no cases in which nonincremental

policy proposals are well understood. Most policy-making will occur within the realm of

normal incrementalism, where understanding is low and policy-makers conﬁne themselves to

incremental proposals.

Nonincremental policy change is possible in the remaining quadrant, which Braybrooke and

Lindblom term the realm of wars, revolutions, and grand opportunities. On a less sweeping

level, policy-makers may opt for nonincremental change any time a particular policy is judged

to be broken beyond repair. Under such circumstances a “calculated risk” is warranted, and

Table 22.1 Four quadrants of policy change

Incremental Nonincremental

High understanding Rational decision-making

is possible for small,

technical problems

This quadrant is empty;

no cases occur here

Low understanding The realm of normal

incrementalism;

most cases ﬁt here

Nonincremental change

possible here;

realm of “wars, crises, and grand

opportunities”

Source: Hayes (1992) Incrementalism and Public Policy, New York: Longman

Incrementalism

289

policy-makers may opt for major policy change even though all involved recognize the limits of

their inability to predict the consequences of their actions (Dahl and Lindblom 1953: 85).

Aroused mass public opinion as a catalyst for nonincremental change: Charles O. Jones’s case study of

the Clean Air Amendments of 1970 showed that nonincremental policy outcomes can result

where mass public opinion is aroused on an issue, forcing policy-makers to satisfy a “pre-formed

majority.” Although incrementalism provided a good description of policy-making on the air

pollution issue from 1941–1967, the increased salience of the issue in 1970 produced a very dif-

ferent kind of policy process, necessitating the development of an alternative model to describe

events (Jones 1974, 1975).

Under normal “majority-building incrementalism,” as understood by Jones, mass public opinion

is typically inattentive and ill-informed. Policy development takes place within policy commu-

nities, consisting of congressional committees or subcommittees with jurisdiction over the issue,

executive agencies with responsibility for administering laws within the issue area, and clientele

groups with a stake in the policies developed by these policy communities. As bills work their way

through multiple veto points, the need for bargaining and compromise virtually guarantees that

proposals will emerge weaker than they were initially. Jones terms this process “tapering demands

from the optimal down to the acceptable.”

Majority-building incrementalism could not explain air pollution policy-making in 1970,

however. Public opinion polls pointed to a dramatic rise in public concern over the pollution

issue, and the sharp rise in issue salience triggered the formation of many new interest groups

representing environmental interests. This public pressure for dramatic federal action constituted

a kind of “pre-formed majority,” obviating the normal need for tapering down. Policy-makers

who had shown little interest in the issue before (including the president) now saw it as having

electoral potential. This expanded circle of policy-makers scrambled to satisfy the highly aroused,

pre-formed majority, and air pollution legislation grew steadily stronger as it moved through

the House and Senate through a process Jones termed “ policy escalation.”

The end result, according to Jones, was a nonincremental policy change: Congress gave the

federal government sweeping new authority and set technology-forcing, health-based air quality

standards. However, the strong new law was not the product of any expansion in the scientiﬁc

knowledge base available to policy-makers but rather was driven primarily by the need to

appease an aroused mass public. For this reason, Jones characterized the 1970 policy outcome as

“legislating beyond capability.”

Attainment of the conditions for rational decision-making: throughout all his works on increment-

alism, Lindblom identiﬁed two preconditions for rational decision-making: agreement among

participants on objectives and the possession of a knowledge base suﬃcient to permit accurate

estimation of the consequences associated with various alternatives. In Lindblom’s view, incre-

mentalism describes how policy is made under normal circumstances because these two conditions

are almost never met.

While Lindblom is doubtless correct in asserting that these two conditions are not satisﬁed for

most policy problems, the failure to meet these two conditions should not be taken for granted.

For example, participants may agree on values without engaging in a comprehensive review of

alternative value mixes. The attainment of consensus on objectives changes the policy process sig-

niﬁcantly even where it results from the failure of policy-makers to examine any other goals or

the inability of interests with contrary values to mobilize. In the same way, policy-makers may

understand some problems better than others, or at least think that they do. While a complete

understanding of cause-and-eﬀect relationships is almost never attained, there are nevertheless issues

for which participants have achieved a consensus on basic facts. Robert Rothstein has termed

this condition “consensual knowledge”:

Hayes

290

a body of belief about cause-eﬀect and ends-means relationships among variables (activities,

aspirations, values, demands) that is widely accepted by the relevant actors, irrespective of

the absolute or ﬁnal “truth” of these beliefs.

(1984: 736)

As with consensual objectives, the policy process is fundamentally altered whenever policy-

makers agree on how the world works for a given issue whether this shared understanding of

cause-and-eﬀect relationships is ultimately vindicated or disconﬁrmed.

Combining these two dimensions produces a typology of four quadrants that diﬀers sig-

niﬁcantly from the one advanced earlier by Braybrooke and Lindblom. It should be noted that

where issues ﬁt within the typology is properly determined by their placement on the under-

lying dimensions and not by the names attached to the categories. Thus initial placement of a

particular issue within one of the four cells in no way precludes its movement over time into

one of the other cells. Issues will move from one category to another whenever the degree of

agreement on values or the adequacy of the knowledge base is aﬀected by events.

Where conﬂict over objectives is combined with an inadequate knowledge base that leads to

disagreement over policy consequences (conﬂictual knowledge), as in cell A, incrementalism is

virtually inevitable. The policy process will exhibit all the characteristics of normal increment-

alism and policy outcomes will necessarily be incremental. Lindblom is almost surely right in

assuming that this will be the normal case.

By contrast, cell B consists of what might be termed pure problems of value conﬂict. Here

the knowledge base is not the problem and policy-makers exhibit what Rothstein called con-

sensual knowledge. While policy outcomes are likely to be incremental at best in this cell, the

process by which these outcomes are determined will be much more conﬂictual than Lindblom

envisioned for normal incrementalism. Within normal incrementalism, problems are complex

and the self-interest of various participants will be poorly understood, thus tempering conﬂict to at

least some extent. By contrast, where problems are well understood, the clarity of the stakes for all

involved will intensify conﬂict. For example, social security is a diﬃcult issue for policy-makers

precisely because all reforms have immediate and readily understood redistributive consequences.

The ﬁght is over who will bear the costs of repairing the system.

The policy process will be diﬀerent again for pure problems of knowledge base (cell C). Here, a

consensus on objectives exists and rational policy-making is precluded only by an inadequate

understanding of the problem. Policy-making here will monitor information from the envir-

onment and adapt to changing circumstances through feedback. Macroeconomic policy-making

provides a good example here,

Finally, where consensual knowledge is combined with consensual objectives, as in cell D,

rational decision-making is attainable. While many of the issues for which these two conditions

Table 22.2 A typology of decision environments

Conﬂictual objectives Consensual objectives

Conﬂictual or contested knowledge (A) Realm of

normal incrementalism

knowledge base

Consensual knowledge (B) Pure problems of

value conﬂict

(D) Realm of rational

decision-making

Source: Hayes (2001) The Limits of Policy Change: Incrementalism, Worldview and the Rule of Law. Washington, DC:

Georgetown University Press

Incrementalism

291

are satisﬁed may involve relatively small technical or administrative problems, as Lindblom has

suggested (Braybrooke and Lindblom 1963: 61–79), nonincremental policy departures are also

possible here. Where policy-makers agree on objectives and share a consensual knowledge base,

there is nothing to preclude large change if large change seems warranted.

The life cycle of policies: when are the conditions for rational decision-making most likely to be met?

All policies go through a predictable life cycle, and the preconditions for rational decision-making

are most likely to be attained late in the life cycle of policies.

Lindblom saw policy-makers as typically building on past policies. This presumes the exis-

tence of past policies to build on. Most policy problems have in fact been around for many

years, taxing the capacities of policy-makers and resisting any ﬁnal solution. However, all policies

began as brand new policies at some point in time. Building on the earlier research of Lawrence

D. Brown (1983), we may call legislative enactments establishing a new federal responsibility

federal role breakthrough policies. This is the ﬁrst stage in the life cycle of policies.

Over time, policy communities develop around these new programs. Congressional committees

are assigned responsibility for legislative oversight, executive agencies are given statutory authority

to administer the new law, and aﬀected interest groups establish close ties with these legislators

and administrators. Because new policies are unlikely to be well-understood or to generate con-

sensual goals, policy-making will typically exhibit characteristics of normal incrementalism for

many years. Brown termed policies aimed at reforming or improving government programs

already in existence “rationalizing policies.” Because such rationalizing policies tend to exhibit

the characteristics of normal incrementalism for many years, we may call this second stage of the

life cycle the stage of incremental rationalizing policies.

For many policies, this will be the ﬁnal stage in the life cycle. For at least some policies, how-

ever, workable solutions to problems may emerge after many years. As policy-makers come to

understand problems better over time through accumulated experience with policy imple-

mentation, a consensus on both objectives and knowledge base may develop. Where this occurs,

the conditions for rationality are met, permitting the development of what we may term ratio-

nalizing breakthroughs. Such initiatives are rationalizing policies because they seek to reform existing

government programs, but they also constitute breakthrough policies of a very special sort

inasmuch as they represent a dramatic shift in thinking and policy design that permits eﬀective

action in addressing enduring problems.

Thus the conditions for rational policy-making are more likely to emerge at the end of the

life cycle than the beginning. While nonincremental policy change may result from crisis con-

ditions or in response to an aroused mass public opinion, it will be most eﬀective and lasting

where it constitutes a rationalizing breakthrough in the ﬁnal stage in the life cycle of an issue.

Policies do not reach this ﬁnal stage without going through the two prior stages, however.

There are no shortcuts to rational decision-making.

Can incrementalism be trusted to yield good policies?

Is incrementalism the best way to make policies, as Lindblom asserts? There will never be complete

agreement on this point because attitudes toward incrementalism depend on underlying assump-

tions about human nature and what it is possible to achieve through analysis. In this regard,

Hayek (1948), Sowell (1987), and Spicer (1995) identiﬁed two distinct worldviews, which may

be termed rationalist and anti-rationalist (see also Hayes 2001: 7–50).

Rationalists reject Lindblom’s thesis that rational-comprehensive analysis is unattainable. They

are optimistic about human nature and believe that all social problems can be solved through

the exercise of reason. By contrast, anti-rationalists are pessimistic about human nature and see

Hayes

292

distinct limits on what can be achieved through human reason (Hayes 2001: 8–11). Where ration-

alists see solutions as emerging from analysis, anti-rationalists see good public policies as emer-

ging from a policy process that is pluralistic and typically conﬂictual. In Sowell’s terms (1987),

rationalists believe in articulated rationality while anti-rationalists believe in systemic rationality.

Wildavsky (1979) makes this same distinction, characterizing the alternatives as intellectual cogitation

vs. social interaction.

Rationalists reject incrementalism for a variety of reasons. They dislike the muddled com-

promises that emerge from pluralistic political processes, eschew piecemeal and gradual social

reforms in favor of sweeping policy proposals, and decisively reject any strategy for building on

a status quo they view as inherently ﬂawed. By contrast, anti-rationalists embrace incremental-

ism, accepting incrementalism’s core tenet that good policies are more likely to result from

pluralistic processes and endorsing incrementalism’s stress on learning through seriality.

Critics have advanced a variety of indictments against incrementalism (see the various authors

reviewed in Weiss and Woodhouse 1992: 252–68). While some of these indictments are essen-

tially rationalist in orientation, others are anti-rationalist, accepting the need for policy-making

through social interaction but pointing to inherent ﬂaws in the policy process. By and large, the

rationalist critiques can be easily dismissed; the anti-rationalist critiques are more compelling.

Rationalist indictments: here there are four main criticisms: (1) that incrementalism is insuﬃ-

ciently goal-oriented; (2) that it is hostile to analysis; (3) that it is inherently reactive rather than

proactive; and (4) that it cannot anticipate “sleeper eﬀects.”

1 Critics of incrementalism want policy-making to be purposive, or goal-oriented. This criti-

cism assumes it is actually possible to identify agreed upon ends. For example, Dryzek

(1987: 424) deﬁnes the concept of instrumental rationality as “the capacity to devise, select,

and eﬀect good means to clariﬁed ends” (emphasis added). By contrast, Lindblom sees

multiple and conﬂicting objectives as the normal case. While participants may agree on the

desirability of all the ends, agreement on objectives implies agreement on priorities—a

much higher standard to satisfy.

In this vein, rationalists and anti-rationalists tend to diﬀer in their handling of trade-oﬀs

(Sowell 1995: 135–42). Anti-rationalists recognize that desired goals are often in conﬂict

and cannot be satisﬁed simultaneously. On the contrary, to gain more of one value, you

must give up some of the other. Understood in this way, agreement on objectives implies

a consensus among participants on exactly how to make these trade-oﬀs, a consensus that

is typically unattainable. By contrast, rationalists tend to focus on one goal at a time,

ignoring or denying the existence of trade-oﬀ relationships with other goals. Any attempt

at rational analysis that achieves agreement on objectives by focusing on a single goal may

facilitate coherent, purposeful decision-making but only at the expense of neglecting

alternative values that do not go away just because they are ignored.

2 A related criticism charges that incrementalism is hostile to analysis. This charge misrepresents

how incrementalism works. Anti-rationalists, including Lindblom, expect participants to

employ analysis both in pursuing their self-interest and in attempting to identify workable

solutions to public problems. Anti-rationalists see distinct limits on the ability of any one actor

to formulate comprehensive, workable solutions to public problems. Those limits ﬂow out of

inherent limitations on man’s cognitive capacities and the dispersion of pertinent information

throughout the population. Accordingly, good policies are more likely to emerge from the

interplay of multiple actors, each of whom can be expected to marshal the best arguments they

can in defense of their positions. Policy-making by social interaction (systemic rationality)

is not just a balance of power among participants; it is also a competition among ideas.

Incrementalism

293

3 The third criticism holds that incrementalism is inherently reactive rather than proactive.

According to Lindblom, policy-making is reactive by necessity. Where values are multiple

and in conﬂict with one another, as is usually the case, proactive policy-making is virtually

impossible. Fortunately, policy-makers do not need to secure agreement on goals. To the

contrary, they can consider ends and means together through the adjustment of objectives

to policies. Because diﬀerent options embody diﬀerent value trade-oﬀs, to choose one alter-

native over others is to choose a value trade-oﬀ, and conﬂict over which option to choose

represents a conﬂict over value trade-oﬀs. Policy-making is meliorative as a consequence,

moving away from concrete problems rather than toward abstract ideals.

4 The ﬁnal rationalist critique stresses the the inability of incrementalism to deal with what are

called sleeper eﬀects. These are negative policy consequences that remain invisible through

many policy cycles only to show up suddenly at a point when it has become diﬃcult, if not

impossible, to change course. A contemporary example would be global warming; here

reliance on fossil fuels as an energy source has created a greenhouse eﬀect. The consequences

of this greenhouse eﬀect for climate change only began to be understood by scientists

when an irreversible threshold was about to be crossed.

Although sleeper eﬀects clearly pose a serious problem for incrementalism, it is not clear how

late-blooming negative consequences could be identiﬁed earlier through eﬀorts at rational-

comprehensive analysis (Weiss and Woodhouse 1992: 265). Indeed, it is hard to see how any

decision-making method could anticipate policy consequences which are beyond the available

knowledge base by deﬁnition. In theory, at least, sleeper eﬀects should pose less of a threat to a

process in which policy changes occur gradually, in small steps. If sleeper eﬀects can be likened to

lemmings running oﬀ a cliﬀ (Dryzek 1983), it would seem better to approach the cliﬀ at reduced

speed. In the case of global warming, the science has taken decades to develop, and under-

standing is still incomplete. How would nonincremental policy-making magically solve the

problem of an inadequate knowledge base?

The rationalist critiques just reviewed all treat incrementalism as a strategy for policy-making

that can be chosen or discarded at will. However, policy-making is not pluralistic and con-

ﬂictual because participants have chosen incrementalism over rationality as a strategy for policy-

making. Rather, it is pluralistic and conﬂictual because one or more of the conditions for

rational decision-making have broken down, as is the case in three of the four quadrants identiﬁed

earlier.

Anti-rationalist criticisms: anti-rationalist critiques acknowledge the superiority of policy-making

by social interaction but point to inherent ﬂaws in the policy process. Here the main criticism is

that incrementalism is inherently conservative, ﬁrst because it encourages policy-makers to focus

exclusively on small increments, and second because the interplay of multiple interests will

necessarily favor organized elites (especially corporations) with disproportionate political power.

First, critics indict incrementalism for rejecting fundamental changes in the status quo and

focusing instead on gradual policy change through incremental steps (Weiss and Woodhouse 1992:

260–2). This argument ignores the potential for continual improvement in policies through

modiﬁcation in subsequent policy cycles. Seriality allows policy-makers to converge on a solution

through trial and error. While any given outcome may be incremental, signiﬁcant policy change

is possible through an accumulation of small changes over time. In fact, signiﬁcant change may

occur sooner through a rapid succession of small steps than through an occasional nonincremental

policy

change (Lindblom 1979:

520–1).

If signiﬁcant change is clearly possible through seriality, can incrementalism be legitimately

charged with a conservative bias? Admittedly, it makes a degree of sense to view rationalists as

Hayes

294

liberals because of their optimistic view of what it is possible to achieve through politics. Similarly,

anti-rationalists might be characterized as conservative to the extent that they are more skeptical

regarding what it is possible to achieve through politics. Sowell (1995) characterizes the two

worldviews in precisely this way.

This simple, dichotomous distinction is misleading, however, as incrementalism may be employed

to pursue liberal policies. As Karl Popper observed:

This method [incrementalism, or “piecemeal social engineering” in Popper’s terms] can be

used … in order to search for, and ﬁght against, the greatest and most urgent evils of

society, rather than to seek, and to ﬁght for, some ultimate good … But a systematic ﬁght

against deﬁnite wrongs, against concrete forms of injustice or exploitation, and avoidable

suﬀering such as poverty or unemployment, is a very diﬀerent thing from the attempt to

realize a distant ideal blueprint of society.

(1994: 91–2)

Clearly, the fact that incrementalism is meliorative rather than utopian, and that it pursues policy

change through seriality rather than once-and-for-all solutions, does not make the strategy inher-

ently conservative. In this regard, Hayes (2001) distinguishes between rationalist liberals (which

he terms utopian visionaries) and anti-rationalist (or meliorative) liberals, as well as between

nostalgic (rationalist) and adaptive (anti-rationalist) conservatives. Within this scheme, both

Lindblom and Popper would qualify as meliorative liberals (ibid.: 28–41).

Second, incrementalism has also been characterized as conservative because the interplay of

multiple interests will tend to favor organized elites with disproportionate power (Weiss and

Woodhouse 1992: 260–2). Two conditions in particular must be met for incrementalism to

yield good public policies. First, all aﬀected interests must succeed in mobilizing. Lindblom was

overly optimistic on this point in his early writings, assuming that every important social interest

would have its watchdog (Lindblom 1959: 85). The free rider problem prevents many groups

from organizing at all, however (Olson 1970), creating a systemic bias in favor of small groups

over large diﬀuse groups like consumers or taxpayers. Inasmuch as membership groups face a

free rider problem that never goes away, they will be at a disadvantage vis-à-vis institutions, which

are already organized for other purposes and thus face no free rider problem. Finally, groups

mobilizing aﬄuent constituencies will have an advantage over groups attempting to mobilize

the poor inasmuch as the aﬄuent can more easily aﬀord membership dues (on these points, see

the sources cited in Hayes 2001: Chapter 4).

The second condition that must be met for incrementalism to operate eﬃciently is a rough

equality in power across groups. Just as economic markets work badly whenever some ﬁrms

have market power, incrementalism works badly whenever some actors possess disproportionate

political power. Lindblom (1977, 1982) argued that corporations will occupy a privileged

position within capitalist societies. Because corporations can deliver or withhold prosperity, they

have unparalleled legitimacy to the extent that policy-makers equate the class interest of busi-

ness with the public interest (Miliband 1969). This legitimacy gives corporations both a veto

power over policies that threaten them and a form of indirect inﬂuence (“control without trying”),

as corporations receive favorable tax or regulatory beneﬁts from policy-makers that they have

not actively sought.

Lindblom is not advancing a power elite thesis here; rather, he sees the policy process as

pluralistic but skewed toward the interests of corporations. His primary concern is that business

power hampers the proper operation of incrementalism, keeping some items oﬀ the agenda

entirely and channeling disproportionate beneﬁts to business groups. Although this concern is

Incrementalism

295

valid, disproportionate business power would plague any policy-making method: “[I]t is the socio-

political power structure that favors organized elites, not incrementalism or any other decision

strategy” (Weiss and Woodhouse 1992: 262).

While corporations may occupy a privileged position in capitalist societies, Lindblom’s ana-

lysis is overstated. Business groups do not win all the time, and their ability to block policies

they oppose is overrated (Hayes 2006: 63–79). While business groups do win more than any

other group in contemporary American politics, they also suﬀer defeats, and there are many

issues on which they take no position at all. The larger question is whether the American system

features dispersed or cumulative inequalities (Dahl 2005). In a system of dispersed inequalities,

the money or manpower of some groups might be oﬀset by the legitimacy or expertise of other

groups. In a system of cumulative inequalities, by contrast, some groups will rank high on all or

most dimensions of inﬂuence while others will rank low across the board. Right now, corpora-

tions possess vast wealth and unparalleled legitimacy. They also have an advantage in expertise

on technical regulatory issues, and, with their ability to deliver or withhold prosperity, they

occupy a strategic position within the economy.

While this business advantage in resources may seem insurmountable, it is worth recalling

that Dahl (2005) found a surprising evolution in New Haven over time from a system of cumu-

lative inequalities to one of dispersed inequalities. This occurred through the natural develop-

ment of new sources of power that none of the original power-holders could have anticipated

and which they were powerless to block. Policy-makers bent on dispersing inequalities could

take action both to facilitate mobilization of previously unorganized interests and to ameliorate

inequality in tangible resources across groups through a system of public ﬁnancing for congres-

sional campaigns (Hayes 2001: 162–6). Although such reforms would mitigate the problem of

cumulative inequalities, they would not eliminate completely the privileged position corporations

will always occupy within any capitalist society—a problem that would plague any decision-making

method, as noted above.

Special problems plaguing incrementalism in the American context: two additional anti-rationalist

critiques may be oﬀered that would apply with special force in the United States. The ﬁrst stems

from our system of checks-and-balances. The second ﬂows out of the potential for divided

government in an era of polarized political parties.

First, in Lindblom’s view (and mine), there is no viable alternative to incrementalism under

most circumstances. The question is not whether policy-making should be incremental, but rather

whether incrementalism is functioning eﬃciently. Lindblom is right that more can be achieved

through a rapid succession of small steps than through an occasional major policy change.

However, aspects of the political system may preclude revisiting policies on a regular basis, thus

preventing the kind of rapid succession of small steps Lindblom envisions. Under such circum-

stances, incrementalism can be properly accused of a conservative bias. Lindblom (1979) believes

our system of checks-and-balances impedes the proper operation of incrementalism in precisely

this way by making it extremely diﬃcult to pass any kind of legislation. Too often, incrementalism

in the United States features small policy changes a generation or more apart. This problem has

become especially acute in recent years with the development of the party ﬁlibuster and the 60-vote

Senate (more on this below).

Second, for any system of policy-making through social interaction to function properly, all

aﬀected interests must be eﬀectively represented. We have already seen how unrealistic this is.

The development in recent years of a system of polarized parties makes the representation of a

broad range of interests even more problematic. While our contemporary parties help voters

hold politicians accountable by functioning as responsible parties, they also hamper the eﬀective

operation of incrementalism. To the extent that each party represents a narrow range of

Hayes

296

interests, uniﬁed government could produce a policy equilibrium that is necessarily incomplete,

representing only the range of views held by members of the majority party.

Seen in this light, divided government might seem to force consideration of a more complete

range of viewpoints by giving eﬀective leverage to both parties. Where the two parties are as

polarized as ours are today, however, they are more likely to be completely opposed on most

major issues, making gridlock more likely than cooperation. This tendency toward gridlock may

be reinforced by electoral incentives, as neither party wants to cooperate on issues that might

otherwise be used against the other party in the next election. In a world of polarized parties,

then, uniﬁed government may be a necessary condition for passing controversial legislation. As

we have seen in recent years, the 60-vote Senate gives the minority party some leverage even

under conditions of uniﬁed government.

Under such circumstances, incremental policy changes a generation apart may be all we can

hope for. While meliorative liberals like Lindblom would view this as the worst of all policy-making

worlds, it is not inconsistent with the vision of the framers. While the checks-and-balances built

into the constitution may foster incrementalism, they were never intended to foster eﬃcient

policy-making via a rapid succession of small steps.

In conclusion, incrementalism is clearly not a utopian method of policy-making. While the

rationalist critiques are rooted in a misunderstanding of how the policy process operates, the anti-

rationalist critiques point to real problems with the method. It must be noted, however, that

Lindblom never characterizes incrementalism as functioning to yield optimal policies in every

instance. Rather, he argues that incrementalism will outperform misguided eﬀorts to employ

rational-comprehensive analysis. Indeed, as noted in an earlier section of this chapter, once

rationality breaks down, there is no longer any criterion for determining whether policy out-

comes are good or bad. One can only hope that good policy processes will yield good policy

outcomes, at least most of the time. Ultimately, Lindblom’s case for incrementalism as a policy-

making method is meliorative rather than utopian. While corporate power and the failure of

some interests to mobilize make for a “rigged competition” (Weiss and Woodhouse 1992: 262),

incrementalism—however imperfect—is still the only viable way to make policy when the

conditions for rationality break down.

Bibliography

Braybrooke, David, and Charles E. Lindblom. 1963. A Strategy of Decision: Policy Evaluation as a Social

Process. New York: Free Press of Glencoe.

Brown, Lawrence D. 1983. New Policies, New Politics: Government’s Response to Government’s Growth.

Washington, DC: Brookings Institution.

Dahl, Robert A. 2005. Who Governs? Democracy and Power in an American City, 2nd edn. New Haven, CT:

Yale University Press.

——and Charles E. Lindblom. 1953. Politics, Economics, and Welfare: Planning and Politico-Economic Systems

Resolved into Basic Social Processes. New York: Harper & Row, Harper Torchbooks.

Dryzek, John S. 1983. ‘Present choices, future consequences: A case for thinking strategically, World

Futures 19: 1–19.

—— 1987. ‘Complexity and rationality in public life’, Political Studies 35: 424 – 42.

Hayek, Friedrich. 1948. Individualism and Economic Order. Chicago: University of Chicago Press.

Hayes, Michael T. 1992. Incrementalism and Public Policy. New York: Longman.

—— 2001. The Limits of Policy Change: Incrementalism, Worldview, and the Rule of Law. Washington, DC:

Georgetown University Press.

—— 2006. Incrementalism and Public Policy. Lanham, MD: University Press of America.

Jones, Charles O. 1974. ‘Speculative augmentation in federal air pollution policy-making’, Journal of Politics

36(May): 438–64.

—— 1975. Clean Air: The Policies and Politics of Pollution Control. Pittsburgh, PA: University of Pittsburgh Press.

Incrementalism

297

Lindblom, Charles E. 1959. ‘The science of “muddling through’, Public Administration Review 19 (Spring):

79–88.

—— 1965. The Intelligence of Democracy: Decision-Making Through Mutual Adjustment. New York: Free Press.

—— 1977 Politics and Markets: The World’s Political Economic Systems. New York: Basic Books.

—— 1979. ‘Still muddling, not yet through’, Public Administration Review 39(6): 517 –26.

—— 1980. The Policy-Making Process, 2nd edn. Englewood Cliﬀs, NJ: Prentice Hall.

—— 1982. ‘The market as prison’, Journal of Politics 44: 324–36.

Miliband, Ralph. 1969. The State in Capitalist Society. New York: Basic Books/Harper Colophon Books.

Olson, Mancur, Jr. 1970. The Logic of Collective Action: Public Goods and the Theory of Groups. New York:

Schocken Books.

Popper, Karl. 1994. The Poverty of Historicism. London and New York: Routledge and Kegan Paul.

Rothstein, Robert L. 1984. ‘Consensual knowledge and international collaboration: Some lessons from the

commodity negotiations’, International Organization 38: 733–62.

Sowell, Thomas. 1987. A Conﬂict of Visions: Ideological Origins of Political Struggles. New York: William

Morrow, Quill Books.

—— 1995. The Vision of the Anointed: Self-Congratulation as a Basis for Social Policy. New York: Basic Books.

Spicer, Michael. 1995. The Founders, the Constitution, and Public Administration. Washington, DC: Georgetown

University Press.

Weiss, Andrew and Woodhouse, Edward. 1992. ‘Reframing incrementalism: A constructive response to

the critics, Policy Sciences 25: 255–73.

Wildavsky, Aaron. 1974. The Politics of the Budgetary Process. Boston: Little, Brown.

—— 1979. Speaking Truth to Power: The Art and Craft of Policy Analysis. Boston: Little, Brown.

—— 1992. The New Politics of the Budgetary Process, 2nd edn. New York: HarperCollins.

Hayes

298

Models for research into

decision-making processes

On phases, streams, rounds and tracks

of decision-making

Geert R. Teisman and Arwin van Buuren

Models for the reconstruction of decision-making

Public administration scholars agree that decision-making has become more complex. Several

reasons can be identiﬁed for this increased complexity. Two important ones are increased uncer-

tainty about the dynamics and interdependencies in global networks and the power-sharing

characteristics or that ‘network society’. In networks, local circumstances can change quickly

due to networks dynamics. At the same time, decisions are taken in a society where nobody is

in charge (Bryson and Crosby 1992; Kickert et al. 1997). These two system changes do make

public decision-making complex. Situations can change quickly and any decision taken interacts

with decisions taken by others.

This chapter is based on the assumption of increased complexity. As such, it ﬁts in with many

others. Often the focus is on the consequences of and challenges for governments or govern-

ance systems. Complexity is clearly taken into account in the wider discussion on governance in

networks. This was already the case before 2000 (for example: in Germany Marin and Mayntz 1991

and Scharpf 1997; in France Crozier and Friedberg 1980; in Great Britain Rhodes 1996b; and

in the United States E. Ostrom 1990; Smith 1998). After 2000, many other publications have ela-

borated the complexity-government/governance relation (Strand 2002; Teisman and Klijn

2002; Duit and Galaz 2006; Wagenaar 2007; Gerrits 2008, Teisman et al. 2009).

Complexity, however, also raises questions for researchers of complexity. What kinds of models

can researchers in decision-making apply to describe, analyse and evaluate decision-making and

how do these models deal with complexity? Attention was paid to this topic by Mintzberg

(1973), Kingdon (1984), Butler (1991) and Teisman ([1992, 1995] 1998). They generated images

of decision-making as phases, as streams and as rounds. After 2000, van Buuren and Gerrits (2008)

and Haynes (2008) contributed to the thinking about this topic by introducing concepts as

tracks and traces as well as timelines and rates of change to indicate dynamics in processes. These

authors pay special attention to the compounded, erratic and unpredictable characteristics of

299

decision-making processes. By looking at processes they focus far less on the decision and more

on the emergence of outcomes from a variety of more or less coupled decisions of a variety of

actors. To indicate this conﬁguration of ecology of decisions as an important object of study

Teisman, van Buuren and Gerrits (2009: 231–49) introduce the concept of process systems.

This concept underlines the embeddedness of individuals and their decisions in groups of indi-

viduals and their decisions into an organizational and institutional context where many more (in

terms of content unrelated) decisions are taken that can emerge into important ones for the

progress in decision-making on the level of individuals and groups.

This chapter focuses on the question of how to depict decision-making which evolves in

complex network structures. We will concentrate on the characterization of successive decision-

making activities and of concurrent decision-making activities. We will also discuss the assumed

relations between the activities, in sequential as well as parallel combinations. The following

research questions can be formulated:

1 Which assumptions are made with respect to chains of activities in decision-making processes?

We will analyse diﬀerent criteria used to distinguish between strands of activities, typolo-

gies of decision-making as a whole and elements of this process, and signs of progress in

decision-making.

2 What sorts of assumptions are made with respect to simultaneous activities in decision-

making processes? Speciﬁcally, we will distinguish here between (more or less relevant)

actors, the relationship between problems and solutions, and the content of decision-making

in terms of ambitions, frames and facts.

To analyse decision-making, the researcher needs to make a reconstruction of the study

object. Such a reconstruction will be selective in nature. Observation is not simply an eﬀort to

learn what is going on. Rather it is a process where observations are made to conform to sets of

assumptions (Edelman 1971). The gathering and classiﬁcation of empirical observations into

meaningful information is based on the a priori images of decision-making used. We cannot

depict decision-making without making assumptions about its appearance. Various terms are

used to describe such a set of assumptions: model, image, metaphor, referential framework or

methodology. In this chapter the term ‘model’ will be used. Models help us to understand

decision-making in distinctive yet partial ways (Morgan 1997: 4).

Four models will be discussed in this chapter. Two of these are generally accepted and

respected, namely the so-called phase model and the stream model. The phase model is the most

common approach, both in science (e.g. Anderson 1979; Bryson and Crosby 1992) and in policy

practice (procedures are often based on the concept of phasing). Decision-making is represented

in terms of a number of distinct stages (Mintzberg 1976). Phase models distinguish between (at

least) policy formation, policy adoption and policy implementation. Each phase has its speciﬁc

characteristics and participants. Ministries, for instance, are often divided into departments that

are responsible for policy formation and others responsible for implementation.

The stream model depicts decision-making as a combination of three separate concurrent

streams (Kingdon 1984). One stream consists of problems, another of policies/solutions, and a

third one of politics/participants. Like the phases, streams have their own characteristics, but

they exist side by side. A decision results out of the coincidence of streams.

The third model, the round model, has become more and more accepted as a conceptual

approach to analyse decision-making in the context of complex networks (van Bueren et al. 2003;

Howlett 2010). In this so-called rounds model decision-making is assumed to consist of diﬀerent

decision-making rounds. In all sets of rounds, the interaction between diﬀerent actors results in

Teisman and van Buuren

300

one or more deﬁnitions of problems and solutions. In addition to Teisman (2000) we will present a

fourth approach of complex decision-making, elaborated by van Buuren (2006) based upon the

assumption that we can interpret decision-making as three co-evolving tracks: of fact-ﬁnding,

framing and will-forming (cf. van Buuren and Gerrits 2008; van Herk et al. 2011).

In the round model the interaction between varieties of actors is the primary focus of the

researchers. All participants can score points in each round, in terms of a leading deﬁnition of

the problem and the (preferred) solution. By doing so, they deﬁne the beginning of a next

round. But at the same time each new round can change the direction of the match: new players

can appear, and in some cases the rules of the game can even be changed. The rounds model

was developed during years of research in the ﬁeld of urban and infrastructural planning. The

ﬁrst author was fascinated by the long duration of decision-making processes and by the changes

in course these processes often take. Another remarkable result of his research was that the

actors involved in decision-making often did not agree on the classiﬁcation of a certain stage in

the process, in terms of formation, adoption and implementation. Although he was able to identify

several oﬃcial decisions taken by ministers, Parliament, Parliamentary committees, etc., none of

these decisions could be clearly depicted as the moment of adoption. Some of the oﬃcial decisions

were followed by actions which could not possibly be seen as the implementation of these

decisions. This brought the ﬁrst author to a fascinating conclusion: in complex decision-making

results are achieved without the existence of the decision that marks the switch from the pre-

paration phase into the implementation phase. Furthermore, the distinction between what the

problem is and what the solutions are proved to be more complicated than assumed in the stream

model. What was a solution for one actor could easily be a problem for another. Participants

bring along closely intertwined problems and solutions.

In order to understand these dynamics and variety in perceptions, Teisman gradually altered

his conceptual deﬁnition of decision-making. Due to the fact that the central decision could not

be found, he began to collect all the decisions that were taken in a certain case. Second, he no

longer assumed that the decisions were arranged on the basis of an a priori order and hierarchy.

On the contrary, it was the task of research to clarify the empirical relation between decisions.

In order to do so, decision-making was redeﬁned as an intertwined ‘clew’ of a series of decisions

taken by various parties, leading to a new analytical model. Progress is described in terms of

successive rounds.

The second author (2006) took this insight as point of departure to answer the question: but

how can we understand the evolution of the content of decision-making processes when it comes

to the issues at stake? What are the various elements which constitute the content of decision-

making? And how do they inﬂuence each other? To answer this question there are numerous

points of departure to be chosen. Many authors (implicitly or explicitly) depart from a distinc-

tion between elements of ambition, interest and stake, versus elements of knowledge, analysis

and information. This distinction can be traced back to Lindblom and Woodhouse (1992) who

distinguish between activities to gather information and activities to exert inﬂuence, and to Heclo

(1974) who uses the concepts of powering and puzzling to depict the same phenomena.

More recently, much literature has appeared which deals with a third element, the role of

interpretation and thus of frames or belief systems in decision-making processes (Feldman 1989;

Schön and Rein 1993; Fischer 2003; Sabatier 1999). Frames are the normative and subjective

perceptions of actors within decision-making processes. Frames are important in decision-making

because they determine which knowledge questions are taken into account, but also which

ambitions are seen as ﬁtting in a speciﬁc problem context.

Each actor brings in his own facts, frames and ambitions within a decision-making process

and negotiates with other actors to agree about them. These elements evolve in a dynamic and

Research models

301

erratic way and inﬂuence each other. Decisions can be seen as temporal selections of facts,

frames and ambitions. This temporal ‘equilibrium’ can be easily challenged when the content of

its constituting tracks changes.

The tracks model (elaborated in Figure 23.1) is based upon the assumption that it is helpful to

describe the evolution in these three tracks separately, to be able to explain (the consequences

of) their mutual interaction and their inﬂuence on the decisions realized.

After having examined all four models, they will be compared with each other. A four-fold

analysis of decision-making about the Betuwe freight railway line will be carried out, and ﬁnally

we shall draw some conclusions regarding the added value of the tracks and rounds models

compared to the phase and streams model.

Concepts within and principles of the phase model

In policy process analyses ‘policy can be understood and examined as a combination of several

processes, which are interrelated but can still be conceived as distinct components that are deter-

minants of government actions’ (Sato 1999). The phase model assumes that decision-making is

‘the succession of diﬀerent situations in the formulation, adoption, implementation and eva-

luation of a policy’ (Bryson and Crosby 1992: 57–66)

. Often formation is divided into a phase

of problem deﬁnition and a phase wherein solutions are presented. The ﬁrst stage of the tradi-

tional policy process, problem deﬁnition, involves the emergence and recognition of some pro-

blem or crisis. In the second stage a policy to address speciﬁc problems is formulated by various

governmental and non-governmental actors such as legislators, executive branch oﬃcials, the

courts, citizens and special interest groups. Special policy proposals are adopted in the third stage.

The fourth stage is policy implementation, wherein the adopted alternatives are executed by

administrative units. Finally, in the policy evaluation stage, policy-makers determine whether

the policy has achieved its goals (Altman and Petkus 1994).

Policy formulation is described as ‘the collecting and analysing of information and the for-

mulating of advice regarding the policy to be followed’. Parts of this phase include recognition,

diagnosis, search for information, design and evaluation of the diﬀerent alternatives that are designed.

Policy adoption involves the ‘taking of decisions about the contents of a policy’. During imple-

mentation the chosen means are applied: ‘Decision-making is a sequence of steps which, if

followed, should lead to the best solution; that is, to action which optimises the decision maker’s

utility’ (Butler 1991: 43–4).

Analysts using the phase model are aware of the fact that reality does not reﬂect the assumption

of the model: ‘Planning in shared-power situations hardly ever follows a rigidly structured sequence

from developing problem deﬁnitions and solutions to adopting and implementing proposals.

Figure 23.1 A depiction of four models for the analysis of decision-making processes

Source: Teisman (2000) ’Models for research into decision-making processes: On phases, streams

and decision-making rounds’, Public Administration 78(4): 937–56

Teisman and van Buuren

302

Serious diﬃculties arise when people try to impose this rigidly sequential approach on situations

in which no one is in charge’ (Bryson and Crosby 1992: xiv). ‘Nonetheless, to be steadily

eﬀective, it is essential to have an organised approach of some sort’ (ibid.: xiv). Several scientists

reach the same conclusion (Hoogerwerf 1982; Mintzberg et al. 1976). They are aware that empiri-

cism deviates from this, but feel it is worthwhile to reconstruct policy-making as though it was

taking place in phases. The phase metaphor allows scientists to develop diﬀerent theories regarding

the various stages. To reconstruct the policy formation phase, concepts such as ‘problem deﬁ-

nition’, ‘generating alternative solutions’ and ‘policy design’ are used. To analyse the adoption

of policies, analysts look for the central decision which demarcates the transition from formation

to implementation. During this phase, the policy proposal needs to be determined that is opti-

mally suited to achieve the set objectives. Many parties are involved in this phase, but in the end

only one or two actors determine which means are used to achieve the objectives. During the

implementation phase, the researcher focuses on how chosen means were used and how any

opposition was handled.

Problems preceding the search for and choice of solutions

Scientists using the phase model assume that decision-making is problem-oriented (Scharpf

1997). There is, or at least there should be, one actor whose decision supersedes those of others

and who therefore determines the problem and the policy. ‘We therefore deﬁne public policy

as substantive decisions, commitments, and actions made by those who hold or aﬀect govern-

ment positions of authority, as they are interpreted by various stakeholders’ (Bryson and Crosby

1992: 63). Even though Bryson and Crosby (ibid.: 159) agree with Rittel and Webber (1973)

that public problems are ‘wicked’ in the sense that they have no deﬁnite formulation, cannot be

solved immediately and are unique, they still assume that problems do exist and should be known

before a search for solutions can begin. Developing a ‘problem deﬁnition to guide action’ is

seen as the ﬁrst and most important phase in decision-making, which should be carried out before

the adoption of a policy by the policy-maker can take place. Bryson and Crosby also assume that a

solution can be formulated at a single point and place (see Figure 23.2). It is on these points that

the stream and rounds models use diﬀerent assumptions.

Concepts within and principles of the stream model

Some researchers assume that the horizontal division of activities is a more crucial distinction to

analyse processes than the vertical division used in the phase model. In their view, the analysis of

the various phases does not result in speciﬁc theories on policy formation, policy adoption and

implementation. Based on this point of view, the so-called stream model was developed in 1972

by Cohen et al. In 1984 Kingdon elaborated this model further. His model was based on the

idea that policy-making consists of three streams: problems, solutions/policies and politics. As

opposed to the phase model, here decision-making is dissociated from a speciﬁc participant. The

idea is that decision-making consists mainly of a stream in which problems are discussed, a stream

in which solutions are discussed and a stream consisting of things such as the attitude of the public,

campaigns by pressure groups, and ideological contributions (Kingdon 1984: 152). Politicians

can determine the problems and solutions they wished to concentrate on. For this reason, they

are likely to rush from one combination of problem and solution to another. As a result of this,

the level of participation in decision-making is likely to vary strongly. This is partly because

these processes develop unpredictably (March and Olsen 1976: 10–23). Thus, the temporal

sequence of the phase model is replaced by the postulate of simultaneousness (Koppenjan 1993: 26).

Research models

303

The three streams exist simultaneously. ‘They are largely independent of one another, and each

develops according to its own dynamics and rules’ (Kingdon 1984: 20). There are three separate

worlds where speciﬁc products are developed and transformed into their own dynamics and there-

fore are not linked in any temporal sequence. ‘While there are indeed diﬀerent processes, they

do not necessarily follow one another through time in any regular pattern’ (Kingdon 1984: 83).

Actors with solutions in the policy stream are looking for problems and political commitment,

while politicians are looking for both solutions and problems with which they can ‘score’. According

to this conceptual model, major policy changes are likely to occur only if the three streams

become linked. Such linkages can occur especially if there is a favourable momentum, a so-called

‘policy window’ (Kingdon 1984: 174; Anglund 1999). The researcher can make decision-making

transparent by investigating to what extent links are forged and why they are forged. Thus

decision-making is not primarily separated into vertical strands, in the sense of consecutive steps

over time, but rather in horizontal strands, in the sense of streams existing simultaneously side

by side (see Figure 23.3).

The rounds model of decision-making: interacting decisions

In the rounds model, actors are once again the focal point of analysis. The assumption here is

that solutions/policy and problems are relevant to a policy process, insofar as they are presented

by an actor during this process (Scharpf et al. 1978; Teisman 1998) In contrast to the phase model,

here the researcher assumes that problems and solutions are not linked to a single actor (policy-

maker) and are therefore not ﬁxed at the single moment at which the policy is adopted. Many

actors are involved in decision-making, and they will introduce their own perceptions of

Figure 23.2 The concept of decision-making used in the phase model

Source: Teisman (2000) ‘Models for research into decision-making processes: On phases, streams

and decision-making rounds’, Public Administration 78(4): 937–56

Teisman and van Buuren

304

relevant problems, possible solutions and political judgement. In order to understand decision-

making, the researcher focuses on the variety of actors, objectives and solutions, their dynamics

as well as the interaction between these elements. Complex decision-making involves many

policy-makers who take decisions.

The rounds model can be seen as an interactive approach (Scharpf 1997). Policies, in terms of

the actual interventions that take place in society, do not stem from an intended course of

action formulated by one actor, but result from of a series of decisions taken by diﬀerent actors

(see Figure 23.3, from Teisman 2000).

Political scientists … should be interested in the fact that many or most of the well-

designed policy proposals will never get a chance to become eﬀective. The reason is that

public policy is not usually produced by a unitary actor with adequate control over all required

action resources and a single-minded concern for the public interest. Rather it is likely to

result from the strategic interaction among several or many policy actors, each with its own

understanding of the nature of the problem and the feasibility of particular solutions, each

with its own individual and institutional self-interest and its own normative preferences,

and each with its own capabilities or action resources that may be employed to aﬀect the

outcome.

(Scharpf 1997: 11)

Figure 23.3 The concept of decision-making used in the stream model

Source: Teisman (2000) ‘Models for research into decision-making processes: On phases, streams

and decision-making rounds’, Public Administration 78(4): 937–56

Research models

305

The focus, therefore, should be on the interaction among purposeful actors. To gain insight

into policy-making, the researcher depicts which actors are participating at what time. Actors

are units capable of developing a recognisable course of action (individuals, groups or collective/

corporate entities).

To separate strands of decision-making, the train of thought of the phase model is combined

with that of the stream model. On the one hand, a vertical classiﬁcation of decision-making is

made, by looking at series of decisions that were taken in that time. On the other hand a horizontal

classiﬁcation is applied by looking at interactions concerning the same subject, even if actors are

unaware of each other’s decisions at the moment they take these decisions. The division into

time periods diﬀers from that of the phase model in a number of respects. It is not the feature of

the time period as such (i.e. this is preparation and this is implementation) that is being deter-

mined, but rather the starting and concluding points of a certain period. Such a period is called

a ‘decision-making round’. The researcher demarcates decision-making rounds by determining

the most crucial decisions of decision-making in retrospect. This concerns particularly the choice

of decisions that in a later period of decision-making serve as an important point of reference

for the behaviour of the actors that are present at the time (Teisman 1998).

The application of the rounds model yields a picture of decision-making, particularly focusing on

the ability of parties to handle their dependency on other parties in interaction (see Figure 23.4).

The tracks model of decision-making: looking at the ingredients of

interaction processes

The rounds model gives us an impression of how we can understand the intermittent development

of decision-making processes and the way in which consecutive decisions emerge. It explains

Figure 23.4 The concept of decision-making used in the rounds model

Source: Teisman (2000) ‘Models for research into decision-making processes: On phases, streams

and decision-making rounds’, Public Administration 78(4): 937–56

Teisman and van Buuren

306

how strategies of individual actors result in collective outcomes. At the same time, it has less

relevance for the content of a decision-making process, an element which is quite central to

Kingdon’s streams model and is also important in the phase model when it comes to the

rational approach of policy-making and the central role of policy analysis (Quade 1972).

The tracks model emerged out of the recognition that decision-making has to do with three

ingredients brought in by participants (cf. van Buuren and Gerrits 2008). First of all, they bring

in their ambitions, based upon their material interests or normative goals. They participate because

they have something to win or lose. But second, these actors also have the opportunity to

mobilize factual evidence in favour of their ambitions (Sabatier 1988). They criticize evidence

when this is not in accordance with their point of view. Due to the fragmentation within the

knowledge domain, they do have many opportunities to mobilize counter-evidence to attack

the evidence used by formal policy initiators. But third, implicitly or explicitly, all participants in

decision-making processes do have their own problem perceptions, mental frame or belief system.

They use a speciﬁc interpretation of the problem at stake and also a speciﬁc interpretation of the

future they are convinced to be desirable (Feldman 1989). We can ﬁnd a comparable idea in

Vickers (1965) who distinguishes between action judgements, reality judgements and value

judgements.

Facts, frames and ambitions are the ingredients of interaction processes. During these inter-

action processes these ingredients evolve, merge and disappear, due to endogenous and exo-

genous developments. The evolution of these elements is inﬂuenced by internal dynamics but

also by their mutual interaction: facts can inﬂuence frames as can ambitions and vice versa. Actors

negotiate about which facts are authoritative, which meta-frames are constructed and which

ambitions can be combined or traded. New facts are mobilized in order to break through a

situation of frame ﬁxation and frame reﬂection can be initiated when negotiations stagnate. The

tracks model thus analyses the evolution of these tracks separately as well as by their mutual

interaction.

Figure 23.5 The concept of decision-making used in the tracks model

Source: Author

Research models

307

From this perspective decisions are seen as (temporal and dynamic) selections of facts, frames

and ambitions. These selections are only temporary: developments within one of the tracks (for

example new facts or changing ambitions) can distort such an equilibrium and can trigger new

decisions (van Buuren and Gerrits 2008; van Buuren et al. 2009). In more consensus-oriented

decision-making processes, the extent to which there is consensus within these three tracks is an

important selection criterion, while in more adversarial processes this selection depends upon

the outcome of a power struggle between actors. Actors easily can mobilize new facts, construct

new frames or air other ambitions to challenge this equilibrium.

A comparative analysis to highlight the relative value of the four models

In the previous sections we have examined four models. The phase model provides an insight

into the subsequent stages a focal actor goes through. The stream model focuses on links between

problems, solutions and politics. The rounds model provides an insight into the interaction

between actors. Mutual adjustment (by way of co-operation, conﬂict or avoidance) leads to policy

results (see Table 23.1). And ﬁnally, the tracks model explains the way in which the content of

decisions evolves.

In order to argue what the added value of the rounds model can be, three themes will be dealt

with. The ﬁrst looks at the issue of actors, problems and solutions, the second with policy adoption

as a yardstick, event or result and the third with criteria for the evaluation of decision-making.

Finally, the rounds model will be considered as a useful tool for developing the governance

theories further.

Table 23.1 Comparative perspectives on the phase model, the stream model and the rounds model

Phase model Stream model Rounds model Tracks model

Criteria for the

separation of

strands of

activities

Stages a focal

organization goes

through

Different concurrent

streams of

problems, solutions

and politics

Rows of decisions

taken by actors,

creating rounds

through interaction

Tracks of different

types of content,

brought in by

various actors

Characterization

of decision-

making

Sequence of

formation, adoption

and implementation

Coincidental or

organized links

between streams

Interaction between

decisions taken by

various actors

Co-evolving tracks

of fact-ﬁnding,

framing and

will-forming

Assumptions

about the

nature of the

process

One moment of

policy adoption

holds way over

other decisions and

guides the process

A simultaneous

stream of problems,

solutions and

politics, linked more

or less at random

Decisions that

conclude a round

and initiate a new

round, without

ﬁxing its progress

Decisions as

temporary

constellations of

facts, frames and

ambitions

Assumptions

about the

content of the

process

A focal actor adopts

a dominant

deﬁnition of the

problem solution,

creating

governmental

policy

Dynamics within

and links between

streams determine

major policy

changes

Interdependent

actors take decisions

separately or jointly,

leading to

governance policies

Content evolves

dynamically due to

the evolution within

and between the

three tracks

Source: Authors

Teisman and van Buuren

308

Actors, problems, solutions and interaction

It is generally accepted that various actors are involved in decision-making. The phase model,

however, places certain actors in a more central position than others on an a priori basis. The

rounds model refrains from doing this. When dealing with complexity, it is not sensible to

exclude actors in advance or to assume that certain actors can be characterized on a priori

grounds as policy designers, adopters or implementers. Rather, it is advisable to start out with all

the (potential) participants who perform in decision-making. In the rounds model, decision-

making is not about a single issue, nor about separated streams of problems, solutions and par-

ticipants, but about dynamic combinations of sets of problems and solutions represented by

diﬀerent actors. Complications in decision-making often appear when a solution adopted by

one or more actors constitutes a problem for others. Progress is often made when a solution is

provided which deals with sets of problems and the ambitions of several of the actors involved.

For this reason the rounds model focuses on the interaction between actors, during which they

can negotiate acceptable combinations of problems and solutions. To understand this interaction

in more detail, the tracks model focuses on the evolution of the ingredients of decisions during

this interaction process.

Policy adoption, yardstick, event or result

The phase model assumes that policies are set at a certain moment. The stream model assumes

that policy streams from an event in which the streams coincide. The rounds model, in contrast,

assumes that this moment does not exist. Policies result from a series of decisions taken by

various actors. The dynamics of combining problems and solutions and the relation between the

two account for the course of decision-making. A round of decision-making begins and ends

with the adoption of a certain combination of a problem deﬁnition and a (virtual) solution by

one or more actors.

The assumption is that the actors assess to what extent other actors share their deﬁnition of

reality and proceed to interact on this basis. In contrast to the phase model, none of the deﬁ-

nitions are seen as ﬁnal or permanent. Research based on the rounds model will focus on per-

ceived problems and solutions and will subsequently analyse whether and how actors have

managed to combine perceptions to such an extent that they are willing to support a joint solution

(Termeer 1993: 44, 48–51). Adoption then becomes the consolidation of a problem-solution

combination (in terms of a speciﬁc constellation of facts, frames and ambitions) over a longer

period of several decision rounds.

Policy evaluation and evaluation criteria, shifting from a government to a

governance approach

In the phase model, the public interest as deﬁned by the focal organization is often used as

guideline for evaluation. Evaluation focuses on the ‘ﬁt’ between policy result and policy for-

mulated in advance by the focal organization. The rounds model questions this criterion. Policy

intentions at the beginning of a process are not necessarily the best indicator of public interest. It

could be argued that a posteriori opinions and judgments are at least as good an indicator.

Second, it is assumed that network society as a whole, and the government in particular, are

fragmented. As a result, the representation of the public interest is distributed over various parts

of government and organizations outside the government. Policy evaluation becomes more

meaningful if all these intentions are taken into account in the analysis. A ﬁtting overall concept

Research models

309

is joint interest (Teisman 1992: 91–2). Evaluation then no longer focuses on the question

whether the policy result agrees with a single policy intention, but whether it is based upon

agreed facts between participants, a shared problem frame and an attractive combination of the

various ambitions, and to what extent it responds to the objectives of all the parties involved at

the moment policy eﬀects can be distinguished.

The shift in orientation from government to governance

By doing so, the rounds model can contribute to the discussion about the shift from government

to governance within several scientiﬁc communities involved in public administration.

In response to empirical studies by Mackintosh (1977) and Jones (1975), which undermined

some of the assertions of the Westminster model, Rhodes (in Rhodes and Dunleavy 1995)

suggested a new framework of analysis. This framework should focus on the complex web of

institutions, networks and practices surrounding the prime minister, Cabinet, Cabinet committees

and their oﬃcial counterparts, namely the less formal ministerial ‘clubs’ or meetings, bilateral

negotiations, interdepartmental committees, and so on (Rhodes 1995: 12). From this insight it

was a logical next step to focus on complexity of interactions between actors (see Smith 1998:

47). Government is depicted as an institutional framework, not as an actor. Governance is about

steering without presuming the presence of hierarchy (Rosenau 1992: 14). It refers to relational

contracting, organized markets in group enterprises, clans, networks, trade associations and

strategic alliances (Jessop 1995). Departments have to be seen as one of many competing centres

of authority (Rhodes 1996a, 1996b).

It is within this perspective that the rounds and tracks model must be placed. Consideration

should be given to the idea of using the rounds model in empirical research focusing on gov-

ernance. The model oﬀers a way to reconstruct a basically unlimited complexity of events that

can be combined into a decision-making process. This will provide us with additional, more

detailed insights into these processes, thereby creating a basis for more management theories

about networks (Kickert et al. 1997).

A fourfold analysis of the Betuwe railway line

In this section we will present an in-depth case study of decision-making processes regarding

the construction of a new freight railway line between Europe’s largest port, Rotterdam, and

Germany. The so-called Betuwe line is one of the most disputed projects of the last decade in

the Netherlands. The parties involved have even called it a ‘battle’.

Despite all the problems

and opposition, construction of the railway line is still continuing. Therefore, we may consider this

to be a successful decision-making process, in which the ministry, the cabinet and the parliament

were able to undertake a strategic project that ﬁtted in with their policies.

At the same time, however, many questions can be raised. The expected growth of freight

transport by rail did not take place. It is far from clear whether or not the German section of the

line can support the number of trains needed to make the Betuwe railway line viable. There are

no private parties that want to invest in the line or exploit it, even though this was assumed to

be the case by the cabinet when the policy was being adopted.

Studies have questioned the positive environmental impact of transport by rail compared

with that by road and waterways. The costs of the line have risen from 2.5 billion initially to

9.35 billion in 1998. Recently the transport minister decided to forgo the construction of the

Northern branch of the Betuwe line in order to avoid any further increase in expenditure. A

fourfold analysis of this case is presented below.

Teisman and van Buuren

310

Application of the phase model: a central actor, a deﬁned problem and a

good solution

In the Netherlands, the central government, speciﬁcally the Ministry of Traﬃc and Waterways,

is responsible for the planning of new infrastructure. To properly organie decision-making, the

ministry uses a formal procedure known as the Planning Central Decision Procedure (in Dutch:

PKB). At the end of the 1980s, the ministry was faced with congestion on existing freight con-

nections from Rotterdam to the hinterland, combined with an expected growth in freight trans-

port. In previous policy memoranda, the Government had appointed the distribution sector as a

core activity for the Dutch economy. The two main Dutch ports, Schiphol Airport and the port of

Rotterdam, should be enabled to function optimally and the main transport lines should be

provided with suﬃcient capacity. The capacity of the hinterland connection could be increased

by three diﬀerent means: road, waterways or a new railway line. Because road traﬃc was seen as

having a negative impact on the environment and transport by water was not seen as a viable

option, the ministry decided in 1989 that a new railway link was needed.

The ministry decided that the railway should run through the rural region of the Betuwe and a

freight line was due to be completed before 1997. To implement this policy, the ministry set up the

necessary procedures to determine the line’s trajectory and its incorporation into the landscape. But

while going through these procedures, the ministry encountered opposition from civilians and local

authorities. In order to delay implementation, regional actors asked for additional studies to be car-

ried out into the possible construction of an underground tunnel. Although the ministry indicated

that this was not a realistic plan, the Second Chamber let itself be persuaded to investigate this option.

This was the signal for other parties to present the Chamber with other new alternatives, such as

the submersion of the railway line within a cutting. The involvement of parliament resulted in a

considerable series of amendments and an increase in costs. While the ﬁrst plans were expected

to cost 4 billion Dutch guilders, subsequent calculations found that with the addition of the new

line the project would have cost more than 9 million guilders. At the end of 1993 the project

was approved by parliament. Decision-making was complicated, however, by the parliamentary

elections of 1994. One supporter of the project left the cabinet and was replaced by a minister

who opposed the project. To deal with this situation, the new cabinet appointed a Committee of

Experts chaired by a member of the opposition. In January 1995 this Committee declared itself in

favour, thus persuading the opposition to revise their point of view. In early 1997 all complaints

were rejected by the Council of State. The construction works could begin.

Application of the stream model: insights into how solutions, problems and

politics come together

The ﬁrst additional insight is that the need for a freight railway line had existed long before being

placed on the cabinet’s political agenda in 1989. In the early 1980s this idea had already been

discussed in Rotterdam. Transport of goods to the German hinterland was only possible via road

and inland waterways. Road transport was vulnerable due to environmental restraints, while any

growth in water transport was questioned due to its rigid institutional nature. The attractive third

solution, the construction of a freight railway line, however, was at that time incompatible with

problems at the national level.

Years later, it was taken up again by the new enterprise responsible for freight transport by rails.

This organisation resulted from the division of the Dutch Railway Company, a monopoly, into

two separate companies, one for passengers and one for freight. The latter saw the railway line as a

solution to its own unproﬁtable position. The line was deﬁned as a prerequisite for the growth

Research models

311

of freight transport by rail. However, the political argument for it was not made, and in 1988

the Transport Ministry turned down the proposal.

Suddenly in 1989 the ministry became interested in the solution, even though its character-

istics had not changed. An explanation for this can be found within the ministry itself. It was

preparing a strategic plan for transport investments in the next decade, and became aware in 1989

that a major key project was lacking in this plan, and other projects were faced with delays. A

Ministerial Committee was set up which advised the minister to approve the Betuwe railway

line. Due to budget deﬁcits, ﬁnancial problems remained. These were solved in two ways. First,

it was assumed that it would be co-ﬁnanced by private parties, and second, that exploitation of

the line would be proﬁ table. It was this combination of problems and solutions that convinced

parliament in 1993. The policy was approved.

Application of the rounds model: additional insights into internal and

external dynamics

In the rounds model, attention is paid to streams of decisions taken by several actors. The fol-

lowing groups of actors can be distinguished: the Rotterdam Port Authority, the freight railway

company NS Cargo, the Transport Ministry, parliament, local and national environmental groups,

the German Railway Company, and the European and German governments. Five diﬀ erent

rounds are distinguished. During the ﬁrst round, which started in the beginning of the 1980s,

supporters of the Rotterdam port called for an additional freight link to the hinterland, in addition

to the existing roads and waterways. They initiated the process. The aim was to increase the

amount of transport modalities that could be used. They were less interested in the use of rail as

such, but rather in increasing the range of options. This idea was discussed locally and shelved

after some time, partly due to the negative reaction of the Chamber of Commerce in Rotter-

dam. The ﬁrst round was terminated for this reason, and it seemed as though the proposal had

disappeared into a desk drawer for good.

A second round was started in 1987 by NS Cargo. It had recently split from its parent, the

Dutch Railway Company, and was faced with a weak market position. To strengthen its posi-

tion, it proposed to build a line speciﬁcally for freight transport. This had not been proposed in

the past, for the reason that no powerful actor had been responsible thus far for freight transport

by rail. In the former Dutch Railway Company freight had always been subordinate to passenger

transport. With the birth of NS Cargo the railway line had obtained its own policy entrepre-

neur. However, NS Cargo did not have the legal or ﬁnancial means to construct the railway

line and began a lobbying process. Until 1989 the ministry showed no interest.

In that year, a third round of decision-making was begun in which the ministry now played a

leading role. The line was placed high on the government’s agenda and has held this position

ever since. Decision-making was accelerated had already by the minister, and one year later a con-

crete route was proposed and the fourth round of decision-making had begun. In this round the

ministry still played an important role, but was at times outclassed by Parliament. This stemmed

directly from the role played by local and regional governments and environmental groups. At

the beginning of the process, national environmental groups supported the choice in favour of

transport by rail. Due to the speeding-up of decision-making, however, and a total lack of com-

munication with local and regional governments, resistance grew quickly in 1992 and 1993, and

national environmental groups felt compelled to withdraw their support. The Betuwe line became

the Dutch example of a closed and autistic central government giving priority to the traditional

economy (port development, transport and infrastructure) instead of the environment and quality

of life, or a more modern type of economy (knowledge infrastructure, services). The ministry rejected

Teisman and van Buuren

312

all the alternatives presented by the local authorities and other groups in society. Parliament was

more accommodating, however, and a range of amendments resulted, raising the cost to almost

10 billion guilders. The Betuwe line appeared to have been approved by the government.

The ﬁfth round began with the parliamentary elections and the change in the coalition of

governing parties, as has been described above. A high-level committee (the Hermans Com-

mittee) was created to provide independent advice about the Betuwe line. This committee advised

in favour of the Betuwe line but found it diﬃcult to formulate their advice. They received

contradictory reports from consultancy ﬁrms, and various local and regional stakeholders tried to

inﬂuence the committee. More interesting here is that under the surface of party-political

manoeuvring, a lot of interaction was going on in terms of ﬁnancing the project, exploiting the

line when in operation and creating a European competing network of freight railway services.

These decisions would be crucial to the success of the Betuwe line.

Several rounds of decision-making were yet to come. Decisions needed to be taken about

ﬁnancial support from private sources, exploitation of the line, harmonizing European working

arrangements, and so on. These decisions would determine whether or not a parliamentary inquiry

should be set up in the next century to investigate why so much public money had been

invested in something thatwas a failure. Thus far, decision-making on this issue had already had

two important unforeseen but desired results: the expected competition by the railway forced the

inland waterways shipping sector to make institutional changes and adopt a more competitive

approach. This had already led to an enormous increase in inland shipping. Second, freight trans-

port by road was becoming ecologically more sound. Ironically, it was then argued that both

these developments made rendered Betuwe railway line superﬂuous.

Application of the tracks model: additional insights into internal and

external dynamics

The tracks model enables us to analyse in more detail the composition of the various decisions

marking the succession of the various rounds in the decision-making process.

Within the ﬁrst round especially, the port authorities were convinced about the shortcomings

of the existing railway’s capacity. Their ambition to construct a railway was based upon empirical

knowledge about the weaknesses of the other possibilities: road transport and inland shipping.

This ambition matched that of the freight railway company perfectly.

However, for a long time the ministry and parliament did not recognize the urgency of the

perceived problem. Only in 1989 was the Minister of Transport convinced of the necessity to

develop better railway connections, but this was not based upon much factual evidence. The new

1980 cabinet (Christian Democrats and Socialists) decided to invest in sustainable economic growth

and support the Betuwe line project. A breakthrough was came about by framing the Betuwe

line in such a way that it was compatible with the policy of sustainable economic growth.

After two years of further investigation, the Dutch parliament hesitated to ratify the Planning

Central Decision Procedure. A second opinion was sought and compensatory measures proposed

to meet the wishes of local and regional stakeholders. Parliament only approved the Betuwe

line after combining its main policy with a number of regional policies.

The Hermans Committee brought together much additional factual evidence but advised again

in favour of the Betuwe line. Increasingly the line was framed as a long-term strategic invest-

ment and as the only viable alternative for the harmful environmental impact of the existing

means of transport. This advice was an important building block for parliament to ratify the deci-

sion to start constructing the Betuwe line. Five years later, doubts arose about the short-term

eﬀectiveness of the Betuwe line, and it came to be regarded as a potential planning disaster.

Research models

313

Table 23.2 Events in decision-making on the Betuwe line and a fourfold research result

Empirical events Interpretations based on

the phase model

Interpretations based on

the stream model

Interpretations based on

the rounds model

Interpretations based on

the tracks model

The Rotterdam Port

Authority proposes a

railway link to Germany

(1985)

Irrelevant

(Ministry did not respond to

proposal)

Solution was unable to ﬁnd a

problem and sufﬁcient

participants

Start of ﬁrst round; the port

authority aims to increase the

range of transport modalities

Based upon a strong interest to

remain competitive, port

authorities perceive hinterland

connections as insufﬁcient

New enterprise for railway

freight transport adopts

Betuwe railway line (1987)

Lobbying activity resulting in

a ministerial committee in

1989 (still not relevant)

Problem of a lack of proﬁt

was combined with the

solution of a new line

The start of second round of

decision-making, aiming to

create a viable freight train

enterprise

Private ambitions are connected

and stakeholders starts to work on

a convincing factual underpinning

and problem frame

A ministerial committee

is set up to investigate the

viability of the Betuwe line

(1989/1990)

Minister begins

decision-making by setting

up a committee (starting

point for analyses)

The department is faced with

a lack of major key projects

in its new strategic plan

Start of third round leading to

the adoption of the project as

part of the strategic plan of the

ministry

Ambition of cabinet to invest in

sustainable growth is connected to

problem frame of Betuwe line

coalition

The minister points out a

concrete route (1992),

starts Planning Central

Decision Procedure and

organizes popular

participation

The ofﬁcial start of the

Planning Central Decision

Procedure

(policy formation)

Choosing a concrete route

was seen by the policy

community as a solution to

solve the problem of slow

decision-making

Start of fourth round, aiming to

speed up decision-making, but

actually generating and

activating opposition against

project

The EIA and positive transport

scenarios contribute to the

problem frame that railway

connections are the weak link

Parliament adopts proposal

(1993)

Central decision is taken

(policy adopted)

Coupling of problem, solution

and participants

End of fourth round Parliament accepts Betuwe line

based upon second opinion and

with compensatory measures

Parliamentary elections

resulting in a new Cabinet

(1994)

The central decision is

reconsidered in an opaque

political arena

Second coupling in which

new participants introduce

new policy deﬁnitions

Fifth round helps party to

change position and used by

societal groups to get additional

proposal accepted

Hermans Committee decides to

use data about existing capacity to

advise in favour of Betuwe line.

Parliament enforces additional

compensatory measures

Table 23.2 (continued)

Empirical events Interpretations based on

the phase model

Interpretations based on

the stream model

Interpretations based on

the rounds model

Interpretations based on

the tracks model

The Administrative Court

overrules all written

objections (1997)

Ofﬁcial start of construction

activities; the implementation

really starts

Conﬁrmation of the existing

coupling between political,

policy and problem stream

In sixth round, opponents

activate potential coalition in

business, criticizing added

value of project

New factual evidence fuels framing

of Betuwe line as potential

planning disaster but is not

combined with powerful

ambitions that can alter new

ofﬁcial frame in which Betuwe line

is ‘strategic choice for the

long-term.’

The Cabinet decides not

to build the northern

branch of the Betuwe

railway line (1999)

Partial policy termination due

to new information

New linking of politics with

alternative problems or

solutions

In seventh round, minister is

convinced by inﬂuential actors

that public money is needed

elsewhere.

New insights in the added value of

the northern branch and local

resistance against it, contribute to

the decision not to build this

branch

Source: Authors

There was also ﬁ erce regional resistance. All these considerations led to the decision not to go

ahead with the construction of the northern branch.

Conclusion

Decision-making has become increasingly complex during the past decades. There is a growing

variety of relevant actors and deﬁnitions of problems and solutions. In this chapter the question

of how to conceptualize decision-making in order to generate useful insights for the under-

standing of complex dynamics in decision-making processes has been answered. Four con-

ceptual models were presented: the phase model; the stream model; the rounds model; and the

tracks model. In the phase model, the focus is on decisions taken by a focal actor, targeting a

speciﬁc problem. In the stream model, the focus is on the linking of three more or less inde-

pendent streams, namely problems, solutions and politics. The rounds model focuses on the

interaction between the various decisions taken by diﬀerent actors. And ﬁnally, the tracks model

clariﬁes how decisions are composed out of speciﬁc selections of facts, frames and ambitions,

where the separate tracks evolve erratically due to internal and external dynamics. If we analyse

reality from these four models, diﬀerent images of reality are obtained, each of which provides a

partial insight into reality.

Thanks to the use of the phase model, we now know a great deal about how a single actor

deﬁnes policy in terms of problems and solutions, policy adoption, implementation and eva-

luation. Attention is paid to a focal actor, often the central government, and the way in which

this central actor organizes its own policy processes. The phase model focuses mainly on the

intended policy of this focal actor. This is both its strength and its weakness. The case study on

the Betuwe railway line presents a national government that is dedicated to the project and was

able to defend it successfully against outside attacks.

The stream model concerns the fragmentation of the decision-making world into problems,

solutions and politics as separate streams. If these streams come together by chance, progress can

be made. It is not the intended policy of one actor, but the links between the streams that are

crucial. The case study shows that diﬀerent problems were linked to the Betuwe line solution.

The lack of alternative solutions was an important reason for this project being placed on the

central government’s agenda. New fundraising initiatives are still a threat to the project.

The rounds model focuses on the interaction between interdependent actors. Every new actor

entering decision-making introduces new problems and solutions. Particularly on these points,

the rounds model oﬀers a number of additional possibilities to enhance insight into decision-

making. By not just accepting that there are various actors, but also accepting that all these actors

contribute to the decision-making process and can even inﬂuence the results of this process

helps to explain the complexity and gain insight into interaction patterns which are used for

governance of the network society. The model emphasizes that achieving satisfactory results

depends not only upon decisions taken by individual actors, but to an increasing extent upon

the interaction between decisions taken by several actors. By using the rounds model, the researcher

can analyse the interactions between decisions. All actors involved in governance can beneﬁt

from these insights.

As a further elaboration of this model, the tracks model highlights the evolution of the content

of decision-making within the various rounds. It focuses upon the question: what is exchanged

in the various interactions between actors? By distinguishing between facts, frames and ambi-

tions and by analysing the way they inﬂuence each other, we enrich our understanding of how

the results of decision-making processes are formed out of the complex interaction between a

variety of partial and sometimes partisan knowledge claims, subjective problem perceptions and

Teisman and van Buuren

316

competing or complementary interests. Because of the fact that it combines attention to develop-

ment in actor ambitions with attention to knowledge dynamics and framing/learning processes, it

goes one step beyond traditional models which normally focus on only one or two of these issues.

New insights on traces, tracks, time lines and process systems indicate the need for a model

able to accommodate a wider perspective.

Notes

1 This is an updated and extended version of Geert R. Teisman (2000) ‘Models for research into decision-

making processes: On phases, streams and decision-making rounds, Public Administration 78(4): 937–56.

2 In this article we will not deal with the question of evaluation.

3 Between 1991 and 1997 the project director wrote a book on the subject entitled Slag om de Betuwe

Route (The Battle of the Betuwe Line) (Boom 1997).

Bibliography

Anderson, J. 1979. Public Policy Making. New York: Praeger.

Altman, J.A. and E. Petkus. 1994. ‘Towards a stakeholder-based policy process: an application of the social

marketing perspective to environmental policy development’, Policy Sciences 27: 37–51.

Anglund, S.M. 1999. ‘Policy feedback: The comparison eﬀect and small business procurement policy’,

Policy Studies Journal 27(1): 11–27.

Bache, I. and M.V. Flinders. 2004. Multilevel Governance. Oxford: Oxford University Press.

Boom, H. 1977. Slag om de Betuwe Route. Leuven: Balans.

Börzel, T.A. 1998. ‘Organizing Babylon: On the diﬀerent conceptions of policy networks’, Public Administration

765: 253–73.

Bryson, J.M. and B.C. Crosby. 1992. Leadership for the Common Good, Tackling Public Problems in a Shared-Power

World. San Francisco: Jossey-Bass.

Butler, R. 1991. Designing Organizations: A Decision-making Perspective. London: Routledge.

Cohen, M.D, J.G. March and J.P. Olsen. 1972. ‘A garbage can model of organizational choice’, Administrative

Science Quarterly 17(1): 1–25.

Crozier, M. and E. Friedberg. 1980. Actors and Systems: The Politics of Collective Action. Chicago: University

Press.

Duit, A. and V. Galaz. 2006. ‘Governance and complexity: Emerging issues for governance theory’, Governance

23(3): 11–35.

Dunleavy, P. and Gillian Peel (eds). 1990. Developments in British Politics 3. Basingstoke: Palgrave Macmillan.

Edelman, M. 1971. The Symbolic Uses of Politics: Mass Arousal and Quiescence. Chicago: Markham Publishers.

Feldman, Martha. 1989. Order Without Design: Information Production and Policy Making. 1st edn. Stanford,

CA: Stanford University Press.

Fischer, F. 2003. Reframing Public Policy: Discursive Politics and Deliberative Practices. Oxford: Oxford University

Press.

Gerrits, L. 2008. The Gentle Art of Coevolution: A Complexity Theory Perspective on Decision Making over

Estuaries in Germany, Belgium and the Netherlands. Rotterdam: Erasmus University.

Hanf, K. and F.W. Scharpf (eds). 1978. Interorganizational Policy Making: Limits to Co-ordination and Central

Control. London: Sage.

Haynes, Philip. 2008. ‘Complexity theory and evaluation in public management’, Public Management

Review 10(3): 401–19.

Heclo,

H. 1974. Modern Social

Politics in Britain and Sweden: From Relief to Income Maintenance. New Haven,

CT: Yale University Press.

Heritier, A., C. Knill and S. Mingers. 1996. Ringing the Changes in Europe: Regulatory Competition and

Redeﬁnition of the State. Berlin: De Gruyter.

Hoogerwerf, A. (ed.). [1982, 1985]1989. Public Policy: An Introduction to the Policy Science (Overheidsbeleid:

een inleiding in de beleidswetenschap), 2nd, 3rd, and 4th edns. Alphen a/d Rijn: Samsom Tjeenk Willink.

Jessop, B. 1995. ‘The regulation approach, governance and post-Fordism’, Economy and Society 24, 307–33.

Jones, G. 1975. ‘Development of the Cabinet’, in W. Thornhill (ed.) The Modernisation of British Government.

London: Pitman.

Research models

317

Judge, D. 1993. The Parliamentary State. London: Sage.

Kingdon, J.W. 1984. Agendas, Alternatives, and Public Policies. Boston: Little, Brown.

Kickert, W., E.-H. Klijn and J. Koppenjan. 1997. Managing Complex Networks. London: Sage.

Koppenjan, J. 1984. Management of Policy Processes (Management van de beleidsvorming). The Hague: Vuga.

Lindblom, Charles E. and Edward J. Woodhouse. 1992. The Policy Making Process, 3rd edn. Englewood

Cliﬀs, NJ: Prentice Hall.

Mackintosh, J.P. 1977. The Politics and Government of Britain. London: Hutchinson.

March, J.G. 1988. Decisions and Organisations. Oxford: Blackwell.

—— and J.P. Olsen. [1976]1979. Ambiguity and Choice in Organisations, 1st and 2nd edns. Bergen:

Universitetsforlaget.

Marin, B. and R. Mayntz (eds). 1991. Policy Networks: Empirical Evidence and Theoretical Considerations.

Frankfurt: Campus Verlag.

Mintzberg, H. 1973. The Nature of Managerial Work. New York: Harper & Row.

—— 1976. ‘Planning on the left side and managing on the right’, Harvard Business Review 54: 49–58.

——, D. Raisinghani and A. Therot. 1976. ‘The structure of unstructured decision processes’, Administrative

Science Quarterly 21(2): 246–75.

Morgan, G. 1997. Images of Organisation, 2nd edn. London: Sage.

Ostrom, E. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge:

Cambridge University Press.

Quade, Edward S. 1972. Analysis for Public Policy Decisions. New York: Rand Corporation.

Rhodes, R.A.W. 1995. ‘From prime ministerial power to core executive’, in R.A.W. Rhodes and P. Dunleavy

(eds) Prime Minister, Cabinet and Core Executive. London: Macmillan.

—— 1996a. ‘Governing without governance: Order and change in British government’, inaugural lecture,

University of Newcastle, 18 April.

—— 1996b. ‘The new governance: Governing without government’, Political Studies 44: 652–67

——and P. Dunleavy (eds). 1995. Prime Minister, Cabinet and Core Executive. London: Macmillan.

Rittel, H.W.J. and M.M. Webber. 1973. ‘Dilemmas in a general theory of planning’, Policy Sciences 4:

155–69.

Rosenau,

J. 1992. ‘Governance, order

and change in world politics’, in J.N. Rosenau and E.O. Czempiel

(eds) Governance without Government, Cambridge: Cambridge University Press.

Sabatier, P.A. 1988. ‘An advocacy coalition framework of policy change and the role of policy-oriented

learning therein’, Policy Sciences 21(2–3): 129–68.

—— 1999. Theories of the Policy Process, Colorado: Westview Press.

Sato, H. 1999. ‘The advocacy coalition framework and the policy process analysis: The case of smoking

control in Japan’, Policy Studies Journal 27(1): 28–44.

Scharpf, F. 1997. Games Real Actors Play: Actor-centered Institutionalism in Policy Research. Boulder, CO:

Westview Press.

——, B. Reissert and F. Schnabel. 1978. ‘Policy eﬀectiveness and conﬂict avoidance’, in Hanf, K. and

F.W. Scharpf (eds) Interorganizational Policy Making: Limits to Co-ordination and Central Control. London:

Sage.

Schön, D.A. and M. Rein. 1994. Frame Reﬂection: Toward the Resolution of Intractable Policy Controversies,

New York: BasicBooks.

Smith, M.J. 1998. ‘Reconceptualizing the British state: Theoretical and empirical challenges to central

government’, Public Administration 76(1): 45–72.

——, D. March and D. Richards. 1993. ‘Central government departments and the policy process’. Public

Administration 71: 567–94.

Strand, R. 2002. ‘Complexity, ideology, and governance’, Emergence 4(1–2): 164–83.

Teisman, G.R. [1992, 1995]1998. Complex Decision-making, A Pluricentric View (Complexe Besluitvorming),

1st, 2nd, and 3rd edns. The Hague: Vuga.

—— 2000. ‘Models

for research into decision-making processes: On phases, streams and decision-making

rounds’, Public Administration 78(4): 937–56.

——with

E.H. Klijn. 1997. ‘Strategies and games in networks’, in W. Kickert, E.H. Klijn and J. Koppenjan

(eds) Managing Complex Networks. London: Sage, pp. 98–118.

——and E.H. Klijn. 2002. ‘Partnership arrangements: Governmental rhetoric or governance schema?’,

Public Administration Review 62(2): 197–205.

——, A. van Buuren, and L. Gerrits. 2009. Managing Complex Governance Systems, Dynamics, Self-Organization

and Coevolution in Public Investments. London: Routledge

Teisman and van Buuren

318

Termeer, C. 1993. Dynamics and Inertia in Manure Policy (Dynamiek en inertie in mestbeleid). The Hague:

Vuga.

van Bueren, E.M., E.H. Klijn, and J.F.M. Koppenjan. 2003. ‘Dealing with wicked problems in networks:

Analyzing an environmental debate from a network perspective’, Journal of Public Administration Research

and Theory 13(2): 193–212.

van Buuren, A. 2006. Competent Decision-making. Making Multiple Knowledge for Spatial Development (in

Dutch), Boom/Lemma, Den Haag.

——and L. Gerrits. 2008. ‘Decisions as dynamic equilibriums in erratic policy processes’, Public Management

Review 10(3): 281–399.

——, A. Gerrits, P. Marks. 2009. ‘Public policy-making and the management of co-evolution’, in Ties-

man, G., van Buuren, A., Gerrits, L. (eds) Managing Complex Governance Systems, London: Routledge,

134–153.

van Herk, S., C. Zevenbergen, J. Rijke, and R. Ashley. 2011. ‘Collaborative research to support transition

towards integrating ﬂood risk management in urban development’, Journal of Flood Risk Management

4(4): 306–17.

Vickers, Sir Geoﬀrey. 1965. The Art of Judgment: A Study of Policy Making. New York: Basic Books.

Wagenaar, H. 2007. ‘Governance, complexity, and democratic participation: How citizens and public

oﬃcials harness the complexity of neighborhood decline’, The American Review of Public Administration

37: 17–50.

Research models

319

The garbage can model and the

study of the policy-making process

Gary Mucciaroni

Origins and description of the GCM

Kingdon imported the garbage can model (GCM) into political science and adapted it to study

public policy-making from the seminal work of Cohen et al. (1972) who used the “garbage

can” metaphor to describe decision-making in universities and other complex organizations. As

Nikolaos Zahariadis (2007: 66) points out, the GCM “theorizes at the systemic level, and it

incorporates an entire system or separate decision as the unit of analysis.” Kingdon applied the

GCM to the study of the federal government and conducted panel interviews with more than

240 participants in the policy-making process, primarily in the areas of transportation and health

policies. Borrowing from Cohen et al., Kingdon described the federal agenda-setting process as

a mix of order and chaos, or “organized anarchy.” The “ can” is the set of more permanent and

predictable features of the policy-making process, such as institutional arrangements and

procedures that shape and structure behavior.

The main focus of the GCM, however, is on the “garbage” in the can because the items that reach

the agenda at any given time arise out of the mix of contents in the can. The can’s contents consist

of three “streams”—“problems,”“solutions” and “politics”, which operate largely independent of

one another. According to Kingdon (1984: 93), an issue reaches the agenda when “a problem is

recognized, a solution is available, and the political climate makes the time right for change.” Policy-

makers and others recognize problems when social indicators, such as crime or unemployment rates,

indicate that an undesirable condition is worsening or reaching a critical threshold, or, when a

“focusing event”—a salient and usually dramatic episode or crisis—draws attention to the problem

and lends it a sense of urgency. Solutions include “the gradual accumulation of knowledge and

perspectives among specialists in any given area.” Specialists generate and disseminate solutions,

provide evidence and arguments in support of them that are intelligible to non-specialists, and help

to package them into concrete proposals. “Politics” refers to the wide variety of actors and forces

that constitute the political environment, including the “national mood,” the organization and

mobilization of interest groups and social movements, the partisan and ideological composition of

Congress and the executive branch, and other aspects of the political context in which agendas are set.

The streams operate largely independently because diﬀerent groups of people or forces control

each of them. Problems come and go given changing conditions in the world, without regard

320

for either the status of solutions that might address the problems or the state of the political

context. Groups of experts and advocates conceive, develop and disseminate solutions over the

long-term, without necessarily reacting to the immediacy of problems or the tenor of the political

environment. And political change is governed by processes, like elections, budgeting, and the

mobilization of interests that operate on their own schedules and trajectories.

An item (problem or solution) reaches the agenda (or is adopted) only when the three streams

are aligned properly. Problem X or solution Y will not get on the agenda (or be chosen) unless

the problem is recognized, a solution is considered to be available, and political conditions are

conducive for giving the problem and solution serious attention. If any of these three elements

is missing, the item will not get on the agenda (or the policy adopted). Problem recognition,

solution availability and political opportunity, thus, each constitute necessary conditions for an

item to reach the agenda (or a policy adoption).

At particular junctures, a policy entrepreneur with a commitment to a particular issue and the

ability to draw attention to it, will couple a problem with a solution. Entrepreneurs wait for an

auspicious time, such as after a dramatic focusing event and when Congress and the president are

controlled by oﬃcials who are interested in addressing the issue. These “policy windows ” often

open unpredictably and only for a short while. Thus, entrepreneurs must act expeditiously and

eﬃciently to get their pet item on the agenda (or get it adopted) before the window closes,

when the political environment changes and attention shifts to other issues.

Criticisms and limitations of the GCM

Scholars have pointed to a number of important problems and deﬁciencies with the GCM. In

this section, we review several of these criticisms.

Falsiﬁability. The model, at least as a whole, is not falsiﬁable. Can we imagine a set of cir-

cumstances in which some condition in society was not considered a problem receiving serious

consideration in government? Or, a situation in which an item got on the agenda (or a policy

adopted) even though the political environment was not favorable to getting it on? Since the

agenda, by deﬁnition, is composed of conditions that some key group of decision-makers con-

siders “problems,” it would be impossible for non-problematic conditions to get on the agenda.

Since the political environment refers to those actors who are in power or who have inﬂuence

over those in power, a problem or solution could not receive serious attention unless they

permitted it. Non-falsiﬁ ability arises, as in most cases, because the model is formulated very broadly

and ambiguously. One strategy for making the model more falsiﬁable would be to develop

categories of problems, solutions and political contexts from which we could derive testable

hypotheses about how particular kinds of problems and solutions reach the agenda given the

existence of particular political conditions (Mucciaroni 1992: 463–5).

Bendor et al. (2001) have critiqued the original GCM developed by Cohen et al. (1972),

arguing that the model is more like a metaphor than a model. As a result, many applications of

the GCM “do little more than describe some parts of an organization as garbage cans … but it

is not at all clear that they provide real … explanations.” According to them, the GCM merely

provides a scheme for describing and labeling phenomena without explaining them.

Situational versus structural levels of analysis. Before we can model agenda-setting, we need to be

clear about what we are trying to explain related to agendas. It makes a great deal of diﬀerence

whether we are asking why certain items reach the agenda at particular moments in time versus

why certain items reach the agenda ever. The GCM is better at explaining when the agenda

changes than explaining how and why the agenda changes. Much of the reason that the GCM is

better at explaining the timing of agenda (or policy) change rather than give a full accounting of

The garbage can model

321

it, arises because the model conceptualizes constraints on agenda-setting (or decision-making)

mainly at what I call the “situational” level of analysis—aspects of the policy-making environment

that are relatively temporary and changeable. Problems (e.g. crime, airline safety) “get hot” and

then cool oﬀ as attention drifts to other problems and new crises. Solutions (e.g. the “ﬂat tax,”

“deregulation”) snowball into fads but eventually turn stale and their popularity fades. Congres-

sional majorities and presidential administrations come and go every two, four, or eight years;

the “public mood” varies with the vagaries of the economy and other conditions. But political

actors and forces that are more permanent, that change slowly or hardly at all from decade to decade,

also shape agendas and decisions. It would be as much of a mistake to overlook the impact of

these forces on agendas and policy decisions as it would be to ignore the impact of old age or

genetics on human health.

One of kind of structural constraint is culture, political traditions and dominant ideologies.

For example, the problems of anti-gay discrimination and poverty were oﬀ the agenda for many

decades before they appeared in the 1960s in the case of poverty, and after 1970 in the case of

gay rights. They were kept oﬀ the agenda by deeply ingrained, taken-for-granted beliefs that

either nothing could be done about poverty or that it was the result of personal failings and that

discrimination against gays was justiﬁed because they were “ sick,”“predators,” or “sinners.”

Similarly, institutional arrangements hinder and facilitate problems and solutions from reaching

the agenda. For example, during the Great Depression of the 1930s, policy-makers in some

countries considered Keynesianism a plausible solution to the crisis while those in other nations

did not. But whether Keynesian ideas became accepted partly depended upon which institutions

had responsibility for making economic policy and whether institutions provided Keynesian

economists with access to decision-makers (Hall 1989; Skocpol and Weir 1985). And some

solutions have a bad “ﬁt” with particular institutions. For example, regardless of the backgrounds

of particular judges, or other relatively ﬂuid conditions, solutions that do not ﬁt into the logic of

legal reasoning and the language of “rights” may not get a hearing or may be rejected.

Kingdon was aware of the importance of structural features of the policy-making process. He

mentions, for example, that the budget process has particular procedures and timetables that

constrain agenda-setting, and that American political culture ﬁlters out solutions that are incompa-

tible with it before they can reach the agenda. The problem is that the GCM does not theorize

structural constraints, leaving them largely outside its parameters. Because they are not incorpo-

rated in the model, the GCM is signiﬁ cantly incomplete. Admittedly, all models are incomplete,

but these are large omissions in light of the GCM’s theoretical ambitions.

Lack of historical causality. Much of what happens in the policy-making process, how it hap-

pens, and to whom, is in some fashion rooted in the past. The GCM neglects much of history

because it focuses solely upon immediate or near-term changes in problems, solutions, and

politics. Kingdon justiﬁes overlooking historical explanations for agenda and policy change on

the grounds that historical investigation would simply lead us down a fruitless path of “inﬁnite

regress”: in searching for the relevant antecedent events and conditions that shaped the agenda,

we would be compelled to search for the antecedents to the more proximate antecedents, and

so forth, endlessly. Tracing historically relevant antecedents to explain later choices and condi-

tions need not, of course, end up being a slippery slope into an unceasing quest for initial

causes. If that were the case, historical studies would be impossible.

The lack of historical perspective contributes to the overall randomness of the GCM. With-

out examining antecedent choices and conditions that might have given rise to, or facilitated, a

problem or solution from getting on the agenda, many items seem to appear and disappear

unpredictably and suddenly. Problems may indeed “suddenly

get hot” but

the problems them-

selves and the reasons they come to be perceived as problems are rooted historically. For

Mucciaroni

322

example, under the GCM, we might observe that unemployment rises sharply within a few

months and that the public and policy-makers take notice. But whether we regard unemploy-

ment as a problem suitable for government attention, how high unemployment must be to be

considered “unacceptably high,” and why the rate went up, we can only understand adequately

in historical perspective. Similarly, to understand why “supply-side tax cuts” became perceived

as a plausible solution to unemployment or sluggish economic growth, one has to have some

historical understanding of the evolution of economists’ theories and opinions, in what sense

American economists came to understand “Keynesianism,” its rise and fall as a theory and policy

prescription, and the lessons that economists and others drew from experiences with previous

eﬀorts to cut taxes. It is not enough to know that at a particular point in time Keynesianism

came to be seen as a solution to a particular economic problem. We want to know why and how

it became a viable solution.

Items often reach the agenda and come up for decision because past eﬀorts to put them there

have failed or never materialized. Or, a problem reached the agenda and a policy was adopted

to address it, which in turn, gave rise to new problems that eventually gained agenda status.

Scholars working from a historical perspective have developed a number of concepts to help us

to understand history as a causal process. The development and dissemination of solutions, for

example, is often done through learning and lesson-drawing, which occur over time (Heclo 1975;

Rose 1991). Public policies are also often the product of “critical junctures,” which produce

distinct “legacies” that would not have occurred if the critical juncture had not occurred (Collier

and Collier 1991). Alternatively, the legacies may be the result of “antecedent conditions” and

“constant causes” that predated the hypothesized critical juncture and account for the legacy.

Policymaking is also subject to path dependency. Political events and policy choices at one point in

time often put nations and organizations on particular trajectories that close out or reduce the

probability of making other choices in the future because the original events and choices create

dominant frames, institutions and interests that become entrenched long after the conditions

that established the path have ended (Pierson 2000).

The criticism that the GCM is insuﬃciently grounded in historical analysis would be less

problematic if the model was considered exclusively a model of agenda-setting. Since an “agenda”

is conventionally deﬁned as consisting of those items that a government (or set of decision-

makers in any complex organization) gives serious attention at a point in time, paying attention

only to contemporary, immediate, situational factors is understandable. An organization’s atten-

tion to particular problems is usually short-term and often intermittent. But if we extend the

GCM beyond agenda-setting to policy formulation and adoption, as many scholars have done, we

have a harder time justifying the lack of historical analysis and appreciation of structural con-

straints. Policy-making is more than attention-getting and a deeper understanding must go

beyond situational factors.

Randomness and the separation of the streams: Probably the most frequent criticism of the GCM

is that it portrays decision-making as overly chaotic and random. Even defenders and users of

the GCM admit that the model “shares common ground with chaos theories in being attentive

to complexity, in assuming a considerable amount of residual randomness, and in viewing systems

as constantly evolving and not necessarily settling into equilibrium” (Zahariadis 2007: 66). The

feature of the GCM that contributes most to the model’s randomness is the assumption the

loose and independent relationship exists among the three separate streams. While changes within

each stream reﬂect purposive behavior and patterns, the core of the model—the simultaneous

occurrence of a salient problem, available solution and favorable political conditions—gives rise

to outcomes that are largely fortuitous and unpredictable. The coupling of the three streams is a

purposive, non-random process, but it occurs after the alignment of the three streams. Kingdon

The garbage can model

323

(1984: 93) stipulated that only “hints of connection” existed among the streams. But is the lack

of connectedness among the streams justiﬁed? An alternative model would stress the importance

of changes in one stream contributing to changes in the others and would trace out the inter-

actions and interdependencies among the streams. Pursuing an approach in which the streams

were more interconnected would reduce the currently high level of indeterminacy and ran-

domness in the GCM and uncover a greater degree of behavior that is collaborative, purposive,

and strategic. Indeed, those items that reach the agenda may be precisely those that exhibit

stronger linkages and congruence among the streams, where advocates and their allies do not

leave congruence to chance, but actively promote it. In this model, changes in one stream give

rise to and reinforce changes in the others, increasing the probability that entrepreneurs will

attempt to couple the streams and that their eﬀorts will succeed.

Perhaps the closest links are those between solutions and politics. Solutions that reach the

agenda are those that have a good ﬁt with policy-makers’ goals, such as reelection, inﬂuence,

and promoting the adoption of “good” policy (Fenno 1973). Obviously, policy-makers choose

solutions that have the best chance of promoting their goals. At the same time, experts and

advocates craft and promote their favorite solutions with at least some attention to policy-

makers’ goals. The type of policy or program promoted, its speciﬁc provisions, the evidence and

arguments mustered for it, how it is packaged and framed––all are geared toward building support,

muting opposition, and persuading the undecided and ambivalent that the proposed course of

action is desirable, or at least plausible. Since policy-makers’ goals are usually often broad, and

their views on how to achieve them are malleable, those who advocate on behalf of particular

solutions have some room to persuade them.

Politics and problems are also closely connected. Perhaps the best-known example of this

phenomenon is when incumbent politicians are ousted from oﬃce, or threatened with being

outsted, when economic indicators, particularly the unemployment rate, worsen or fail to

improve during a recession. Many less salient examples also exist. For example, when fuel prices

rose sharply in the mid-1970s, businesses that shipped their goods by truck, including farmers,

manufacturers and many other businesses, formed a coalition with liberal consumer groups to

push for trucking deregulation (Robyn 1987: 96–7, 234–5). In other cases, previous policy

choices create new problems or exacerbate existing ones that, in turn, create political opportu-

nities for elected oﬃcials. For example, the creation of very generous tax loopholes in 1981

resulted in large revenue losses. By the mid-1980s, a bipartisan coalition emerged in favor of

closing the loopholes and converting them into across-the-board rate cuts for taxpayers gen-

erally (Conlan et al. 1990). Republicans regained control of the House and made gains in the

Senate in the 2010 elections because the combination of a devastating recession, war spending

and growing entitlement spending created a “debt crisis” that allowed Republicans to push debt

reduction as the central issue on the agenda after the election. In both cases, the perceived wor-

sening of problems contributed to conditions that were conducive to getting the problems on

the agenda.

Finally, problems and solutions are linked. The GCM rightly uncovered instances when,

counterintuitively, experts developed solutions before they found the problems to which they

were eventually applied. At the same time, experts often develop solutions with very clear pro-

blems in mind at the outset and their linkage of a solution to that particular problem often

carries through the policy-making process. Both patterns are at work, perhaps often in the same

proposal. For example, the original rationale that tax experts in the Treasury Department put

forward the argument for reforming the tax code (by reducing the number of special provisions

in the code reducing taxpayers’ liability) was to increase the fairness of the income tax. Eventually,

policy-makers and others took up reform to address other problems (the complexity of the code

Mucciaroni

324

and economic ineﬃciency created by distorted market incentives), but the original problem that

the experts were trying to address endured as a principal rationale for reform.

Problems and solutions are linked in more subtle ways as well. As problems get worse,

plausible solutions become attractive and are more likely to gain agenda status. Solutions that

might have a harder time getting political support (for any number of reasons) when a problem

is perceived to be under control or imposing only modest costs, will have an easier time as the

problem worsens and the expected windfall from applying the solution grows. In the tax reform

example again, as the problems of unfairness, revenue loss, complexity and ineﬃciency grew, so

did the attractiveness of closing the loopholes because the costs of the existing policy rose at the

same time that the expected beneﬁt from closing the loopholes also rose (in the form of lower

tax rates, for example).

Studies that have employed the GCM after Kingdon

As of June 25, 2011, Google Scholar registered 6,650 citations for Kingdon’s Agendas book.

Many of the citations are from political science and policy studies as well as other disciplines,

ﬁelds and professions. At the same time, according to Bendor et al. (2001), the GCM is not used

extensively in political science research or in related disciplines. Indeed, the number of studies in

public policy that actually use the GCM or test hypotheses derived from it is surprisingly small

given the attention that Kingdon’s book received and its status as a staple on public policy syl-

labi. Zahariadis (2007: 79) acknowledges that the GCM, or “multiple streams model,” as he

prefers to call it, “does appear to be an argument that many scholars quote but few explicitly

use.” Bendor et al. attribute the GCM’s relative lack of use in empirical studies to weaknesses in

the model that render it a poor theoretical guide for research. Zahariadis counters that the problem

has more to do with “paradigmatic problems” that have plagued the model. Following Sabatier

(1999), he suggests a number of reasons for why the GCM does not guide empirical research in

policy studies more frequently: the model’s principal proponent and popularizer—Kingdon—

did not vigorously promote its usage; many people consider the GCM only relevant for under-

standing agenda-setting rather than policy formulation and adoption; and, researchers working

on subnational levels of government identify the GCM with agenda-setting at the national level.

These reasons are plausible. Kingdon retired from academe not that long after he published the

second edition of his book. And, in conﬁning the application of the GCM to the agenda-setting

stage, Kingdon may have limited its perceived relevance given that conceptualizing the policy

process as a set of distinct stages may be a useful heuristic device, but not very realistic.

The GCM has appeared prominently in a modest number of public policy studies that came

after Kingdon ’ s book was published. Most of these works have sought to broaden the applicability

of the GCM (or elements of it), reﬁne its central concepts, and use it to interpret case study ﬁnd-

ings. Zahariadis, for example, has sought to show how the GCM (or “multiple streams ” model,

as he prefers to call it) is useful for understanding policy-making outside the United States, speciﬁ-

cally the privatization of public enterprises in Britain and France (1995) and Greek foreign

policy (2005). An important aim of Zahariadis’s (1995, 9) study of privatization was to extend the

GCM beyond agenda-setting, to the entire process of policy formulation and decision-making.

Zahariadis argues that Cohen, March and Olsen formulated the GCM originally to explain decision-

making, not just agendas, and that agenda-setting, policy formulation and decision-making are

integrally related. These arguments are sensible, but extending the GCM to the policy-making

process as a whole makes it even harder to maintain the notion that problems, solution, and politics

are largely separate until a ﬁnal coupling. Once an item reaches the agenda, it is implausible to

believe that the independence of streams remains relevant during policy formulation and

The garbage can model

325

decision-making. Unless we are trying to answer the narrow question of “why did policy-makers

make any decision rather than none at all?” the GCM would ﬁnd it diﬃcult to explain why policy-

makers decided on one speciﬁc policy and rejected others. Second, extending the GCM beyond

agenda-setting does nothing to ameliorate other problems with the model, such as its failure to

ground its explanations suﬃciently in the historical and structural aspects of policy-making.

Framing his explanation in terms of the coming together of the three streams of the GCM,

Zahariadis (1995) argues that Britain and France adopted privatization when the following con-

ditions converged: (1) policy-makers perceived the solution—selling oﬀ state-owned enter-

prises—as technically feasible and compatible with their values; (2) social indicators revealed a

problem—that government budgets were under stress and the public was resistant to new taxes;

and (3) the occurrence of a key change in the political environment—the rise to power of

political parties that were not identiﬁed with public enterprise and that possessed the ideological

commitment and political capacity to work to end it. In some of cases, as Kingdon had found

earlier, specialists developed the solution (privatization) before the speciﬁc problem for which it

eventually was adopted had emerged.

In trying to account for Greek foreign policy-making over the span of ﬁve years on the issue

of Macedonia, Zahariadis (2005) argues that the GCM is more useful than two other explana-

tory frameworks (“rational internationalism” and “two-level games”). The substance of Zahar-

iadis’s explanation focuses on the importance of policy elites’ manipulation of emotive symbols

and domestic politics on foreign policy decisions. If the manipulation of emotion and con-

siderations of domestic politics are what matter in his explanation of Greed policy, then the

need for the GCM in Zahariadis’s account is not readily apparent. According to Zahariadis

(2005: 172), “as multiple streams [i.e. GCM] hypotheses, domestic agendas make a big diﬀer-

ence in the making of foreign policy.” But what rationale exists for equating the GCM (or

“multiple streams”) with “domestic politics”? And again, how does the GCM contribute to an

explanation that the variables “emotion” and “domestic politics” do not already contribute on

their own? A compelling case that the GCM helps to explain Greek foreign policy would

require showing how elements of the GCM are essential causal mechanisms that facilitate or

shape the inﬂuence of emotion and domestic politics on foreign policy.

McLendon (2003) employed the GCM in his study of how the issue of higher education

decentralization reached the agenda in the US states in the 1980s and 1990s. He compared the

GCM with incremental and rational-comprehensive theories of policy-making to assess their

usefulness for explaining how the issue reached the agenda and for elucidating which actors in

the policy-making process exerted the most inﬂuence. Analyzing an eclectic mix of data,

including 61 semi-structured interviews, from in-depth case studies of three states, McLendon

found that agenda setting on this issue resembled the GCM far more than the other two

models. Like Kingdon, McLendon found “clarity, order and pattern” within the structure of

each stream, and at the same time, “elements

of ambiguity, anarchy

and unpredictability … in

the speciﬁc coupling of problems, solutions and politics.” He also found support for several of

the features of the GCM as Kingdon formulated them: diﬀerent streams of activity that exhib-

ited a “surprising degree of independence” (though they were “not entirely unrelated”); “con-

ﬂict over autonomy and control within the education policy domain” was weakly connected to

the reasons that decentralization reached the agenda; the development of the decentralization

solution occurred before, not in response to, the problem policy-makers eventually addressed;

the rapid, non-incremental movement of the issue to the top of the agenda; and the signiﬁcant

role of policy entrepreneurs in helping to push the item on to the agenda.

Instead of employing the entire GCM, other studies have explored and elaborated upon

particular elements of it. Birkland (1997: 2004) has clariﬁed and developed the concept of

Mucciaroni

326

“focusing events” and investigated under what empirical conditions focusing events help to

push issues on to the agenda and lead to the adoption of new policies. In his study of natural

disasters, oil spills, and nuclear accidents, Birkland (1997) stipulates that focusing events are harmful,

unexpected, relatively rare, and aﬀect large numbers of people who learn about them simulta-

neously. Focusing events favor groups (particularly relatively weak ones) that seek changes in

public policy because the events provide valuable opportunities for the groups to point out

problems with the status quo and relieve them of the task of convincing the mass media to pay

attention to their issue. Not all focusing events lead to changes in the agenda or in policy.

Birkland argues that certain characteristics of focusing events and of the political environment in

which they occur help or hinder them from having an impact on policy. How much “focal

power” an event has depends, ﬁrst, upon the media attention that it receives. The level of media

attention, in turn, depends upon the scope of the event (or, how many people it aﬀects), its

rarity, the size of the media market in which it occurs, and the visibility and tangibility of the

damages that it produces. Focusing events also have greater impact if the policy domain is char-

acterized by a high level of organization, whether the community of experts support change,

and in domains with less opposition and polarization among advocacy coalitions. In a follow-up

study, Birkland (2004) examined the impact of the September 11 attacks on the United States

He hypothesized that highly signiﬁcant focusing events like 9/11—in which the level of damage,

potential future damage, and media attention were greater than typical focusing events—would

lead to windows of opportunity that were kept open much longer and would inﬂuence a

broader set of policy areas, initiatives, and agencies.

Finally, a few studies have tried to test hypotheses derived from the GCM. Even though the

GCM as a whole is not falsiﬁable, parts of it generate testable hypotheses. Robinson and Eller

(2010) tested hypotheses related to one of the model’s central and most contested features: the

independence of the three streams. Robinson and Eller collected data from surveys of Texas

school district superintendents related to participation in local, school district policy-making

concerning violence prevention programs. They hypothesized that if the GCM is correct, we

should observe diﬀerent sets of participants involved in the problem and solution streams.

Observing a participant in one stream should reduce the probability of ﬁnding the same parti-

cipant in the other stream. Since Kingdon also suggested that special groups of experts dominate

the solution stream, Robinson and Eller also tested the hypothesis that the solution stream

would be dominated by elite participants rather than ordinary citizens and non-specialists. Contrary

to the GCM, they found that participation in one stream dramatically increased the likelihood that

a participant would participate in the other stream and that elite participation did not “crowd

out” participation by “mass” participants (e.g. teachers and parents). One diﬀerence between

Robinson and Eller’s study and Kingdon’s, of course, is that one focused at the local and the other

at the federal level. Perhaps diﬀerences in policy-making at the two levels accounts for the

diﬀerent results. Obviously, much more research needs to be done to see if Robinson and Eller’s

results hold up at the local and state levels, in other policy areas and on other issues, and if they

do, what might account for the diﬀerences found when we compare degree of independence of

streams across levels of government.

Conclusion

The GCM has been, and will continue to be, among a handful of approaches to the study of

policy-making that all students in policy studies will learn about, discuss and use in their empirical

work. Undoubtedly, the debate over its strengths and weaknesses, its promise and limitations for

elucidating the real world of policy-making, will persist as well.

The garbage can model

327

Note

1 http://scholar.google.com/scholar?q=kingdon&hl=en&btnG=Search&as_sdt=1%2C39&as_sdtp=on

Bibliography

Bendor, Jonathan, Terry M. Moe, and Kenneth W. Shotts. 2001. ‘Recycling the garbage can: An assessment

of the research program’, American Political Science Review 95(1) (March): 169–90.

Birkland, Thomas A. 1997. After Disaster: Agenda Setting, Public Policy, and Focusing Events. Washington,

DC: Georgetown University Press.

——. 2004. ‘“The world changed today”: agenda-setting and policy change in the wake of the September

11 terrorist attacks’, The Review of Policy Research 21 (2) (March): 179–200.

Cohen, Michael, James March and Johan Olsen. 1972. ‘A garbage can model of organizational choice’,

Administrative Science Quarterly 17(March): 1–25.

Collier, Ruth Berins and David Collier. 1991. Shaping the Political Arena: Critical Junctures, the Labor Movement

and Regime Dynamics in Latin America. Princeton, NJ: Princeton University Press.

Conlan, Timothy J. 1990. Taxing Choices: The Politics of Tax Reform . Washington, DC: Congressional

Quarterly.

——, Margaret T. Wrightson and David R. Beam. 1990. Taxing Choices: The Politics of Tax Reform.

Washington: CQ Press.

Fenno, Richard F. 1973. Congressmen in Committees. Boston: Little, Brown.

Hall, Peter. 1989. (ed.) The Political Power of Economic Ideas: Keynesianism Across Nations. Princeton, NJ:

Princeton University Press.

Heclo, Hugh. 1975. Modern Social Politics in Britain and Sweden. New Haven, CT: Yale University Press.

Kingdon, John. 1984. Agendas, Alternatives, and Public Policies. Boston: Little, Brown.

McLendon, Michael K. 2003. ‘Setting the governmental agenda for state decentralization of higher education’,

The Journal of Higher Education 74(5) (September/October): 479–515.

Mucciaroni, Gary. 1992. ‘The garbage can model and the study of policy making: A critique’, Polity 24(3)

(Spring): 459–82.

Pierson, Paul. 2000. ‘Increasing returns, path dependence, and the study of politics’, American Political Science

Review 94(2) (June): 251–67.

Robinson, Scott E. and Warren S. Eller. 2010. ‘Participation in policy streams: Testing the separation of

problems and solutions in subnational policy systems’, Policy Studies Journal 382 (May): 199–216.

Robyn, Dorothy. 1987. Braking

the Special Interests.

Chicago: University of Chicago Press.

Rose, Richard. 1991. ‘What is lesson drawing?’, Journal of Public Policy 11(1): 3–30.

Sabatier, Paul A. 1999. Theories of the Policy Process. Boulder, CO: Westview Press, 117–66.

Skocpol, Theda and Weir, Margaret. 1985. ‘State structures and the possibilities of “Keynesian” responses

to the depression in Sweden, Britain and the United States’, in P. Evans, D. Reuschemeyer and Theda

Skocpol (eds). Bringing the State back in. Cambridge and New York: Cambridge University Press.

Zahariadis, Nikolaos. 1995. Markets, States and Public Policy: Privatization in Britain and France. Ann Arbor,

MI: University of Michigan Press.

—— 2005. Essence of Political Manipulation: Emotion, Institutions and Greek Foreign Policy. New York: Peter Lang.

—— 2007. ‘The multiple streams framework: Structure, limitations, prospects’, in Paul A. Sabatier, Theories

of the Policy Process. Boulder, CO: Westview Press, 65–92.

Mucciaroni

328

Part VII

Understanding the

implementation process

Bureaucracy and the policy process

Ora-orn Poocharoen

Introduction

Bureaucracy is a concept that has been widely written about. Scholars in various ﬁelds of social

sciences, including sociology, political science, economics, anthropology, history, public admin-

istration and public policy, have produced a vast amount of literature pertaining to almost every

angle of bureaucracy one could ever imagine. This chapter focuses only on studies from the

ﬁeld of public administration and public policy. The aim is to provide a description of impor-

tant debates, theories and frameworks that explain the relationship between bureaucracy and the

public policy process. The chapter starts with deﬁnitions of the bureaucracy, followed by expla-

nations of bureaucratic roles in policy-making and policy implementation. In the latter half of

the chapter, the discussion focuses on the role of bureaucracy in the era of governance and the

chapter ends with the topic of heterogeneity of bureaucrats in the present.

Every country has some form of bureaucratic apparatus to perform government functions,

thus it is a subject of wide interest for scholars and practitioners who are involved with the public

sector. The term originates from a cloth covering the desk of French oﬃcials in the eighteenth

century called a ‘bureau’ (Niskanen 1973). In earlier works on public administration, scholars

used the word bureau synonymously with bureaucracy (e.g. ibid.; Downs 1967). Since German

sociologist Max Weber (1864–1920) wrote on the ideal type of bureaucracy, our understanding

of the bureaucracy has evolved in various ways. The term has attracted diﬀerent sentiments ranging

from being powerful and eﬀective to being slow and ineﬃcient. The bureaucracy can be seen as

the enforcer, the implementer, the administrator, the facilitator, the manager, or the mediator of

government. We ﬁrst turn to the original works on bureaucracy to explore its deﬁnitions.

Deﬁnitions of bureaucracy

The word bureaucracy can be referred to in two ways. First, it refers to the bureaucracy as a

form of organization, where there are structures, rules, resources and goals. The organization

can be in any sector: public, private, non-proﬁt or a combination of these. Second, it refers to

the bureaucracy as administrative machinery that is part of the government. Its primary function

is to execute the will of the state or in other words to implement public policies. It is worth

331

noting that earlier scholars have identiﬁed the deﬁnition of bureaucracy to capture a spectrum

of meanings that fall within the two categories of organizational form and as part of the gov-

ernment. An often-cited author is Albrow, who suggested seven deﬁnitions of bureaucracy: (1)

bureaucracy as rational organization; (2) bureaucracy as organizational ineﬃciency; (3) bureaucracy

as rule by oﬃcials; (4) bureaucracy as public administration; (5) bureaucracy as administration by

oﬃcials; (6) bureaucracy as the organization; (7) bureaucracy as modern society (Albrow 1970).

Deﬁnitions 1, 6 and 7 emphasize bureaucracy as an organizational form, whereas deﬁnitions 2,

3, 4 and 5 refer to bureaucracy as part of the government. When the word bureaucracy is used

in everyday language, often it is a combination of these two deﬁnitions. In addition, sometimes

the term bureaucracy is used synonymously with its agent, the bureaucrats. Typically, these are

the people who work as civil servants in the public sector and who would be involved in public

policy processes in various degrees. Next, we shall look at these two deﬁnitions in more detail.

Bureaucracy as an organizational form

Organization theory is a subﬁeld that cuts across many disciplines including public administration,

business management and sociology. Organizations have various forms, structures, processes,

cultures, goals and values. Bureaucracy is nothing but one of the organizational form. The subﬁeld

of organization theory provides rich research on the bureaucracy and bureaucratic behaviour.

Examples include March and Herbert (1958), Blau and Scott (1962), Downs (1964), Hall

and Quinn (1983), Wilson (1989), Scott (1992), Hall (1996), Christensen (2007) and Shafritz

et al. (2010).

The classic deﬁnition of the bureaucracy as deﬁned by Max Weber often forms the basic fun-

damental traits of public sector organizations. In his seminal work, Weber oﬀers an explanation

of the ideal type of bureaucracy that has six characteristics (1946).

There are ﬁxed jurisdictional areas that are determined by rules that are laws or administrative

regulations:

 There is hierarchy, where there is supervision of subordinates by supervisors.

 The management of the oﬃce is based upon written documents.

 There is oﬃce management.

 Full working capacity of the oﬃcial is expected.

 The management of the oﬃce follows general rules that are stable and can be learned.

These characteristics were drawn from a study of various public organizations in Europe and

non-Western civilizations, including Egypt, Rome, China and the Byzantine Empire. This ideal

type describes the legal and rational system of authority that represents a modern bureaucracy. It

is diﬀerent from the traditional authority system that relies more on personal ties and the char-

ismatic authority system that relies on strong leadership. Weber’s work was translated into English

only in 1940 and since has become the ‘departure for all further analyses’ of the bureaucracy

(Shafritz and Hyde 2008).

This legal-rational foundation of the bureaucracy allows for institutions to gain and maintain

expertise, power and legitimacy for governance overtime. Public organizations, whether they

are ministries, bureaus, departments, or agencies, usually would be set up according to laws and

would employ full-time professionals, who are categorized into diﬀerent hierarchical ranks. The

work of bureaucracies would be based on documentation, where records are kept and rules are

followed. This form of organization or system of administration is also found in private com-

panies, non-proﬁt organizations, and international organizations where complex and large-scale

Poocharoen

332

administrative tasks take place. Kamenka provides a concise deﬁnition of bureaucracy that is

applicable to all organizations and not just governmental organizations as ‘a centrally directed,

systematically organized and hierarchically structured staﬀ devoted to the regular, routine and

eﬃcient carrying out of large-scale administrative tasks according to policies dictated by rulers or

directors standing outside and above the bureaucracy’ (Kamenka 1989: 157). Bureaucracy was

and is still is, in some ways, understood to be the most eﬃcient form of organizational

arrangement and is inevitable in all societies. Elliot Jacques alluded to this when he wrote,

[Bureaucracy] is one of the primary social institutions for any society which seeks the demo-

cratic enrichment and economic security which large-scale social, political, educational and

production technologies seem at their best to be able to provide. The simple fact is that if

we decide to proceed with the development of industrialized societies, then bureaucracies

on a large scale are here to stay.

(1976: 13)

Bureaucracy as part of the government

This view of the bureaucracy equates all public sector organizations ‘the bureaucracy’. Here, the

bureaucracy is made up of public oﬃcials who are appointed through recruitment processes and

not elected. They are in contrast to electoral institutions such as politicians, parliament, and the

executive branch. The bureaucracy, as a legitimate actor of the state, has power to govern according

to laws and rules. Governments deﬁne the scope and size of their bureaucracy diﬀerently. Some

would include the military and judiciary, while others would not. In federal systems, the bureau-

cracy could refer to either national level or subnational levels of government agencies. Some

scholars would use the term ‘public bureaucracy’ to emphasize that the focus is on the public sector.

However, most of the time, writers do use ‘bureaucracy’ and ‘public bureaucracy’ synonymously

and they usually mean bureaucracy in the public sector.

In conjunction with this deﬁnition, the bureaucracy is the institutional structure whereby the

bureaucrats are the individuals within that structure to implement policy and enforce laws.

Earlier, Downs wrote, ‘In my analysis, the term “bureaucrat” is in no way derogatory, but because

it is so universally regarded as an insult, I will use the more neutral term “oﬃcial” to describe

the type of person described above’ (Downs 1964: 4). Fifty years on and the general feeling

about the term have not changed. Due to connotations of ineﬃciencies and unresponsiveness, it

is now more favourable to refer to bureaucrats as public managers. It gives the impression that the

bureaucrat is not just someone who pushes paper and complies with standard operating proce-

dures but rather someone who is entrepreneurial and innovative and who can manage complex

situations and problems. Whether they are called bureaucrats, public oﬃcials, civil servants,

administrators, or public managers they are all agents within the bureaucracy.

One of the major debates in the ﬁeld of public administration is the premise that ‘the will of

the state’ and ‘the execution of that will’ should and can be separated. This is known as the

politics-administration dichotomy debate. Earlier scholars who have discussed this dichotomy

include W. Wilson (1887), Goodnow (1900) and White (1926). It is worth noting that their

works did not favour the dichotomy as many later claimed and that this debate has historical

roots in the United States, where the foundation of the state lies in the separation of power

between the executive, legislative and judicial branches of government (Svara 2001). This debate

is closely linked to the policy-making and policy implementation discussion of whether in

reality the two processes can be separated or not (e.g. Appleby 1949; Peters 1995). We can also

argue in the normative sense of whether or not the two processes should be separated. These

Bureaucracy and the policy process

333

debates shape the normative direction of division of role between politicians and bureaucrats.

The politicians are understood to be the designers of the will of the state, thus they have a clear

role to make policies, while bureaucrats are to execute that will by professionally implementing

policies using the bureaucratic machinery to achieve their objectives. Scholars have drawn

extensively on the debate to analyse political-administrative arrangements in diﬀerent countries.

Examples of studies include Riper (1984), Rabin and Bowman (1984), Montjoy and Watson

(1995), Dunn (1997) Svara (1998, 1999, 2001) and Rutgers (2001). While in practice most would

agree that administrators do play important roles within the policy process and that the dichotomy

does not hold true, this debate continues to serve as a theoretical starting point for reﬁning and

understanding the relationship between bureaucrats and politicians and the bureaucracy’s role in

the policy process.

In addition to the relationship between politics and administration, observation of the bureau-

cracy cannot be separated from the political system it operates in. Students of political science

and public administration would notably agree that political system is one of the most important

variables to explain decision-making processes, choices and outcomes of policy. Studies on the

bureaucracy have pointed to the inherent tension between bureaucracy and democracy (Mosher

1968; Beetham 1987; Smith 1988; Gormley and Balla 2004; Goodsell 2004). Albrow suggests

that this debate dates back to the nineteenth-century (Albrow 1970). Bureaucracy needs to have

a strong capacity to be eﬀective in administration but at the same time it needs to be controlled

to ensure responsiveness to the public and higher authorities (Garvey 1995). Etzioni-Halevy argues

that bureaucratic power generates a dilemma for democracy (1983). It poses a threat to democratic

political structure. He also argues that democracy generates a dilemma for bureaucracy. Bureau-

cracy is expected to be independent and subservient to elected politicians at the same time. This

rule is self-contradictory. Last, he argues that the dilemma of bureaucracy is an important source

of political friction especially between senior bureaucrats and senior politicians. Many have used

this debate as a point of departure and have either rebutted or supported these claims.

Some scholars see bureaucracy as not in conﬂict with democracy but as an important vehicle

to push for better democratic governments, in addition to political parties, elections, and the

legislatures (Thompson 1965). For example, Kim has empirically demonstrated that democratic

control can and does go hand-in-hand with bureaucratic autonomy (2008). He emphasized

how bureaucratic behaviour is inﬂuenced by multiple political institutions that have relation-

ships with the public agency. Since the mid-1990s, scholars have begun to discuss the rela-

tionship between public bureaucracy and the use of performance management systems in the

evolving democratic processes (e.g. Gormley and Balla 2004; Peters 2010). Performance systems

can be seen as instruments to control the bureaucracy by electorates and citizens in democratic

settings. Rather than focusing only on the bureaucracy, Meier points to the need to balance our

focus on the electorates (1997). He argues that most problems in the United States stem from

the inability of politicians to formulate good policies. These debates illustrate the ongoing need

for students of political science and public administration to understand the two groups in relation

to each other. In sum, bureaucracy continues to be an important component of democratic

systems, despite inherent tensions between the two concepts. Needless to say, the bureaucracy

has always been important for non-democratic systems especially in authoritarian and commu-

nist systems where the state has an overarching role in many aspects of citizens’ lives. In general,

bureaucracy in the twenty-ﬁrst century is inevitably becoming more democratic as citizens con-

tinue to demand higher levels of accountability and responsiveness, regardless of the type of

political system they are in.

As political systems and economic systems are intertwined, observation of the bureaucracy

cannot be separated from the economic system it operates in. There are numerous studies of the

Poocharoen

334

relationship between bureaucracy and the market mechanism. There are three ways of viewing

this relationship. One is to treat the bureaucracy as a system existing within a larger economic

system. Weber’s studies on the bureaucracy stemmed from his question of the consequences of

modern capitalism. He infers that the capitalist system provided the best chance for preservation

of individual freedom and creative leadership in a bureaucracy (Mommsen 1974: xv). Marxist

schools of thought equally debated and discussed the role of the bureaucracy, speciﬁcally on the

bureaucracy and class and implications for society as a whole (e.g. Chattopadhyay 1993; Burn-

ham 1941; Westoby 1985). In this camp, bureaucracy presented the new elites with managerial

and technical power that would dominate the working class (Westoby 1985). The second view

sees the co-existence of the bureaucratic system and market mechanisms. Recent debates have

extended the focus to observe the co-existence between market, bureaucratic and network

mechanisms in public service delivery (e.g. Olsen 2006; Koliba 2011). This view does not treat

the bureaucracy as working in a certain economic system but rather as a system operating side-

by-side with other systems. The third view is to treat bureaucracy and the market as diﬀerent

mechanisms for organizational control or management (e.g. Ouchi 1980). Similarly, accountability

systems, as a form of organizational control, can also be divided into market and bureaucratic or

administrative frames, aside from the democratic frame (Koliba 2011).

Now that the two views of the deﬁnition of bureaucracy have been elaborated, next we will

discuss the role of the bureaucracy in the policy process, which is an extension of this second

deﬁnition that bureaucracy is part of government. From here onward I will use the term

“bureaucracy” synonymously with the term ‘public bureaucracy’.

Role of the bureaucracy in the policy process

The bureaucracy plays a very important role in the entire policy process: setting agendas and

formulating, implementing and evaluating policies. This understanding of the dominant role of

the bureaucracy in the policy process is not new and has been recognized for many years by

prominent scholars. Examples include Our Wonderland of Bureaucracy (Beck 1932) and The

Managerial Revolution (Burnham 1941) as well as more recent works such as Breaking through the

Bureaucracy (Barzelay 1992). In his work to introduce the concept of ‘bureaucratic government’,

Peters writes, “It is by now almost trite to say that bureaucracy and administration are an increas-

ingly signiﬁcant – if not the most signiﬁcant – feature of modern policy–making” (1981: 56).

Peters identiﬁes subsets of literature pertaining to bureaucracy and public administration. They

are the positivist theorists that borrow economic rationality through the public choice approach

to explain bureaucrats to be power-maximizing actors (Downs 1964, 1967). The other group of

scholars, called descriptive theorists, explain bureaucrats’ behaviour as political actors and not the

non-political technicians stressed in the Weber’s ideal-type bureaucracy. Bureaucracies are given

wide latitude to include details of actual implementation of legislation and to spell out all the

rules and guidelines that need to accompany the policy or law.

Peters reiterated in his often-cited book The Politics of Bureaucracy that administrative systems

do inﬂuence policy outputs of the political system (1995). Bureaucracies are always either

explicitly or implicitly included in theoretical discussions of the public policy process. Charles

Lindblom’s well-known incrementalism theory explains that due to standard operating proce-

dures ingrained in the bureaucracy, there is always a tendency for governments to preserve the

status quo or to change very minimally from the status quo. The bureaucracy is understood to

be one of the main actors in other models of public policy whether it is the elite model, the

group model or the institutionalist model. The traditional rational model of policy-making, by

emphasizing objective and technical knowledge in decision-making is inevitably supporting the

Bureaucracy and the policy process

335

role of the bureaucracy (see Parsons 1995). Thus it is vital to understand the bureaucratic admin-

istration system alongside political, economic and social systems in order to understand public

policy. Most public policy textbooks would not be complete without mentioning the role of

the bureaucracy and bureaucrats in the policy process. See, for example, Wu et al. (2010),

Howlett and Ramesh (2003), Parsons (1995), Considine (1994) and Gupta (2001).

There are several reasons why bureaucrats can and do inﬂuence policy processes at various

stages. First, the policy process can only be divided in theory but not in practice. In practice, the

process is often a mix of the policy stages because policies are constantly being changed or

adjusted based on implementation and evaluation experiences. Second, because the bureaucracy

often has expert and long-term institutional knowledge of public issues, it is often involved at

the very early stage of agenda-setting. Sometimes it is the bureaucracy that initiates policies. A

clear example is the bureaucracy’s involvement in the national budget decision-making process.

Line ministries are asked to submit a draft of their budget proposal each year. This often occurs

without the direct involvement of ministers or politicians. Wildavsky has written extensively on

this topic (e.g. Wildavsky 1979, 1986; Wildavsky et al. 2006) and this topic was extended to

include assertions that bureaucrats tend to maximize their interests through the budget process

(Blais and Dion 1991; Peters 1995).

In addition, bureaucrats usually stay in their posts longer than politicians. They have the

opportunity to see through policies. They are sometimes known for taking a longer-term view

of public problems. Furthermore, bureaucrats are recruited on the basis of their expertise and

profession, while politicians are elected on popularity and big policy promises. The bureaucracy

is often perceived to speak more creditably about the technical aspects of policies. This is a form

of knowledge-based power that bureaucrats exercise. Finally, bureaucrats usually have access to

more data and information regarding speciﬁc policies. By being the implementers of policies,

bureaucrats have many opportunities to collect data and information that would be valuable for

policy-making and evaluation. The level of inﬂuence that bureaucrats do have in the policy pro-

cess is also determined by other factors, including historical legacies, governance structures or

political systems, the Constitution or the most important laws. In one country, the level of

inﬂuence can also change over time depending on the arrangement and preferences at diﬀerent

points in time. In sum, it has long been recognized that the bureaucracy plays an important role

in all stages of the policy process. Next, we will focus on the policy-making process.

Bureaucrats as policy-makers

Bureaucrats can and do greatly inﬂuence policy-making. They can be the source of ideas. They

can be the advocate of ideas. They can take part formally and informally in the policy-making

process. They can also be the sole actor or leader in the policy-making process.

Bureaucracy can directly and indirectly shape the discourse of public policies. Indirectly,

during the policy formulation stage, bureaucrats can inﬂuence the discourse of broad policies

through their communications with politicians and the wider public. After broad goals have

been set, the bureaucracy plays an important role in outlining policy details through the design

of speciﬁc programmes. It is precisely this that will largely inﬂuence the overall policy outcome

and citizens’ perceptions of the policy and the government. Peters suggests that there are soft

and hard versions of agency ideologies (1995). Soft ideology refers to when ministers are being

inﬂuenced by bureaucrats’ preferences in order to maintain certain policies, that is to say they

are the source of ideas and are indirectly inﬂuencing policy.

Hard ideology refers to when bureaucrats actively seek to alternate or initiate policies. This

latter version sees bureaucracy as an active participant in the political arena of the policy process.

Poocharoen

336

It accepts that bureaucrats can act similarly to pressure groups, namely they are the advocates of

ideas and are directly inﬂuencing policy. This could take the form of lobbying, trying to

inﬂuence discourse through media and possibly also through participating in protests. This is

because bureaucrats have their own sets of ideologies and preferences. This view diﬀers from

seeing the bureaucracy as being politically neutral and acting as a buﬀer between the state and

interest groups. A similar idea is found in the concept of representative bureaucracy, where bureau-

cracy is accepted as a political actor in the policy sphere and is expected to help shape public

policies (Krislov 1974; Seldon 1997; Dolan and Rosenbloom 2003). To put it positively, bureau-

crats can also be seen as policy entrepreneurs. In systems where the bureaucrat is stationed in

one agency for their entire career, as is the case in the United States and Scandinavian countries,

for example, they tend to be policy advocates because they hold such deep expertise and opinions

on their policy area (Peters 1995). On the other hand, if bureaucrats are centrally recruited, and

are rotated around various agencies, such as in France and the United Kingdom, they might be

less attached to certain policy opinions (ibid.). This is in stark contrast to before the 1950s when

bureaucrats were understood to discreetely shape policy behind the scenes (Svara 2001).

Furthermore, the bureaucracy sometimes even takes the leadership role with very little

inﬂuence from other actors, including politicians (Peters 1995). This resonates with Mosher’s

idea of the ‘professional state’ where professionals in public agencies through their expertise and

knowledge do not only generate new policy ideas but are the main actors to formulate the ﬁne

details of policies as well (Mosher 1968). In relation to seeing the bureaucracy as taking the leading

role in speciﬁc policy areas, many scholars have pointed to the bureaucracy’s critical role in the

entire administration of developing countries. For example, scholars have associated the suc-

cessful development of some Asian countries and regions with the idea of a strong state, namely

strong bureaucracy. They include the so-called four tigers of Taiwan, South Korea, Singapore, and

Hong Kong (for more on East Asia see, for example, Wade 1990; World Bank 1993; Henderson

2011). Japan is also known to have a strong bureaucracy with greater inﬂuence than politicians and

which has dominated policy-making since World War II (Ramseyer and Rosenbluth 1993).

Bureaucracies that are core agencies, regulatory agencies and planning agencies are often in a

position to inﬂuence policy-making and setting the policy direction for the country. The term

‘super bureaucrat’ depicts the prominent role of high-ranking bureaucrats in the policy process

(Campbell and Szablowski 1979). These bureaucrats are usually in the central ministries or the

peak policy agencies such as Thailand’s National Economic and Social Development Board,

Singapore’s Economic Development Board, Taiwan’s Council for Economic Planning and

Development, Japan’s Ministry of Economy, Trade and Industry, and Malaysia’s Prime Minis-

ter’sOﬃce. Similarly in the area of public administrative reform policy, only a handful of key

central agencies would be involved in the design of the policy. These agencies have been called

‘system designers’ (Olsen and Peters 1996). On the other hand, in some cases the line ministries

might dominate policy-making. Mostly these are areas where great technical expertise is required.

They are, for example, policies related to energy, health, transportation, food regulation, just to

name a few. Examples of studies include the Medicare movement (Marmor 1973) and com-

munity mental health programmes (Foley 1975) in the United States and foreign policy-making

(Ahn 1998) and telecommunication policy (Eiji 2001) in Japan.

In a bureaucracy, it is not only the bureaucrats who can inﬂuence policy-making. In pre-

sidential systems where there are political appointees in the bureaucracy, such as the United

States and the Philippines, studies conﬁrm that career and non-career administrators contribute

signiﬁcantly to policy-making (Aberbach et al. 1981; Rourke 1984; Dolan 2000). In parlia-

mentarian systems, such as the United Kingdom, it is also known that oﬃcials whether selected

by appointment with lifetime tenure or politically selected for the duration of a government,

Bureaucracy and the policy process

337

can also have signiﬁcant inﬂuence on policy-making as long as they are part of the bureaucracy

(e.g. Bevir and Rhodes 2006).

In the debate on the appropriate role of the bureaucracy in the policy process, some see that

the bureaucracy can stabilize governance where the political party system is weak. In many

countries, when the political executive is weak, the bureaucracy often gains more importance

and strengthen its hold on policy decisions (Peters 1981: 68). For some developing countries in

the early to middle stages of development, the bureaucracy often have control over policy

directions of the country. For example, during the years 1932 to 1973 the bureaucrats in

Thailand were considered to be elites who had overthrown the absolute monarchy. They led

the government in an authoritarian fashion, setting the tone of a semi-democracy that still

persists many years after (Bowornwathana 2011). In turn, bureaucracy can also dominate policy

processes to the detriment of political institutions and actors. Heady calls this ‘imbalance thesis’

(2001). Similarly, Riggs has argued that bureaucracy is mostly made up of the ruling class that

impedes the development of strong political systems (2001).

Naturally, this bureaucratic power in policy-making has never been a comfortable fact for

politicians. Studies have shown how political leaders have intentionally excluded high-ranking

bureaucrats from policy-making processes (Cole and Caputo 1979; Bevir and Rhodes 2006).

Parliamentarians keep a check on the public bureaucracy by scrutinizing their policies and imple-

mentation in parliament debates (e.g. Ho 2000 on Singapore). The new public management

(NPM) movement that originated in the Anglo-Saxon countries in the 1980s is an attempt to

strengthen the checks on bureaucracies and bureaucratic power. On the contrary, when the rela-

tionship between bureaucrats and politicians becomes too cozy, it is labelled ‘iron triangle’ and

that is when there is a monopoly of policy-making power among the politicians, the bureau-

cracy and industry or the interest groups. Some of these sectors include defence in the United

States (Adams 1981) and agriculture in the European Union (Hix 1999). These relationships can

easily lead to corruption and a conﬂict of interest for the government.

Another aspect of the bureaucracy that inﬂuences policy-making is structure. Bureaucratic

structure can be both an independent and dependent variable in relation to policy. Policy can be a

determinant to explain bureaucratic structures. Frequently we witness governments setting up

new bureaucracies in order to take charge of new policies. These are usually new policies that

existing agencies do not cover. For instance, they could be policies related to climate change,

food security, crisis management, terrorism, non-traditional security and population or migration.

Bureaucratic structure is a reﬂection of politics (West 1997). It shows how scarce resources are

allocated. The structures are nothing but evidence of institutional choice and power in the public

sector. However, bureaucratic structure can be a determinant of policy direction and policy

choice. Examples include studies on bureaucratic structure and personality by Robert K. Merton

(1957), Allison (1971) and Egeberg (1999). Bureaucratic structure determines the ﬂow of

information processing, which would inﬂuence the type and value of information that is ﬁltered

up and down the bureaucracy. Bureaucratic structure also dictates the dynamics of politics among

bureaucracies. And scholars have long argued that policy decisions are dominated by the dynamics

of bureaucratic politics (Allison and Zelikow 1999). Following Allison’s (1971) seminal work on

foreign policy other works have expanded our understanding of how bureaucratic politics govern

policy processes (e.g. Peters 1995; Bowornwathana and Poocharoen 2010). There are many exam-

ples of when decisions on administrative reform policies are shaped by the struggle between key

central agencies to maintain or gain more power during the reform and after the reform.

Bureaucracies can also indirectly inﬂuence domestic policy-making from the international

arena. It is suggested that bureaucrats often take part in professional policy networks, where

they pick up ideas and good practices and then try to implement those policies and practices to

Poocharoen

338

their own agency, community or government (Peters 1995). International organizations often

provide platforms for such exchanges, such as the model of community of practice in the World

Bank or Asian Development Bank. Many examples can be found in the EU especially in the

standardization of various social services (ibid.). There are also studies that point to how domestic

bureaucracies can inﬂuence policies made at the international level (Hopkins 1976). Lastly, aside

from focusing on domestic bureaucracies, more and more attention is being focused on the role of

international bureaucracies. In a highly globalized world, we cannot ignore the signiﬁcance of

international bureaucracies and their impact on global governance and domestic policy choices

(e.g. Mouritzen 1990; Barnett and Finnemore 1999; Muldoon et al. 2010). In sum, international

bureaucracies are also important actors in the policy-making process both domestically and

internationally. Next we will discuss bureaucracy in the policy implementation process.

Bureaucrats as policy implementers

The main function of the bureaucracy is to implement public policies, programmes and pro-

jects. In the process of implementation, the bureaucracy inﬂuences the direction of policy in

several ways. Policies are often very broad and vague due to the complexity of most public

policy problems. For example it is diﬃcult for governments to have certain explicit policies on

climate change issues, gender equality or poverty reduction because often the responsibility to

solve these problems is diﬀused among many bureaucracies. There is no one bureaucracy to

take ownership of these complex problems. Thus, during the implementation stage, depending

on the level of discretionary power, diﬀerent bureaucracies have room to interpret policies to ﬁt

their own views and maintain practicality. The bureaucracy also adjusts rules and instructions to

suit local contexts. This process of interpretation is where intentionally or unintentionally bureau-

crats can and do inﬂuence policies. Sometimes, the original intent of the policy, as set by policy-

makers, is altered due to the bureaucrats’ interpretation of the policy (Pressman and Wildavsky

1973). Also, during the implementation stage, as policies are being further deﬁned and designed,

bureaucracy can come up with bureaucratic standard operating procedures or process-oriented

policies that can also aﬀect the policy outcome. At this stage, politicians and other stakeholders

might also intervene and inﬂuence how implementation processes are set (Furlong 1998).

Studies about street-level bureaucrats give a clear illustration of how bureaucracies inﬂuence

policy outcomes at the implementation stage (Lipsky 1980). Street-level bureaucrats are those

who have face-to-face communication with citizens. Examples include police, teachers, immi-

gration oﬃcers, social workers and those who work in service delivery agencies such as birth

and marriage registration or passport and driver’s license issue. Street-level bureaucrats have to

constantly interpret and reinterpret policy goals. At the same time, they often have individual

agendas, expectations, values and capacity. Studies have found that street-level bureaucrats’ lack

of conformity to procedures and decision-trees set by top administration is a stable norm of

policy implementation (Mashaw 1983; Mashaw and Marmor 1988). They often feel frustration

due to the need to face ongoing conﬂict between responding to clients or citizens and the need

to properly implement policies. Put positively, street-level bureaucrats’ discretion can help policies

to be more adaptive and enhance government’ s ability to respond to individual cases. On the other

hand, street-level bureaucrats can also change the course of the policy entirely.

Aside from street-level bureaucrats, there are also studies that explain how mid-level bureaucrats

also inﬂuence policies. In their study of mid-level civil servants in the United Kingdom Page

and Jenkins identify three types of policy roles: (1) a production role in making policy drafts and

documents; (2) a maintenance role in tending and managing policies; and (3) a service role in

oﬀering knowledge and skills to those involved in the policy process (2006).

Bureaucracy and the policy process

339

Aside from large- and medium-sized states, some scholars have pointed to the uniqueness of

bureaucracy of small island states. Similarly to many fragile states, these bureaucrats are often

implementing not only the national policy but also policies directed by international donors

(Baker 1992; Garcia-Zamor 2004). Often bureaucrats struggle to achieve policy goals set by

international donors and international organizations because they are confronted with many

challenges on the ground during the implementation stage.

During the policy monitoring and evaluation stage, bureaucrats often play the most impor-

tant role in providing information to external evaluators, if the evaluation is not done in-house.

As mentioned in the previous section, the bureaucracy has access to sensitive data and institu-

tional information that outsiders do not have. Thus they have some control over the evaluation

studies or reports of most policies and programmes. It is often diﬃcult to determine the validity

of evaluation reports when done in-house by the bureaucracy. Naturally they would be inclined

to not publicize policy failures. Overall, there is limited scholarly work on the issues pertaining

to bureaucracy and policy evaluation. In general, evaluation frameworks tend to focus less on

who is doing it but rather on how to do it. This topic should be further expanded in the future.

In sum, we can see that the bureaucracy is very inﬂuential in the policy process. Bureaucrats

can be seen as policy owners, policy advocates, policy managers, policy interpreters and policy

technicians. Next we shall discuss the evolving role of bureaucracy in the era of governance.

Bureaucracy and governance

Between the 1980s to 2000s, we witnessed the spread of the NPM movement across many

countries. As bureaucracies grew in size and power over the years between the 1940s and the

1980s, the rhetoric of the need to curb bureaucratic power gained momentum. Many govern-

ments went through phases of downsizing the public sector, tightening control over bureau-

cracies, and introducing market-based competition for traditional public services. The discourse

emphasized how bureaucracies were ineﬃcient and too big. Proponents of such ideas came

from public choice theorists (Niskanen 1974), New Institutional Economic theorists in New

Zealand (Boston 1991), and New Right ideologies in the 1980s (Saint-Martin 2004; Pollitt and

Bouckaert 2004). The book, Reinventing Government by Osborne and Gaebler (Osborne and Gaebler

1993), which made an impact on the United States reform movement, was widely read by

practitioners of public administration and public policy in other countries.

The NPM movement highlighted that bureaucrats should only implement policies and be

accountable to politicians and citizens for policy outcomes. Bureaucrats should not inﬂuence policy-

making. The movement had a goal to change bureaucratic structure and culture to be more

eﬃcient and responsive to public interests (Barzelay 1992). Proponents of NPM used economic

theory, incentive theories and market-driven decisions as a basis for solutions. A prime example

of a tool that emerged was the signing of contracts between ministers and permanent secretaries to

promise certain deliverables, such as in Australia and the United Kingdom. Another example

was the separation of policy-making agencies from service delivery and production bureau-

cracies. Owing to NPM, some have pointed to the decline of bureaucratic power in the policy-

making process (Wilson 2008). However, there are also a number of studies that explained how

NPM was business as usual and that the bureaucracy’s role had not diminished in any way (e.g.

Cheung and Scott 2003).

Since the early 2000s, especially since 9/11 in the United States, scholars have referred less to

the NPM paradigm and now refer more to the governance paradigm. The rapid process of

globalization, the emergence and spread of the Internet and other communication media have

changed the way that we understand and approach problems of governance. The meaning of

Poocharoen

340

governance can be broadly deﬁned as ‘structures and processes by which people in societies make

decisions, set rules, and share power’ (Folke et al. 2005: 444). Rhodes oﬀers six approaches to

the deﬁnition of the term: governance as corporate governance; governance as New Public

Management; governance as ‘good governance’; governance as international interdependence;

governance as a socio-cybernetic system; and governance as networks (Rhodes 2000).

Despite the changes and reforms during the NPM era, in the era of governance, the bureaucracy

continues to be an integral part of the policy process. Some argue that the bureaucracy is

stronger than before and that there is a need to continue to build up bureaucratic capacity. Thus

it is necessary to re-emphasize the bureaucracy in the study of policy processes (e.g. Howlett

and Ramesh 2003; Krause and Meier 2005). In the governance paradigm, the role of the

bureaucracy and the bureaucrats is understood to be diﬀerent from before. Governance stresses

the value of bureaucrats among other actors in creating public value and the bureaucracy as an

important part of issue networks and service delivery networks. While most literature on public

administration regards the evolution of the bureaucracy as a linear process, some have argued

the contrary. Lynn argues that the traditional bureaucratic paradigm showed far more respect for

law, politics, citizens and values than the NPM paradigm (Lynn 2001). Thus there is more resem-

blance between the traditional bureaucratic paradigm and the governance paradigm. The gov-

ernance paradigm values the role of the bureaucracy, unlike the NPM paradigm, yet at the same

time it is pushing the bureaucracy to improve and evolve further.

Within the governance paradigm the notion of networks and collaborations is prominent. There

are now more studies on the bureaucracy and its relations and power in various types of networks.

One example is the notion of policy networks. Policy networks refer to a group of various

people or organizations who are connected or are held together by common interests in certain

policy problems. Policy networks try to inﬂuence policy decisions (Marsh and Rhodes 1992; Marsh

1998). A bureaucracy is one actor in the complex web of actors in the society that make up policy

networks. Eﬀective bureaucracies would be able to manage and manoeuvre policy networks in

the direction that is best for the general public. This is deﬁned as network governance which is the

act of designing, managing, coordinating strategies, structures and processes of inter-organizational

relations in order to aﬀect public policies (Baker 1992; Provan and Kenis 2007). Bureaucrats need

to have a diﬀerent set of skills to maneuver in this kind of context, whether it is to eﬀectively

take part in a policy network or to exercise network governance (Edelenbos and Klijin 2005).

Public managers are co-creators of public value, which in turn sets the direction for bureaucracies

and public policies. Moore’s concept of public value has gained momentum in the public sector and

has implications on how we understand the role of the bureaucracy in the policy process (Moore

1995). As bureaucracies are working in a networked environment, they are consistently shaping

public value through networks. Concepts that have gained attention include co-production and

co-creation where the bureaucracy partners work with the private sector and non-proﬁt orga-

nizations to provide public goods and services, and to set regulations for governance. These

ideas fall under the rubrics of collaborative governance (e.g. Ansell and Gash 2008; O’Leary and

Bingham 2009) and also sound governance (Farazmand 2004). All these frameworks have one

common focus on partnership as a form of participatory governance. This shifting role of the

bureaucracy will have implications on how it can inﬂuence the policy process. In one view, this

could lead to a more restricted role for bureaucracies. This is due to the increased role and

power of non-state actors in networks. On the other hand, bureaucracy might gain more power

and control because of their continual advantage over public resources and the opportunity to

shape public values through public policies.

In sum, the bureaucracy continues to be the centre of study no matter how paradigms have shifted

in public administration, whether it is the traditional public administration, new public management,

Bureaucracy and the policy process

341

or governance paradigms (Koliba 2004; Jreisat 2012). Its role in the policy process will continue to

change depending on the political, economic and social changes that each country goes through.

For now, it continues to be one of the most important actors in all stages of the policy process.

Future studies

To successfully understand this topic in greater depth, there need to be more studies of the

diﬀerent types of bureaucracy and of bureaucrats, in other words the ‘heterogeneity of bureaucracy

and bureaucrats’. Bureaucracies are currently not homogeneous. Many have little resemblance

to Weber’s original ideal type. We should compare diﬀerent types of bureaucratic behaviour,

structure and culture and their inﬂuences on policy processes. For example, they could be studies

of comparisons between central and local bureaucracies, line -ministries versus central ministries,

higher, middle and lower ranks of bureaucrats, developed and developing countries, and

bureaucracies of diﬀerent sectors such as health and education. The literature on decentraliza-

tion has begun to focus on bureaucratic capacity but not so much on how local bureaucracy

shapes policy at the local and national levels. Lastly the literature on multilevel governance has

started to focus on the relationship between national and local bureaucracy with supra-national

bureaucracy such as that of the EU. But there is much more room for research.

Another phenomenon worth noting is the expansion of semi-bureaucrats, in other words,

those who are employed by the public sector but who are treated diﬀerently from a career civil

servant or bureaucrat. Semi-bureaucrats are usually employed based on market-mechanisms.

Examples of this group include contract employees, state enterprise employees, experts, profes-

sionals, advisors, consultants and ad hoc public managers. It is also important to understand how

their role diﬀers or is the same compared to the typical bureaucrat.

In the past, scholars could study one agency within a ministry and could claim that the study

represented that entire bureaucracy. Today, uniqueness is the new norm. Each agency can be

quite unique due to information technology, leadership style, organizational culture, and hier-

archical arrangements of the organizations. Aside from the above mentioned semi-bureaucrats,

many public agencies are frequently using more matrix approaches to management rather than

relying solely on hierarchical control. These semi-bureaucracies or hybrid organizations have

mixed management instruments between bureaucracy, the market and networks. They have a

variety of structures for decision-making such as the use of task forces, inter-departmental com-

mittees, and ad hoc committees. Thus, in addition to understanding the semi-bureaucrats’ roles,

we should also understand how these new structures of the bureaucracy eﬀect policy processes.

In sum, the Weberian model continues to be a reference point for any discussion on the

bureaucracy and its role in the policy process. However, we have moved quite far from that

starting point. Bureaucracies have achieved great success in some countries as well as failing

miserably in others. Despite being understood as being rigid and hierarchical, they continue to

evolve and adapt to changes in political, economic and social systems. In this era of governance

we recognize the bureaucracy as public organizations and bureaucrats as public managers, who

have to manoeuvre in the complex web of multiple actors in all stages of the policy process.

bureaucracy has played an important, if not the most important role, in policy-making and policy

implementation throughout the history of its existence and it will continue to do so.

Bibliography

Aberbach, Joel D., Robert D. Putnam and Bert A. Rockman. 1981. Bureaucrats and Politicians in Western

Democracies. Cambridge, MA: Harvard University Press.

Poocharoen

342

Adams, Gordon. 1981. The Iron Triangle: The Politics of Defence Contracting. New York: Council on Economic

Priorities.

Ahn, C.S. 1998. ‘Inter-ministry coordination in Japan’s foreign policy making’, PaciﬁcAﬀairs 72(1): 42–60.

Albrow, Martin. 1970. Bureaucracy. London: Pall Mall.

Allison, G. 1971. Essence of Decision: Explaining the Cuban Missile Crisis. Boston: Little Brown.

——and P. Zelikow. 1999. Essence of Decision: Explaining the Cuban Missile Crisis. New York: Longman.

Ansell, C. and A. Gash. 2008. ‘Collaborative governance in theory and practice’, Journal of Public Administration

in Research and Theory 18(4): 543–71.

Appleby, Paul. Policy and Administration. 1949. Tuscaloosa, AL: University of Alabama Press.

Baker, R. 1992. Public Administration in Small and Island States. West Hartford, CT: Kumarian Press.

Barnett, Michael N. and Martha Finnemore. 1999. ‘The politics, power, and pathologies of international

organizations’, International Organization 53: 699– 732.

Barzelay, Michael. 1992. Breaking Through Bureaucracy. Berkeley, CA: University of California Press.

Beck, James M. 1932. Our Wonderland of Bureaucracy: A Study of the Growth of Bureaucracy in the Federal

Government, and Its Destructive Eﬀect upon the Constitution. New York: Macmillan.

Beetham, David. Bureaucracy. 1987. Milton Keynes: Open University Press.

Bevir, Mark and R.A.W. Rhodes. 2006. Governance Stories. Routledge: London.

Blais, Andre and Stephane Dion (eds). 1991. The Budget-Maximizing Bureaucrat: Appraisal and Evidence .

University of Pittsburgh Press, Pittsburgh, PA.

Blau, Peter M and W. Richard Scott. 1962. Formal Organizations: A Comparative Approach. San Francisco:

Chandler.

Boston, J. 1991. ‘The theoretical underpinnings of public sector restructuring in New Zealand’, in J. Boston,

J. Martin, J. Pallot and P. Walsh (eds) Reshaping the State: New Zealand’s Bureaucratic Revolution. Auckland:

Oxford University Press,

Bowornwathana, Bidhya. 2011. ‘History and political context of public administration in Thailand ’,in

Evan Berman (ed.) Public Administration in Southeast Asia: Thailand, Philippines, Malaysia, Hong Kong and

Macau. New York: CRC Press, Taylor & Francis, 29–52.

——and Ora-orn Poocharoen. 2010. ‘Bureaucratic

politics and administrative

reform: Why politics matters’,

Public Organization Review 10: 303–21.

Burnham, James. 1941.The Managerial Revolution. New York: John Day.

Campbell, C. and G. Szablowski. 1979. The Superbureaucrats: Structure and Behaviour in Central Agencies.

Toronto: Macmillan.

Chattopadhyay, Paresh. 1993. ‘The (ex) Soviet economy: Towards a reassessment of contending theses’,

Economic and Political Weekly 28(5): 13–22.

Cheung, Anthony B.L. and Ian Scott. 2003. Governance and Public Sector Reform in Asia: Paradigm Shifts or

Business as Usual? London and New York: RoutledgeCurzon.

Christensen, T. 2007. Organization Theory and the Public Sector: Instrument, Culture and Myth. London:

Routledge.

Cole, Richard and David Caputo. 1979. ‘Presidential control of the senior civil service’, American Political

Science Review, June: 399–413.

Considine, Mark. 1994. Public Policy: A Critical Approach. South Melbourne: Macmillan Education Australia

Pty Ltd.

Dolan, Julie. 2000. ‘The senior executive service: Gender, attitudes, and representative bureaucracy’, Journal of

Public Administration Research and Theory 10: 513– 30.

——and David H. Rosenbloom. 2003. Representative Bureaucracy: Classic Readings and Continuing Controversies.

New York: M.E. Sharpe.

Downs, Anthony. 1964. Inside Bureaucracy. Santa Monica, CA: RAND Corporation.

—— 1967. Inside Bureaucracy. Boston: Little, Brown.

Dunn, Delmer D. 1997. Politics and Administration at the Top: Lessons from Down Under. Pittsburgh, PA:

University of Pittsburgh Press.

Edelenbos, J. and E.H. Klijin. 2005. ‘Managing stakeholder involvement in decision-making: A comparative

analysis of six interactive processes in the Netherlands’, Journal of Public Administration Research and Theory

16: 417–46.

Egeberg, Morten. 1999. ‘The impact of bureaucratic structure on policy-making’, Public Administration 77:

155–70.

Eiji, Kawabata. 2001. ‘Sanction power, jurisdiction, and economic policy-making: Explaining contemporary

telecommunications policy in Japan’, Governance 14(4): 399– 427.

Bureaucracy and the policy process

343

Etzioni-Halevy, E. 1983. Bureaucracy and Democracy: A Political Dilemma. London: Routledge and Kegan Paul.

Farazmand, Ali. 2004. Sound Governance: Policy and Administrative Innovations. New York: Greenwood

Publishing Group.

Foley, Henry A. 1975. Community Mental Heath Programs: The Formative Process. Lexington, MA: D.C.

Health.

Folke, C., T. Hahn, P. Olsson and J. Norberg. 2005. ‘Adaptive governance of social-ecological systems’,

Annual Review of Environment and Resources 30: 441–73.

Furlong, F.R. 1998. ‘Political inﬂuence on the bureaucracy: The bureaucracy speaks’, Journal of Public

Administration Research and Theory 8: 39–65.

Garcia-Zamor, Jean-Claude. 2004. ‘The struggle of small bureaucracies to develop traditional ethical

policies’, in Ali Farazmand (ed.) Sound Governance: Policy and Administratiive Innovations. Westport, CT:

Praeger, 290–308.

Garvey, Gerald. 1995. ‘False promises: The NPR in historical perspective’, in Donald F. Kettl and John

J. DiIulio, Jr., Inside the Reinvention Machine: Appraising Governmental Reform. Washington, DC: The

Brookings Institution, 87–106.

Goodnow, Frank J. 1900. Politics and Administration: A Study in Government. New York: Macmillan.

Goodsell, Charles T. 2004. The Case for Bureaucracy: A Public Administration Polemic. Washington, DC: CQ

Press.

Gormley, William T. Jr. and Steven J. Balla. 2004. Bureaucracy and Democracy: Accountability and Performance.

Washington, DC: CQ Press.

Gupta, Dipak K. 2001. Analyzing Public Policy: Concepts, Tools, and Techniques. Washington, DC: Congressional

Quarterly Inc.

Hall, Richard H. 1996. Organization: Structure, Processes and Outcomes. Englewood Cliﬀs, NJ: Prentice Hall.

——and Robert E. Quinn. 1983. Organization Theory and Public Policy. Thousand Oaks, CA: Sage.

Heady, Ferrel. 2001. Public Administration: A Comparative Perspective. New York: Marcel Dekker.

Henderson, Jeﬀrey. 2011. East Asian Transformation: On the Political Economy of Dynamism, Governance and

Crisis. London: Routledge.

Henry, J. 2006. ‘ Educating managers for post-bureaucracy: The role of the humanities’, Management

Learning 37: 267–81.

Hix, Simon. 1999. The Political System of the European Union . London: Macmillan.

Ho, Khai Leong. 2000. The Politics of Policy Making in Singapore.

Oxford: Oxford University

Press.

Hopkins, R.F. 1976. ‘The “international role of “domestic” bureaucracy’, International Organization 30(3):

405–32.

Howlett, Michael and M. Ramesh. 2003. Studying Public Policy: Policy Cycles and Policy Subsystems. Toronto:

Oxford University Press.

Jacques, E. 1976. A General Theory of Bureaucracy. London: Heinemann.

Jreisat, Jamil E. 2012. Globalism and Comparative Public Administration. Boca Raton, FL: CRC Press.

Kamenka, Eugene. 1989. Bureaucracy. Oxford: Blackwell.

Kaufman, H. 1981. ‘Fear of bureaucracy: A raging pandemic’, Public Administration Review 41: 1–9.

Kim, Doo-Rae. 2008. ‘Political control and bureaucratic autonomy revisited: A multi-institutional analysis

of OSHA enforcement’, Journal of Public Administration Research and Theory 18(1): 33–55.

Kjaer, Anne Mette. 2004. Governance. Oxford: Polity Press.

Koliba, C. 2004. ‘Is service-learning contributing to the downsizing of American democracy? Learning our

way out’, The Michigan Journal of Community Service Learning Spring: 57–68.

—— 2011. ‘Administrative strategies for a networked world: The educational imperative for inter-

governmental relations in 2020’, in K. Thurmaier and J. Meek (eds) Network Governance: Implications for

Intergovernmental Relations in 2020. New York: Sage.

——, Jack W. Meek and Asim Zia. 2010. Governance Networks in Public Administration and Public Policy.

New York: CRC Press.

Krause, George A. and Kenneth J. Meier. 2005. Politics, Policy, and Organizations: Frontiers in the Scientiﬁc

Study of Bureaucracy. Ann Arbor, MI: University of Michigan Press.

Krislov, Samuel. 1974. Representative Bureaucracy. Englewood Cliﬀs, NJ: Prentice Hall.

Lipsky, Michael. 1980. Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. New York;

Russell Sage.

Lynn, Laurence E. Jr. 2001. ‘The myth of the bureaucratic paradigm: What traditional public administration

really stood for’, Public Administration Review 61(2): 144–60.

March,

James G and

A. Simon Herbert. 1958. Organisations. New York: Wiley.

Poocharoen

344

Marmor, Theodore R. 1973. The Politics of Medicare. Chicago: Aldine and Atherton.

Marsh, David. 1998. Comparing Policy Networks. Milton Keynes: Open University Press.

——and R.A.W. Rhodes. 1992. Policy Networks in British Government. Oxford: Clarendon Press.

Mashaw, Jerry. 1983. Bureaucratic Justice: Managing Social Security Disability Claims. New Haven, CT: Yale

University Press.

——and T.R. Marmor. 1988. Social Security: Beyond the Rhetoric of Crisis. Princeton, NJ: Princeton University

Press.

Meier, K.J. 1997. ‘Bureaucracy and democracy: The case for less democracy and more bureaucracy’, Public

Administration Review 67(3): 193–99.

Merton, Robert K. 1957. Bureaucratic Structure and Personality. Glencoe, IL: Free Press.

Mommsen, Wolfgang J. 1974. The Age of Bureaucracy: Perspectives on the Political Sociology of Max Weber.

Oxford: Blackwell.

Montjoy, Roberts S. and Douglas J. Watson. 1995. ‘A case for reinterpreted dichotomy of politics and

administration as a professional standard in council-manager government’, Public Administration Review

55(3): 231–9.

Moore, Mark Harrison. 1995. Creating Public Value: Strategic Management in Government. Cambridge, MA:

Harvard University Press.

Mosher, Frederick C. 1968. Democracy and the Public Service. Oxford University Press: New York.

Mouritzen, Hans. 1990. The International Civil Service: A Study of Bureaucracy: International Organisations.

Aldershot: Dartmouth,

Muldoon, James P., Joann F. Aviel and Richard Reita. 2010. The New Dynamics of Multilateralism: Diplomacy,

International Organizations, and Global Governance. Boulder, CO: Westview Press.

Niskanen, W.A. 1973. Bureaucracy: Servant or Master? Lessons from America. Washington, DC: The Institute

of Economic Aﬀairs.

—— 1974. Bureaucracy. Chicago: Aldine Press.

O’Leary, Rosemary and Lisa Blomgren Bingham. 2009. The Collaborative Public Manager: New Ideas for the

Twenty First Century. Washington DC: Georgetown University Press.

Olsen, Johan P. 2006. ‘Maybe it is time to rediscover bureaucracy’, Journal of Public Administration Research

and Theory 16(1): 1–24.

——and G. Peters. 1996. Lessons from Experience. Oslo: Scandinavian University Press,

Osborne, David and Ted Gaebler. 1993. Reinventing Government: How the Entrepreneurial Spirit is Transforming

the Public Sector. New York: PLUME.

Ouchi, William G. 1980. ‘Markets, bureaucracies, and clans ’, Administrative Science Quarterly 25: 129 – 41.

Page,

C. and B.

Jenkins. 2006. Policy Bureaucracy: Government with a Cast of Thousands. Oxford: Oxford

University Press.

Parsons, D.W. 1995. Public Policy: An Introduction to the Theory and Practice of Policy Analysis. Cheltenham:

Edward Elgar.

Parsons, Wayne. 1995. Public Policy: An Introduction to the Theory and Practice of Policy Analysis. Northampton,

NH: Edward Elgar.

Peters, B.G. 1981. ‘The problem of bureaucratic government ’, The Journal of Politics 43(1): 56.

—— 1995. The Politics of Bureaucracy, 4th edn. New York: Longman Publishers USA,

—— 2010. ‘Bureaucracy and democracy in the modern state’, Public Administration Review 70(4): 642–3.

——and John Pierre. 1998. ‘Governance without government? Rethinking public administration’, Journal

of Public Administration Reseach and Theory : 223–43.

Pollitt, Christopher and Geert Bouckaert. 2004. Public Management Reform: A Comparative Analysis. Oxford:

Oxford University Press,

Pressman, Jeﬀrey L and Aaron B. Wildavsky. 1973. Implementation: How Great Expectations in Washington are

Dashed in Oakland. Berkeley: University of California Press,

Provan, Keith G and Patrick Kenis. 2007. ‘Modes of network governance: structure, management, and

eﬀectiveness’, Journal of Public Administration Research and Theory 18: 229–52.

Rabin, Jack and James S. Bowman. 1984. Politics and Administration: Woodrow Wilson and Public Administration.

New York: Marcel Dekker.

Ramseyer, J.M and F.M. Rosenbluth. 1993. Japan’s Political Market Place. Cambridge MA: Harvard University

Press.

Rhodes, R.A.W. 2000. Transforming British Government. London: Macmillan,

Riggs, Fred W. 2001. ‘Bureaucratic links between administration and politics’, in Ali Farazmand (ed.)

Handbook of Comparative and Development Public Administration. New York: Marcel Dekker.

Bureaucracy and the policy process

345

Riper, Paul P Van. 1984. ‘The politics–administration dichotomy: Concept or reality’, in Jack Rabin and

James S. Bowman (eds) Politics and Administration: Woodrow Wilson and Public Administration. New York:

Marcel Dekker, 203–18.

Rourke, Francis E. 1984. Bureaucracy, Politics and Public Policy. Boston: Little, Brown.

Rutgers, Mark R. 2001. ‘Splitting the universe: On the relevance of dichotomies for the study of public

administration’, Public Administration Administration and Society 33(1): 3–20.

Saint-Martin, Denis. 2004. Building the New Managerialist State: Consultants and the Politics of Public Sector

Reform in Britain, Canada and France, 2nd edn. Oxford and New York: Oxford University Press.

Scott, W. Richard. 1992. Organizations: Rational, Natural, and Open Systems. Englewood Cliﬀs, NJ:

Prentice Hall.

Seldon, Sally Coleman. 1997. The Promise of Representative Bureaucracy: Diversity and Responsiveness in a

Government Agency. New York: M.E. Sharpe.

Shafritz, J.M. Jr. and A. Hyde. 2008. Classics of Public Administration, 6th edn. Boston: Wadsworth.

——, S. Ott and Y.S. Jang. 2010. Classics of Organization Theory. Boston: Wadsworth Cengage Learning.

Smith, B. 1988. Bureaucracy and Political Power. Brighton: Wheatsheaf.

Svara, James H. 1998. ‘The politics-administration dichotomy model as aberration’, Public Administration

Review 58(1): 51–8.

—— 1999. ‘Complementarity of politics and administration as a legitimate alternative to the dichotomy

model’, Administration and Society 30(6): 676–705.

—— 2001. ‘The myth of the dichotomy: complementarity of politics and administration in the past and

future of public administration’, Public Administration Review 61(2) (March/April): 176–83.

Thompson, J.A. 1965. ‘Bureaucracy in a democratic society’, in R. Martin (ed.) Public Administration and

Democracy. Syracuse, NY: Syracuse University Press, 210–12.

Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian

Industrialization. Princeton, NJ: Princeton University Press,

Weber, Max. 1946. Max Weber: Essays in Sociology, H.H. Gerth and C. Wright Mills (ed. and trans.).

London: Oxford University Press,

West, William F. 1997. ‘Searching

for a theory of bureaucratic structure’, Journal of Public

Administration

Research and Theory: 591–614.

Westoby, Adam. 1985. ‘Introduction’, in Bruno Rizzi (ed.), Adam Westoby (trans.) The Bureaucratization of

the World: The First English Edition of the Underground Marxist Classic that Analyzed Class Exploitation in the

USSR. New York: Free Press, 1–33.

White, Leonard D. 1926. Introduction to the Study of Public Administration. New York: Macmillan.

Wildavsky, Aaron. 1979. The Politics of the Budgetary Process. Boston: Little, Brown.

—— 1986. Budgeting: A Comparative Theory of Budgetary Processes. New Brunswick, NJ: Transaction Publishers.

——, Brendon Swedlow and Joseph White. 2006. Budgeting and Governing, Brendon Swedlow (ed.). New

Brunswick, NJ: Transaction Publishers

Wilson, James Q. 1989. Bureaucracy: What Government Agencies Do and Why they Do it. New York: Basic Books.

—— 2008. American Government Brief Version. Boston: Wadsworth.

Wilson, Woodrow. 1887. ‘The study of administration’, The Academy of Political Science: 197–222.

World Bank. 1993. The East Asian Economic Miracle: Economic Growth and Public Policy. New York: Oxford

University Press,

Wu, Xun, Ling Hi Yan, Wen Sun Tung and Allen Lai. 2010. ‘Public administration research’,inReform

and Transition in Public Administration Theory and Practice in Greater China: 1978–. Hong Kong: University

of Hong Kong.

Poocharoen

346

Disagreement and alternative

dispute resolution in the

policy process

Boyd Fuller

Introduction

Disagreement is a natural feature of the policy process, and negotiation is one of the common

processes for resolving it. Both are facts of life in the interdependent relationships among gov-

ernment and non-government actors involved in and impacted by the policy process. However,

the degree and kind of disagreements can have very diﬀerent impacts on the eﬃcacy and outcomes

of the policy process and its negotiations. Similarly, not all negotiation behaviors and processes

are equally eﬀective and they need to be chosen strategically to suit the context and kind of

disagreement. Some negotiations are organized, visible, inclusive, and are managed using carefully

constructed processes. In other cases, they can be disorganized, hidden, exclusive, and ad hoc.

Whatever the case, where parties seek to cooperate, the quality of their negotiations and its

contribution to public value can be improved by using the best knowledge from the alternative

dispute resolution literature.

Studies on policy-making and the policy process tend to focus mostly on negotiations that are

convened to ﬁnd agreement among stakeholders who have varying interests and, where the policy

has become the subject for signiﬁcant disagreement among stakeholders, to resolve a dispute. As

such, mediation and consensus building theory have received the most attention. Negotiation

theory, however, provides much of the basis for them and so this chapter covers it too.

Negotiation theory starts with the premise that parties’ publicly stated positions mask their

needs and interests and that an eﬀective negotiation must uncover the interests on all sides. After

that, interests can be met creatively through a multiplicity of diﬀerent possible solutions, which

often also gain value through skillful trading across interests that the parties prioritize diﬀerently.

More detail on negotiation theory is given in the next section.

Consensus building and mediation theory tell us about when parties can be brought together,

by whom, for what purposes, and using what kind of processes. Getting the process right is

important, and often challenging. In public policy settings, there are many processes and forums

in which stakeholder representatives can pursue the goals of their organization, and thus it may

be diﬃcult to get stakeholders to authentically participate in negotiations until they are con-

vinced that the other venues cannot produce a better solution. Even when stakeholders do send

347

representatives to the negotiating table, it is often diﬃcult to truly represent the constituencies,

which are often large, divided, and thus inﬂexible. This inﬂexibility increases when one or more

constituencies believe that fundamental values are at stake. Knowledge provides no assistance,

because it is often uncertain and contested. These challenges and the ideas and practices in the

consensus building literature that address these challenges are discussed below.

Finally, possible criteria for evaluating consensus building processes employed in the policy-

making process, the shortcomings in the literature’s current data and the trials facing those

seeking to monitor and evaluate such processes in the future are summarized.

Negotiation theory as a starting point

Alternative dispute resolution theories about negotiation, mediation, and consensus building touch

upon all stages of the public policy process as long as there are interdependent parties who need

to cooperate to achieve a solution. In this section, I introduce the fundamental building blocks of

negotiation theory, which is the key building block for much of mediation and consensus building

theory as well. Negotiation, mediation, and consensus building theory approach a similar problem,

cooperation among parties with diﬀerent interests, from slightly diﬀerent perspectives. All are

focused on the challenges to and opportunities for improving the outcomes of negotiations.

Negotiation theory focuses on the barriers, strategies and tools relevant to individual parties

who are preparing for or participating in negotiations, and thus the ideas and practices proposed

are evaluated according to what beneﬁt they will bring one of the parties.

These parties can be

an individual, an organization, a nation, or some kind of grouping of these as long as they seek

to coordinate their actions together. However, the ﬁeld generally assumes that parties can do

better when they also help the other party(-ies) achieve results that are better than what the latter

might achieve when there is no cooperation. Otherwise, their counterpart will seek to meet

their interests elsewhere. (Fisher and Ury 1991)

Perhaps the most important building block of much of current negotiation theory is the

separation of interests and positions, which in practice are often conﬂated by negotiators.

Negotiation theory deﬁnes positions as the stated wants and solutions that one party seeks to get

the other to accept. The underlying interests, on the other hand, are the wants and needs that

motivate the positions negotiators take. So, for example, an organization may demand that certain

best management practices for water management be required for farmers. That is their position.

Their interests, on the other hand, may be to preserve an ecosystem or particular species, to gain

political advantage in an upcoming election, or another want or need that can be fulﬁlled

should their position be accepted by the other parties.

To be able to negotiate interests rather than positions, it is important that the negotiating parties

also (a) identify the interests of their counterpart(s) in pre-negotiation research and in-negotiation

discussions and (b) determine how each party, including themselves, prioritizes their interests.

First, it is important to consider the interests of counterpart(s) because any idea we propose

should meet their interests as well or they will not agree. Second, not all interests are equally important,

and sometimes value can be gained when a negotiator agrees to a counterpart’s proposal on one

issue that impacts on less important interests as long as they get a more preferred option on

another issue which touches more important interests.

Getting interests on the table is also important for problem deﬁnition. The purpose of a

negotiation is, after all, to ﬁnd a solution that meets the interests of the parties as eﬀectively as

possible. So, for example, in California, parties locked in a decades-long conﬂict about water

allocation among agricultural and environmental uses were able to make progress when they

redeﬁned the problem. The conﬂict had for many years revolved around two contentious

Fuller

348

questions: (a) which universal best management practices should be used on farms throughout

the state; and (b) and how much water could be saved by agricultural water users? The parties

were able to engage in more productive negotiation when they redeﬁned the problem by

asking, simply, how can downstream environmental objectives be met in a way that allows

farms to choose what means they employ for doing so? This deﬁnition touched more directly

on the interests of the various parties (preservation of environmental systems, self-determination

in managing farms and water, proﬁt) and, once found, resulted in negotiations that reached an

unexpected and welcome agreement.

Once a negotiator clearly deﬁnes the interests of the parties to the negotiation, he or she

must then identify diﬀerent possible solutions for meeting his or her own interests eﬀectively.

Here, negotiation theory broadly distinguishes two kinds of solutions: options and alternatives.

Options are the set of possible solutions that the negotiating party thinks (a) are beneﬁcial

to themselves and (b) also meet the interests of their counterparts in the negotiation. The nego-

tiator is encouraged to prepare multiple options that can be combined to make deals possible; they

must also be willing to brainstorm additional ones as they learn more about the interests of the

other party(-ies). Often, value can be created on both sides if the right trades can be made,

especially those based on diﬀerently prioritized interests, so that each side is trading something

that they want less for something that they want more. Alternatives, on the other hand, are the

set of possible solutions by which the party can meet its interests without the cooperation of

their counterpart. The most attractive of these alternatives, or the best alternative to a nego-

tiated agreement (BATNA), then becomes the minimum that a party will accept in the nego-

tiation, because no party should accept an agreement that produces less value for them than what

they could achieve elsewhere.

Interests, options, and alternatives are three elements of the Seven Elements Framework which

was developed by the Harvard Negotiation Project.

The other elements are objective criteria,

commitment, communication, and relationships. Objective criteria are external standards that

parties agree to as the basis for a decision about how value will be divided among them.

Com-

mitment is an important element in public policy and relates to the connection between the

negotiators, the constituencies they represent, the broader institutions of governance, and the

general public. Essentially, negotiators need to consider what authorization will be required for

proposed deals to be implemented. Such authorization can come from the support of key

individuals, but it may also require a change in regulation or law, the creation of a necessary

organization, or some other change in the policy, rules, or another institution. The relationship

element reminds negotiators that it is important to maintain and improve relationships with

their counterparts and that this relationship building should be done in parallel with the sub-

stantive elements of the negotiation, rather than depending on its outcomes. Finally, the commu-

nication element is where a negotiator will consider what information will be shared and sought,

and how he or she would like that process to proceed. Again, this is important in public policy

settings where information must be managed among multiple, complex constituencies, the broader

public, the media, and government decision-makers.

In the next section, this chapter introduces some of the challenges that make negotiation and

consensus building in policy-making diﬃcult.

Why negotiations (and disagreements) in the policy process are

more difﬁcult

Negotiations in the policy processes are marked by several factors that can make the management

of disagreement within them complex including challenges posed by managing the complexity

Disagreement and alternative dispute resolution

349

and scale, representing complex or unorganized constituencies, stakeholders’ contested knowledge,

and divergent values, frames, and belief systems. I describe each in more detail in this section.

Complexity and scale

As policies move through the policy process, discussions, debates, and negotiations occur in

many forums, some concurrently and some sequentially. Policy-making interactions can occur,

for example, in state and federal legislatures, courtrooms and administrative tribunals, the public

media, global and local multi-stakeholder processes and, of course, various backroom negotia-

tions. Each of these forums has its own rules and norms, includes diﬀerent parties, and produces

diverse kinds of outcomes.

More importantly for negotiation, these forums can restrict or

encourage the participation of various stakeholders and other parties (e.g. experts and mediators).

Such choices are often made to “set the table” by promoting or discouraging certain interests,

ideas, and styles of deliberation (e.g. emotive vs. less emotive) as well as inﬂuencing what resources

are brought to the table, how much and what information is shared (e.g. people are less likely

to disclose sensitive information to parties they mistrust), and the methods by which the support

of various constituencies be gained.

In addition, the visible processes are often not the only ones occurring nor are they always

the most important. In rooms out of view, some parties may enter in negotiations that shape

who gets standing, what issues are discussed, and other elements of the more oﬃcial processes.

Similarly, the bigger, more visible negotiations impact the backroom ones. These backroom

negotiations may occur among fewer parties and exclude some stakeholders (de Sousa Briggs

2008), and there may be many smaller negotiations instead of one larger, multiparty one. In some

cases, the convening of less visible processes is on purpose, providing space for exploring new

ideas away from the inﬂexible gaze of warring constituencies (Martinez et al. 2000; Fuller

2009b, 2009c)

Keeping track of all the parties and processes can become overwhelming and so it is not a

surprise that parties tend to reduce the complexity. Often they do so by forming coalitions

organized around common positions and belief systems. However, while coalitions can reduce

complexity, they also introduce some additional challenges. As coalitions form, they must

negotiate what they stand for. Often, these negotiations are settled by diﬃcult compromises on

a common position that becomes inﬂexible. However, negotiation theory argues that negotiating

from positions leads to unwise outcomes because of their inﬂexibility and because they hide the

true wants and needs of the parties that could be solved in other ways more acceptable to their

counterparts.

In two-party negotiations, skilled negotiators can build relationships and manage commu-

nication with their counterparts to get at the interests and look for creative options and trades to

meet them concurrently with their own. However, as new ideas emerge from negotiations, each

side must consider quickly the impact of these ideas on their interests. Individuals can do so

quickly, but as the complexity of each entity increases, so does the time required to analyze the

new ideas, particularly when the parties in the constituency need to reopen negotiations among

themselves about what their joint response will be—e.g. accept, reject, modify for the other party’s

consideration.

Representing complex or unorganized constituencies

The classic problem of divergent interests between principal and agent presents one well-known

challenge.

Principal-agent or representative-constituency relationships are much more complex

Fuller

350

in public policy settings, however. First, constituencies are often too complex to be eﬀectively

characterized as a single entity. Instead, constituencies are often divided by their internal

diﬀerences in interests, values, and frames—e.g. when the constituency is an alliance of

multiple organized actors—or unorganized populations that have not yet formulated and

articulated what they stand for. In both cases, negotiators of these constituencies will ﬁnd it

diﬃcult to negotiate well. They may become inﬂexible, especially if their negotiating position is

the result of a hard-won compromise among the factions of the constituency, who would

naturally be reluctant to reopen their internal negotiations when the negotiators ask them to be

more ﬂexible.

Where constituencies are unorganized, it may be easy to ignore their interests and values at

ﬁrst. However, agreements can then be undermined if the unorganized feel that their interests

and values have been ignored. Mansbridge (1996) notes that policy systems often go through

cycles in which decisions are made which impact unorganized and uninterested groups, who

then organize to change the policies, which then can impact other old and new parties, and the

cycle repeats itself.

Of course, constituencies are not the only groups to which the negotiation and its parties are

beholden. In the public realm, any deals must also respect the institutions of governance. Where

there are sunshine laws which require transparency in public participation processes, for instance,

then the meetings must be open to the public, and thus negotiators may play to their home crowd

rather than seeking to make a potentially diﬃcult deal. Whatever agreement is constructed must

also be allowable given the laws, policies and regulations of the government(s) involved or there

must be a plan to change the appropriate institution, perhaps along with a contingency plan

should that change turn out to be impossible.

In the next two sections, this chapter covers the challenges that arise when facts are contested

and stakeholders believe values are at stake.

Contested knowledge

Knowledge is one of the pillars for problem deﬁnition and the formulation of solutions. When

it is contested, identifying, developing, and deciding among multiple viable problem deﬁnitions

and solutions can be diﬃcult. And yet, without knowledge, policy-making is blind, both in

what solution it chooses and also in what it learns from the implementation and evaluation of

that solution.

Contested knowledge arises from apparently competing claims about what data and ﬁndings

are correct. Knowledge can diﬀer across groups for two kinds of reasons. The ﬁrst kind relates

to the diﬀerent frames and practices that each group employs in creating, validating, and dis-

seminating knowledge. The second kind occurs when parties choose which knowledge they

share for strategic rather than factual reasons: lying, withholding, and re-interpretation are all

possible activities here. I start with the ﬁ rst kind.

Parties holding diﬀerent frames may ask diﬀerent questions about the environment and hold

to diﬀerent facts and theories about how things work. Such disputes are hard to resolve in most

policy disputes. Data is often incomplete, uncertain, and often associated with one side or the

other. The problem is further exacerbated by how easily knowledge claims can be decon-

structed in the diﬀerent policy forums. Much knowledge relies on a set of imperfect practices that

are nonetheless accepted within knowledge communities. The problem arises when such sci-

ence enters other forums which have diﬀerent methods for generating and validating knowl-

edge. In the forums of public policy—legislatures, courtrooms, scientiﬁc advisory committees,

and so on—it is easy to argue that the knowledge of other parties is invalid. The standing of

Disagreement and alternative dispute resolution

351

diﬀerent stakeholders as knowledge holders can be challenged. Assumptions and methodologies

can be questioned. The amount of uncertainty and risk that is acceptable for decision-making can

be contested. For example, local communities collect signiﬁcant data, but the method of its col-

lection (everyday practice and experience) and the form of its storage (often stories) is not

considered valid by many scientiﬁc communities (Fischer 2000; Corburn 2005; Fuller 2009a).

Those same communities, however, argue that much of expert science lacks the data that local

communities have, and that they overly simplify the complex situations on the ground and use

values and criteria for analyzing the meaning and impact of the data that do not reﬂect the

values of the community. Furthermore, whosoever counts as an expert may be contested as

well, and a person’s standing is often crucial for the authority of his analysis in addition to

whatever methods they use. Even the questions that the knowledge holders ask can be quite

diﬀerent, and thus the data that one side collects may be seen by another side as the invalid

because its holder was asking the wrong question.

Furthermore, it would be naïve to neglect the strategic considerations behind contested knowl-

edge. A party that feels under threat may construct false data and ﬁndings to protect its interests

(Forester 2005). Given the degree of uncertainty in today’s knowledge about the many complex

systems being considered in public policy, such knowledge claims can be diﬃcult to completely

refute. While their authenticity may be called into question, the claimers can also easily undermine

the knowledge claims of others. This again leaves policy analysts and makers with no one

legitimate set of facts that they can use to make authoritative policy.

Values, frames, and belief systems

Perception, interpretation, and the formulation of solutions are shaped by the lens through

which we view the world. Schön and Rein (1994) and Lewicki et al. (2003) argue that many

policy conﬂicts are driven by the divergence among actors’“frames,” which they describe as

normative/prescriptive stories that actors use to interpret the world, deﬁne problems, and choose

actions. The frames that individuals, organizations, and coalitions use contain causal perceptions

that link actions and impacts as well as values about right and wrong. Similarly, Sabatier and

other students of the Advocacy Coalition Framework

argue that many long-term policy

disputes are driven by the presence of two or more advocacy coalitions that are organized around

divergent belief systems and who disagree about what should be done around a particular

policy issue.

Conﬂicting values and frames pose several problems to eﬀective negotiation. First, the diﬀerent

interpretations and meanings actors attribute to actions, words, and phenomena often lead to

misunderstanding, which over time can lead to each side attributing hostile intentions to their

counterparts. Even the most well-meaning actions by one party may antagonize the other(s). In

some cases, leaders of one side may aggravate the situation by strategically misinterpreting the

other’s actions, since the creation and maintenance of a hostile enemy discourages the exploration

of internal disagreements and diﬀerences.

Furthermore, the parties with divergent frames may deﬁne problems quite diﬀerently. Some

might, for example, see that slums are an unsightly blight which will then lead them to think of

removing the slums and placing its inhabitants elsewhere. Others might see slums as a place where

poor people work as a community to generate resources for self-reliance and see the problem

more as one of providing the right policy environment for that innovation to ﬂourish. Often,

the diﬀerences in stakeholders’ stories of problem deﬁnition and appropriate solutions are missed

by analysts and negotiators, who then attribute negative characteristics to the other side (unwilling

to listen, not intelligent, aggressive) instead of understanding that the two sides are having two

Fuller

352

mostly independent conversations in which they argue about solutions (positions) without ﬁrst

clarifying what problem each side is talking about and what problem they should be solving

as a group.

The problem is not only one of interpretation. When parties start speaking about values and

rights, negotiation becomes even more diﬃcult. Conﬂicts about values and rights are notoriously

more diﬃcult to resolve than interest-based disputes (Tribe et al. 1976; Lord 1979; Susskind and

Field 1996; Fuller 2009b, 2009c). Once a party invokes one (or both), it becomes more diﬃcult

for them to consider solutions that are diﬀerent from their public positions since no one wants

to “compromise their values.” This is especially true when the negotiators represent constituencies

who have access to the meeting or its records. In these instances, negotiators tend to prefer

tough language, inﬂexible positions, and disparaging language to maintain their reputation

within their constituency. All of these, unsurprisingly, provoke the other side to negotiate in a

similar fashion.

In public policy contexts, in which policy processes can extend over multiple years and even

over a decade, these problems of framing become hardened as negotiators can fall into unhelpful

routines and habits. In many cases, constituencies no longer believe that they have anything to

gain from cooperating with the other side, and will punish representatives who start exploring

those possibilities. Group polarization can occur, in which each side demonizes the other side,

attributing the worst intentions to their actions while suppressing their own internal diﬀerences

(Northrup 1989; Sunstein 2002). As hostility grows, the ability of each side to negotiate becomes

more diﬃcult as agreement is seen as a “deal with the devil” (Benjamin 1990; Mnookin 2010).

In the light of these challenges, it is not surprising that some authors propose that the parties

need to increase their understanding of one another. Schön and Rein argue that policy delib-

erations should help the parties “put the other’s shoes on,”

which means that each party can see

the problems, solutions, and other elements of their discussion using the same frames or belief

systems of their counterparts. In doing so, it is understood that the negotiating parties (or analysts)

should be able to reframe the negotiation in such a way that the problems of their divergent

belief systems—diﬀerent and conﬂicting interpretations of communication and phenomena,

problem deﬁnitions, and knowledge—will be resolved. What is not spelled out is how the parties

use this shared understanding and how they will make their solutions acceptable to the constituencies

who have not undergone this epiphany.

In the next section, I introduce and summarize consensus building and mediation theory, and

show what ideas and recommendations they put forward to manage and move through these

challenges.

Consensus building and mediation theory

Consensus building and mediation theory relate to the convening and management of nego-

tiation processes. In particular, they provide ideas, understandings, and guidance related to the

challenges of and better approaches to designing processes, convening stakeholders, managing

their negotiations, and promoting eﬀective agreements. Both provide guidance about how the

parties can approach the challenges described in the previous section. Consensus building theory

focuses more on negotiations convened by actors other than the courts or other adjudicative

bodies. However, for the purposes of this chapter, there is enough similarity between the two

theories that I will hereafter refer to both as consensus building theory.

Negotiation theory is the building block for consensus building theory. Like negotiation

theory, consensus building theory emphasizes interests instead of positions, the desirability of

creating and investigating multiple options to meet interests, and so on. As a starting point,

Disagreement and alternative dispute resolution

353

consensus building theory promotes certain steps as part of an eﬀective strategy, which for the

readers will seem mostly similar to the policy stages, save that evaluation is missing. I cover the

evaluation of these processes in the next section. For consensus building, the recommended

steps are

(with the equivalent policy stage in brackets):

 convening the relevant parties (agenda-setting);

 clarifying the rules and responsibilities for the group and its participants;

 deliberating to develop options and then packages of options that seek to meet the needs of

the participating stakeholders as well as the public interest (policy formulation);

 making decisions that have the support of all or almost all the stakeholders (decision-making);

 implementing agreements and other commitments (implementation).

In the rest of this section, I describe the steps and highlight where they address the challenges

described above.

Convening—getting the problem and parties right

Convening in consensus building is about the iterative design of the consensus building process.

Essential in convening are two tasks: getting the problem and parties right. Convening starts

with identifying the key players, including both nongovernment and government actors. Some

of these parties will need to participate in the proposed consensus building process. It often helps to

have certain key agencies or players openly support the process—for example, a prominent

leader might talk to certain people to gain their participation. Others may provide staﬀ support,

particular expertise, monetary resources, or just observation. (Carlson 1999)

To get the problem right and the right parties at the table, consensus building theory argues

that a conﬂict assessment should be conducted. A conﬂict assessment occurs before the process is

oﬃcially convened and is crucial both for process design, problem deﬁnition and for motivating

stakeholders to attend. In practice, a conﬂict assessment is usually conducted by a neutral third

party who is often but not always the mediator. The assessor’s ﬁrst task is to identify and interview

potential stakeholders and their representatives. In these interviews, the conﬂict assessment has

two basic tasks. The ﬁrst is to gather information about the parties, including their interests,

ideas for solutions, and their willingness to participate in the dialogue. The second task is to tell

them, and answer questions, about the conﬂict assessment process and the proposed negotiation.

The latter is a part informing and part trust-building task. It allows the stakeholders to develop

more comfort about the process (Susskind and Thomas-Larmer 1999).

After the interviews, the conﬂict assessor should put together a report, usually without attri-

bution, that summarizes what he or she has discovered about the parties and what process she

suggests for their negotiation. In some cases, the assessor might also recommend that the

negotiation should not proceed, especially if he or she believes that the parties are unlikely to

reach agreement. This report is shared with the parties who are asked to provide feedback on it.

Once it is completed, it will be handed over to the proposed convenor of the negotiations.

Again, this feedback is not only intended to improve the document, but also to build trust with

stakeholders. If they approve of the suggestions, they are more likely to attend when the

negotiations are convened.

As an agenda-setting exercise, conﬂict assessment and the broader convening process is a proac-

tive means of determining who the parties are and what issues should be discussed. Participation

in a consensus building process is also voluntary and so the conﬂict assessor will seek some

agenda of issues that attracts the interests of participants and decision-makers and will often seek

Fuller

354

an overarching framing of the issue that all sides agrees is a good start for the discussions. This

reframing and agreement on the agenda is an important ﬁ rst step to dealing with problems of

divergent frames and beliefs and in keeping constituencies apprised and supportive of the

deliberations.

In some cases, for example, the conﬂict assessor may propose that stakeholders can be grouped

into categories based on similar interests so that the common perspective can be spoken for by a

few selected representatives. The process can also be set up other forums in which the repre-

sentatives of large constituencies can meet with their selected representatives to debrief current

developments in the consensus building group’s emerging ideas for vetting and feedback.

While consensus building theory’s ideas about convening and conﬂict assessment provide

some traction on the challenges discussed above, implementing them in practice is no trivial

task. For example, identifying the right stakeholders can be diﬃcult when some of the interests

are held by populations and communities who are unorganized (Mansbridge 1996; Laws 1999).

Their members may not have realized that they do constitute a group or at least they may not

have chosen a representative. Where an organization has chosen a representative, that person may

not be the best choice for the negotiation because they are, for example, inﬂexible and unwilling

to learn, poor negotiators, or unable to forge good working relationships with other partici-

pants. In other cases, some stakeholders may be reluctant to send a representative because they

feel that they have little to gain—for example, because they distrust each other, believe that the

convenor is unwilling or unable to translate their agreement into action, or feel that they can

achieve their interests and protect their values more eﬀectively in other policy forums.

Clarifying rules and responsibilities

At the beginning of the process, the various parties should agree about the responsibilities of the

group and the various participants (stakeholders’ representatives, mediators, staﬀ, experts, etc.).

For example, they need to determine what purpose the group is supposed to achieve, what pro-

ducts it seeks to produce, what roles and responsibilities, the mediator, negotiating representa-

tives, experts, and any other participating parties will take on, and so on. Ground rules will also

have to be established. Some will cover what parties can and cannot do during the delibera-

tions. Ground rules start with behavior at the table, but they can be extended to cover behavior

away from it as well. Examples include whether parties will pursue their interests in other

forums and whether and how the individual parties can share information with non-participants,

including constituencies, other oﬃcials, the media, and so on.

Clarifying the responsibility of the group and its individuals is especially important when

parties believe values and rights are at stake (Forester 1999). A relatively safe environment for

dialogue and negotiation needs to be created when there is signiﬁcant mistrust among the par-

ties or about the process. By using rules to moderate the dialogue, and establishing norms of

good listening, mediators can help to create the conditions needed where parties can explore

each other’s claims about values in order to test their genuineness and to acknowledge them as

being legitimate. This inquiry and acknowledgment need not mean agreement.

Also, ground rules are important in terms of moderating how information is shared with

others outside the group. Rules favoring transparency, for example, are common in some

countries but while they have normative appeal, they may hamper communication and pro-

blem-solving in cases where the parties do not trust each other. Where constituencies distrust

representatives who make overtures to the other side, it may make more sense for some of the

negotiations to take place out of the public eye. During the less visible negotiations, the

representatives can listen more carefully and explore some new ideas without fear of censure.

Disagreement and alternative dispute resolution

355

However, when this is done, there should be clear ground rules and a strategy for how infor-

mation will be shared with constituencies, decision-makers, and the public (Martinez et al.

2000; Fuller 2009c).

Deliberating to ﬁnd and package options

To develop a solution, the parties need ideas that go beyond the initial positions that each bring

to the table. These ideas, when combined, need to produce potential agreements that all parties

see as being better than that which they could achieve in another forum (i.e. their best alter-

native to a negotiated agreement—BATNA). These ideas start as discrete solutions to diﬀerent

parts of the puzzle that the group is trying to solve. Some may provide value to all parties;

others may beneﬁt only some of the parties. This packaging is made more eﬀective when parties

prioritize among their interests, and then look for trade-oﬀs in which each party gives up

something less important in exchange for something more important. As parties move toward a

solution, they will begin to package these ideas together more and more, with each party

examining the emerging packages to determine whether their set of interests is being met more

eﬀectively in the proposed deal than in their best alternative.

Often this packaging is done using single texts. Single texts—such as draft agreements, charts,

and maps—can be used to synthesize and present certain understandings, draft ideas, and mile-

stones during the process. They can be records created by the mediator that the parties agree

clarify areas of emerging agreement and ongoing contention, including possible solutions, facts,

and so on. Stakeholders can use draft agreements to organize their deliberation, by adding,

subtracting, and modifying text in the single text until they can agree to what is stated there.

Making decisions

In negotiations between two individuals, the decision is simply made by the parties at the table.

In any other negotiations, decision-making becomes more complex, especially as the number of

parties and constituencies increases. Public policy involves many stakeholders and so most nego-

tiations will include more than two parties. Since not all these negotiations will require unani-

mity to make a decision, a decision rule must be agreed upon at the beginning of the negotiation.

Such a decision rule could be simply 50 percent plus one or some kind of supra-majority such as 66

percent. It might also be a weighted majority. For example, the Kyoto Protocol decision rule

required that at least 55 percent of the countries approve the deal and that the aggregate percentage

of emissions represented by those countries must exceed 55 percent of the world’s total emissions.

While the parties may make a decision at the table, the work of decision-making is not done.

The transition to implementation often requires signiﬁcant eﬀorts to gain support from key

members of constituencies and other parties who will be crucial to the decision’s implementa-

tion. In international cases, the decision will likely have to be formally ratiﬁed, which requires

an additional set of negotiations within the constituency.

Implementation

Decision-making and implementation are in fact intertwined in consensus building. Decisions

should not be made until the parties are fairly sure that they will get the support needed from

decision-makers (legislators, community leaders, agency oﬃcials, etc.) who were not at the table.

In well-managed processes, draft ideas and agreements will be shared with such leaders to get

their feedback and seek their backing. Where success is unlikely, the negotiators should consider

Fuller

356

other options even if the parties agree the ﬁrst idea is superior. Where success is possible, the

parties then consider what additional mechanisms and responsibilities should be built into the

recommendations to improve their implementation. For example, the parties might agree to

lobby particular parties together to gain support for the recommendations.

Sometimes, the negotiations may also ignore various additional forums to test the emerging ideas

with constituencies and the wider public. The city of Chelsea, Massachusetts, for instance, created

a new city charter using a consensus building process in which developments were continually

revealed to the public for their feedback and then support using radio and other media.

Often the convenor of a negotiation or consensus building process has the major role in

implementing the recommendations. The negotiating groups themselves rarely have formal

authority to make decisions. Instead, they make recommendations that they expect the convening

actor to translate into an agenda, set of solutions, or a policy or program to be implemented as

part of the policy process. In reality, the conveners may choose to follow only some or even

none of the recommendations, though they put their legitimacy at peril when they do so. Or

they may adopt them all but struggle to implement them in practice.

Finally, as the literature on public policy has also found, implementation brings about its dif-

ferent interpretations of the agreement, as well as new misunderstandings and disagreements. An

eﬀective agreement often includes provisions for how the parties will manage their disagreements

and disputes during the implementation (Podziba 1999).

Evaluating negotiations and mediations in the public sector

Negotiations that occur in policy processes can be both easy and very diﬃcult to assess. At ﬁrst

brush, they are easy to evaluate because as long as the parties are satisﬁed, then the negotiations

can be deemed to be a success. If eﬃciency is a concern, one could look at the interests of the

diﬀerent actors and see from a dispassionate viewpoint whether one or both of the parties could

have done better without any party losing gains (Raiﬀa 1982; Raiﬀa et al. 2002).

However, scholars of the policy process can and should ask more demanding questions of

such negotiations (Coglianese 2003). We ask that policy processes be eﬃcient in time and

resources, wise in the use of the best knowledge, and fair in how the process is conducted and

how resources are allocated. In this light, Bingham (2003) proposes that there are three possible

kinds of beneﬁts from negotiations in the public sector: substantive, process, and relationships.

Substantive beneﬁ ts revolve around improved outcomes. One common claim is that the out-

comes of negotiations are better than that achievable by each party without their agreement and

cooperation (their BATNA). Similarly, some point to the creativity sparked by options, in

which new ideas emerge from the deliberations among the participants. Proponents of public

dispute resolution also argue that when parties do negotiate and their recommendations are imple-

mented by the convening actor, stakeholders are more likely to support, or at least not actively

oppose, the subsequent policy. This logic is also used to argue that the policies developed by

such negotiations are better than those developed through other processes.

Process beneﬁts include reduced time and resources to reach an implementable policy. A

negotiation can also increase the perceived fairness of the process by providing opportunities for

new actors to participate and by giving each actor an equal chance to present and advocate for

their interests. A negotiation may also be considered more fair and wise when multiple forms of

knowledge are brought to the table and when joint fact-ﬁnding allows the group as a whole to

decide what facts they will accept.

Relationship beneﬁts can include better personal relationships among the representatives who

negotiated or between the groups who were represented. Groups can learn that the other side is

Disagreement and alternative dispute resolution

357

reasonable even if they have come to diﬀerent conclusions. They may share more information

and thus be better informed. Future negotiations and other deliberative and policy-making

processes may go more smoothly because the parties now understand one another better.

Unfortunately, the research on evaluating such processes is still in its infancy and faces several

key challenges. Data can be hard to access because the group agreed to conﬁdentiality. Many studies

of negotiations in the public sector are individual, descriptive case studies rather than hypothesis

testing, higher-n inquiries. These cases tend to be custom-designed to ﬁt each context and they

occur in a wide variety of settings, both of which makes them diﬃcult to compare. The cases

tackled through negotiation are suﬃciently diﬀerent from other cases in that it is diﬃcult to

compare the eﬀectiveness of traditional policy-making processes with those supplemented by

negotiation (Emerson et al. 2003).

In addition, many of the evaluations are written up by mediation practitioners, which raises

the possibility of bias. Finally, negotiations convened in and around policy processes are not dis-

crete phenomena. Even a process that fails to reach an agreement may produce information that

improves the policy process or result in improved relationships, thus reducing the amount of

conﬂict manifest in other stages of the policy process.

Conclusions

Negotiation itself has always been a part of policy-making. However, the push in recent dec-

ades for more inclusive and science-based processes has brought negotiation out of the back

halls and legislatures to include a wider group of people who set the agenda, seek inclusive and

creative solutions, and make and implement agreements. Such negotiations can touch any part

of the policy process. They may, for example, be used to set the agenda, to create ideas for policy

formulation, to make recommendations that inform decisions, and to aid in the implementation

of policies.

Negotiation and consensus building rely on the parties to share information and to be open

to new ideas. To make the negotiations productive in public policy settings, much care must be

taken to help the stakeholders to be willing and able to engage one another even when they

feel that values are at stake or that they cannot trust the other parties. Such negotiations must be

carefully convened, usually through an interactive conﬂict assessment that helps the convenor

and mediator determine the agenda and process design and otherwise bring the parties together

ready to talk about a solvable problem. For the negotiation to be credible and eﬀective, it must

be strategically linked to the ongoing policy-making process and to decision-makers outside

government as well.

Knowledge is often contested and so mediators must help the parties to engage in joint fact-

ﬁnding so that they have a common set of agreed-upon facts for their deliberations. Joint fact-

ﬁnding does not necessarily result in one agreed-upon theory or interpretation of the data, but

it is a common foundation for discussions about what is happening and what might happen if

changes are made.

Notes

1 This summary of negotiation theory draws upon several sources, including Fisher and Ury (1991), Lax

and Sebenius (2006), and Raiﬀa (1982).

2 For more information on water conﬂicts in California and the challenges and opportunities of nego-

tiating them, see Fuller (2009b, 2009c) and the special issue on the CALFED Bay Delta Program in the

Environmental Science and Policy Journal 12(6).

3 See Fisher and Ury (1991) and the Program on Negotiation website: http://www.pon.harvard.edu.

Fuller

358

4 For example, negotiators might refer to pollution standards in other areas with similar conditions and

industries as a fair way of setting the standard for their own region.

5 Some outcomes are enforceable decisions, some provide input or analysis, and some have no real

outcomes even as they try to increase the legitimacy of the process—for example, courtroom decisions

have legal backing that encourages implementation while global agreements rely on national legislatures

to ratify and enforce agreements.

6 Principles and agents’ interests are not always aligned, and thus agents may act in ways that serve the

agent’s interests rather than the interests of the principle who hired them.

7 For more information about the Advocacy Coalition Framework, see Chapter 10 in this volume and

Sabatier and Weible (2007).

8 See Schön and Rein (1994) and Rein and Schön (1996).

9 See Susskind (2005) and Susskind and Cruikshank (2006).

10 In fact, where parties, problems, and solutions are many and complex, negotiating groups often focus

most of their eﬀorts on developing, modifying, and make decisions by focusing mostly on packages, as opposed

to the discrete trades often discussed and analyzed in the negotiation literature (Innes and Booher 1999,

Fuller 2009c).

Bibliography

Benjamin, M. 1990. Splitting the Diﬀerence: Compromise and Integrity in Ethics and Politics. Lawrence, KS:

University Press of Kansas.

Bingham, G. 2003. ‘Foreword’,inR.O’Leary and L.B. Bingham (eds) The Promise and Performance of

Environmental Conﬂict Resolution. Washington, DC: Resources for the Future.

Carlson, C. 1999. ‘Convening’, in L. Susskind, S. McKearnan and J. Thomas-Larmer (eds) The Consensus

Building Handbook: A Comprehensive Guide to Reaching Agreement. Thousand Oaks, CA, Sage.

Coglianese, C. 2003. ‘Is satisfaction enough? Evaluating public participation in regulatory policymaking’,

in R.O’Leary and L.B. Bingham (eds) The Promise and Performance of Environmental Conﬂict Resolution.

Washington, DC: Resources for the Future.

Corburn, J. 2005. Street Science: Community Knowledge and Environmental Health Justice. Cambridge, MA:

MIT Press.

de Sousa Briggs, X. 2008. Democracy as Problem Solving: Civic Capacity in Communities Across the Globe.

Cambridge, MA: MIT Press.

Emerson, K., T. Nabatchi R. O’Leary and J. Stephens 2003. ‘The challenges of environmental conﬂict

resolution’,inR.O’Leary and L.B. Bingham (eds) The Promise and Performance of Environmental Conﬂict

Resolution. Washington, DC: Resources for the Future.

Fischer, F. 2000. Citizens, Experts, and The Environment: The Politics of Local Knowledge. Durham, NC: Duke

University Press.

Fisher, R. and W. Ury. 1991. Getting to Yes: Negotiating Agreement Without Giving In. New York: Penguin

Books.

Forester, J. 1999. ‘Dealing with deep value diﬀerences’, in L. Susskind, S. McKearnan and J. Thomas-

Larmer (eds) The Consensus Building Handbook: A Comprehensive Guide to Reaching Agreement. Thousand

Oaks, CA: Sage, 463–93.

—— 2005. ‘Policy analysts can learn from mediators’, in J.T. Schloz and B. Stiftel (eds) Adaptive Governance

and Water Conﬂict: New Institutions for Collaborative Planning. Washington, DC: Resources for the Future,

150–63.

Fuller, B. 2009a. ‘Local knowledge and consensus building: from access to impact’,

LKY School of

Public

Policy Working Papers, SPP09-08.

—— 2009b. Moving Through Value Conﬂict: Consensus Building and Trading Zones for Resolving Water Disputes .

Wiesbaden: VDM Verlag Dr. Mueller e.K.

—— 2009c. ‘Surprising cooperation despite apparently irreconcilable diﬀerences: Agricultural water use

eﬃciency and CALFED’, Environmental Science and Policy 12(6): 663–73.

Innes, J.E., Booher, D.E. 1999. ‘Consensus building as role playing and bricolage: Towards a theory of

collaborative planning’, Journal of the American Planning Association, 65(1): 9–26.

Laws, D. 1999. ‘Representation of stakeholding interests’, in L. Susskind, S. McKearnan and J. Thomas-

Larmer (eds) The Consensus Building Handbook: A Comprehensive Guide to Reaching Agreement. Thousand

Oaks, CA: Sage, 241–85.

Disagreement and alternative dispute resolution

359

Lax, D.A. and J.K. Sebenius. 2006. 3-D Negotiation: Powerful Tools to Change the Game in your Most Important

Deals. Boston, MA: Harvard Business School Press.

Lewicki, R.J. and B. Gray. 2003. ‘Framing of environmental disputes’, in R.J. Lewicki, B. Gray and M. Elliott

(eds) Making Sense of Intractable Environmental Conﬂict: Concepts and Cases. Washington, DC: Island Press.

Lord, W.B. 1979. ‘Conﬂict in federal water resource planning ’, Water Resources Bulletin 15(5): 1226–35.

Mansbridge, J.J. 1996. ‘Using power/ﬁghting power: The polity’, in S. Benhabib (ed.) Democracy and Diﬀerence.

Princeton, NJ: Princeton University Press.

Martinez, J., L.E. Susskind, et al. 2000. ‘Parallel informal negotiation: An alternative to second track

diplomacy’, International Negotiation 5(3): 569–86.

Mnookin, R. 2010. Bargaining with the Devil: When to Negotiate, When to Fight. New York, Simon and

Schuster.

Northrup, T.A. 1989. ‘The dynamic of identity in personal and social conﬂict’, in L. Kriesberg, T. A.

Northrup and S. J. Thorson (eds) Intractable Conﬂicts and their Transformation. Syracuse, NY: Syracuse

University Press.

Podziba, S. 1999. ‘“Case 3”. The Chelsea Charter consensus process’, in L. Susskind, S. McKearnan and

J. Thomas-Larmer (eds) The Consensus Building Handbook: A Comprehensive Guide to Reaching Agreement,

Thousand Oaks, CA: Sage.

Raiﬀa, H. 1982. The Art and Science of Negotiation. Cambridge, MA: Belknap Press of Harvard University

Press.

——, J. Richardson and David Metcalfe 2002. Negotiation Analysis: The Science and Art of Collaborative

Decision Making. Cambridge, MA: Belknap Press of Harvard University Press.

Rein, M. and D. Schön. 1996. ‘Frame-critical policy analysis and frame-reﬂective policy practice’, Knowledge

and Policy: The International Journal of Knowledge Transfer and Utilization 9(1): 85–104.

Sabatier, P.A. and C. Weible. 2007. ‘The advocacy coalition framework: Innovations and clariﬁcations’,i

P.A. Sabatier (ed.) Theories

of the Policy Process. Boulder, CO, Westview Press.

Schön, D.A. and M. Rein. 1994. Frame Reﬂection: Toward the Resolution of Intractable Policy Controversies.

New York: Basic Books.

Shafritz, Jay M., J. Steven Ott, and Yong Suk Jang. 2010. Classics of Organization Theory. Boulder, CO:

Wadsworth Cengage Learning.

Sunstein, C.R. 2002. ‘The law of group polarization’, Journal of Political Philosophy 10(2): 175–95.

Susskind, L.E. 2005. ‘Consensus building and ADR: Why they are not the same thing?’,inM.Moﬃtt and

R. Bordone (eds) The Handbook of Dispute Resolution. San Francisco, Jossey-Bass.

—— and P. Field. 1996. Dealing with An Angry Public: The Mutual Gains Approach to Resolving Disputes.

New York: Free Press.

—— and J. Thomas-Larmer. 1999. ‘Conducting a conﬂict assessment’, in L. Susskind, S. McKearnan and

J. Thomas-Larmer (eds) The Consensus Building Handbook: A Comprehensive Guide to Reaching Agreement.

Thousand Oaks, CA: Sage.

——and J.L. Cruikshank. 2006. Breaking Robert’ s Rules: The New Way to Run Your Meeting, Build Consensus,

and Get Results. New York: Oxford University Press.

Tribe, L.H., C.S. Schelling and John Voss 1976. When Values Conﬂict: Essays on Environmental Analysis,

Discourse, and Decision. Cambridge, MA: Ballinger Publishing Company.

Fuller

360

Governance, networks and

intergovernmental systems

Robert Agranoff, Michael McGuire and Chris Silvia

Governments today operate within systems of governance that include interacting govern-

mental bodies and nongovernmental organizations (NGO). These interconnections have

increased dramatically in recent decades because of increased intergovernmentalization and

government externalization. The former refers to grant, regulatory, and related vehicles of

connecting central and subnational governments and the latter to moving governmental func-

tions and services delivery outside governments’ boundaries through contracting, vouchers for

goods and services, and related means of “privatization.” This has led to the need to capture the

essence of public activity as governance, “a mix of all kinds of governing eﬀorts by all manner

of socio-political actors, public as well as private; occurring between them at diﬀerent levels, in

diﬀerent governing modes and orders” (Kooiman 2003: 3). Such interactive governance

increasingly occurs within networks of government oﬃcials and NGO actors, that is “structures

of interdependence involving multiple organizations or parts thereof, where one unit is not

merely the formal subordinate of the others in some larger hierarchical arrangement” (O’Toole

1997: 45).

Collaboration

In governance networks, the actors depend on the collaboration imperative, that is the actions

of oﬃcials and managers crossing the boundaries of their organizations working in dyadic, tria-

dic, or networked relationships that are transactional. These actions bring on the need for col-

laborative actions beyond cooperation, working jointly in some fashion, creating or discovering

solutions within given constraints, for example knowledge, time, money, legal authority, or other

resources. As such, “collaborative management is a concept that describes the process of facil-

itating and operating in multi-organizational arrangements to solve problems that cannot be

solved, or solved easily, by single organizations” (Agranoﬀ and McGuire 2003: 4). Collaborative

activity among the actors within governance makes their home organizations more conductive,

organizing its work for external as well as internal activity. A conductive organization is one “that

continuously generates and renews the capabilities to achieve breakthrough performance by

enhancing the quality and ﬂow of knowledge and by calibrating its strategy, culture, structure

and systems” externally (Saint-Onge and Armstrong 2004: 213).

361

The transformation of intergovernmental relations

The era of governance networks has also transformed intergovernmental relations (IGR) beyond

its traditional scope as involving “combinations of interdependencies and inﬂuences among

public oﬃcials” working on ﬁnancial, policy, and political issues (Krane and Wright 1998: 1168) to

incorporate waves of collaborative operationalism external to governments as well as between

them. Indeed, IGR has evolved through four distinct waves of development including the

current network era (Agranoﬀ 2010).

The ﬁrst, law and politics, emerged with the building of integral nation-states, primarily in

the eighteenth and nineteenth centuries, where legal distinctions of governmental isolation and

“jurisdiction” held true. Generally, responsibilities were separated and isolated. The dual feder-

alism doctrine in the United States held that the national government and the states were each

sovereign in their respective spheres and that between them exist areas of activity in which neither

can enter. While there is reasonable evidence that completely separate spheres never existed in

practice (Elazar 1962), the legal distinction and intergovernmental norms of separation held for

some time, despite the fact that by the mid-1800s modern communications, which required a series

of local delivery units (postal units, roads, canals, railways) and cross-links, pushed jurisdictional

separation to its limits.

This situation changed from the early twentieth century to roughly the 1960s when the

welfare state ushered in the second epoch. It marked a time of growth and more professional

central governments that also included an increasing interdependency that linked subnational

and central governments (Skowronek 1982). In particular, the welfare state was very much a

top-down eﬀort that enhanced the ﬁscal and program strength of national governments every-

where (Loughlin 2007: 389). Most central social policies (welfare, social services, employment,

economic development) were polity-wide because central governments were suspicious of uni-

versal local commitment, so national ﬁnancing and programming were linked with local com-

mitment (Ashford 1988: 18). Later in the era came newer social welfare programs, for example

drug abuse, child and family abuse, mental disabilities, and new eﬀorts in community develop-

ment, which were less suited to central organizations, bringing on the need for “important

intergovernmental adjustments” (ibid.: 19). The resulting impact was considerable jurisdiction

overlap (Watts 1999: 38).

By the middle of the twentieth century governments began to recognize the gradual intro-

duction of organized actors outside government as also involved in funding and programming,

particularly in Western Europe and the Anglo-American countries. NGOs became agents and

partners of the state (Rhodes 1997). Through some grants but predominantly by contracts, gov-

ernments linked with nonproﬁt service agencies and for-proﬁt vendors of services. In the case of

nonproﬁts, they had been around for decades but the boundaries of the state expanded to

include them in various forms of externalized direct services delivery, a sort of government “for

hire” (Smith and Lipsky 1993: 5). For-proﬁts have always been part of government procurement,

and certain basic services like building security and road building were regularly contracted out,

but now such direct government services as public health care, services for the disabled, voca-

tional rehabilitation, mental health, substance abuse, and family violence, along with ﬁnance and

accounting services and other management functions, are contracted out. Just as the welfare state

philosophy once expanded the role and number of involved governments, beliefs in the primacy of

market forces, reduction and importance of the public sector, deregulation of state controls, and

abandonment of the principle of equality led to a prevailing political view of a more “minimalist”

state, with less direct government intervention in the economy and society (Loughlin 2007:

390). “Market superiority” that could either provide for the needs of people or a market-model

Agranoff et al.

362

of government services that could provide greater eﬃciencies than by public operation prevailed

as an attitude. This also was the era of the New Public Management (NPM), with heavy bor-

rowing from the private sector with its benchmarking, performance targeting, competitive

bidding, outsourcing, and the like, all of which reinforced IGR by government/NGO part-

nerships (Hood 1991). This led to new sets of alliances between governments at all levels and a

host of public and private bodies. Thus, “the public administration problem has spread well

beyond the borders of the government agency” (Salamon 1995: 2).

The fourth and currently developing IGR network era gradually emerged out of the pre-

vious era and became fully acknowledged in the ﬁrst decade of the twenty-ﬁrst century. It has

been dubbed a world where everything is connected in networks (Castells 1996). Public agen-

cies and NGOs now network for purposes of exchanging information, enhancing one another’s

capabilities, smoothing services interactions, and solving policy/program problems (Agranoﬀ

2007). In some ways, IGR networking began as a parallel activity to contracting, where funders

and their agents began to build contractor-government networks (Brown and Potoski 2004;

Van Slyke 2007), building on prior networks of local government, business associations, and

economic development agents who have worked among themselves at the community level for

some four or ﬁve decades, and these entities have had extensive links with higher level gov-

ernments in order to secure support to promote local economies (Agranoﬀ and McGuire 2003;

Eisinger 1988; McGuire 2002). The emergent set of intergovernmental networks—and what

makes them diﬀerent—is the way oﬃcials from the diﬀerent levels of governments and NGOs

representing the nonproﬁt sectors are challenged to sit down with one another at the same table

to deal with uncertainties and discuss, explore, negotiate, and solve issues (Koppenjan and Klijn

2004; Radin et al. 1996).

The network construct

While just one of a number of vehicles in the panorama of governance (Thompson 2003: 150,

237; Salamon 2002: 11), networks involve a cluster or complex of organizations connected to

each other by resource dependencies and are distinguished from other clusters or complexes by

breaks in the structure of resource dependencies (Rhodes and Marsh 1992). In order to broaden

the discourse on networks (Thompson 2003: 9) into a potential theory in this area, it is important

to understand “diﬀerentiation of governance capacity” when the potential role of networks is

considered “to construct a more nuanced understanding of the relative contributions which

these actions now make to the governance of society” (Peters 2000: 47, 50). This implies that

both network opportunities and barriers are equally important to understanding any outcomes

and process elements (Scharpf 1978: 346) in any set of “macro-level theories of the state”

(Rhodes and Marsh 1992: 203).

To focus on governance networks requires the drawing of parameters around use of the term

“network.” As Rhodes (2003: 21) suggests, “network” is an everyday term used by consumers

and managers (and perhaps politicians) to describe the web of relationships in which they are

embedded, giving rise to diﬀerent meanings that need to be captured. Here we focus on the

construction of meanings given by government managers and their interlocutors as they work

within organized entities that involve parts of organizations nonhierarchically (O’Toole 1997).

As such, we are not directly focusing on the broader issue of “democratic network governance,”

that is, cross-sectoral and based in civil society involvement where issues of deliberation and

citizen responsibility, as well as questions of equity, accountability, and democratic legitimacy to

serve public purposes are at stake (Bogason and Toonen 1998; Skelcher 2004; Sorenson and

Torﬁng 2007). Although we clearly realize that network management is a central part of

Governance, networks and intergovernmental systems

363

public network governance and impacts it in many ways (for example, it could limit participa-

tion and representation), we choose here to focus more directly on managerial arrangements

and behaviors that contribute to policy theory within larger issues of democratic participation

and governance. We thus use the term “network” to refer to structures involving multiple

nodes—agencies and organizations—with multiple linkages, ordinarily working on cross-boundary

collaborative activities.

Networks and their work

Some networks are constructs based more or less on forms of linkages and are intermittently

interactive and cooperative, that have been identiﬁed as serendipitous, whereas others are goal-

directed and are based on more regularized interactions (Kilduﬀ and Tsai 2003: 91), that is the

action of networking (Alter and Hage 1993). The latter can be either chartered or nonchartered

in character. Both types of networks share certain characteristics: permanent status; regular formal

meetings; a deﬁnable communication system; leaders and participants; taskforces or work groups;

identiﬁable operating structures; identiﬁable partners; and some form of division of labor/task

allocation. Chartered networks are formally established as organized entities. In the United States

they are often established by intergovernmental agreement, registration as a 501c(3) nonproﬁt

organization, by act or resolution of a state legislature, a governor’s executive order, and/or through

corporate registration with a state government representative, such as the secretary of state.

Nonchartered networks have no such formal-legal status, but their continuing presence and

operations, regular meetings, concrete problem-solving actions, websites, newsletters, and the like

are testimony to their existence. Many longstanding networks are non-chartered. Nonchartered

networks are often harder to locate in telephone directories or on websites than those that have

been formalized, but those without chartered status prove to be equally viable bodies.

Networks need to be understood in the context of how they operate. Five basic characteristics

of their functioning provide a context for understanding formation, operation, and value-adding.

First, not all networks are alike; whereas some exchange information, some build partner capacity,

some “blueprint” strategies and process interorganizational programming, and some make

policy/program adjustments. Some do more than one or more. Second, networks are non-

hierarchical but organized into collaborarchies that blend today’s conductive bureaucracies with

voluntary organization-like structures. The key is not the oﬃcial leaders but their champions,

vision-keepers, technical cores, and staﬀ. They are mainly organized around work and working

groups as communities of practice. Third, the most important function of these communities is

to discover, organize, and engage in knowledge management (KM). The KM process binds the

network as it approaches problems of a multi-organization, multijurisdictional orientation. Fourth,

networks are overlays on the hierarchies of participating organizations. They inﬂuence but do not

control home agency decisions. The core work of the public agency goes on but in an

increasingly conductive manner. Fifth, networks do add public value. To varying degrees they

help multiple organizations to engage in problem identiﬁcation and information exchange,

identify extant knowledge, adapt emergent technologies, engage in KM, build capacity, develop

joint strategies and programs, and adjust policies and programs (Agranoﬀ 2007).

Governance networks and public policy

The ﬁrst group of scholars to deal with the network concept in IGR were those who dealt with

policy implementation (Marsh and Smith 2000). As IGR expanded through welfare state pro-

grams and beyond the problem became, as Sundquist (1969: 12) observes, “how to achieve

Agranoff et al.

364

goals that are established by the national government, through the actions of other governments.”

In most cases, he concludes, governments choose to rely “upon systems of mutual adjust-

ment … rather than through the exercise of hierarchical authority” (1969: 19). Implementation

then began to focus on the interaction of actors down the line as they attempted to make

programs work (Ingram 1977; Pressman 1975; Pressman and Wildavsky 1973; Williams 1980).

The network of actors, particularly at the bottom rung of the chain of implementation, became

a focal point of research (Elmore 1985; Hjern and Porter 1981; Lynn 1981). In one of the early

studies of implementation by network, Scharpf concludes that:

it is unlikely, if not impossible, that public policy of any signiﬁcance could result from the

choice process of any single uniﬁed actor. Policy formulation and implementation are inevi-

tably the result of interactions among a plurality of separate actors with separate interests,

goals, and strategies.

(1978: 347)

As such, he concludes that research should go beyond the strategies of policy formulation and

implementation but include the more structured and stable relations between organizations.

Network formation

The policy implementation connection

The formation of networks is, in many cases, driven organically by the simple need to combine

resources in ways that facilitate the implementation of a program. Networks are also formed by

legislation or are mandated by others. Hall and O’Toole (2000) examined the US institutional

arrangements incorporated into the legislation enacted by the eighty-ninth and the one hundred

and third Congresses. They found that the majority of signiﬁcant new legislation prescribed the

involvement of collaborative structures for policy implementation. Their ﬁndings

demonstrate that the institutional settings in which public programs are placed, at least via

national legislation, are multiorganizational and networked rather than unitary and unam-

biguously hierarchical. Most new and amended programs encounter implementation set-

tings of the more complicated varieties—with intergovernmental and/or cross-sectoral

involvement more the rule than the exception. The arrays for program execution also

require signiﬁcant and skilled public management, particularly because the extent of interunit

coordination required for program success is often substantial … The networked reality of

today’s public management, furthermore, is important but clearly not unprecedented.

Subsequent research by the authors showed that post-legislation rule-making by implementing

federal agencies also led to collaborative administrative arrangements (Hall and O’Toole 2004).

Pressman and Wildavsky (1973) were among the ﬁrst to discuss public policy in terms of

shared administration, suggesting the networked nature of policy implementation. Based on an

empirical investigation of the US Economic Development Administration’s attempts in the

1960s to address unemployment of minorities in Oakland, California, their ﬁndings describe the

multiplicity of participants and perspectives from all levels of government pursuing policy goals

that in practice may be conﬂicting. The evidence from Europe is the same: collaborative

structures used for implementing manpower training in Germany and Sweden in the 1970s

were characterized at that time in terms of multiple power centers with reciprocal relationships,

Governance, networks and intergovernmental systems

365

many suppliers of resources, overlapping and dynamic divisions of labor, diﬀused responsibility

for actions, massive information exchanges among actors, and the need for information input

from all actors (Hanf et al. 1978). More than three decades ago, Hjern and Porter (1981)

described implementation structures that operate with representatives of diﬀerent agencies and

exercise considerable discretion in practice. Many policy studies in the 1980s revealed the extent

of collaboration in public policy implementation (Hull with Hjern 1987; Mandell 1984;

O’Toole 1985). There is thus a rich history of networked policy implementation.

Activating networks

Activating a network for the purposes of policy planning and/or implementation requires a

heavy dose of developmental activity on the part of some partners who act either as champions

or promoters (or “vision keepers”). Exclusivity or limited involvement leads to information and

support gaps, as well as lost potential in interagency adjustment, so network managers need to

know who has the critical policy-making resources: money, technology, information, expertise,

time, and other necessary commodities (Agranoﬀ 2003). Even if networks are formed through

mandated action, there is still the need to recruit constantly by enlisting necessary partners who

possess particular resources, expertise, and support for the network’s activities. All networks

depend on one or a small number of champions not just to help establish the network, but also,

perhaps most importantly, to maintain commitment among network partners.

Networks are based in at least four dimensions. First, all network structures have a technical

core, or knowledge about “how to do it” (Agranoﬀ 2007). For example, internal and external

expertise is a mainstream source of technical knowledge. Most participants will ask staﬀ scientists

or specialists inside and outside governments, along with university-based researchers, to share their

technological knowledge with the network. In environmental and natural resource networks,

engineering knowledge that deals with ﬂooding and ﬂoodplain concerns, water supply, water

quality, agricultural use, and recreation and wildlife management are all at the forefront of partici-

pating agency needs. In local economic development, city administrators, development directors,

and NGO development corporation executives work to revise development plans, requiring

expertise in engineering and planning; they negotiate ﬁnancing mechanisms with local business

groups, ﬁnancial institutions, and state and federal governments; and they handle issues of water

capacity and waste water with the state environmental agency. The network activities become a

technical basis of intergovernmental/interorganizational exchange.

Second, there is also a legal dimension that deﬁnes whether a particular networked action can

be undertaken. Laws governing each entity aﬀect the ability of any single organization to act.

An added overlay is the set of legal norms governing the management activity, whether it is law

determining the formal structure (e.g. task force, committee, coalition) or the general legal fra-

mework governing any project that a network might undertake. Control of operations may be

achieved through the use of government regulations that constrain the actions of network

partners, through the distribution of funds or manpower, or through regulations that stipulate

the powers and responsibilities of certain actors or specify what conditions potential actors must

meet before they can qualify to be members of a collaborative structure.

Third, ﬁnancial feasibility is another important dimension of networks. “Can we aﬀord it?” is

a central question in any public undertaking. Perhaps the most contentious, diﬃcult, but necessary

element of networking is pooling ﬁnancial resources. Such sharing or pooling is implicit in

Barbara Gray’sdeﬁnition of collaboration in which she asserts that it is the pooling of appreciations

and/or tangible resources, e.g., information, money, labor, and so on, by two or more stake-

holders to solve a set of problems which neither can solve individually (Gray 1985). However,

Agranoff et al.

366

there is no magic formula for determining the extent to which public sector organizations and

NGOs exchange, share, or pool resources.

The fourth important dimension of network management is political. The politics of activities

between and among organizations and their managers permeate the managerial process. Public

managers not only engage in politics within their own organizations, but also must strike political

(in addition to operational) bargains with other partners. In many contexts, network manage-

ment consists of conﬂict resolution using the important political skills of bargaining, negotiation,

diplomacy, and consensus building (Gray 1989; Ingram 1977).

Network operation

Building blocks

Network management behaviors are employed to help to frame the structure and the norms

and values of the network as a whole. Managers cannot draw up an organizational chart in a

network as is done in single organization structures, but they do try to inﬂuence the roles that

each participant may play at any given time and the perceptions one has about the common

purpose of the network. Managers do this by facilitating agreement on leadership roles; helping

to establish an identity and culture for the network; assisting in developing a working structure

for the network (e.g. committees, network “assignments”); and altering the perceptions of parti-

cipants to understand the unique characteristics of working with persons in contexts without

organizational mechanisms based in authority relations.

Seeking support from external groups and stakeholders for collaborative operations has proven

to be a major component of collaboration eﬀectiveness. Those within the collaboration must

foster the buy-in of key stakeholders, particularly those whose power will be relied upon to

signal support of the collaborative throughout their home agency. Managing the external net-

work environment thus is important for the collaborative public manager. Encouraging support

from and keeping a collaborative network in good standing with the higher governmental

authority ranked very high in an empirical study of network leadership (Silvia and McGuire 2010).

Similarly, encouraging support from and keeping the network in good standing with stakeholders

both inside and outside the government was viewed as being very signiﬁcant. Such behavior

helps to establish the legitimacy of the network (Milward and Provan 2006) and acts as a “mobi-

lizer” to develop commitment and support for network processes from network participants and

external stakeholders (McGuire 2002).

Managers must also employ behaviors intended to create an environment and enhance the

conditions for favorable, productive interaction among network participants. One critical behavior

of network management is to build relationships and interactions that result in achieving the

network purpose. The strategies of each network participant and the outcomes of those strategies

are inﬂuenced by the patterns of relations and interactions that have developed in the network.

Such behaviors include facilitating and furthering interaction among participants, reducing com-

plexity and uncertainty by promoting information exchange, and facilitating linkages among

participants. Successful network management will thus achieve cooperation between actors

while minimizing and removing informational and interactional blockages to the cooperation.

Klijn and Edelenbos (2008) draw a distinction between process design and management and

institutional design for describing various network management strategies. Process design involves

agreements regarding the nature of the interaction process and the ground rules for participating

in the network. The authors assert that there is no standard, all-encompassing design type for

networks; it is situational. Process management includes the direct, hands-on application of the

Governance, networks and intergovernmental systems

367

behaviors and tasks discussed above. Institutional design, on the other hand, is indirect. Such

design strategies are typically focused on changing the formal and informal rules that “inﬂuence,

guide, and limit” the behavior of the network participants (2008: 207). The strategies include

changing the network composition, inﬂuencing network outcomes, and aﬀecting network

interactions.

Learning and network management

Public managers in network settings face a number of both constraints and opportunities, most

of which exist due to the primacy of government in many such settings. Such actions are driven

by the knowledge that develops as implementers work through the process and learn what

works and what does not work. The public knowledge that is generated by the process is thus

informed by the outcomes of networks. Knowledge is that major step beyond information; it

provides the capacity to act (Sveiby 1997). It is the

ﬂuid mix of framed experience, values, contextual information, and expert insight that

provides a framework for evaluating and incorporating new experiences and information …

In organizations, it often becomes embedded not only in documents or repositories but

also in organizational routines, practices, processes, and norms.

(Davenport and Prusak 2000: 5)

The importance of learning together in networks to create knowledge is captured by the work

of Koppenjan and Klijn (2004). Joint action by interaction is seen in part as

searches wherein public and private parties from diﬀerent organization (levels of) govern-

ment and networks jointly learn about the nature of the problem, look at the possibility of

doing something about it, and look at the characteristics of the strategic and institutional

context within which the problem-solving develops.

(2004: 10)

Cooperation, then, presupposes structured learning between actors: “It requires numerous

skills, tacit knowledge of the network and negotiation skills since the adopted strategies are

implemented in a situation where singular hierarchical relations are lacking” (Koppenjan and

Klijn 2004: 11). Thus, learning in multi-actor situations is crucial.

Informal structures that develop around shared knowledge are namely communities of practice

and the epistemic communities that they often lead to. The former are self-organizing systems

that share the capacity to create and use knowledge through informal learning and mutual

engagement (Wenger 2000). Most communities are self-organized and bring in new knowledge

bearers when needed, from wherever they can be found. Maintenance of communities of

practice requires eﬀorts to keep diﬀerent types of knowledge bearers in, by challenging busy people

with solving important public problems, and by calling on their experience and know-how in an

interdisciplinary manner. Epistemic community can be facilitated by mobilizing a multi-agency

group of professionals from diﬀerent disciplines because they often share common outlooks and

similar solution orientations. They also tend to share causal beliefs, notions of validity, and a

common policy experience. An epistemic community normally produces consensual knowledge.

Epistemic communities also can be important knowledge sustainers, as they can have a dispropor-

tionate eﬀect on organized learning and behavior. Bringing together these communities for

enhanced deliberation is among the emergent networked tasks of public management.

Agranoff et al.

368

Governance networks add value

The ability to create or add value has long been a central concern of the public sector. As a result of

public managers attempting to “discover, deﬁne, and produce public value” (Moore 1995: 20),

the utilization of collaborative networks for public sector service delivery and policy implementa-

tion has increased notably over the last two decades. It is now quite common for governments

at all levels to rely upon other governments, private sector actors, and non-proﬁt organizations to

plan, implement, and manage “government” programs and services (Silvia 2011). However, as

Bardach (1998) concludes, the question is not whether networks exist, but rather, whether networks

increase the public value beyond which the agency would have in the absence of the network.

It has been argued that the frequency of the establishment of networks and the sustained

utilization of networked approaches in the public sector provides evidence that such organiza-

tional structures do add value (Cropper 1996; Howlett 2002; Thomson et al. 2008). Howlett

(2002) found that in four policy sectors networks matter in policy change as ideas and interest

are transmitted through policy subsystems. Nevertheless, collaboration is, after all, not easy. The

presences and gravity of collaborative challenges, including time costs, the logistical issues, varying

degrees of interorganizational and/or interpersonal trust, the integration of multiple organiza-

tional cultures, the sharing of control over resources, and the disparate goals and objectives held

by the diﬀerent collaborative partners, have caused some to dissuade parties from collaborating

unless it is necessary (Huxham and Vangen, 2005). However, the inability of single organizations

to solve the wicked problems they face often requires a collaborative approach. Thus, despite

the challenges posed by collaboration, networked arrangements for addressing shared problems

has “added important value to the public undertaking that undoubtedly would not have otherwise

occurred” (Agranoﬀ 2007: 4).

To the network administrators, the rationale for investment in the network entails more

than serving the collective public purpose, vaguely understood … but also includes certain

advantages the network can bring to their organization’s mission and functioning and to

the managers as professionals involved in public programs.

(Agranoﬀ and Yildiz 2007: 337)

Thus, this added value can be seen from the perspective of the individual collaborative partner, from

the perspective of the participating organization, as a result of the network process, and as a

result of network outcomes (Agranoﬀ 2007, 2008; Agrano ﬀ and Yildiz 2007).

From the perspective of the individual collaborative partner, value is created through a number

of diﬀerent mechanisms. In his studies of 14 public management networks, Agranoﬀ (2007,

2008) described ﬁve mechanisms through which individual collaborators beneﬁt from their

involvement in the network. The ﬁrst relates to the knowledge gained and the skills learned

through working with others. Second, networks are often interdisciplinary in nature. Thus, col-

laborators learn how to work with and learn from those from diﬀerent organizations, back-

grounds, and functional areas. In addition, the opportunity to work with and/or manage individuals

from diﬀerent organizations enhances the ability to manage and work within intergovernmental

arenas is a third advantage. The fourth area for potential beneﬁt for the individual is the increased

networking opportunities. Working with representatives of other organizations allows network

members to build relationships with and gain access to other entities and individuals. Finally,

individuals perceive that networks add value to them personally as networked governance struc-

tures strengthen their public service motivation and provide them with additional opportunities

for public service.

Governance, networks and intergovernmental systems

369

Policy design and implementation via a networked structure also provides a number of

advantages to the organizations participating in the joint venture. The value added is realized by

the participating organizations as they attempt to achieve collaborative advantage, or the synergy

created as organizations collaborate (Huxham 1996; Huxham and Vangen 2005). Collaboration

has been conceived of as a “strategy for increasing the value that a single organization is able to

create on its own” (Kim 2010: 112) in part because such arrangements expand the capabilities

and capacity of the individual organizations involved. By working together, organizations are

able to increase their capacity to serve their constituents by accessing their collaborative partners’

knowledge, information, connections, personnel, material, and monetary resources. In addition,

the participating organizations also realize the value added via collaboration as risk is shared

among all of the partners, as eﬃciencies are realized as networks are able to take advantage of

economies of scale and sectoral diﬀerences, as the organizations are able to learn from each

other, and as the delivery of services is coordinated (Cropper 1996; Huxham and Vangen 2005).

There is also value added as a result of the collaborative process. While not realized by any one

person or organization, the beneﬁts of the collaborative process are realized by the network as a

whole and, more importantly, often felt by those for whom the network is intended to serve. If

the “value” that the network is trying to create is the public value, then perhaps the diﬀerent

perspectives voiced while collaborating can help to determine what the public value is. This value

creation may in essence be the product of the interorganizational, intersectoral, and interideological

conversation aimed at problem identiﬁcation and resolution between government agencies and

its interlocutors. As Innes and Booher remind us, “a process that is inclusive, well informed, and

comes close to achieving consensus is more likely to produce an implementable proposal than one

lacking these qualities” (1999: 420). In the absence of the network, a siloed approach to value

creation occurs, whereby individual organizations view the problem from their own, isolated per-

spective. Networked approaches to service provision and program implementation, on the other

hand, can result in a process whereby the agencies involved engage in dialogues that they otherwise

would not have, approach the issues from a less partisan perspective, create a synergy where

individual members exchange ideas, and build communal knowledge to work toward a solution.

The added value resulting in improved collaborative outcomes is, generally speaking, the

driver behind the formation of collaborative networks. Bardach (1998: 8) deﬁ ned collaboration

as any joint activity by multiple entities “that is intended to increase public value by their

working together rather than separately.” Here, this public value refers to the outcomes pro-

duced by the collaborative eﬀorts. Networks, in the course of addressing the problems for

which they were formed to solve, generate collaboratively produced products, such as reports,

publications, conferences, training programs, datasets, and so on, that can be used not only by

the networks and its constituent members, but also by others outside the network and even

outside the area served by the network. Collaboration within a networked governance structure

also commonly results in changes to old programs and the formation of new ones. These jointly

constructed and implemented service delivery programs are often more holistic and compre-

hensive in their approach. In the process of working with each other on the issue that brought

them together, networks also often engage in spin-oﬀ projects as discussion between network

members bring previously unaddressed or unrealized issues to the forefront.

Conclusion

Networks are notable information age responses to the most diﬃcult of policy and adminis-

trative problems but they are by no means replacing the authority of government as they share

involvement with NGOs. First, while networks and the NGOs work with public agencies,

Agranoff et al.

370

inﬂuence policy and have a role in management, those organizations and agencies that comprise

these networks work with governments as the latter continue to maintain some important abil-

ities that guide public action. The agencies retain their authority under law. Second, while an

interactive interdependency has emerged within networks crossing many boundaries, public agency-

NGO organization connections seem to overlay the hierarchy rather than act as replacements

for government action. Networks are not the exclusive mode of collaborative government-

nongovernmental relations. Many of the more standard tools of government—for example, grants,

loans, regulatory programs, insurance guarantees, and cooperative agreements—tie governments

together and public agencies to NGOs without employing networks. Finally, not all govern-

mental agency administrators are totally or to a high degree in the business of working in net-

works. With the exception of certain “boundary spanners” who spend full-time in cross-agency

work, many administrators spend a small proportion of their total work time in collaborative

activity, including participating in networks.

Networks have not eclipsed or displaced the power or centrality of government agencies. The

empirical evidence is too mixed to support such a contention. Thus, despite the very broad range

of network activity, in the United States it was found that the “ability [of nongovernmental

actors] to inﬂuence the public agency domain is real but quite limited in scope … accommodations

are made, decisions are inﬂuenced, strategies are altered, resources are directed, intensive groups

exert undue inﬂuence, and public responsibility is indirectly shared (Agranoﬀ 2007: 219). The

reality is that neither networks (or the for-proﬁt and nonproﬁt partners within them) nor gov-

ernment agencies dominate, and in many cases the government agencies prove to be the key

leaders within the networks.

Bibliography

Agranoﬀ, Robert. 1986. Intergovernmental Management: Human Services Problem-Solving in Six Metropolitan

Areas. Albany, NY: SUNY Press.

—— 2003. Leveraging Networks: A Guide for Public Managers Working across Organizations. Arlington, VA:

IBM Endowment for The Business of Government.

—— 2007. Managing Within Networks: Adding Value to Public Organizations. Washington, DC: Georgetown

University Press.

—— 2008. ‘Enhancing performance through public sector networks: Mobilizing human capital in

communities of practice’, Public Performance & Management Review 31(3): 320–47.

—— 2010. ‘Towards an emergent theory of governance within IGR’, in Edoardo Ongaro, Andrew

Massey, Marc Holzer, and Ellen Wayenberg (eds) Governance and Intergovernmental Relations in the European

Union and the United States. Cheltenham: Edward Elgar.

——and Michael McGuire. 2003. Collaborative Public Management: New Strategies for Local Governments.

Washington, DC: Georgetown University Press.

——and Mete Yildiz. 2007. ‘Decision making in public management networks’, in Goktug Morcol (ed.)

Handbook of Decision Making. New York: CRC, Taylor & Francis, 319–45.

Alter, Catherine and Jerald Hage. 1993. Organizations Working Together. Beverly Hills, CA: Sage.

Ashford, Douglas E. 1988. ‘Decentralizing welfare states: Social policies and intergovernmental policies’,

in Bruno Dente and Francesco Kjellberg (eds) The Dynamics of Institutional Change: Local Government

Reorganization in Western Democracies. London: Sage.

Bardach, Eugene. 1998. Getting Agencies to Work Together. Washington, DC: Brookings.

Bogason, Peter and Theo A.J. Toonen. 1998. ‘Networks in public administration’, Public Administration

76(Summer): 205–27.

Brown, Trevor L. and Matthew Potoski. 2004. ‘Managing the public service market’, Public Administration

Review 64(6): 656–68.

Castells, Manuel. 1996. The Rise of the Network Society. Oxford: Blackwell.

Cropper, S. 1996. ‘Collaborative working and the issue of sustainability’, in C. Huxham (ed.) Creating

Collaborative Advantage. London: Sage.

Governance, networks and intergovernmental systems

371

Davenport, Thomas H. and Laurance Prusak. 2000. Working Knowledge: How Organizations Manage what

they Know. Cambridge, MA: Harvard Business School Press.

Eisinger, Peter K. 1988. The Rise of the Entrepreneurial State. Madison, WI: University of Wisconsin Press.

Elazar, Daniel J. 1962. The American Partnership: Intergovernmental Cooperation in the Nineteenth Century

United States. Chicago: University of Chicago Press.

Elmore, Richard F. 1985. ‘Forward and backward mapping: Reversible logic in the analysis of public

policy’, in Kenneth Hanf and Theo A.J. Toonen (eds) Policy Implementation in Federal and Unitary Systems.

Dordrecht: Martinus Nijhoﬀ Publishers.

Gray, Barbara. 1985. ‘Conditions facilitating interorganizational collaboration’, Human Relations 38: 911–36.

—— 1989. Collaborating: Finding Common Ground for Multiparty Problems. San Francisco: Jossey-Bass.

Hall, Thad E. and Laurence J. O’Toole. 2000. ‘Structures for policy implementation: An analysis of

national legislation, 1965–66 and 1993–94’, Administration and Society 31(6): 667–86.

—— 2004. ‘Shaping formal networks through the regulatory process’, Administration and Society 36(2):

186–207.

Hanf, Kenneth, Benny Hjern and David O. Porter. 1978. ‘Local networks of manpower training in the

Federal Republic of Germany and Sweden’, in Kenneth Hanf and Fritz W. Scharpf (eds) Interorganizational

Policy Making: Limits to Coordination and Central Control. London: Sage.

Hjern, Benny and David O. Porter. 1981. ‘Implementation structures: A new unit for administrative

analysis’, Organizational Studies 3: 211–37.

Hood, Christopher. 1991. ‘A public management for all seasons?’, Public Administration 69(1): 3–19.

Howlett, Michael. 2002. ‘Do networks matter? Linking policy network structure to policy outcomes:

evidence from four Canadian policy sectors 1990–2000’, Canadian Journal of Political Science 35(2): 235–67.

Hull,

Christopher J. with

Benny Hjern. 1987. Helping Small Firms Grow: An Implementation Approach.

London: Croom Helm.

Huxham, Chris. 1996. ‘Collaboration and collaborative advantage’, in C. Huxham (ed.) Creating Collaborative

Advantage. Thousand Oaks, CA: Sage, 1–18.

——and Siv Vangen. 2005. Managing to Collaborate: The Theory and Practice of Collaborative Advantage. New

York: Routledge.

Ingram, Helen. 1977. ‘Policy implementation through bargaining: The case of federal grants-in-aid ’, Public

Policy 25(4): 499–526.

Innes, Judith E. and David E. Booher. 1999. ‘Consensus building and complex adaptive systems: A framework

for evaluating collaborative planning’, Journal of the American Planning Association 65(4), 412–23.

Kilduﬀ, Martin and Wenpin Tsai. 2003. Social Networks and Organizations. Los Angeles, CA: Sage.

Kim, S. 2010. ‘Collaborative leadership and local governance’,inR.O’Leary, D.M. Van Slyke and S. Kim

(eds) The Future of Public Administration Around the World: The Minnowbrook Prespective . Washington, DC:

Georgetown University Press, 111–15.

Klijn, Erik-Hans and Jurian Edelenbos. 2008. ‘Meta-governance as network management’, in Eva Sor-

ensen and Jacob Torﬁng (eds) Theories of Democratic Network Governance. London: Palgrave Macmillan,

199–214.

Kooiman, Jan. 2003. Governing as Governance. London: Sage.

Koppenjan, Joop F.M. and Erik-Hans Klijn. 2004. Managing Uncertainties in Networks. London: Routledge.

Krane, Dale and Deil S. Wright. 1998. ‘Intergovernmental relations’,inInternational Encyclopedia of Public

Policy and Administration. Boulder, CO: Westview Press, 1168 – 70.

Loughlin, John. 2007. ‘Reconﬁguring the state: Trends in territorial governance in European States’,

Regional and Federal Studies 17(4): 385–404.

Lynn, Lawrence E. 1981. Managing the Public’s Business: The Job of the Government Executive .

New York:

Basic Books.

McGuire,

Michael. 2002. ‘Managing networks: Propositions on what managers do and why they do it’,

Public Administration Review 62(5): 426–33.

——and Robert Agranoﬀ. 2010. ‘Networking in the shadow of bureaucracy’, in Robert F. Durant (ed.)

The Oxford Handbook of American Bureaucracy. New York: Oxford University Press.

Mandell, Myrna P. 1984. ‘Application of network analysis to the implementation of a complex project’,

Human Relations 37(8): 659–79.

Marsh, David and Martin Smith. 2000. ‘Understanding policy networks: Towards a dialectical approach’,

Political Studies 48(1): 4–21.

Milward, H. Brinton and Keith G. Provan. 2006. A Manager’s Guide to Choosing and Using Collaborative

Networks. Arlington, VA: IBM Center for the Business of Government.

Agranoff et al.

372

Moore, Mark H. 1995. Creating Public Value: Strategic Management in Government. Cambridge, MA: Harvard

University Press.

O’Toole, Laurence J. 1985. ‘Diﬀusion of responsibility: An interorganizational analysis’, in Kenneth Hanf

and Theo. A.J. Toonen (eds) Policy Implementation in Federal and Unitary Systems . Dordrecht: Martinus

Nijhoﬀ, 201–25.

—— 1997. ‘Treating networks seriously: Practical and research-based agenda in public administration’,

Public Administration Review 57(1): 45–52.

Peters, B. Guy. 2000. ‘Policy instruments and public management: Bridging the gaps’, Journal of Public

Administration Research and Theory: J-PART 10(1): 35 – 47.

—— 2002. ‘Globalization, institutions, governance’, in B. Guy Peters and Donald J. Savoie (eds) Governance

in the Twenty-First Century: Revitalizing the Public Service. Montreal: McGill-Queen’s University Press.

Pressman, Jeﬀrey L. 1975. Federal Programs and City Politics: The Dynamics of the Aid Process in Oakland.

Berkeley, CA: University of California Press.

——L. and Aaron Wildavsky. 1973. Implementation. Berkeley, CA: University of California Press.

Radin, Beryl A., Robert Agranoﬀ, C. Gregory Buntz, Ann O’M. Bowman, Barbara Romzek and Robert

Wilson. 1996. New Governance for Rural America: Creating Intergovernmental Partnerships. Lawrence, KS:

University of Kansas Press.

Rhodes, R.A.W. 1997. Understanding Governance: Policy Networks, Governance, Reﬂexivity and Accountability.

Buckingham: Open University Press.

—— 2003. ‘Putting people back into networks’, in Arie Salminen (ed.) Governing Networks. Amsterdam:

IOS Press.

——and David Marsh. 1992. ‘New directions in the study of policy networks’, European Journal of Political

Research 21(2): 181–205.

Saint-Onge, Hubert and Charles Armstrong. 2004. The Conductive Organization. Amsterdam: Elsevier.

Salamon, Lester M. 1995. Partners in Public Service.

Baltimore, MD: Johns

Hopkins University Press.

—— 2002. ‘The new governance and the tools of public action’, in Lester M. Salamon (ed.) The Tools of

Government. New York: Oxford.

Scharpf, Fritz W. 1978. ‘Interorganizational policy studies: Issues, concepts, and perspectives’, in Kenneth

Hanf and Fritz W. Scharpf (eds) Interorganizational Policy Making. London: Sage.

Silvia, Chris 2011. ‘Collaborative governance concepts for successful network leadership’, State and Local

Government Review 43(1), 66–71.

——and Michael McGuire. 2010. ‘Leading public sector networks: An empirical examination of integrative

leadership behaviors’, The Leadership Quarterly 21(2): 264–77.

Skelcher, Chris. 2004. ‘The new governance of communities’, in Gerry Stoker and David Wilson (eds)

British Local Government into the 21st Century. Basingstoke: Palgrave Macmillan.

Skowronek, Stephen. 1982. Building a New American State: The Expansion of National Administrative Capacities,

1877–1920. Cambridge: Cambridge University Press.

Smith, Steven R. and Michael Lipsky. 1993. Nonproﬁts for Hire: The Welfare State in the Age of Contracting.

Cambridge, MA: Harvard University Press.

Sorenson, Eva and Jacob Torﬁng. 2007. ‘Theoretical approaches to democratic network governance’, in Eva

Sorenson and Jacob Torﬁng (eds) Theories of Democratic Network Governance. London: Palgrave Macmillan.

Sundquist, James L. 1969. Making Federalism Work . Washington, DC: Brookings.

Sveiby, K.-E. 1997. The New Organisational Wealth: Managing and Measuring Knowledge-Based Assets. San

Francisco: Berrett-Koehler.

Thompson, Grahame F. 2003. Between Markets and Hierarchies: The Logic and Limits of Network Forms of

Organization. Oxford: Oxford University Press.

Thomson, A.M., James L. Perry, and T.K. Miller 2008. ‘Linking collaboration processes and outcomes’,in

L.B. Bingham and R. O’Leary (eds) Big Ideas in Collaborative Public Management. Amonk, NY: M.E.

Sharpe, 97–120.

Van Slyke, David. 2007. ‘Agents or stewards: Government nonproﬁt social service contracting relationships’,

Journal

of Public Administration

Research and Theory 17(2): 157–87.

Watts, Ronald L. 1999. Comparing Federal Systems in the 1990s, 2nd edn. Kingston: Institute of Inter-

governmental Relations, Queen’s University.

Wenger, E. 2000. ‘Communities of practice: The key to knowledge strategy’, in E.L. Lesser, M. A. Fontaine

and J.A. Slusher (eds) Knowledge and Communities. Boston: Butterworth-Heinemann.

Williams, Walter W. 1980. Government by Agency: Administering Grants-in-Aid Programs. New York: Academic

Press.

Governance, networks and intergovernmental systems

373

Development management and

policy implementation

Relevance beyond the global South

Derick W. Brinkerhoff and Jennifer M. Brinkerhoff

As a subset of the discipline of public administration, development management (also referred to

as development administration) has traditionally concentrated on the organizational and managerial

problems confronting the countries of the global South. Development management has been

long associated with international foreign assistance, and debates regarding what it is, how to do

it, and what it achieves are frequently embedded within larger arguments about foreign policy,

development assistance, and power imbalances between the global North and South. Reﬂecting

this association, development management is often distinguished as concerning the implementation

of donor-funded policies, programs, and projects intended to promote socio-economic devel-

opment in low-income countries. Development management has been swept by the same tides

that have shaped foreign assistance over the decades since the post-World War II era, and that have

revisited notions of development and what is needed to make it happen (Brinkerhoﬀ 2008).

Some analysts distinguish between management of development—referring to the manage-

ment of the “mechanics” of foreign aid procedures, programs, and projects—and management

for development, which addresses the underlying substance of development as a transforma-

tional process that improves the well-being of citizens in poor countries (see Thomas 1999). Yet

this distinction can be diﬃcult to maintain. Donor countries pursue a wide variety of foreign

policy objectives under the broad rubric of foreign assistance. The publicly espoused aim of most

donor foreign assistance is to contribute to development in recipient countries. Thus, oﬃcial

development policy statements reﬂect an assumption that management of the development

projects and programs that are designed to implement those policies is tantamount to, or at least

aspires to be, management for development. As numerous critical studies of foreign assistance

reveal, however, donor countries use foreign aid in the service of political, diplomatic, economic,

humanitarian, and security aims that may sideline or subvert development. Further, the mod-

alities of delivering foreign assistance are accused of contributing to these perverse impacts (see

Easterly 2006; Picard and Buss 2009). Radical critiques of development management tar the

discipline with the same antidevelopment brush (see Dar and Cooke 2008).

This chapter opens by brieﬂy reviewing the range of perspectives on development management.

The discussion is framed around our model of development management, which deﬁnes the

374

discipline in terms of institutional agendas, tools, values, and processes (Brinkerhoﬀ and Brinkerhoﬀ

2006). We explicitly recognize the complexities, trade-oﬀs, and tensions inherent in the foreign

assistance policy environment where many development managers operate. We then turn to policy

implementation and examine what the development management discipline has to say about

how managers can transform policy reform intent into outcomes.

Perspectives on development management

Deﬁning development management remains problematic for several reasons. We have already

noted the diﬃculty of separating it from its foreign assistance policy environment. Second is the

question of what development is. Third is what, then, is being managed. Fourth is the issue of

what we mean by management. The answers to these questions have occupied theorists and

practitioners for decades, so we can only touch on them in the most cursory manner here.

Regarding the development question, the predominant view is informed by modernization

theory, which (to oversimplify) posits that development is a process that moves countries and

their citizens progressively along a path to Western-style liberal economies and democratic

governance. Modernization involves societal transformations from rural to urban, ascriptive/

personalistic roles to merit/position-based ones, impoverished to wealthier, low levels of health

and education to higher ones, and so on. Besides these material advances (“having more”),

development is widely seen as “being more ”; that is, enhancing capacities, fulﬁlling potentials,

expanding freedoms, and exercising rights (e.g. Sen 1999). In a much-cited shorthand, the end-

state toward which this modernizing development path leads is “Denmark,” that is, a stable,

prosperous, and well-managed liberal democracy (Pritchett and Woolcock 2004).

As the dichotomy between developed and developing countries has become less distinct, and

as a middle range of countries such as China, India, and Brazil has emerged, the modernization

paradigm has been increasingly recognized as an artifact of outdated thinking. There are mul-

tiple paths to development, and issues of politics—contestation for power and patronage among

elites—and of identity—how people deﬁne themselves relative to language, religion, family,

and culture—ﬁgure strongly in determining what those paths are and where they lead. The innate

superiority of Western-style modernization has been increasingly questioned, particularly in light

of the global economic crisis, climate change, and the perceived failure of the United States and

its allies to achieve quick success in rebuilding Iraq and Afghanistan.

As for the third question, whether these twin transformations of “having and being more”

can be managed depends upon the conception of change that underlies them. Donor approaches

to development, implicitly and explicitly, reﬂect the idea that societal change is both amenable

to planning and intentional direction, and can be usefully guided and supported by external actors.

The alternative view is that change is historically and institutionally embedded in a given society’s

individual trajectory, and is in essence emergent and not easily subject to manipulation. Man-

agement is only possible, at best, at the margins, and is not accessible to outsiders except in the most

superﬁcial ways. The experience of the international community in fragile and post-conﬂict

states, such as Liberia, Afghanistan, and Iraq, has provided traction for this standpoint.

Regarding what is management in the international development context, in line with the

modernization paradigm, early answers focused on the technical tasks associated with Weberian

bureaucracies; for example, the classic public administration tasks encapsulated in the acronym,

POSDCORB (planning, organizing, staﬃng, directing, coordinating, reporting, and budgeting),

coined in 1937 by Luther Gulick and Lyndall Urwick, and on designing and operating the

organizational systems to enable managers to carry them out (see the overview in Esman 1991).

Later conceptions of management were, and remain, heavily inﬂuenced by the New Public

Development management and policy implementation

375

Management (NPM), again drawing on models from industrialized countries and transferring

them to developing societies (see Brinkerhoﬀ 2008). In the donor-funded international develop-

ment arena, the tools and techniques of data gathering, project/program design (e.g. the Logical

Framework), and monitoring and evaluation are well recognized deﬁning features of management.

With the replacement of modernization’s single path to development by what Woolcock

(2009) terms “multiple modernities” and the focus on development as an indigenous, emergent

phenomenon, management has expanded beyond administrative technologies and routines, to

address reform, change processes, participation, learning, and local adaptation. Fowler (1997),

writing about nongovernmental organizations engaged in international development, talks about

management as “striking a balance” between the demands of donor-driven technocratic pro-

cedures and requirements, and the rights and needs of the poor and marginalized. In another

example of an expanded conception of management, Andrews et al. (2010) consider leadership

for change management as paramount, and see the leader’s role as creating space where change

can happen.

A multidimensional deﬁnition of development management

As the above brief review demonstrates, a single, agreed-upon deﬁnition of development man-

agement would be diﬃcult to come up with, and in fact, this is the case. Previously, we have

sought to clarify the meaning of development management by characterizing it as encompassing

four dimensions: a means to institutional agendas; a set of values; a planning and management

toolkit; and processes that both support the agency of development actors and reconcile competing

interests (Brinkerhoﬀ and Brinkerhoﬀ 2006).

A means to institutional agendas

While donor agendas may predominate, due to their associated funding, they exist alongside

those of national and local public oﬃcials, citizen groups, and private sector actors. Each of these

interest groups may apply development management tools and processes to enact their particular

values and further their particular agendas. Thus development management is not restricted solely

to donor-designed and -funded initiatives. This is a particularly important observation when

considering how development management relates to policy implementation.

A set of values

With the expanded notions of what constitutes development and the inclusion of identity in its

deﬁnition, the values dimension of development management becomes especially salient. Values

associated with development management include empowerment, participation, a focus on the

poor and disadvantaged, self-determination (commonly operationalized as “country ownership”),

and, more recently, sustainability. Responsiveness to the needs and desires of recipient country

governments and their citizens also requires respect for and incorporation of national values,

culture, and identity. These increasingly overlap with aspirations encompassed in liberal values

such as human rights and basic freedoms (e.g. speech, assembly, religion).

Planning and management toolkit

Development management as a toolkit incorporates all of the basic tools supportive of public policy

and public administration, as well as those more speciﬁcally designed to support development

Brinkerhoff and Brinkerhoff

376

processes. These include analytic tools drawn from political science and human behavior, as well

as management tools deriving from strategic management, budgeting, and evaluation. Tools, in

and of themselves, are value-neutral instruments. Through their application in the service of

particular institutional agendas and values, however, they lose their neutrality.

Processes

The results of the application of management tools depend largely on the processes through

which they are applied. That is, tools may be selected and modiﬁed to support particular agendas

and values, but it is in their application that values, in particular, will be made manifest. Thus,

the process dimension of development management self-consciously considers the politics of

competing values and agendas and power diﬀerentials. The process dimension is speciﬁcally struc-

tured to bring into consideration the desires, needs, and values of those aﬀected by development

plans, programs, and investments.

Development management debates

Because of the complexities in deﬁning development, management, and development man-

agement, there are competing perspectives on the discipline. Our multidimensional framework

recognizes these complexities, and seeks to capture the tensions and disconnects among them.

The key disconnect is between development management as a means to implement institutional

agendas and the other three dimensions. Our view is that through a commitment to pro-poor

and empowerment values, and the application of participatory processes and interest mediation,

this disconnect does not necessarily undermine the possibilities for achieving positive policy out-

comes. Development management can serve the aims and intents of reformers and activists. Donor

practices and procedures are not always the straitjacket that critics perceive them to be

(Brinkerhoﬀ and Ingle 1989).

This view places us in the reformist camp, which recognizes the political realities of foreign

assistance, globalization, and national politics, and the limitations of externally supported socio-

economic change. The reformist outlook critiques development management in practice, while

still believing in its actual and potential contributions to desired development outcomes. The

radicalist/rejectionist camp, on the other hand, holds that development management is so inti-

mately connected with neocolonialist power relations and “managerialist” control that actions

taken by development managers are invariably detrimental to the interests of developing

countries, and particularly of the poor residing in those countries (see Gulrajani 2010; Mowles

2010). Table 28.1 illustrates these two competing perspectives on development management

using our four development management dimensions.

For each of the dimensions of development management, the radicalist/rejectionist perspec-

tive focuses on their oppressive nature, and their contributions to reinforcing political and

bureaucratic power imbalances and inequalities. The history of foreign assistance conﬁrms that

“bad things” have indeed been done in the name of international development, and that tech-

nocratic managerial hubris has been a contributing factor. We argue, however, that the con-

clusion to be drawn is not to throw the management “baby” out with the development

“bathwater,” but precisely to apply development management in the service of those actors

who seek to confront the ills identiﬁed by the radicalist/rejectionist ﬁnger pointers.

We are not alone in making this argument. The results of a survey of development man-

agement scholars and practitioners reinforces the activist orientation in favor of empowering

local actors to pursue country-led development strategies and objectives, and of seeking to

Development management and policy implementation

377

inﬂuence donor agency practices and procedures to correct power imbalances (Brinkerhoﬀ and

Brinkerhoﬀ 2010). Gulrajani (2010) elaborates the contours of “non-managerialist” develop-

ment management, and both Mowles (2010) and Abbott et al. (2007) stress the role of joint

learning, values, negotiation, and adaptation in dealing with the power imbalances inherent in

development and its emergent, unpredictable nature.

We see some encouraging signs of changes. For example, current foreign assistance policies

(e.g. the Organisation for Economic Co-operation and Development’s Paris Principles that stress

country ownership and alignment with indigenous systems) and practices (e.g. cash-on-delivery

aid, and budget transfers) are reshaping the management environment for development, placing

more emphasis on the role of country actors as development managers. While the Paris Principles

have been criticized in some quarters as largely rhetoric, nonetheless there have been some

operational changes on the ground in response to them. For example, in Liberia’s health sector,

the health minister has gotten buy-in from a range of international donors for his sector plan,

and the donors have committed to a pooled funding arrangement where the minister retains

control over the funds in the implementation of the programs that they support.

Development management and policy implementation

The evolutionary trajectory of scholarship and practice related to the policy process and policy

implementation in the global South closely tracks that of development management. The resulting

conﬂuence provides the rationale for development management’s relevance in implementing

policy. Three parallels are pertinent to the discussion here. First is the source of policy concepts,

models, analytic tools, and practical advice. The policy literature has originated largely in indus-

trialized countries, principally the US and the UK, which raises the same transferability issue

that development management has debated. The evolution in both literatures has been a gra-

dual recognition of the fallacies of institutional isomorphism and of the need for contextual

understanding and adaptation to individual circumstances.

Second, very similar to thinking about development planning, where the transformation of

plans into actions and results was given short shrift, the classic policy literature’s stages heuristic

treated implementation as (a) a separate step in the policy process, and (b) a “black box” that

Table 28.1 Competing perspectives on development management

Development Management

Dimensions

Reformist perspective Radicalist/rejectionist perspective

Institutional agenda Equitable growth, MDGs, good

governance, rights-based

approaches, enhanced

capacities

Neo-colonialist replication of

Northern dominance over the global

South

Values Bottom-up empowerment,

pro-poor, pluralism, country

ownership

Top-down imposition of Western

paradigms and universal

performance metrics

Tools Responsive planning and

management for efﬁciency and

effectiveness

Technocratic, managerialist

instruments of control

Process Participation, agency, social

accountability

Co-optation, lip-service to disguise

power imbalances

Source: Authors

Brinkerhoff and Brinkerhoff

378

faithfully translates policy objectives into outcomes. Pressman and Wildavsky’s classic book (1973)

started to unpack the black box, and helped to launch the US policy implementation literature.

Within a few years, international donors discovered and debated the implementation and

sustainability gaps in development projects and programs.

The third parallel is the eventual acknowledgment that, while the technical content of policies is

important to achieving desired policy outcomes, their implementation is strongly inﬂuenced by

bureaucratic and political factors. Having the right policy is not suﬃcient, and in fact what may be

technically optimal (“right’) is in many cases likely to be bureaucratically and/or politically

infeasible (e.g. Mazmanian and Sabatier 1989). This realization gave rise to the literature on

policy advocacy, coalitions, and networks. In the international development arena, the mismatch

between technical and politico-bureaucratic factors became increasingly clear during the period

beginning in the 1980s when the World Bank and the International Monetary Fund were pursuing

structural adjustment policies and encountering problems. Development management scholars

and practitioners focused on these factors as well (Grindle and Thomas 1991; Brinkerhoﬀ and

Crosby 2002; Brinkerhoﬀ 2008).

Framing international development policy implementation

The focus on the policy implementation gap revealed some important features that distinguish

implementing policies from project and program management. The following discussion draws

upon Brinkerhoﬀ and Crosby (2002), whose book encapsulates the ﬁndings of an extensive

applied research eﬀort on policy implementation in developing countries.

 Policy implementation is not a linear, coherent process. With policy implementation, change is often

multidirectional, fragmented, frequently interrupted, and unpredictable. How to sequence

actions, what to pay attention to, and who to include can be hard to determine, and can vary

over the life of the policy change process.

 Most policies require the concerted actions of multiple agencies and groups. Even if one of them is

nominally the lead agency, in reality no individual entity is “in charge” of policy implementation.

Authority and responsibility are dispersed among the actors involved.

 Policy implementation is often highly political. Policies usually involve the imposition of costs on

some societal groups as well oﬀering advantages to others, which creates winners and losers, and

makes reforms highly contentious. Current policy equilibriums favor the powerful that beneﬁt

from them; these actors are usually well positioned to defend the status quo and resist change.

 The resources required to implement policies may not be readily available. Projects and programs

have dedicated budgets, but policies—particularly at the start of a reform—often lack the

resources needed for implementation. Making progress means lobbying for new funding,

identifying existing sources of implementation support, and negotiating for resource reallocation.

In response to these identiﬁed characteristics of policy implementation and building on the

policy implementation literature, plus action research and technical assistance in approximately

30 countries, Brinkerhoﬀ and Crosby (2002) developed a framework that parses implementation

into six roughly sequential tasks. These are posited as an iterative succession through which

managers need to cycle repeatedly during the life of a given reform:

1 Creating legitimacy, getting the policy accepted as important, desirable, and worth achiev-

ing, is the ﬁrst implementation task. Legitimization means getting buy-in from the right

people in the country to push the reform process forward. An important outcome of this

Development management and policy implementation

379

task is the emergence of a policy champion (an individual or group who believes in the

policy) to take on leadership for the subsequent implementation tasks.

2 Building constituencies, or gaining active support for a proposed policy from groups that see

the reform as desirable or beneﬁcial, is the second implementation task. This support needs to

translate into commitment to take actions that will help to achieve the policy’s objectives. As

well as establishing coalitions of supporters, constituency building seeks to reduce or deﬂect

the opposition of groups that see the proposed reform as harmful or threatening. This task must

often be pursued throughout implementation to assure ongoing support and to resist derailment.

3 Accumulating resources means ensuring that present and future budgets and human resource

allocations are suﬃcient to support implementation requirements. Accomplishing this task

can involve, for example: lobbying constituencies to contribute resources, negotiating with

ministries for budget line-item funding, designing new resource allocation systems, and/or

setting up public-private partnerships.

4 Modifying organizational structures is the fourth policy implementation task. This entails

adjusting the objectives, procedures, systems, and structures of the agencies responsible for

policy implementation. Sometimes this task can also include establishing new organizations

to coordinate the various entities with a role in implementation.

5 Mobilizing resources and actions builds upon the favorable constituencies assembled for the

policy (Task 2) and marshals their commitment and resources (Task 3) to engage in concrete

eﬀorts to make change happen. Its focus is on identifying, activating, and pursuing action

strategies. It brings together mobilized constituencies and resources, and within the orga-

nizational structures created, develops and carries out the steps necessary to translate intent

into results.

6 Monitoring impact, or setting up systems to monitor implementation progress, is the ﬁnal

policy implementation task. Monitoring systems not only alert decision-makers to imple-

mentation snags, but also inform them of the intended and unintended impacts of

implementation eﬀorts.

Managing policy implementation

Brinkerhoﬀ and Crosby argue that each of these tasks calls for explicit managerial attention, and

that failure to do so is a major contributor to the slippage between policy intents and imple-

mented outcomes. Development management’s dimensions are visible in each of the implementa-

tion tasks, though one particular dimension may be more signiﬁcant to some tasks than others.

Institutional agendas and values, for example, come to the fore in the legitimization task. If

disagreements around these are not suﬃciently resolved, they complicate or may even preclude

the constituency building and resource accumulation necessary to policy implementation.

Managing each of the tasks necessarily confronts competing institutional agendas and values

whose eﬀective resolution most often occurs through process interventions aimed at reaching

agreement and specifying shared objectives. Where top-down decision-making is prevalent, two

outcomes are likely: ﬁrst, agreements and objectives may lack the broad buy-in necessary for

eﬀective implementation follow-through, and/or second, agreements and objectives may need

to be revisited later in the implementation task sequence, resulting at best in delays or, at worst,

an impasse. An important consideration for development managers is expanding access to decision-

making, reﬂected in sensitivity to which stakeholders gain a “seat at the table” and whose

priorities and purposes are ultimately served. Today’s development policy emphasis on the

Millennium Development Goals and poverty alleviation highlights equity, representativeness,

and responsiveness.

Brinkerhoff and Brinkerhoff

380

Development managers use a variety of diagnostic and analytic tools to pursue the policy

implementation cycle. They assess constituency preferences, identify organizational and political

constraints, ﬂag points of agreement/disagreement, develop work plans, and so on (e.g. using

stakeholder mapping, strengths-weaknesses-opportunities-threats (SWOT) analysis, responsibility

charts, etc.). In keeping with the values of empowerment and ownership, development man-

agers apply the tools through participatory processes (see, for example, Blackburn and Holland

1998). Participation, as both a value and a management process, is a core feature of development

management.

Participation in policy implementation, however, becomes problematic in the particular context

of developing countries, with their varying degrees of dependence upon international donors

for resources, policy advice, and technical assistance, where external actors play a major role by

virtue of their ability to drive policy agendas and the resources they command. This feature

brings us back to the radicalist/rejectionist critique discussed above, which holds that managing

policy implementation under these conditions replicates existing international power relations,

reinforces the dominance of the industrialized West, and is antithetical to the interests of recipient

countries and their poorest populations. Our reformist perspective characterizes the elements of

this critique as risks rather than certainties, and argues that the risks can be mitigated through

attention to our development management dimensions during policy implementation to increase

the chances that:

 policies incorporate institutional agendas that balance donor and country interests, including

those of poor and marginalized citizens;

 the values policies enact reﬂect national values and identities;

 management tools serve to support responsiveness to country-driven policy priorities as well

as donor accountability;

 processes of engagement facilitate putting local actors in the driver’s seat and building their

capacities.

The predominant inﬂuence of donor-driven development policy design and implementation

makes such risk-mitigating eﬀorts challenging to say the least. As Mowles (2010: 156) observes,

“Strategic planning, log frames, and ideas of transformational management are so ubiquitous

that it would be impossible to proceed without engaging with them.” The trick for develop-

ment managers is to employ the tools and the language of the dominant discourse to recalibrate

the power distributions among the actors involved. These distributions are not simply between

donors and country/local actors, but among local interest groups as well. Which local actors, for

example, should gain access to the driver’s seat? Experience shows that in many countries the

more powerful and better-oﬀ groups tend to occupy the available societal space in ways that

preclude change (see Andrews et al. 2010). Policies intended to be pro-poor are likely to be

watered down so as to include some beneﬁts for the better-oﬀ, and/or to have those beneﬁts

diverted away from the poor during implementation (see Gillespie et al. 1996). The dilemma is

not just a matter of striking a balance between the power and interests of donors and local

actors, but in reconciling the “multiple modernities” and associated interests of a broad range of

local actors.

Conclusion

Our starting point for this chapter was a characterization of development management as geo-

graphically bounded through its association with the poor countries of the global South, and

Development management and policy implementation

381

operationally circumscribed through its connection to foreign assistance and international donor

practices. Yet, poverty, social exclusion, pandemic diseases, food insecurity, political instability,

and debt crises are recognized as aﬀecting all countries to varying degrees. Policies and programs

to address these issues no longer distinguish the developed from the developing world, leading to

what some consider a convergence between the global North and South (see Mosley and

Dowler 2003). The implication for development management is that its relevance is no longer

geographically restricted; and its tools, processes, and values, and its explicit attention to whose

agendas policy implementation serves render it broadly applicable beyond the international

donor-funded realm.

In this sense, we see development management as having come full circle. From its birth as a

subdiscipline of US/UK-based public administration, and drawing from that body of scholarship

and practice to application in developing countries, it now oﬀers experience, tools, and lessons

that can inform poverty-focused change management and policy implementation in countries in

the global North. As industrialized countries have increasingly turned to the non-proﬁt and

private sectors for social services through contracting out, the dynamics of the design and ﬁnancing

of those services more closely resemble the donor-funded international development arena of

projects and programs than meets the eye.

Bibliography

Abbott, D., S. Brown, and G. Wilson. 2007. ‘Development management as reﬂective practice’, Journal of

International Development 19(2): 187–203.

Andrews, M., J. McConnell and A. Wescott. 2010. Development as Leadership-led Change: A Report for the

Global Leadership Initiative. Washington, DC: World Bank.

Blackburn, J. and J. Holland (eds) 1998. Who Changes? Institutionalizing Participation in Development.

London: Intermediate Technology Publications.

Brinkerhoﬀ, D.W. 2008. ‘The state and international development management: Shifting tides, changing

boundaries, and future directions’, Public Administration Review 68(6): 985–1002.

—— and M. Ingle. 1989. ‘Integrating blueprint and process: A structured ﬂexibility approach to development

management’, Public Administration and Development 9(4): 487 – 503.

—— and B.L. Crosby. 2002. Managing Policy Reform: Concepts and Tools for Decision-Makers in Developing

and Transitioning Countries. Bloomﬁeld, CT: Kumarian Press.

—— and J.M. Brinkerhoﬀ. 2006. ‘International development management in a globalized world’,inE.Otenyo

and N. Lind (eds) Comparative Public Administration: The Essential Readings. Amsterdam and Oxford:

Elsevier Press, 831–63.

Brinkerhoﬀ, J.M. and D.W. Brinkerhoﬀ. 2010. ‘International development management: A northern

perspective’, Public Administration and Development 30(2): 102 – 15.

Dar, S. and B. Cooke (eds) 2008. The New Development Management: Critiquing the Dual Modernization.

London: Zed Books.

Easterly, W. 2006. The White Man’s Burden: Why the West’sEﬀorts to Aid the Rest Have Done So Much Ill

and So Little Good. New York: Penguin Press.

Esman, M. 1991. Management Dimensions of Development: Perspectives and Strategies. West Hartford, CT:

Kumarian Press.

Fowler, A. 1997. Striking a Balance: A Guide to Enhancing the Eﬀectiveness of Non-Governmental Organizations

International Development. London:

Earthscan Publications.

Gillespie, P., M. Girgis, and P. Mayer. 1996. ‘“This great evil”: Anticipating political obstacles to development’,

Public Administration and Development 16(5): 431–53.

Grindle, M.S. and J.W. Thomas. 1991. Public Choices and Policy Change: The Political Economy of Reform in

Developing Countries. Baltimore, MD: Johns Hopkins University Press.

Gulrajani, N. 2010. ‘New vistas for development management: Examining radical-reformist possibilities

and potential’, Public Administration and Development 30(2): 124– 36.

Mazmanian, D.A. and P.A. Sabatier. 1989. Implementation and Public Policy. Lanham, MD: University Press

of America.

Brinkerhoff and Brinkerhoff

382

Mosley, P. and E. Dowler (eds) 2003. Poverty and Social Exclusion in North and South: Essays on Social Policy

and Global Poverty Reduction. London: Routledge.

Mowles, C. 2010. ‘Post-foundational development management: Power, politics and complexity’, Public

Administration and Development 30(2): 136–49.

Picard, L.A. and T.F. Buss. 2009. A Fragile Balance: Re-examining the History of Foreign Aid, Security, and

Diplomacy. Sterling, VA: Kumarian Press.

Pressman. J.L. and A.B. Wildavsky. 1973. Implementation. Berkeley, CA: University of California Press.

Pritchett, L. and M. Woolcock. 2004. ‘Solutions when the solution is the problem: Arraying the disarray in

development’, World Development 32(2): 191–212.

Sen, A. 1999. Development as Freedom. New York: Random House.

Thomas, A. 1999. ‘What makes good development management?’ Development in Practice 9(1–2): 9–17.

Woolcock, M. 2009. ‘The next 10 years in development studies: From modernization to multiple modernities,

in theory and practice’, European Journal of Development Research 21(1): 4–9.

Development management and policy implementation

383

Part VIII

Understanding the

evaluation process

Six models of evaluation

Evert Vedung

Evaluation deﬁned

Evaluation is an activity aimed at distinguishing the precious from the worthless, the acceptable from

the unacceptable, the bene ﬁcial from the detrimental. In present-day public sector management,

however, evaluation has acquired more speciﬁc and narrow meanings. Here, evaluation is a mechanism

for monitoring, systematizing, and grading ongoing or just ﬁnished government interventions

(organizations, policies, programs, projects, activities, their eﬀects, and the processes preceding these

eﬀects, perceptions of intervention content included) so that public oﬃcials and other stakeholders

in their future-oriented work will be able to act as responsibly, creatively, equitably and economically

as possible. In this chapter, the following minimal deﬁnition of evaluation will be adopted:

Evaluation is careful assessment of the merit, worth, and value of organization, content,

administration, output, and eﬀects of ongoing or ﬁnished government interventions, which

is intended to play a role in future, practical action situations.

Merit criteria that evaluations use

In order to perform evaluation, descriptions, however exact, of the phenomena to be assessed

are not enough: evaluators also need criteria of merit, worth and value.

Suppose that a municipal anti-smoking campaign targeted at secondary school students adopts

as its goal that 5 percent of smoking students shall quit smoking within three months after the

completion of the campaign. In addition, suppose that an evaluation is launched four months after

the end of the campaign to ﬁnd out whether the campaign did achieve its own goals. In this

case, the merit criterion is the campaign’s own goal: 5 percent of the smoking students shall quit

smoking within three months after the completion of the campaign. An evaluation model taking

goals as evaluative criteria is using the goal-attainment model.

The evaluation models included in this chapter are the most important ones in the ﬁeld. A

taxonomy is presented below:

1 goal-attainment model;

2 side-eﬀects model;

387

3 relevance model;

4 client-oriented model;

5 stakeholder model;

6 collegial models: peer review, self-evaluation.

Evaluation models organized according to their criteria of merit constitute a fairly varied group.

In addition to the standard goal-attainment model, the taxonomy embraces the side-eﬀects model,

the relevance model, the client-oriented model, the stakeholder model, and collegial models

combining peer review with self-evaluation.

One obvious set of models will not be covered here: economic models. In addition to substantive

values, economic models focus on cost aspects of public interventions. The two basic variants

among economic models are the productivity and the eﬃciency models. Eﬃciency models are

divided into cost-eﬀectiveness and cost-beneﬁt. However, for reasons of space they are excluded

from this chapter.

Adiﬀerent possibility would be to organize the survey according to methodology to be used

in ascertaining intervention eﬀects. This methods-driven point of view is pursued by the present

movement for evidence-based policy and public administration. Models are ordered according

to strength of designs for impact assessment (experiments, quasi-experiments, time-series, process

tracing, etc). I have not chosen this option either.

Another signiﬁcant dimension is subject matter to be appraised. Since the focus here will be

on public sector interventions, subject matters are commonly divided into intervention results

and intervention processes. Intervention results mean outcome eﬀects, i.e. those consequences

that the intervention, at least to some extent, directly or indirectly has produced in society or

nature. In addition to members of the target group actually reached, results include target group

responses to the intervention in terms of, e.g. attitudes, factual and valuational utterances, and

actions suggested or actually taken. Intervention processes usually include procedures and practices

preceding results, occurring in implementation between intervention adoption and intervention

results. Naturally, processes do not include targets reached, target responses and consequences

beyond those responses, because all that is included in intervention results.

There are many other possible dimensions that might be used to classify models. Data assembly

methods is one (documentary, interrogatory or observational methods). Here, these possible

dimensions will be discussed now and then under each evaluation model but they are not used

to classify them.

My review will treat the models in the order shown above.

Goal-attainment model (effectiveness model)

An approach on long standing is goal-achievement evaluation, also known as goal-attainment

evaluation and eﬀectiveness evaluation. The two basic ingredients of goal achievement evalua-

tion are goal-attainment measurement and intervention impact assessment. In goal-attainment

measurement the key question is: are the results in accord with intervention goals? And the impact

assessment issue can be formulated thus: are the results produced by the intervention?

Goal-attainment evaluation is a paragon of simplicity and lucidity. Its ﬁrst step involves

identifying the intervention goals, teasing out their actual meaning and rank order, turning

them into measurable objectives, and determining to what extent they are realized in practice.

The second step implies ascertaining the degree to which the intervention has promoted or

dampened goal realization. The plain anatomy of the goal-attainment model is outlined in

Figure 29.1.

Vedung

388

The strength and shortcomings of the goal-attainment model

In earlier literature, public sector evaluation was goal-attainment appraisal, period. Since then,

however, quite a few other merit criteria have been suggested and used. There are, however,

several worthwhile reasons in favor of goal-attainment assessment of public sector interventions.

The argument from representative democracy is one.

In a democracy, all power belongs to the people. Yet, due to lack of competence and time,

the people cannot make all the complicated decisions concerning the well-being of citizens.

The people do not have time to participate in hundreds of thousands of decisions. And they do

not have the necessary competence to make wise decisions on, for instance, placement of

patients in line for surgery, or day-to-day care for ailing senior citizens in public sector homes

for the elderly. For these reasons, the citizenry must elect political representatives to make the

decisions for them. But representatives in political assemblies do not have the time or compe-

tence to make all decisions. They must delegate their power to governments to make decisions

for them. But governments do not have time and the speciﬁc knowledge necessary, so they in

turn have to delegate to civil servants and professionals to take decisions, and so on. The public

sector is made up of long chains of principal-agent relationships.

If an agency adopts a program in order to reach some goals, these goals derive their legiti-

macy from the fact that the agency’s decision-making authority has been delegated to it by the

government and that the government, in turn, has received its authority to do so from parlia-

ment, and parliament, in turn, from the people. It is a merit of the goal-attainment model that

it recognizes this democratic aspect of public sector goals.

In sum, the goal-attainment model scores an important point with respect to its tilt toward

representative democracy and the parliamentary chain of control. On the other hand, the goal-

attainment model also suﬀers from persistent shortcomings. The most signiﬁcant general reasons

against the goal-attainment model are the haziness argument, and the unintended side-eﬀects

argument.

The haziness argument maintains that intervention goals are deﬁcient as criteria of merit due to

their obscurity. There are two kinds of goal obscurity: goal indeterminateness and goal catalogs.

Occasionally, programs are based on indeterminate goals. Particular goals may be ambiguous and

carry two or more meanings. Yet ambiguity in this sense of dual meanings is exceptional in

policy and administrative language, and barely bothers evaluators. More uncertainty is caused by

vagueness. A goal is vague if it does not delineate clearly cases where it is or is not applicable.

The outer border delimiting the extension of a vague word is so fuzzy that within a certain

range it is impossible to know what is included in the extension and what is not. Rampant in

political rhetoric, vagueness is one favorite expedient to settle political conﬂicts through

semantic formulas without really resolving them.

Figure 29.1 Goal-attainment model

Six models of evaluation

389

The second major obscurity is produced by goal catalogs. Most large social reforms contain

impressive directories of diverse goals. While a single goal may be hailed as the major one, it is

often maintained that this one must be balanced against all the others, maybe including poten-

tially conﬂicting ones. But the necessary trade-oﬀs between the several goals are not indicated,

which makes it impossible to elicit from such lists of goals one distinct, transparent, expected

outcome. Thus, program goals do not oﬀer any safe guidance for continued data assembly. They

are neither speciﬁc nor lucid enough to be usable as value criteria against which to measure

intervention successes and failures.

The goal-haziness argument reveals an important misﬁt between the requirements of the

goal-attainment model and the way public policies, programs, and activities are often com-

posed. If elected oﬃcials and program planners have not speciﬁed individual goals into measurable

objectives, and if they have not balanced the various stated goals into one global outcome or

output measure, the goal-attainment evaluator cannot summarize her ﬁndings into a completely

value-neutral evaluative judgment. She can do so only after she has clariﬁed the goals and

prioritized among them in a fashion that will cast doubts on the objectivity of the whole

enterprise.

The second general counter-argument, about unintended side-eﬀects, is in my view the crucial

one. Public sector interventions, in social security and social work for instance, invariably lead

to consequences which were not foreseen in the original decision situation. Were the evaluators

to conﬁne themselves to ascertaining the achievement of premeditated goals, the search for

serendipitous results or unanticipated side-eﬀects outside the goal area would not be included in

the evaluation process. The evaluation ﬁndings would exhibit a tunnel vision of events, and

produce a biased, if not fundamentally wrong, picture of what the intervention has attained. In

all likelihood, a program generating some interesting spin-oﬀ eﬀects must be better than a

program producing several undesirable spillovers.

In sum, the major strength of the goal-attainment model is grounded in the theory of repre-

sentative democracy. Yet, it has problems with hazy goals and goal catalogs, pervasive as we all

know in public policy. The most compelling rebuttal, however, emanates from the model’ s

blindness to side-eﬀects.

At this point, I would like to present a model that expressly considers the weighty side-eﬀ ects

argument, while retaining the fundamental goal-orientation of the goal-attainment model: the

side-eﬀects model.

Side-effects model

The side-eﬀects model implies a widening of the subject-matter of the goal-attainment model in

the sense that search for results in the target area is supplemented by search outside the target area

for side-eﬀects. The underlying idea is that public interventions may produce other things than

intended results which, in turn, constitute reasons for new interventions. Interventions beget

interventions.

The role of heat pumps in Swedish energy production is an illustrative case of a solution

turned into a problem in need of a solution. From the 1970s, government subsidies were dis-

embursed for the installation of heat pumps for the retrieval of the enormous amounts of waste

energy produced by the paper and pulp industry along the northern coast of the Gulf of

Bothnia as a solution to the oil and nuclear power problem. The rapid dissemination of heat

pumps was a consequence of government subsidization. After some years, it was discovered that

the heating medium, CFCs (freons), in the heat pumps might leak into the atmosphere and,

over time, damage the stratospheric ozone layer protecting the earth from dangerous ultraviolet

Vedung

390

radiation. A small part of this damage can be regarded as an unforeseen and unintended side

eﬀect of Swedish government support of heat pumps. From the mid-1980s, the heat pumps

turned into an environmental problem that needed to be solved politically.

From the interventionist point of view, a side eﬀect can be deﬁned as at least a partial

consequence of the intervention, which cannot be included among the desired main eﬀects.

Main eﬀects are those actual, expected and wanted consequences, which at least are partly

produced by the intervention. Side-eﬀects can be unanticipated but also anticipated and con-

sidered in calculations preceding decisions to adopt policies. They may be beneﬁcial as well as

detrimental.

Public interventions may also create perverse eﬀects. Perverse eﬀects run exactly counter to the

very intentions of the intervention instigators. Since these impacts occur in the target area, they

are not side-eﬀects. Neither are they main eﬀects, since they are not coveted by the policy

instigators. Perverse eﬀects are also diﬀerent from null eﬀects. Null eﬀects mean that interventions

produce no impacts at all on their targeted areas. In the perverse eﬀects case, consequences are

produced but entirely contrary to the ones intended.

Since perverse eﬀects and null eﬀects occur in the targeted areas, the goal-attainment model

with all its attention directed at what happens in these particular ﬁelds has no problem with

handling them. But this also means that the model cannot discover and ascertain side-eﬀects

because they fall outside the targeted sectors. This task is left to the side-eﬀects model.

Evaluators and policy-makers should pay attention to side-eﬀects because by-products,

whether detrimental or beneﬁcial, are crucial factors in every inclusive judgment of the opera-

tion of an intervention. Should it turn out that side-eﬀects, which have been known, discussed,

and positively valued in advance, have not materialized in spite of the fact that the program has

been on the books for the intended period of time, this ought to have consequences for any

appraisal of it.

If the totality of eﬀects of a government intervention inside and outside its targeted area were

to be investigated, the structure of the evaluation on the outcome side might be as shown in

Figure 29.2.

I strongly commend side-eﬀects to goal-attainment evaluation. Indeed, the major rationale

for doing public sector evaluation in the ﬁrst place is that state actions to some extent are

unpredictable and regularly result in side-eﬀects not originally foreseen. It is an important duty

of evaluation to map and assess the worth of these side-eﬀects.

Figure 29.2 Side-effects model with speciﬁed pigeonholes for side-effects

Six models of evaluation

391

Also, in terms of merit criteria, the side-eﬀects model engenders an extension of goal-attainment

in that intervention goals stated in advance are supplemented with merit criteria for side-eﬀects.

If some eﬀects are not foreseen, the criteria and standards for judging merits and demerits of

these eﬀects are not pre-speciﬁed either. Therefore, pre-speciﬁed goals are insuﬃcient as instru-

ments of judging unanticipated side-eﬀects is concerned. A feasible solution is the following.

Aside from mapping the main eﬀect and comparing it with the pre-speciﬁed goals, the eva-

luator may also chart the side-eﬀects but leave it to the commissioners and other users of the

evaluation to ascertain their value and carry out the overall, global assessment of the program ex

post facto. Attention to unanticipated side eﬀects forces evaluators to use other merit criteria than

initially incorporated into the intervention.

Relevance model

In the relevance model, solving the underlying problem is the merit criterion against which the

worth of the intervention is assessed. Is the intervention bearing on, connected with or perti-

nent to the solution to or at least the alleviation of the problematic matter in hand? If all or part

of the problem has been solved by the intervention, the intervention is successful; if not, it is a

failure (more in Vedung 2009: 116ﬀ.).

The underlying problem is a complex criterion. Which underlying problem is at stake? As a

ﬁrst cut, the answer is the problem embedded in the intervention, namely the problematic

situation that the intervention purports to attack. But of course there are others, to be omitted

in this brief overview.

To be relevant, what does it mean? Pertinent to the problematic matter in hand would be one

reasonable clariﬁcation. Yet, pertinence is not enough. To some degree, a relevant intervention

must also be adequate, useful, suﬃcient and proper for the task at hand.

What should be relevant to the underlying problem? The intervention, of course, or more properly

its agreed-upon goals and policy instruments. These relevance issues can be explored early on, that

is once the intervention is adopted but not yet implemented, by reference to scientiﬁc ﬁndings

and tested experience. Another possibility would be outputs. A third would be intervention eﬀects

obtained out there in the community. Eﬀect relevance should be studied when the intervention

has been implemented and in operation for some time. Are the impacts then achieved relevant

to the solving of the problem that the intervention is supposed to attack?

Eﬀect relevance seems very close, if not similar, to goal achievement (eﬀectiveness). Yet, they

are diﬀerent. The change in the state of aﬀairs referred to by a goal need not be identical to a

change in the underlying problem. The goal of an information campaign may be to reduce pur-

chases of alcoholic beverages while the underlying problem is to reduce prevalence of alcohol-

related illnesses. But purchases may be down while alcohol-related illnesses are up due to increased

purchases abroad or on international ﬂights.

Here we leave the models using merit criteria derived from the intervention as such: the

goal-attainment, the side eﬀects and the relevance model. Now we will address actor models.

Speciﬁc to actor models is that their evaluative criteria are taken from the agents concerned. We

will discuss the client-oriented model, the stakeholder model and the collegial model.

Client-oriented model

The client-oriented model allows members of the intervention’s target group to perform the

evaluation on the basis of their own merit criteria. Issues at stake are whether the public inter-

vention (1) in terms of its content, (2) during the process up to its delivery, (3) in the delivery

Vedung

392

proper, (4) as regards client responses, or (5) through its eventual outcome eﬀects achieves the

quality that clients deem reasonable, or require or want to see. Clients are asked to pass judgment

on service accessibility (process up to points of delivery), scope and quality of service provision,

or service eﬀects on the recipients themselves (impact at outcome level). As merit criteria, the

client-oriented model may use participant desires, wishes, requests, demands, goals, concerns,

expectations, and so on.

The term client denotes the recipients of public interventions. Prisoners in jails, pupils of public

schools, patients in county hospitals, elderly people receiving municipal home care, book-borrowers

in municipal libraries and passengers on municipal buses and trains are examples of clients

(addressees, participants, targets, consumers).

Several approaches might be considered client-oriented. Think about a contrived example

concerning client satisfaction with public-sector services. During 2006 60 percent of parents in

the city of Helsinki, Finland, were satisﬁed with the daycare that their children received in

municipal daycare centers. Two years later, the number had dropped to 50 percent. In January 2009

members of the city council expressed concerns about the unexpected plunge. The council

unanimously decided to set 75 percent parent satisfaction as the goal to be achieved in 2011. In

2012 a new study demonstrated that in 2011 65 percent of the parents were satisﬁed. Admittedly,

both the 2008 and the 2012 studies are client-oriented. Are they client-oriented evaluations?

The answer is yes. User satisfaction is a client-oriented value criterion. Interestingly, the 2012

study might be considered a goal-attainment evaluation as well, since 75 percent parent satisfaction

was a goal set by the city council.

There are more advanced forms of client-oriented evaluation where clients are much more

implicated. Let me reason from the case where the evaluation is ﬁrst, commissioned by the admin-

istrators, and second, planned to involve the service users much more than merely asking them in a

questionnaire about their service satisfaction. In this case, the clients themselves are encouraged to

select the aspects of the intervention, its implementation and outcomes on which to pass judg-

ments. For instance, clients may judge intervention output, service availability, service quality, or

even service process and service administration. Is the core service tailored to meet the demands

of the clients? Are the encounters of the clients with the service employees respectful? These are

two questions that might be answered by the service users in their self-reports. The clients may also

choose to raise the causal issue, that is, estimate intervention impacts on themselves or on the

client community in general.

Currently, the client-oriented evaluation model is employed in numerous contexts such as

nursing homes for the elderly, public housing, mental health, public utilities, recreation, and

physical health services, where clientele participation is crucial to the operation of the services.

Client-oriented models are used to evaluate library services, arts, zoos, and museums. It is a favorite

with educators. At universities, students are routinely requested to share their opinions of courses,

reading lists and lectures. They are asked to rate their teachers’ abilities to organize the course

contents, to stimulate and promote altercations, to stir student motivation and critical thinking,

and to show concern and enthusiasm for the students.

Pros and cons of client-oriented evaluation

Client-oriented evaluation is justiﬁed in several ways. Some philosophers ground it in political

ideologies engendering that public administration produces goods and services for customers in

the market place. They claim that customer pressures expressed through attitudes and suggestions

for improvement via evaluations toward service providers, service management and decision-

makers will lead to improvement of not only the core service, but also of service processes,

Six models of evaluation

393

service delivery, service eﬀects, and customer satisfaction. Through client-oriented evaluation,

public services will become more clearly geared toward the wishes and expectations of the users.

Yet, the customer parallel cannot be pushed too far, since the client notion includes partici-

patory and deliberative aspects. The participatory feature suggests that clients are also citizens

who may voice their complaints and desires to the evaluators and service providers, and to some

extent inﬂuence and take responsibility for service content. The deliberative feature engenders a

discursive, reasoning, and learning-through-dialogue countenance, which may educate service

providers to pay heed to client concerns and clients to become better citizens: the consumer as

citizen rather than the consumer as customer.

Client-oriented evaluation may increase the legitimacy of the intervention. If clients are asked

for their opinions and have some inﬂuence on intervention formation, processes and outputs,

their acceptance of the system will probably increase. In addition, client-oriented evaluation may

foster eﬀectiveness and eﬃciency because concentration on clients may force service providers

and managers to do away with many preoccupations besides providing good services.

But evaluators must be aware of the tendency of the clientele to exaggerate complaints in

order to get more service. Clients may also nurture ﬁscal illusions. Greater client involvement in

evaluation may surrender power to groups with narrow vested interests.

The client-propelled model may supplement the previously presented approaches, since it

poses other problems. The requirement that the civil service must be responsive to client concerns

is sound, but within limits. It can never take precedence over the requisite that front-line operators

should follow the directives of their hierarchical administrative superiors, and indirectly political

bodies like parliament, the municipal council, and, ultimately, the citizens as a collectivity, whose

votes have determined the composition and general policy direction of these bodies. The elderly

in a community who enjoy municipal social home aid cannot take decisions that run counter to

the rules of the agents and principals in the representative chain of control. They cannot uni-

laterally lower the service fees, for instance. Evaluation models grounded in representative democ-

racy must take precedence over client-oriented models. Client criteria are reasonable to use, but

within limits; they must be balanced against other criteria like goal-attainment and professional

norms for service excellence.

Stakeholder model

In stakeholder evaluation claims, concerns and issues of the various aﬀected actors serve as merit

criteria when interventions—their contents, processes, outputs, outcomes, and organization—

are assessed and evaluated (Vedung 1998: 73ﬀ.). Stakeholders might be deﬁned as groups or

individual actors that have some interest vested in the intervention to be evaluated or its out-

come eﬀects. Interest may be measured in terms of money, status, power, face, opportunity or

other coin, and may be large or small, as constructed by the groups in question (Guba and Lincoln

1989: 51). This is quite diﬀerent from using preﬁxed objectives as merit criteria, as in goal-

attainment evaluation. Stakeholder evaluation, however, does resemble the client-oriented model,

the major diﬀerence being one of scope: while the client-driven model is basically concerned

with one group of aﬀected interests, the stakeholder model is geared to all of them.

An overview of conceivable stakeholders in local social welfare interventions is presented in

Figure 29.3.

Stakeholder evaluation can proceed in diﬀerent ways. The stakeholders may constitute them-

selves as the evaluation team and carry out the evaluation. The evaluation may also be con-

ducted by particular evaluators, who elicit the views of the stakeholders. In the sequel, I shall

reason from the case that stakeholder evaluation is carried out by particular evaluators.

Vedung

394

Stakeholder-based evaluation starts by evaluator mapping the major groups who are involved

or thought to have an interest in the content, execution, results, and organization of the inter-

vention. The evaluator identiﬁes the people who initiated, hammered out, funded, and adopted

the intervention, that is basically the politicians. She identiﬁes those who are charged with its

implementation: senior, middle, and junior managers, staﬀ, and front-line operators, who actually

deliver intervention output. She singles out the intervention’s primary clients and the clients’

associations. She identiﬁes relatives and relatives’ associations. She may also include lay people.

And she searches for those who know that they have a stake in the intervention but prefer to keep

a low proﬁle and those who are unaware of the stake they hold.

Advocates of stakeholder models nurture a strong penchant for qualitative, interactive data

assembly methodology. The evaluator must talk to the stakeholders to elicit their narrative his-

tories and observational data, which in turn should be allowed to aﬀect the evaluator’s next step

in the search procedure. After a while, she might discover both the purported and the genuine

aims of the intervention, and what concerns various stakeholders nurture regarding it. With time,

the evaluator gets more involved and can start to determine which dimensions and concerns of

the stakeholders should be included in the study. Only then can she take a stand on what the

outline of the evaluation should be.

It is typical for the stakeholder model that the evaluator is permitted to search rather extensively

for the dimensions and the crucial concerns on these dimensions. The idea is that the evaluator

must be responsive to the issues and concerns of the aﬀected people and let these govern the

next step in the investigatory enterprise. Through interactive communication she is supposed to

ﬁnd out which stakeholder dimensions and concerns are to be taken seriously and probed more

deeply. The evaluation design will be gradually determined. Stakeholder evaluation is responsive

evaluation.

To elicit ﬁnal data on stakeholder concerns, advocates of stakeholder models prefer stake-

holder self-observation and sustained interviewing to questionnaires, documentary methods, and

Figure 29.3 Potential stakeholders in local social welfare interventions

Six models of evaluation

395

evaluator observation. In-depth interviewing of individual targets is one favored technique. In

social service and social work, distribution of self-report instruments that clients, parents, relatives

and other stakeholder networks can easily complete themselves is used. In some cases, client-

oriented evaluators endorse focus group interviewing which allows for group deliberations among the

participants and between the participants and the evaluator. The evaluator tries to create forums

of debate to promote deliberative richness. This will support the development of new ideas,

service concepts, solutions, and technologies. It might also, as a side-eﬀect, educate participants

to become better citizens in the future.

After data are amassed and processed, the reporting of ﬁndings, which might vary from one

stakeholder to another, will commence. The key word seems to be “portrayals,” that is, infor-

mation-rich characterizations using pictures, anecdotes, thick descriptions, and quotes. The

comprehensive holistic view mediated through a portrait is important. Normally, several criteria

of merit, standards of performance on these criteria and several comprehensive assessments will

be included.

Upsides and downsides of stakeholder evaluation

Stakeholder models have several advantages of which four will be mentioned here.

The democratic arguments for the stakeholder model depart from participative and deliberative

points of view. True, democracy means that citizens in general elections vote for competing

elites that are supposed to make decisions on their behalf (representative democracy). Yet, the citizenry

should also be able to partake in ﬁnal public decision-making between elections (participative

democracy). Furthermore, discussion, dialogue and debate are also important democratic values

because they help people to form and reﬁne their beliefs and preferences (deliberative democracy).

The stakeholder model satisﬁes these participative and deliberative values somewhat more than

the goal-achievement and the side-eﬀects models.

According to the knowledge argument, avoiding intervention insights, which those involved

indubitably have, would be foolish of evaluators. Stakeholders nurture convictions about inad-

vertent side-eﬀects, sophisticated implementation barriers, and outright cheating which may

furnish evaluators with ideas about topics for further investigation. Since stakeholder-orientation

will bring up more aspects of the subject-matter for discussion, the quality of the evaluation

ﬁndings will increase. All in all, it is easy to agree with the recommendation that almost every

evaluation ought to begin with the determination of relevant actors and rounds of interactive

data assembly.

Findings from goal-attainment and side-eﬀects evaluations, carried out with quantitative metho-

dology and with no interactive involvement of stakeholders, seem have little impact. The stake-

holder approach increases the chances that issues of genuine interest to concerned parties will be

addressed. It brings to light information that meets the real requirements of the diﬀerent stake-

holders, thereby enhancing the probability that the ﬁndings actually will be put to use. This is

the utilization argument in support of stakeholder evaluation.

Finally, the stakeholder models might promote compromises, and forestall political struggle. Sta-

keholder assemblies are consensus-building mechanisms. They are vehicles for shaping agreement

on the results of earlier eﬀorts, and most importantly, proposals for future action. Consensus

building and the rendition of legitimacy to fundamental decisions are considered great advantages

of stakeholder evaluation.

There are also obvious drawbacks with stakeholder models. Stakeholder evaluations are inordinately

impractical and resource-demanding, since several stakeholding constituencies must be contacted and

nurtured.

Vedung

396

Stakeholder models are fuzzy. They provide no authoritative answer to the question of who

the stakeholders are. Furthermore, all the stakeholding audiences, however selected, are treated

as equals. But in a representative democracy, elected politicians must carry more weight than

administrators or experts and even clients on the substantive matters under consideration. The

stakeholder model embodies no priorities among the stakeholders. Like client-oriented evaluation,

stakeholder evaluation must work within the frames ﬁxed by representative democracy.

While hazy in its contours, impractical and resource-demanding, the stakeholder model carries

some important merits. Utilization and compromise arguments speak in its favor. Another strong

reason for it is the knowledge argument: it can be used as a search strategy at the start of the

evaluation in order to get a quick and provisional grasp of the meaning, implementation and

outcome of the intervention.

Finally, a case from deliberative and participatory democracy can be made for stakeholder models.

Through the use of the stakeholder approach, aﬀected interests can deliberate public aﬀairs and

learn to become better citizens in the future. They can participate and inﬂuence the ﬁnal outcome.

On the other hand, this must take place within the conﬁnes set by representative democracy.

Collegial models: peer review, self-evaluation and combinations thereof

Collegial evaluation models imply that personnel of the pertinent agency are entrusted to

evaluate the performance of their colleagues in some other agency or branch of the public

sector using their own professional criteria of merit and standards of quality.

The most celebrated collegial model is the peer review approach, in which the evaluation is

conducted by an external collegium, which by deﬁnition is an assembly of professional equals.

Ideally, these equals should be somewhat better in their area of expertise than the colleagues they

are invited to assess. In this way, lawyers evaluate lawyers, scientists scientists, surgeons surgeons,

nurses nurses, and so on.

Self-evaluation is another form. Then, the professional herself evaluates her own performance

or professionals in an organization together evaluate their organization’ s performance. Often

self-evaluation is combined with external peer review.

Collegial models are constantly used in evaluation of research and higher education. They are

very much based upon dialogue, discussion and deliberation. The procedure usually starts with

self-evaluation. The professionals to be evaluated carry out an appraisal of their own performance,

of their research project, education program, or university department. They discuss strengths,

weaknesses, opportunities and threats (the so-called SWOT scheme). Then, external renowned

scientists and educators of the particular ﬁeld are assigned to assess the quality and relevance of

evaluatee work. These peers base their preliminary assessment on the written self-evaluation and

site visits with dialogues. Then evaluatees are given the opportunity to comment on the eva-

luators’ reports before they are ﬁnalized; all in all, the evaluators listen to the evaluatees and

solicit their opinions.

We might wonder what collegial models have to do with public sector evaluation. These

models seem kilometers away from the arena of policy-making and program enactment. The

answer is embodied in the principle of the profession-driven public sector. In some areas of

public life, goals are so complex and techniques so diﬃcult that political oﬃcials have found it

wise to leave the shaping and debating of them to well-educated professionals. Architects, judges,

professors, doctors, veterinarians, and engineers would be cases in point. Hence, it is also con-

sidered natural to delegate ex post evaluation to the professions. But since these professionals

work in the public sector, peer review must be regarded as an evaluation model on a par with

the other models used in public life.

Six models of evaluation

397

Let me illustrate this evaluation prototype with the Swedish National Agency for Higher

Education former strategy for evaluation of research and education at the state universities in

Sweden. The agency once instituted the following ten-steps evaluation policy:

1 The Board adopts a comprehensive evaluation policy: during the upcoming six-year period,

all universities and colleges in the country will be evaluated on a rotational basis.

2 The Board decides in broad terms about groups of evaluations: e.g. research in all the

university departments of social work will be evaluated next year.

3 For every particular evaluation, the Board appoints a project manager (project group)

within the agency and a number of qualiﬁed external evaluators who receive introductory

information.

4 Each department of social work performs a self-evaluation departing from the national

agency’s special requirements. An important feature of these requirements is that evalua-

tion items ought to be placed under three headings: (i) preconditions for research, (ii)

research processes, and (iii) research ﬁndings. The self-evaluations follow the SWOT

scheme, i.e. strengths, weaknesses, opportunities, and threats.

5 The external experts familiarize themselves with the subject-matter by consulting self-

evaluations, research reports, CVs and other relevant materials from the departments to be

evaluated.

6 External assessments are carried out through site visits. In all, evaluatees, researchers and

research groups make presentations for and participate in discussions with the external

assessors. The expert panels are confronted face-to-face with the evaluated researchers.

The experts observe, listen, ask questions, discuss and take notes.

7 The external assessors compile preliminary assessments for each department, to which the

departments in question are asked to respond.

8Aﬁnal report is prepared by the external evaluators, and the project manager formulates a

ﬁnal assessment on the basis of which the National Agency Board decides on measures

such as cancelation of examination rights.

9 Follow-up conferences are arranged, where each department is informed about the particular

assessments performed by the external reviewers.

10 One year later, the Board carries out a follow-up to verify that recommended measures

are taken.

Central to collegial evaluation is selection and application of evaluative criteria and standards for

estimating the quality of preconditions, processes and ﬁndings. For this, the professionals them-

selves are particularly suitable, because these assessments require specialist knowledge. In addition,

quality criteria are complex by being associated with speciﬁc methodological approaches. Fur-

thermore, due to knowledge growth, yardsticks as well as professional approaches are continually

changing in the disciplines. Add to this that quality criteria are often communicated orally only

in the relevant professions as tacit knowledge. This tacit knowledge must be brought forward

through penetrating dialogues among groups of colleagues.

A second strength is that site visits are short and therefore inexpensive. On the other hand,

the evaluation may still become expensive through consultant fees, travel expenses, and because

it takes time to complete the ﬁnal report.

One weakness with short site visits and self-evaluations is that evaluatees may be tempted to

show up facades and erect Potemkin villages.

Collegial models frequently produce shaky results. Matched panels use widely diﬀerent merit

criteria and performance standards and reach miscellaneous conclusions. However, in technically

Vedung

398

complex ﬁelds, collegial evaluation is probably the ﬁnest method available to judge the quality

of what is produced.

Final note on choice among the six evaluation models

This chapter has provided a broad overview of the evaluation landscape from a narrow but man-

ageable perspective. Since evaluation is a normative, value-laden enterprise, I have organized the

survey by merit criteria used in various evaluation models. Six such models have been reconstructed.

An evaluation model is as a template based on one, two, three and even more dimensions

indicating in a general way how an evaluation can be done

As an innovation in 1950s and 1960s evaluation was equated with the goal-attainment model.

Owing to sharp criticism, this model was pushed into the background in the discourse, but

hardly in practice. Since 1980, however, it has experienced a renaissance in the discussion, partly as

a feature of so-called performance management. In addition, thanks to the so-called evidence-

based wave that presently dominates the evaluation discourse, the goal-attainment model (sometimes

under the designation eﬀectiveness evaluation) has become even more important.

The major downside of goal-attainment evaluation is its inability to take account of unfore-

seen side eﬀects. For this reason the side-eﬀect model is recommended. The side-eﬀects model

is the goal-attainment model expanded into (1) an all-eﬀects model using (2) merit criteria

articulated expost in addition to pre-ordained goals.

In contrast with the goal-attainment model, the relevance model uses solution of underlying

problems as merit criterion. This approach provides evaluators with a critical edge in allowing them

to question the goals in case the intervention achieves its goals without solving the problem.

The stakeholder model has become popular particularly since 1970. Since full objectivity is

impossible to reach in evaluative work, its proponents argue that providing multiple images of

reality from diﬀerent perspectives is an appropriate approach.

Collegial models take the quality criteria of one group of stakeholders—the relevant profes-

sionals—as their bases for assessment. These criteria are diﬃcult to grasp, constantly changing and

often embedded into the professions as tacit knowledge. A collegial evaluation starts with the

evaluatees performing a self-assessment followed by an external evaluation carried out by inde-

pendent highly qualiﬁed colleagues. The collegial models are assumed to capture and appreciate

complex qualities of diﬃcult policy areas like surgery, engineering, basic research, and so on.

Since about 1980 client-oriented evaluation has come to the fore especially within health

care, culture, transportation, services and education. The philosophy is that the citizen is a cus-

tomer whose needs should be considered so that public services become more accustomed to

client wishes. Indirectly this also increases the eﬃciency and legitimacy of the system.

The client-oriented model may conﬂict with the goal-attainment, side eﬀects and relevance models,

all of which derive their legitimacy from representative democracy. Since representative democracy

is the basic form of government in Western nations, the client model as well as the stakeholder

model must work within the limits set by the representative system. In this way the information

produced by the client model may supplement information produced by the other models.

Each evaluation model provides partial perspectives and answers only. For this reason, combinations

of several models are commended.

Notes

1 Besides a new text on the relevance model, the chapter is an abbreviated and reworked version of

Chapter 4 in my book Public Policy and Program Evaluation (New Brunswick, NJ: Transaction Publishers,

Six models of evaluation

399

paperback 2009 (1998), www.transactionpub.com/cgi_bin/transactionpublishers.storefront). For footnotes,

see this book.

2 Several other reasons for and against are presented in Vedung [1998]2009: 40ﬀ.

3 Actually, there is also a school of thought which uses client needs as a point of departure for client-oriented

evaluation. To simplify I have avoided the needs issue here.

4 My extensive experience as an evaluation instructor has taught me to emphasize the diﬀerence between

intervention clients (intervention participants, addressees) and evaluation clients (evaluation users). The

evaluation client is the person, group, or agency that has commissioned the evaluation of an intervention

or is supposed to use its ﬁndings, whereas the intervention client is the intended or actual recipient of

the intervention.

Bibliography

Bardach, Eugene. 2011. A Practical Guide for Policy Analysis: The Eightfold Path to More Eﬀective Problem

Solving. Washington, DC: CQ Press.

Guba, Egon G. and Yvonna S. Lincoln. 1989. Fourth Generation Evaluation, 1st edn. Thousand Oaks, CA:

Sage.

Hansen, Hanne Foss. 2005. ‘Choosing evaluation models: A discussion on evaluation design’, Evaluation

11(4): 447–62.

Hansen, Morten Balle and Evert Vedung. 2010. ‘Theory-based stakeholder evaluation ’, American Journal of

Evaluation 31(3): 295–313. Available at: http://evi.sagepub.com/content/16/3/263.full.pdf+html

Preskill, Hallie and Darlene Russ-Eft. 2004. Building Evaluation Capacity: 72 Activities for Teaching and

Training. Thousand Oaks, CA: Sage.

Rossi, Peter E., Mark W. Lipsey and Howard E. Freeman. 2004. Evaluation: A Systematic Approach, 7th

edn. Thousand Oaks, CA: Sage.

Russ-Eft, Darlene and Hallie Preskill. 2009. Evaluation in Organizations: A Systematic Approach to Enhancing

Learning, Performance, and Change, 2nd edn. New York: Basic Books.

Stuﬄebeam, Daniel L. 2000. ‘Foundational models for 21st century program evaluation’, in Daniel L.

Stuﬄebeam, George F. Madaus and Thomas Kellaghan (eds) Evaluation Models: Viewpoints on Educational

and Human Services Evaluation. Boston: Kluwer Academic Publishers, 33–84.

Vedung, Evert. [1998]2009. Public Policy and Program Evaluation. New Brunswick: NJ: Transaction Publishers.

Available at: www.transactionpub.com/cgi_bin/transactionpublishers.storefront.

——. 2010. ‘Four waves of evaluation’, Evaluation 16(3): 263–77. Available at: http://evi.sagepub.com/

content/16/3/263.full.pdf+html

Wholey, Joseph S., Harry P. Hatry and Kathryn E. Newcomer. 2010. Handbook of Practical Program Evaluation.

3rd edn. San Francisco, CA: Jossey-Bass.

Vedung

400

Policy feedback and learning

Patrik Marier

Introduction

This chapter tackles two subﬁelds that have recently grown in importance in the analysis of public

policy: policy feedback and policy learning. While policy learning is a potential outcome resulting

from policy feedback, the latter is not essential for learning to occur. In fact, relative to the well-

focused literature on policy feedback, this chapter illustrates that the literature on policy learning

is extremely diverse and encapsulates a myriad of concepts, actors, and learning mechanisms.

This chapter is divided into two sections. The ﬁrst deﬁnes, analyzes and explains how policy

feedback and policy learning are employed in the study of public policy. The second section discusses

the mechanisms and tools that facilitate or impede policy learning, with a particular attention

given to commissions.

What is policy feedback and policy learning?

This section discusses both subﬁelds and how they matter when it comes to evaluating policy

and fostering or hindering policy change. Each subﬁeld is deﬁned, analyzed and explained via

these lenses. Special attention is put onto the diﬀerent epistemological roots associated with

these concepts and how this impacts the analysis of public policy.

Policy feedback

Nearing the end of her high school life, Isabelle must decide what subject she will study as a

senior. Should she follow in the footstep of her parents and study chemistry or pursue her

dreams of becoming a historian? She opts for the later and selects a wide range of social science

courses rather than studying chemistry and physics. According to students of policy feedbacks, Isa-

belle has made a crucial decision on her future. If she pursues this path in university, she might

end up being a history professor at a high school or university in the future. Like so many students,

however, she could change her mind. The literature on policy feedback would stress that this

outcome would be far more likely during her ﬁrst year of university than after completing her

graduate studies in history. The costs of reversing back to chemistry increase every year while

401

she remains on the history path. Expressed diﬀerently, a decision taken at t

alters the calculus for

options at t

, which becomes more and more entrenched further in time.

This is, in a nutshell, the logic underpinning policy feedbacks. It explains why we continue

to employ a QWERTY keyboard instead of the vastly superior Dvorak model (David 1985)

and, more controversially, why Americans do not have a universal health system (Hacker 2002).

Each action building on a prior policy decision, akin to Isabelle following more history courses,

generates positive feedback because it strengthens the importance and value of earlier decisions

and, as importantly, increases signiﬁcantly the cost of reversal. Economists refer to this as “increasing

returns” (Pierson 2000; Arthur 1994), while similar phenomena have been discussed in policy

studies under the umbrella of policy succession (Hogwood and Peters 1982) and policy inheritance

(Rose 1990).

In political science, policy feedback owes its origin to the argument that “policies create

politics” (Lowi 1969; Schattschneider 1935) and the historical institutionalism school (Pierson

1993). Policy feedbacks are also more common than in economics for multiple reasons (see

Pierson 2004 for an in-depth discussion). First, policy reversal is more diﬃcult to enact with

political institutions than markets where underperforming ﬁrms and products are replaced by

new ﬁrms and/or products. Consumers can choose the product or service they desire while

public goods tend to be compulsory and exhibit non-excludability (ibid.: 30–1). Second, public

policies tend to empower beneﬁciaries resulting in the creation of interest groups to safeguard

these new beneﬁts (Skocpol 1992). Interest groups are particularly important in political science

since citizens, contrary to consumers, are engaged infrequently with the political system (Pierson

2004). Third, new policies often result in the creation of bureaucratic organization whose very

existence is tied in to the success and expansion of speciﬁc programs (Derthick 1979). This

results in policy professionals highly committed to protecting their program (Ross 2000: 18).

Finally, studies have demonstrated that voters are more likely to remember negative experiences

such as cutbacks than positive news like the creation of a program (Weaver 1986)

With the past playing such a predominant role, particular attention is given to the initial

adoption of policies, formative moments (Rothstein 1992), and critical junctures (Collier and

Collier 1991), which represent exceptional moments where opportunities arise to challenge pre-

existing policy legacies. Although the concept of critical junctures seems reminiscent of King-

don’s (2003) “window of opportunity,” these moments are far more exceptional and infrequent

than what students of agenda-setting suggest (Hacker 2002) since they tend to require events

such as wars, regime changes, or major economic recessions.

Central to the study of policy feedback is the concept of path dependence, which is deﬁned

succinctly by Hacker as being “developmental trajectories that are inherently diﬃcult to reverse”

(2002; 54; see also Kay in this volume). According to Pierson (2004: 44), there are four key

features associated with path dependence processes. First, multiple outcomes are possible at the

initial stage. This is where an actual path is created. Many comparative studies begin with a

similar critical juncture and explain divergent outcomeS as a result of having opted for diﬀerent

paths. For example, Weir and Skocpol highlight that the lack of a cross-alliance between the labor

and agrarian movements in the US explains why the Swedish Social Democratic party, built upon

such a cross-alliance, has been much more successful in implementing a generous and encom-

passing social agenda (Weir and Skocpol 1985). Second, the starting point for a path may be a

relatively minor event. In his study of the American welfare state, Hacker points to a 1974 last-

minute amendment to the Employee Retirement Income Security Act (ERISA) bill that is at the

origin of a weakening of employers’ pension plans (2004: 254–5). Third, when an event occurs

and in which sequence events occur matter greatly. This is because earlier events are far more

important than those occurring at a later date. For example, countries who succeeded in

Marier

402

introducing a value-added tax during the golden age of the welfare state have been able to diversify

their source of revenue and obtained better protection against retrenchment (Kato 2003). Fourth,

once a process is well established, the policy output is likely to remain quite stable and very resistant

to change. In the comparative welfare state literature, multiple debates in the past 20 years have

actually been focused on whether or not the welfare state is an “immovable object” (Pierson 1998).

In policy studies, the logic of policy feedback and the concept of path dependence have enjoyed

tremendous appeal to explain the evolution and strong stability of many social policies such as

pay-as-you-go pension programs. The adoption of social security during the great depression

has actually been utilized to describe and compare six streams of policy feedback (Béland 2010).

Policy feedback from social security has resulted in the creation of future beneﬁciaries who are

now well organized within the AARP (Pierson 1994), in the mobilization of the electorate

(Campbell 2003) and in the expansion of the social security administration, which has played a

key role in promoting and expanding the program (Derthick 1979). Other important elements

point to its stability. Potential reformers not only have to tackle the mobilized beneﬁciaries and

the Social Security Association, but also issues related to the so-called double payment pro-

blem. After years of contributions, there is a strong sense of entitlement embedded with current

and past contributors. With past contributions utilized to provide pension bene ﬁts to the pre-

vious generation, the introduction of privately funded pension plans requires contributions to

be made to both public schemes—to honor prior commitments—and private schemes to build

up assets in the funded scheme (Myles and Pierson 2001). The American political institutions,

riddled with veto points, also makes it extremely diﬃcult to build a coalition to reform social

security (Pierson, 1994).

More recently, the literature has evolved by stressing that a distinction should be made between

positive and negative policy feedback. The later stresses the “the consequences of policies that

tend to undermine rather than reinforce the political, ﬁscal, or social sustainability of a particular

set of policies” (Weaver 2010: 137). Although not identiﬁed as such by many authors, negative

policy feedback constitutes a strong impetus for policy learning (see the section below). For

example, the adoption of monetarism in the UK, which is presented as a case of learning

leading to a paradigm shift (Hall 1993) has also been conceived as a response to negative policy

feedback associated with Keynesianism (Pemberton 2000; Greener 2001). Another interesting

development in the literature on policy feedback has been linked to policy instruments and

policy change. For example, Hacker (2004) develops four kinds of policy change (drift, con-

version, layering, and revision) depending on the number of veto players in the political arena

and the level of policy discretion coupled with its support coalitions. These policy changes are

associated with a diﬀerent use of policy instruments.

Critics of policy feedback have emphasized two key points. First, with an underlying

assumption that policies create politics, the latter is marginalized. While the political orientation

of policy actors may matter when a policy is created, the importance of partisanship decreases to

the point where political leadership, and conﬂict, no longer plays a vital role once a policy or

policy constellation has matured (Ross 2000; Korpi and Palme 2003; Peters et al. 2005). Second,

and reminiscent of previous debates surrounding incrementalism, the policy feedback literature

has diﬃculties accounting for change. Incremental change, which can accumulate and represent

a substantial policy departure, remains largely ignored (Hinrichs and Kangas 2003; Djelic and

Quack 2007; Kay 2005). Most of the attention is given to patterns of persistence and institu-

tionalization, which is often reﬂected in the selection of cases (Peters et al. 2005). The special

attention given to pension policy is a case in point. With high sunk costs, long term commit-

ments, well-established pension agencies and a growing cohort of politically active beneﬁciaries,

it possesses the core attributes to justify the claims of policy feedback arguments.

Policy feedback and learning

403

Policy learning

The analysis of learning has preoccupied researchers in a wide variety of disciplines such as psy-

chology, education, economics, sociology and business. As a result, it is not surprising to notice

that there are multiple conceptualizations and foci surrounding policy learning (Bennett and

Howlett 1992; Dolowitz and Marsh 2000; Marsh and Sharman 2009; Braun and Gilardi 2006;

Radaelli 1995; Evans and Davies 1999). Nonetheless, a common point of departure in public

policy is to deﬁne policy learning as “a process of updating beliefs about key components of

policy” (Radaelli 2009: 1146–7). This deﬁnition encapsulates all kinds of knowledge that can be

acquired by various learning mechanisms, which reﬂects the eclectic environment of policy-

makers. In this brief overview of policy learning, the focus is primarily on studies that employ

the following concepts: lesson-drawing, social learning, organizational learning, policy transfer,

emulation, adaptation, and policy diﬀusion.

Despite the diversity surrounding policy learning, there are core elements unifying the lit-

erature that also distinguishes it from other public policy writings. First, there is a prevalence on

the importance of knowledge within the policy-making process at the expense of other factors

such as politics, institutions, values, or agendas (Radaelli 1995). As stressed by Heclo in his

classical study of social policy-making in Sweden and the UK, “politics ﬁnds its sources not only

in power but also in uncertainty … policy-making is a form of collective puzzlement on society’s

behalf; it entails both deciding and knowing” (1974: 304–5). As such, there is an underlying

assumption that policy-makers can draw lessons—good and bad—from a variety of sources to

replace or improve public policies and programs (Rose 1991). Borrowing originally from the

diﬀusion of technological discoveries, the literature on policy diﬀusion assumes that good inno-

vations travel and it tends to focus on the determinants that facilitate or impede di ﬀusion and

the actual rate of diﬀusion with the so-called S-shape pattern (Walker 1969; Berry and Berry

2007; Weyland 2007).

Second, an underlying assumption in the literature is that learning must occur without coercion

(Freeman 2006), although this is not embraced in some studies on policy diﬀusion and transfer

where learning is not a precondition to the adoption of a policy by a host country (Dolowitz

and Marsh 1996). Thus, learning is a co-operative undertaking, albeit sometime highly com-

petitive (Dunlop and James 2007). While there are diﬀerent forms of learning in public policy,

like social learning (Heclo 1974; Hall 1993), organizational learning (Hedberg 1981; Levitt and

March 1988; March 1991) and even political learning (May 1992), a core assumption remains

that there is an exchange of information between various actors, organizations, and govern-

ments resulting in valuable lessons. As in the classroom, neither the importer nor the exporter of

lessons is passive and both parties engage in meaningful exchanges for learning to be successful.

There are varying degrees of engagements and this results in a broad theoretical and empirical

focus stressing multiple facets of this exchange such as the underlying conditions pushing a

group, organization, or country to seek out a lesson (Rose 1991; Bennett 1997; Bennett 1991),

the mechanism or underlying structural conditions favoring the transfer or diﬀusion of a lesson

(Walker 1969; Berry and Berry 2007; Shipan and Volden 2008), and the motivations of the

exporter of lessons (Orenstein 2008; Brooks 2004).

Third, although rarely expressed so explicitly, policy learning assumes that the knowledge

acquired is fungible (Rose 1993: Chapter 6) and easily accessible (Bennett 1991: 51). Beyond its

intrinsic appeal, policy learning involves the translation of a lesson into practice and it is important

“to distinguish between the knowledge of a (foreign) program, the utilization of that knowl-

edge and the adoption of the same program” (Bennett 1991). A number of contributions have

expanded the boundaries of knowledge utilization recently with, for example, contributions on

Marier

404

the use of second-hand experience (Barzelay 2007) and on the importance of symbolic knowledge

utilization where knowledge serves the purpose of increasing the legitimacy of an organization

rather than improve its performance (Boswell 2008). Adaptation remains particularly important

if one is concerned with policy transfer and diﬀusion. For example, the success of the World

Bank in pension policy is largely attributed to the development of a model that was highly

permutable resulting in its adoption in a high number of countries (Orenstein 2008).

Beyond traditional institutional barriers, the type of policy instruments matters greatly when

it comes to exporting a lesson or policy in another jurisdiction. Contrary to programs with

strong policy feedbacks, like national health care services, regulations are highly fungible. In an

analysis of policy diﬀusion between Sweden and Finland, Karvonen (1981) emphasizes that the

most identical policies adopted from Finnish authorities are regulations and that policy diﬀusion

in the United States are facilitated by their strong preferences for regulations.

The fourth commonality is the broad acknowledgment that not all information is of equal

value or weight in the learning process. This can be as a result of the quality of the information

or simply because of the standing of the individual or group communicating the information,

such as experts in an epistemic community (Haas 1992; Ruggie 1975). There is also evidence

suggesting that the political orientation of a government matters. For example, right-wing gov-

ernments are more likely to acquire information and learn from other right-wing governments

in the American states (Robertson 1991). Gilardi (2010) goes further by demonstrating a diﬀer-

entiation in what is being studied with right-wing governments focusing on policy information

related to their re-election and left-wing governments preoccupied mostly by policy eﬀects.

Organizational learning provides another interesting example of how a certain kind of infor-

mation dominates the learning process. Within organizations, lessons from previous experiences

are deeper than those originating from practice. These are rooted in history, where previous

experiences—and the interpretation of these experiences—shape a collective memory (Levitt

and March 1988). This has a dual impact; it can ensure that failed alternatives are not attempted

again but it can also prevent policy alternatives to emerge.

Finally, there is an underlying assumption that learning leads, at the very least, to a policy

reassessment, but often in policy change. In the case of policy diﬀusion, this also takes the form

of convergence, as more and more jurisdictions adapt an innovation (but see Radaelli 2004). An

extensive search for lessons is often triggered when the status quo is no longer an option (Rose

1991: 10). The importance of knowledge is important for both problem recognition and the

development of an alternative. In ways reminiscent of the policy entrepreneur in Kingdon (2003),

groups empowered by their expertise often seek to demonstrate the failing of a policy arrangement

while actively promoting a solution. These actors can be endogenous or external to the jur-

isdictions receiving the lesson and they include civil servants (Hall 1993; Heclo 1974; Etheredge

and Short 1983), elected oﬃcials (Grossback et al. 2004; Gilardi, 2010; Marier 2008), policy

entrepreneurs (Mintrom 1997), policy networks (Coleman et al. 1997; Haas 1990; Evans and

Davies 1999), international organizations (Cao 2009; Orenstein 2008), economists (Markoﬀ and

Montecinos 1993), and public inquiries (Bradford 1998; Inwood 2005; Marier 2009).

Policy learning also occurs during the normal operations of government (Etheredge and Short

1983), which often results in incremental change. Bureaucrats learn from the policy itself and

make adjustments to improve it (Heclo 1974; Lindblom 1959). Within this context, learning often

results in change of the ﬁrst and second order, which consists of ﬁne-tuning previous policies

and attempting to use new instruments to achieve policy goals (Hall 1993). The long-term

eﬀects of “learning by doing” may result in a “competency trap” where most learning happens

while exploiting current practices at the expense of exploring better or more eﬃcient alter-

natives (March 1991). It is far easier to improve what has been used rather than assimilate new

Policy feedback and learning

405

learning. However, the more one invests in improving current practices, the less likely is an

organization willing to adopt better alternatives especially if the implementation costs are substantial.

The end result is the persistent use of sub-optimal procedures (Levitt and March, 1988).

Interestingly, the inability of a public administration to absorb, gather, and analyze lessons can

be at the origin of a diﬀusion or transfer. This can be as a result of past policy failures, limited

policy capacity or lack of time. In her comparative study of Estonia and Latvia, Tavits (2003)

demonstrates that Latvia embraced the Swedish pension reform because its public administration

did not have the capacity and conﬁdence from the public to enact its own, following past

economic policy failures while Estonian authorities had high conﬁdence in the ability of their

civil servant to devise a good pension system of their own. The adoption of Western policies in

Meiji Japan occurred because there was an urgent need to develop policies quickly (Westney,

1987). In his analysis of pension privatization diﬀusion in Latin America, Weyland (2005) argues

that the newfound knowledge does not even have to be well anchored within the policy elites.

The perception of a highly successful policy in a neighboring jurisdiction may be suﬃcient to trigger

its widespread adoption because of the cognitive heuristics utilized by most policy-makers.

Three important criticisms of the literature on policy learning concerns its diversity, issues

related to the so-called Galton’s problem and the marginalization of politics. First, the multi-

plicity of concepts related to learning and divergence related to the independent and dependent

variables continue to plague the literature (Eyestone 1977; James and Lodge 2003). The ways in

which a policy or lesson is deﬁned is crucial. A closer inspection of programs and instruments

can reveal that a diﬀusion process is no longer present because an adopted policy bears little resem-

blance to the one enacted by leaders (Clark 1985). Underpinning these debates is the continuous

divide between qualitative and quantitative studies (Marsh and Sharman 2009). The quantitative

corpus relies on event history analyses to study the diﬀusion of a policy innovation (Collier and

Messick 1975; Walker 1969; Berry and Berry 2007). The theoretical knowledge and techniques

originating from these studies have also been applied to study the diﬀusion of policies across

nations such as the liberalization of the economy (Simmons and Elkins 2004), the privatization

of pensions (Brooks 2005), and hospital ﬁnancing (Gilardi et al. 2009). As such, most of these

studies are interested primarily in explaining what characteristics account for the adoption of a

lesson or innovation from another jurisdiction. According to Howlett and Rayner, in spite of

improved quantitative techniques, confusion remains with regards to what is changing and what

is being diﬀused (2008: 387) with Rose expressing disbelief about the lack of attention given to

the content of the programs diﬀused (1993: 25).

The body of work relying primarily on qualitative research designs and comparative case

studies has focused mostly on lesson-drawing (Rose 1991), transfers (Dolowitz and Marsh 2000;

Radaelli 2000), emulation (Westney 1987) and adoption (Bennett 1997) of policies, programs

and/or instruments across jurisdictions. Contrary to studies on diﬀusion, these contributions pay

closer attention to the decision-making process, the relationship between the actors involved

and the motivations that push a state to learn from abroad. However, as in the case of policy

diﬀusion, there is also a lack of communality among these studies concerning the conceptualiza-

tion of learning, the dependent and independent variables. The later is quite important since

few studies have stressed that learning from abroad should be considered as an independent

variable to explain policy change.

Second, Galton’s problem where the motivations and sources behind the adoption of a policy

can be endogenous rather than from an independent lesson remains an important issues in policy

learning, but particularly in policy diﬀusion where the adoption of an innovation from an

external source is the primary element under investigation (Braun and Gilardi 2006). There is a

need to demonstrate that a policy change—or a very diﬀerent kind of policy change—could

Marier

406

not have occurred without the presence of learning (Bennett and Howlett 1992: 290) or the

presence of a leader in policy diﬀusion. In a policy environment where information is easily avail-

able, this is increasingly problematic and diﬃcult to disentangle. Thus, the distinction becomes

much more of an issue related to whether knowledge from another source is being utilized or not

and in conjunction with what other sources of knowledge. There are still a lot of improvements

needed to understand how governments are learning from one another (Grossback et al. 2004).

Third, as in policy feedback, critics have mentioned the marginalization of politics. For example,

the epistemic community approach overstates the importance of uncertainty and knowledge

and the ability of policy-makers to act without the help of experts (Dunlop 2009). Politicians

have multiple tools at their disposal to counter or minimize the inﬂuence of an epistemic com-

munity or network of experts, such as adding experts from another ﬁeld in consultative bodies

and empowering the expertise that closely resembles their own policy preferences. The ability

to navigate throughout the policy process is also an important form of knowledge that is often

ignored; this form of knowledge is possessed primarily by politicians. In a nutshell, the translation

of learning into policy remains a highly political aﬀair. A potent combination to enact policy

change includes the expertise in both knowledge and process (Marier 2008) or at least connections

to both kinds of experts (Mintrom 1997).

Defeating the policy feedback and policy learning barriers—the role

of commissions

Both policy feedback and policy learning tend to underscore the capacity to enact policy change.

While these limitations are more obvious in the case of policy feedback, policy learning also

faces important hurdles such as the institutionalization of professional groups who often deﬁne

who is an expert if not expertise itself (Abbott 1988; Eisner 1993), a bias towards learning from

experience in organizations (Levitt and March 1988), and the traditional political barriers asso-

ciated with the policy-making process such as veto players (Tsebelis, 2002) and veto points

(Immergut 1992). This also includes the organizational structure of government, which is not a

monolithic institution but rather a collection of various agencies and departments with their

own organizational culture and operating procedures. Once a practice has been adopted among

various departments and agencies to handle a speciﬁc policy issue, exceptional circumstances are

necessary to alter this relationship.

In order to foster policy learning in this environment and shift discussions away from those

engendered by policy feedbacks, governments may employ other sources of knowledge, such as

the open method of coordination in the case of the European Union (de la Porte and Nanz

2004; Jacobsson 2004; Kerber and Eckardt 2007; Nedergaard 2006), think tanks (Weaver 1989;

Stone and Garnett 1998), and commission of inquiries. This section focuses on the latter.

Commission of inquiries—expanding the learning potential of government

The creation of a commission, deﬁned as “a body to which some task has been referred or

committed by some person or body” (Wheare 1955: 6), by a government can result in a transfor-

mation of the policy landscape if it is set up to encourage policy learning. There are multiple

objectives associated with the creation of a commission that actually runs counter to fostering a

learning environment. A government may actually want to delay a discussion on a controversial

subject (ibid.: 91–2), to drum up support for its own projects (Bulmer, 1983), or even to avoid blame

in the enactment of unpopular measures. Commissions are more potent when they are created

to fulﬁll learning objectives such as enhancing knowledge, facilitating political compromise and

Policy feedback and learning

407

educating the general population (Marier 2009). First, a key underlying theme in the policy learn-

ing literature is the search to curb the level of uncertainty in public policy. Contrary to what is

available within individual ministries under normal circumstances, commissions possess the resour-

ces and the time to perform an in-depth assessment of policies (Olsen 1983; Bradford 1998). In

a nutshell, they operate in a privileged environment, away from the traditional bureaucracy,

allowing them to explore and innovate (Sheriﬀ 1983). As a result, commissions have the capacity

to tackle complex issues and relationships within the policy space and provide recommendations

to resolve ongoing problems (Bulmer 1993).

With these assets, commissions can play a critical role in facilitating learning to improve all

forms of policy knowledge. This goes beyond policy prescriptions since the ideas and paradigms

they promote can remain in public debates for years (Jenson 1994). An excellent example is the

McDonald Commission, which altered profoundly the economic policy goals of the Canadian

government. Prior to its creation, only “a small number of senior government bureaucrats, a

fraction of the business community, and mainstream economists” supported a continental economic

strategy in Canada (Inwood 2005: 309). The McDonald Commission led Canadian authorities

to embrace free trade with the United States and abandon a long-standing national approach to

economic development, which facilitated a paradigm shift (ibid.).

The second learning objective of a commission is to facilitate the negotiation of a political

compromise surrounding a controversial issue. Some political institutions, such as the Joint

Committees in the United States, perform this role to a certain extent. However, commissions

are notorious for granting a voice to multiple groups beyond organized political parties and for

having far more latitude with regard to the elements under discussions. The presence of strong

parliamentary commissions in Norway and Sweden, where experts and politicians interact, have

represented an important forum to generate compromises among political parties, labor unions

and employers’ associations (Heclo 1974; Premfors 1983; Olsen 1983). As such, a government

may want to utilize a commission to test the feasibility of an idea that requires more analysis and

see whether it can be supported by other key stakeholders (Bulmer 1983). A recent example of

a commission created with the purpose of facilitating the acquisition of knowledge and securing

a political compromise is the Swedish working group on pensions, which was composed primarily

of politicians and experts drawn from the civil service (Marier 2008).

Third, commissions can be established to educate the public about controversial policy issues,

such as the use of nuclear energy. Commissions are utilized to debunk beliefs and advertise recent

research in the policy area under study (Chapman 1973: 184). In these instances, a successful

commission alters the policy discourse by marginalizing prior beliefs and reassessing problems and

solutions (Inwood 2005).

The policy impact of commission

A commission has the capacity to expand signiﬁcantly the knowledge possessed by public

authorities. However, the learning experience itself does not guarantee a transformation of public

policies. It is important to distinguish between the acquisition of knowledge and the actual

impact of a commission on policy. A lot of learning can take place and the commission may have

no policy impact. Many variables are involved in the translation of the knowledge acquired

during the workings of a commission into a new policy beyond the institutional barriers gen-

erated by political systems. The ﬁrst key hurdle is the simple fact that commissions’ recom-

mendations are usually nonbinding. Political authorities have the ability to dismiss or ignore the

ﬁndings of a commission. That being said, it is important to note that recommendations from

commissions may be di ﬃcult to ignore since they can set policy discussions moving forward

Marier

408

(Bulmer 1983: 663), especially if there is unanimous support among commission members (Park

1972) and stakeholder groups.

Second, the ability of a commission to reshape the policy space is constrained by the terms of

reference. For example, the inﬂuence of a commission will likely be stronger if a government

asks for measures to reform a policy rather than providing an assessment. The terms of reference

represent an important tool possessed by government to steer discussions, but there is no guar-

antee that commission’s recommendations will please a government since the workings of a

commission can diverge signiﬁcantly from government’s expectations resulting in unexpected

ﬁndings and recommendations (Jenson 1994: 54).

Finally, the number of individuals involved in a commission and their respective aﬃliations

plays an important role in determining whether the solutions proposed will have broad appeal

and whether the commission will have the capacity to challenge existing policies. Based on the

literature on veto players (Tsebelis 2002), a commission with a very large membership repre-

senting very divergent interests is unlikely to yield recommendations that have far-reaching

consequences on current policies. In fact, it is more likely that they would produce a report

with a lot of dissension and minority opinions. Therefore, commissions with a small member-

ship are preferable, if the goal is to stimulate learning and bring about policy recommendations

that diﬀer from current practices. Who is selected in these commissions matters greatly. Gov-

ernments can select members who are clearly identiﬁed with a policy position or associated with

the government. However, such commissions are unlikely to generate the kind of recommen-

dations that would obtain broad political support and succeed in altering the point of view of

key political players. The most potent commissions, in terms of having the highest likelihood of

generation recommendations that could potentially alter the policy landscape, tend to be the

ones with a small independent membership composed of recognized experts in the ﬁeld.

A good example of a recent commission that features all three elements stated above is the

latest British Pension Commission. The three members were not linked to any political party;

they included an individual with ties to the business sector, one with union ties and a university

professor. As a result, they had broad political support for their activities. They worked closely

with governmental oﬃcials to generate an exhaustive assessment of British pension policy and

clear policy directions to resolve its shortcomings. Many of their recommendations were

implemented (Marier 2009).

Conclusion

As demonstrated in this chapter, there is a rich and burgeoning literature on policy feedback and

policy learning. The former has generated multiple studies to explain why policy reform remains

so diﬃcult once a policy or program has been put in place. The emerging literature on negative

policy feedback brings hope that the literature will slowly shy away from focusing primarily on

stability of policy structures (Weaver 2010). The literature on policy learning is highly diversi-

ﬁed, but remains committed to few core elements such as the importance of knowledge in

public policy, the acquisition of knowledge and how individuals and/or organizations learn.

Recent eﬀorts to categorize the literature (Dunlop and Radaelli 2011) and engage in dialogues

to clarify concepts, variables and expectations (Marsh and Sharman 2009) are promising since

the eclectic nature of the literature is both its primary asset and weakness.

Policy feedback and policy learning actually have a lot in common when it comes to describing

the process within which governmental authorities learn from the enactment of policies and

programs and make incremental adjustments as a result. Within the policy learning literature,

organizational learning is consistent with the literature on positive feedback where self-reinforcing

Policy feedback and learning

409

mechanisms entrench programs further at the expense of competing policy proposal. However,

the literature on policy learning assumes that learning can be much broader and emphasizes

policy change while policy feedback stresses mostly the lack of change.

Many of the references discussed in this chapter refer to pension policy because this policy

area illustrates well the diﬀerences between policy feedback and policy learning. On one hand,

there is a wealth of contributions related to the lack of reforms in the pension systems of

advanced industrialized countries. Learning is extremely limited and policy feedbacks are respon-

sible for the limited reforms or the status quo. Pensions represent the case by excellence for

policy feedback with high sunk costs, large upcoming cohorts of beneﬁciaries, and broad pop-

ular appeal. On the other hand, the literature on policy learning, and particularly contributions

on policy diﬀusion, has emphasized rapid change and the adoption of pension privatization in a

growing number of countries, often spearheaded by international forces. However, these pat-

terns of diﬀusion do not apply to advanced industrialized countries (Brooks 2005). One would

be led to conclude that policy learning remains marginal when there are strong policy feedback

eﬀects. However, as demonstrated in the later part of this chapter, the creation of commissions

is one of the tools that can be utilized by policy-makers to enhance learning beyond the incre-

mental options generated by the interpretation of actors involved in managing public programs,

which are strongly associated with policy feedback.

The marginalization of politics and the measurement of policy change remain two core issues

within both the policy feedback and policy learning literature. First, very few contributions

have truly integrated traditional political variables, such as the political orientation of a govern-

ment, within their contributions on learning (but see Gilardi et al. 2009; Robertson 1991) and

policy feedback. The interpretation of a lesson or policy feedback is highly political and who is

in power should matter. For a broader appeal in political science, some notions of political

leadership, choice and interest need to be included in both cases moving forward (for policy

feedback see Ross 2000). Second, a recurring problem in policy studies is to measure change.

The literature on policy feedback has brought back discussions akin to those that were prevalent

in debates surrounding incrementalism (Dempster and Wildavsky 1979) while some contributions

in policy learning have overemphasized the place of learning and/or the extent to which lessons

from abroad are truly exported (Clark 1985; Howlett and Rayner 2008). These issues are unlikely

to ﬁnd a satisfying resolution, but a better acknowledgment of these issues would go a long way

to foster a genuine dialogue among researchers in both policy feedback and policy learning.

Note

1 This section draws from Marier (2009).

Bibliography

Abbott, A. 1988. The System of Professions: An Essay on the Division of Expert Labor. Chicago, IL: University

of Chicago Press.

Arthur, W.B. 1994. Increasing Returns and Path Dependence in the Economy. Ann Arbor, MI: University of

Michigan Press.

Barzelay, M. 2007. ‘Learning from second-hand experience: Methodology for extrapolation-oriented case

research’, Governance 20: 521–43.

Béland, D. 2010. ‘Reconsidering policy feedback: How policies aﬀect politics’, Administration and Society

42: 568–90.

Bennett, C.J. 1991. ‘How states utilize foreign evidence’, Journal of Public Policy 11: 31–54.

—— 1997. ‘Understanding ripple eﬀects: the cross-national adoption of policy instruments for bureaucratic

accountability’, Governance: An International Journal of Policy and Administration 10: 213–33.

Marier

410

—— and M. Howlett. 1992. ‘The lessons of learning: Reconciling theories of policy learning’, Policy Sciences

25: 275–94.

Berry, F.S. and W.D. Berry. 2007. ‘Innovation and diﬀusion models in policy research’, in P.A. Sabatier

(ed.) Theories of the Policy Process, 2nd edn. Boulder, CO: Westview Press.

Boswell, C. 2008. ‘The political functions of expert knowledge: Knowledge and legitimation in European

Union immigration policy’, Journal of European Public Policy 15: 471–88.

Bradford, N. 1998. Commissioning Ideas: Canadian National Policy Innovation in Comparative Perspective, Toronto:

Oxford University Press.

Braun, D. and F. Gilardi. 2006. ‘Taking “Galton’s problem” seriously: Towards a theory of policy diﬀusion’,

Journal of Theoretical Politics 18: 298–322.

Brooks, S.M. 2004. ‘What was the role of international ﬁnancial institutions in the diﬀusion of social

security reform in Latin America?’, in K. Weyland (ed.) Learning from Foreign Models in Latin American

Policy Reform. Washington, DC: Woodrow Wilson Center Press.

—— 2005. ‘Interdependent and domestic foundations of policy change: the diﬀusion of pension privatization

around the world’, International Studies Quarterly 49: 273–94.

Bulmer, M. 1983. ‘Introduction: commissions as instruments for policy research’, American Behavioral Scientist

26: 559–67.

—— 1993. ‘The Royal Commission and departmental committee in the British policy-making process’,in

B.G. Peters and A. Barker (eds) Advising West European Governments: Inquiries, Expertise and Public Policy.

Edinburgh: Edinburgh University Press.

Campbell, A.L. 2003. How Policies Make Citizens: Senior Political Activism and the American Welfare State.

Princeton, NJ: Princeton University Press.

Cao, X. 2009. ‘Networks of intergovernmental organizations and convergence in domestic economic

Policies’, International

Studies Quarterly 53: 1095–130.

Chapman,

R.A. 1973. ‘Commissions in policy-making’, in R.A. Chapman (ed.) The Role of Commissions in

Policy-Making. London: George Allen and Unwin.

Clark, J. 1985. ‘Policy diﬀusion and program scope: research directions’, Publius 15: 61–70.

Coleman, W.D., G.D. Skogstad and M.M. Atkinson. 1997. ‘Paradigm shifts and policy networks: cumulative

change in agriculture’, Journal of Public Policy 16: 273–301.

Collier, D. and R.E. Messick. 1975. ‘Prerequisites versus diﬀusion: Testing alternative explanations of

social security adoption’, The American Political Science Review 69: 1299–315.

—— and D. Collier. 1991. Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime

Dynamics in Latin America. Princeton, NJ: Princeton University Press.

David, P.A. 1985. ‘Clio and the economics of QWERTY’ , American Economic Review 75: 332–7.

de la Porte, C. and P. Nanz. 2004. ‘The OMC: A deliberative-democratic mode of governance? The cases

of employment and pensions’, Journal of European Public Policy 11: 267–88.

Dempster, M. and A. Wildavsky. 1979. ‘On change, or there is no magic size for an increment?’, Political

Studies 27: 371–89.

Derthick, M. 1979. Policymaking for Social Security. Washington, DC: Brookings Institution.

Djelic, M.-L. and S. Quack. 2007. ‘Overcoming path dependency: Path generation in open systems’,

Theory and Society 36: 161–86.

Dolowitz, D. and D. Marsh. 1996. ‘Who learns what from whom: A review of the policy transfer literature’,

Political Studies 44: 343–57.

—— and

D. Marsh.

2000. ‘Learning from abroad: The role of policy transfer in contemporary policy-making’,

Governance 13: 5–24.

Dunlop, C.A. 2009. ‘Policy transfer as learning: Capturing variation in what decision-makers learn from

epistemic communities’, Policy Studies 30: 289–311.

—— and O. James. 2007. ‘Principal-agent modelling and learning: The European Commission, experts

and agricultural hormone growth promoters’, Public Policy and Administration 22: 403–22.

—— and C.M. Radaelli. 2011. ‘Systematizing policy learning: From monoliths to dimensions’, European

Consortium for Political Research, Joint Session Workshop, St. Gallen.

Eisner, M.A. 1993. ‘Bureaucratic professionalization and the limits of the political control thesis: The case

of the federal trade commission’, Governance 6: 127–53.

Etheredge, L.S. and J. Short. 1983. ‘Thinking about government learning’, Journal of Management Studies

20: 41–58.

Evans, M. and J. Davies. 1999. ‘Understanding policy transfer: A multi-level, multi-disciplinary perspective’,

Public Administration 77: 361–85.

Policy feedback and learning

411

Eyestone, R. 1977. ‘Confusion, diﬀusion, and innovation’, American Political Science Review 71: 441–7.

Freeman, R. 2006. ‘Learning in public policy’, in M. Moran, M. Rein and R.E. Goodin (eds) The Oxford

Handbook of Public Policy. Oxford: Oxford University Press.

Gilardi, F. 2010. ‘Who learns from what in policy diﬀusion processes?’, American Journal of Political Science

54: 650–66.

——, K. Fuglister and S. Luyet. 2009. ‘Learning from others: The diﬀusion of hospital ﬁnancing reforms

in OECD countries’, Comparative Political Studies 42: 549–73.

Greener, I. 2001. ‘Social learning and macroeconomic policy in Britain’, Journal of Public Policy 21: 133–52.

Grossback, L.J., S. Nicholson-Crotty and D.A.M. Peterson. 2004. ‘Ideology and learning in policy diﬀusion’,

American Politics Research 32: 521–45.

Haas, P.M. 1990. Saving the Mediterranean: The Politics of International Environmental Cooperation. New York,

Columbia University Press.

—— 1992. ‘Introduction: Epistemic communities and international policy coordination’, International

Organization 46: 1–35.

Hacker, J.S. 2002. The Divided Welfare State. Cambridge, Cambridge University Press.

—— 2004. ‘Privatizing risk without privatizing the welfare state: The hidden politics of social policy

retrenchment in the United States’, American Political Science Review 98: 243 – 60.

Hall, P.A. 1993. ‘Policy paradigms, social learning, and the state: The case of economy policymaking in

Britain’, Comparative Politics 25: 275–96.

Heclo,

H. 1974. Modern Social

Politics in Britain and Sweden. New Haven, CT: Yale University Press.

Hedberg, B. 1981. ‘How organizations learn and unlearn’, in P.C. Nystrom and W.H. Starbuck (eds)

Handbook of Organizational Design. Oxford: Oxford University Press.

Hinrichs, K. and O. Kangas. 2003. ‘When is a change big enough to be a system shift? Small system-shifting

changes in German and Finnish pension policies’, Social Policy and Administration 37: 573–91.

Hogwood, B.W. and B.G. Peters. 1982. ‘The dynamics of policy change’, Policy Sciences 14: 225– 45.

Howlett, M. and J. Rayner. 2008. ‘Third generation policy diﬀusion studies and the analysis of policy

mixes: Two steps forward and one step back?’, Journal of Comparative Policy Analysis 10: 385–402.

Immergut, E.M. 1992. Health Politics: Interests and Institutions in Western Europe. Cambridge Cambridge

University Press

Inwood, G.J. 2005. Continentalizing Canada: The Politics and Legacy of the Macdonald Royal Commission.

Toronto, University of Toronto Press.

Jacobsson, K. 2004. ‘Soft regulation and the subtle transformation of states: The case of EU employment

policy’, Journal of European Social Policy 14: 355–70.

James, O. and M. Lodge. 2003. ‘The limitations of “policy transfer” and “lesson drawing” for public policy

research’, Political Studies Review 1: 179–93.

Jenson, J. 1994. ‘Commissioning ideas: Representation and Royal Commissions’, in S.D. Phillips, (ed.)

How Ottawa Spends 1994–1995: Making Change. Ottawa: Carleton University Press.

Karvonen, L. 1981. ‘Semi-domestic politics: Policy diﬀusion from Sweden to Finland’, Cooperation and

Conﬂict 16: 91–107.

Kato,

J. 2003. Regressive Taxation

and the Welfare State: Path Dependence and Policy Formation. Cambridge

Cambridge University Press.

Kay, A. 2005. ‘A critique of the use of path dependency in policy studies’, Public Administration 83: 553–71.

Kerber, W. and M. Eckardt. 2007. ‘Policy learning in Europe: The open method of co-ordination and

laboratory federalism’, Journal of European Public Policy 14: 227–47.

Kingdon, J.W. 2003. Agendas, Alternatives, and Public Policies. New York, Longman.

Korpi, W. and J. Palme. 2003. ‘New politics and class politics in the context of austerity and globalization:

Welfare state regress in 18 countries’, American Political Science Review 97: 425–46.

Levitt, B. and J.G. March. 1988. ‘Organizational learning’, Annual Review of Sociology 14: 319–40.

Lindblom, C.E. 1959. ‘The science of “muddling through”’, Public Administration Review 19: 79–88.

Lowi, T.J. 1969. ‘The new public philosophy: interest group liberalism’, The End of Liberalism: Ideology,

Policy, and the Crisis of Public Authority. New York: Norton.

March, J.G. 1991. ‘Exploration and exploitation in organizational learning’, Organization Science 2: 71–87.

Marier, P. 2008. ‘Empowering epistemic communities: Specialised politicians, policy experts and policy

reform’, West European Politics 31: 513–33.

—— 2009. ‘The power of institutionalized learning: The uses and practices of commissions to generate

policy change’, Journal of European Public Policy 16: 1204– 23.

Markoﬀ, J. and V. Montecinos. 1993. ‘The ubiquitous rise of economists’, Journal

of Public Policy 13: 37–68.

Marier

412

Marsh, D. and J.C. Sharman. 2009. ‘Policy diﬀusion and policy transfer’, Policy Studies 30: 269–88.

May, P.J. 1992. ‘Policy learning and failure’, Journal of Public Policy 12: 331–54.

Mintrom, M. 1997. ‘Policy entrepreneurs and the diﬀusion of innovation’, American Journal of Political Science

41: 738–70.

Myles, J. and P. Pierson. 2001. ‘The political economy of pension reform’, in P. Pierson (ed.) The New

Politics of the Welfare State. Oxford: Oxford University Press.

Nedergaard, P. 2006. ‘Which countries learn from which? A comparative analysis of the direction of

mutual learning processes within the open method of coordination committees of the European Union

and among the Nordic countries’, Cooperation and Conﬂict 41: 422–42.

Olsen, J.P. 1983. Organized Democracy: Political Institutions in a Welfare State: The Case of Norway. Bergen:

Universitetsforlaget.

Orenstein, M.A. 2008. Privatizing Pensions: The Transnational Campaign for Social Security Reform. Princeton,

NJ: Princeton University Press.

Park, Y.H. 1972. ‘The government advisory commission system in Japan’, Journal of Comparative Administration

3: 435–67.

Pemberton, H. 2000. ‘Policy networks and policy learning: UK economic policy in the 1960s and 1970s’,

Public Administration 78: 771–92.

Peters, B.G., J. Pierre and D.S. King. 2005. ‘The politics of path dependency: Political conﬂict in historical

institutionalism’, The Journal of Politics 67: 1275–300.

Pierson, P. 1993. ‘When eﬀect becomes cause: Policy feedback and political change’, World Politics 45:

595–628.

—— 1994. Dismantling

the Welfare State?

Reagan, Thatcher, and the Politics of Retrenchment. Cambridge:

Cambridge University Press.

—— 1998. ‘Irresistible forces, immovable objects: Post-industrial welfare states confront permanent austerity’,

Journal of European Public Policy 5: 539–60.

—— 2000. ‘Increasing returns, path dependence, and the study of politics’, American Political Science Review

94: 251–66.

—— 2004. Politics in Time: History, Institutions, and Social Analysis. Princeton, NJ: Princeton University Press.

Premfors, R. 1983. ‘Governmental commissions in Sweden’, American Behavioral Scientist 26: 623–42.

Radaelli, C.M. 1995. ‘The role of knowledge in the policy process’, Journal of European Public Policy 2: 159–83.

—— 2000. ‘Policy transfer in the European Union: Institutional isomorphism as a source of legitimacy’,

Governance 13: 25–43.

—— 2004. ‘The diﬀusion of regulatory impact analysis: Best practice of lesson-drawing?’, European Journal

of Political Research 43: 723–48.

—— 2009. ‘Measuring policy learning: Regulatory impact assessment in Europe ’, Journal of European Public

Policy 16: 1145–64.

Robertson, D.B. 1991. ‘Political conﬂict and lesson-drawing’, Journal of Public Policy 11: 55–78.

Rose, R. 1990. ‘Inheritance before choice in public policy’, Journal of Theoretical Politics 2: 263–91.

—— 1991. ‘What

is lesson-drawing?’, Journal of

Public Policy 11: 3–30.

—— 1993. Lesson-Drawing in Public Policy: A Guide to Learning across Time and Space. Chatham, NJ: Chatham

House.

Ross, F. 2000. ‘Interests and choice in the “Not quite so new” politics of welfare’, West European Politics

23: 11–34.

Rothstein, B. 1992. ‘Explaining Swedish corporatism: The formative moment ’, Scandinavian Political Studies

15: 173–91.

Ruggie, J.G. 1975. ‘International responses to technology: Concepts and trends’, International Organization

29: 557–83.

Schattschneider, E.E. 1935. Politics, Pressures and the Tariﬀ: A Study of Free Private Enterprise in Pressure Politics,

As Shown in The 1929–1930 Revision of the Tariﬀ. New York: Prentice Hall.

Sheriﬀ, P.E. 1983. ‘An Anglo-American comparison: Does social science contribute eﬀectively to the work

of governmental commissions?’, American Behavioral Scientist 26: 669–80.

Shipan, C.R. and C. Volden. 2008. ‘The mechanisms of policy diﬀusion’, American Journal of Political Science

52: 840–57.

Simmons, B.A. and Z. Elkins. 2004. ‘The globalization of liberalization: Policy diﬀusion in the international

political economy’, American Political Science Review 98: 171–89.

Skocpol, T. 1992. Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States.

Boston: Belknap Press.

Policy feedback and learning

413

Stone, D. and M. Garnett. 1998. ‘Introduction: Think tanks, policy advice and governance’, in D. Stone,

A. Denham and M. Garnett (eds) Think Tanks Across Nations. Manchester: Manchester University Press.

Tavits, M. 2003. ‘Policy learning and uncertainty: The case of pension reform in Estonia and Latvia’, The

Policy Studies Journal 31: 643–60.

Tsebelis, G. 2002. Veto Players: How Political Institutions Work. New York: Russell Sage Foundation.

Walker, J.L. 1969. ‘The diﬀusion of innovations among the American states’, The American Political Science

Review 63: 880–99.

Weaver, K. 2010. ‘Paths and forks or chutes and ladders? Negative feedbacks and policy regime change’,

Journal of Public Policy 30: 137–62.

Weaver, R.K. 1986. ‘The politics of blame avoidance’, Journal of Public Policy 6: 371–98.

—— 1989. ‘The changing world of think tanks’, Political Science and Politics 22: 563–79.

Weir, M. and T. Skocpol. 1985. ‘State structures and the possibilities for Keynesian responses to the Great

Depression in Sweden, Britain, and the United States’, in P.B. Evans (ed.) Bringing the State Back In.

Cambridge: Cambridge University Press.

Westney, D.E. 1987. Imitation and Innovation: The Transfer of Western Organizational Patterns to Meiji Japan.

Cambridge, MA: Harvard University Press.

Weyland, K. 2005. ‘Theories of policy diﬀusion: Lessons from Latin American pension reform’, World

Politics 57: 262–95.

—— 2007. Bounded Rationality and Policy Diﬀusion. Princeton, NJ: Princeton University Press.

Wheare, K.C. 1955. Government by Committee. Oxford, Oxford University Press.

Marier

414

Randomized control trials

What are they, why are they promoted as the

gold standard for causal identiﬁcation and

what can they (not) tell us?

David Fuente and Dale Whittington

Introduction

Over the past decade scholars, research practitioners, and donors have shown an increasing interest

in using randomized control trials (RCTs) to evaluate the eﬀectiveness of a wide range of social

and infrastructure programs in developing countries. Much as in medicine where randomized

control trials are standard practice, the policy community has turned to RCTs in response to

increasing calls for evidence-based policy. As a result, quasi-experimental approaches to program

evaluation have come under increasing scrutiny and been strongly criticized by advocates of

RCTs. In fact, in some policy and scholarly circles the term RCT has become nearly synonymous

with “rigorous ” and has been hailed as the primary, if not sole, arbiter of “hard” evidence in the

policy process.

The use of RCTs to assess social policy is not new. Fisher (1935) described the use of a

randomized experimental design to evaluate the response of agricultural ﬁelds to a variety of

inputs. In the 1980s labor economists used RCTs to assess the eﬀectiveness of large-scale

workforce training programs in the United States (e.g. Bloom et al. 1993; Doolittle and Traeger

1990; LaLonde 1986). Over the past decade, a new generation of development economists has

adapted and expanded the tools developed by labor economists and applied them in novel ways

to study the impact of a wide range of development programs, including de-worming (e.g. Miguel

and Kremer 2004); educational interventions (e.g. Kremer 2003; Todd and Wolpin 2006); micro-

ﬁnance (e.g. Banerjee et al. 2010; Giné et al. 2006); and point-of-use water treatment (e.g.

Kremer et al. 2009; Ahuja et al. 2010). These advances are well reﬂected in the work of the

Abdul Latif Jameel Poverty Action Laboratory (J-Pal)

based at MIT, with oﬃces in Africa, Europe,

Latin America, and South Asia as well as Innovations for Poverty Action.

These organizations

have collectively executed dozens of RCTs in developing countries across the globe.

Proponents of RCTs often have been vocal critics of past approaches to formulating devel-

opment policy and strong advocates for the broader use of RCTs to identify development

interventions “that work.” For example, Banerjee (2007) has accused the World Bank of “lazy

415

thinking” for not ensuring that its policy advice to developing countries is supported by a strong

(RCT) evidence base. Similarly, Duﬂo (2004) has called RCTs in development “an interna-

tional public good” and been a strong advocate for the broader funding of RCTs by the

international donor community.

While scholars at the forefront of the RCT resurgence have

been a prominent voice in academic journals, they have also taken their message to nontechnical

audiences in a series of books (e.g. Banerjee and Duﬂo 2011; Karlan and Appel 2011) and made

spirited calls for the wider use of RCTs in the popular media (Banerjee 2006; Glennerster and

Kremer 2011; Das et al. 2011).

The recent surge in the use of RCTs to evaluate development programs has improved the

practice of program evaluation and yielded important and interesting empirical results. At the same

time it has spawned a vibrant debate about whether, as their advocates suggest, evidence from

RCTs should occupy a privileged position in the hierarchy of knowledge used to shape public

policy. This has rekindled old debates ignited by LaLonde (1986) about the relative merits of

RCTs compared to other modes of social inquiry.

The most recent incarnation of the RCT

debate has been primarily between economists and has occurred largely in professional meetings

and on the pages of academic journals. Three prominent examples of this debate include the

exchanges between Banerjee (2005), Bardhan (2005), Mookherjee (2005), Kanbur (2005), and

Basu (2005) in the journal Economic and Political Weekly; Angrist and Pischke (2010); Heckman

(2010) and Keane (2010) in the Journal of Economic Literature; and a separate exchange between

Deaton (2010) and Imbens (2010).

To date, the debate among economists about the relative merits of RCTs has largely been

technical in nature. As a result, the broader implications of the debate for policy scholars and

development practitioners are often obscured. This chapter provides an overview of this debate

and oﬀers guidance to policy scholars on the circumstances in which RCTs can be most useful

in policy research.

The current dialog about RCTs is often narrowly pitched as a debate between experimental

and quasi-experimental approaches to program evaluation or structural and reduced form methods

in economics. This chapter does not compare alternative approaches to program evaluation that

are used in economics. Rather, we provide an overview of the arguments typically advanced for

and against the use of RCTs in policy research. Because the arguments for RCTs are typically

more familiar to policy professionals, the chapter spends more time highlighting factors that

complicate claims of RCTs’ superiority. We do this not to advocate for one form of program

evaluation over another, but to provide policy professionals a more nuanced understanding of

both the strengths and limitations of RCTs. Clearly, if a researcher is interested in estimating a

causal impact of a particular intervention, and they have suﬃcient funding to conduct an experi-

ment, they can wait a long time for results, and the intervention can be randomized (ethically

and practically), it is hard to imagine a situation in which the researcher would choose a quasi-

experimental research design rather than an RCT. However, there are many potentially interest-

ing policy questions, types of interventions, and contexts in which a researcher or policymaker

might need to use alternative approaches. This chapter provides researchers and policy-makers

with a framework for thinking about what RCTs can and cannot contribute to important

public policy debates.

The chapter is organized as follows. The second section presents the standard motivation for

RCTs and discusses the core strength of RCTs in relation to other modes of program evalua-

tion. Section three outlines the common critiques of RCTs and draws attention to the type of

policy questions to which RCTs are well suited. Section four concludes with a discussion of

factors researchers, policy-makers, and donors should consider when assessing whether or not

RCTs can best advance their policy or research agenda.

Fuente and Whittington

416

Randomized control trials: what are they, what can they tell us, and why

are they promoted as the gold standard in causal identiﬁcation?

Research and formal program evaluation that seeks to establish causal inference between a program

intervention and observed outcomes faces the classical counterfactual problem— namely what the

outcomes for program participants would have been without the program. For example, to identify a

causal connection between the use of a particular point-of-use (PoU) water treatment technology

in a sample of households and reduced incidence of diarrheal disease, one needs to know the number

of episodes of diarrheal disease the same households would have experienced over the same time

period in the absence of the program. Since in this example an individual household would have

either used the PoU water treatment technology or not, the counterfactual required for causal

inference is not observable. That is, we cannot simultaneously observe the same household both with

and without the PoU technology. The fundamental challenge of program evaluation is to construct

a close approximation of this unobservable counterfactual. This requires a control group, which in

the absence of a program or treatment would have similar outcomes to the treatment group.

In order for the control group to represent a valid approximation of the counterfactual required

to identify a causal relationship, the treatment and control groups must be suﬃciently similar in

both observable and unobservable ways. If this condition is not met, the researcher cannot deter-

mine whether the diﬀerence in observed outcomes between the treatment and control groups is

attributable to the program intervention (treatment) or to preexisting diﬀerences between the two

groups. In general, quasi-experimental research methods (e.g. propensity score matching, regression

discontinuity design, etc.) seek to construct control groups using observable characteristics. RCTs,

on the other hand, construct the counterfactual via randomization into a treatment group.

For example, suppose we are interested in understanding the impact of a particular PoU water

treatment technology on diarrheal disease, measured by the number of episodes of diarrheal

disease per person per month in a household.

Let Y

C be the number of episodes of diarrhea

for a given household i per month in the absence of the PoU water treatment (the control group)

and Y

T the episodes of diarrhea per month in the same household with water treatment (treat-

ment group). To identify the causal impact of the PoU water treatment technology on diarrheal

disease, we would want to measure the average impact of the PoU treatment technology on all

households in the program, which can be written as follows:

E½Y

T  Y

C

ð1Þ

However, we cannot observe household i both with (Y

T) and without (Y

C) the PoU treat-

ment technology at the same time. One might try to study a large sample of households, half

with the PoU treatment technology and half without it (each of these quantities is observable)

and calculate the average number of episodes of diarrheal disease per person per month in each

group. This would provide us the diﬀerence between the average number of episodes of diar-

rheal disease conditional on household i having the PoU treatment technology, E[Y

T|T], and

the average number of episodes of diarrheal disease conditional on household i not having the

PoU treatment technology, E[Y

C|C] represented in equation 2:

E½Y

TjTE½Y

CjCð2Þ

Ideally one would like information on what would have happened to households in the treat-

ment group if they were not provided with the PoU treatment technology. Let E[Y

C|T] be

Randomized control trials

417

the average number of episodes of diarrhea in households in the treatment group if they did not

receive the PoU treatment technology. Simply adding and subtracting E[Y

C|T] from equation

2 this yields:

E½Y

TjTE½Y

CjTE½Y

CjCþE½Y

CjTð3Þ

The second and fourth terms cancel out and therefore do not aﬀect the equation. Because the

expectation operator is linear, we can rewrite equation 3 as:

E½Y

T  Y

CjTþ½E½Y

CjTE½Y

CjCð4Þ

The ﬁrst term of equation 4 is the treatment eﬀect we are interested in measuring. In our example,

it represents the average eﬀect of the PoU water ﬁlter on households in the treatment group.

The second term is selection bias

—namely preexisting or systematic diﬀerences between house-

holds in the treatment and control groups. However, recall that we do not observe E[Y

C|T]

(households in the treatment group without the PoU water ﬁlter). Thus, it is diﬃcult to

determine the magnitude or sign of the selection bias and whether estimates of program impact

reﬂect the true program impact, selection bias, or a combination of both. To get accurate esti-

mates of the program’s causal impact, researchers face the burden of either estimating the

selection bias and correcting it, or arguing that selection bias does not exist.

With a large sample, if households are randomly assigned to treatment and control groups

they will on average (in expectation) be identical with respect to their observable and unob-

servable characteristics. Thus, under ideal conditions, randomization addresses selection bias by

ensuring that the treatment and control groups would have the same outcomes on average in

the absence of the intervention. More formally, in theory randomization should ensure that the

diﬀerence in average observed outcomes between the treatment and control groups equals the

average of the diﬀerence between the observed outcome for households in the treatment group

and what the outcome for those households would have been in the absence of treatment.

Namely that:

E½Y

TjTE½Y

CjC¼E½Y

T  Y

CjTð5Þ

In our example, this implies that the treatment and control groups would experience the same

number of cases of diarrheal disease on average without the PoU treatment technology, i.e. that

the control group is a valid counterfactual for the treatment group.

Equation 5 provides information about the average eﬀect of the treatment on households in

the treatment group (often referred to as average treatment on the treated), but not about the

potential impact of the treatment on households in the control group or about the potential

impact on the population in general.

If the random assignment is executed correctly and if the

research participants comply with their assignments, the average eﬀect of treatment on households

on the treatment group should equal the average eﬀect of the treatment on households in the

control group. Similarly, the estimate of program impact obtained from equation 5 will only be

representative of the potential impact of the program on an individual randomly chosen from

the general population if the study sample is representative of the larger population.

By construction, RCTs provide information on the average (mean) treatment eﬀects of a pro-

gram or intervention. They do not provide information on the impacts on speciﬁc individuals

Fuente and Whittington

418

or on the general population. Rather, RCTs provide an estimate of the average impact of an

intervention on a random person drawn from the study sample.

Under ideal conditions randomization addresses the problem of selection bias. Equation 4

states that selection bias is the average diﬀerence in outcomes between the treatment group and

control group in the absence of the intervention. (Recall that the former is the unobservable

counterfactual and the latter is the observed outcome for the control group.) Randomization

removes selection bias by ensuring that any potential bias in the sample is evenly distributed

between the treatment and control groups (Heckman and Smith 1995: 89).

RCTs provide information on the mean overall impact of a project or intervention after the

program participants and perhaps broader context has dynamically adjusted to the impact. As

noted by Duﬂo et al. (2008) this is distinct from the program impacts holding all else constant,

which is required to estimate the welfare-theoretic consequences of a program.

Under ideal conditions, RCTs provide a means to cleanly estimate the causal impact of a

particular treatment. However, they do not prima facie provide information about the causal

mechanisms that link the treatment to the observed outcome.

Arguments for RCTs

Based on the strong methodological results discussed above, proponents argue that RCTs should

be viewed as the gold standard for identifying causal impact and given special consideration in

the policy arena.

Indeed Banerjee (2006) states “[w]hen we talk of hard evidence, we will

therefore have in mind evidence from a randomized experiment.” Proponents highlight two

main reasons why RCTs are superior to other modes of inquiry. First, they argue that RCTs

have the strongest internal validity of any impact estimation method (Banerjee 2005; Angrist

and Pischke 2010; Duﬂo et al. 2008). Second, they assert that RCTs are based on fewer and less

complicated assumptions than other methods and produce “clean” estimates of average program

impacts RCT (Banerjee and Duﬂo 2009; Duﬂo et al. 2008). Therefore, proponents claim that

RCTs provide information about program impacts that are more accessible to the policy com-

munity than many quasi-experimental methods, which often employ complex econometric

techniques and rely on untestable assumptions. Each of these arguments is considered below.

Strong internal validity

Empirical results from program evaluations are generally judged on two criteria: internal validity

and external validity. Internal validity refers to the degree to which one accurately measures the

true program impact in the study population. External validity refers to the degree to which the

ﬁndings of a study can be generalized to populations not explicitly included in the study, whether

in the same or other geographical locations and cultural contexts. Proponents of randomized

experiments suggest that RCTs provide the highest degree of internal validity of any impact

estimation method (see Duﬂo et al. 2008 for a detailed discussion of the RCTs, quasi-experimental

methods, and internal validity).

Under ideal conditions, randomization ensures that the treatment and control groups are iden-

tical on average with respect to both observable and unobservable characteristics. In contrast,

because all aspects of the treatment and control groups are not observable, quasi-experimental

methods cannot guarantee that the treatment and control groups are on average identical along both

observable and unobservable dimensions. Therefore, unlike RCTs, quasi-experimental methods

cannot ex ante eliminate selection bias and thus may yield estimates of program impact that are

biased by preexisting or fundamental diﬀerences between the treatment and control groups.

Randomized control trials

419

Parsimony

Proponents of RCTs also argue that RCTs are more parsimonious than quasi-experimental

methods and thus are better suited to impact evaluation of public policy programs. They argue

that because RCTs are based on fewer (and more plausible) assumptions than quasi-experimental

methods and report average program impacts, they are more accessible to policy-makers who

may lack the technical training to interpret the results of quasi-experimental studies or discern

the quality of these results. Banerjee and Duﬂo state:

most evaluations of social programs focus exclusively on the mean impact. In fact, one of

the advantages of experimental results is their simplicity: They are easy to interpret because

all you need to do is compare means, a fact that may encourage policy-makers to take the

results more seriously.

(2009: 169)

Under ideal conditions, RCTs are in fact based on fewer assumptions than quasi-experimental

methods. Unlike quasi-experimental methods, which rely upon untestable assumptions about

the independence of the treatment and important factors (individual or contextual) that may

inﬂuence treatment outcomes in the treatment and control groups (i.e. the presence of selection

bias), RCTs rely on the apparently simple concept of randomization to address selection bias.

The arguments in support of RCTs are simple, intuitive, and powerful. This in part explains

why RCTs have received renewed attention from both the academic and policy communities.

Despite the compelling logic behind RCTs, there has been considerable debate regarding the

relative merits of RCTs. The following section provides an overview of the critiques commonly

levied against RCTs.

Critiques of RCTs

Critics of the RCT paradigm typically do not argue that quasi-experimental methods or other

modes of inquiry are theoretically better suited to answer questions related to causal inference than

RCTs. Rather the critiques take two primary forms. First, critics of the RCT paradigm assert

that due to a number of practical issues associated with implementing RCTs or any type of

program evaluation, the ideal conditions from which RCTs derive their relative strength rarely

hold. Critics therefore argue that despite the clear theoretical strengths of RCTs, they do not deserve

a special place in the hierarchy of program evaluation methods. This argument is not unique to

the application of RCTs to social and development policy. Discussing the relative merits of

RCTs in the medical ﬁeld, Kramer and Shapiro state, “[t]he model of methodological rigor

represented by the RCT invites close scrutiny for any departures from the ideal” (1984: 2744).

Second, critics assert that even if the ideal conditions for an RCT can be approximated or even

met in practice, RCTs are well suited to answer only a small subset of questions of potential

interest to policy-makers and scholars. This section ﬁrst considers common practical issues

associated with the implementation of RCTs and the challenges these issues raise for RCTs and

then turns to the scope of problems that RCTs are well suited to answer.

Practical considerations and departures from ideal conditions

In the face of budgetary, cultural, and political constraints, researchers face the challenge of ensuring

that the design and implementation of a program evaluation meet the conditions required for

Fuente and Whittington

420

their methods to yield accurate, reliable results. Thus, researchers must pay attention to any

ways the execution of their research design may deviate from the ideal in both anticipated and

unanticipated ways. Because RCTs share this challenge with other modes of inquiry, critics

claim that the practical diﬃculties associated with implementing RCTs undermine their relative

strength over other methods used to identify causal eﬀects. This sentiment is perhaps best

reﬂected by Deaton who states that:

Conducting good RCTs is exacting and often expensive, so that problems often arise that

need to be dealt with by various econometric or statistical ﬁxes. There is nothing wrong

with such ﬁxes in principle … but their application takes us out of the world of ideal

RCTs and back into the world of everyday econometrics and statistics. So that RCTs, although

frequently useful, carry no special exemption from the routine statistical and substantive

scrutiny that should be routinely applied to any empirical investigation.

(2010: 447)

There are a number of issues that can potentially compromise the internal validity of RCTs,

undermine their apparent simplicity, and limit their external validity. These include, but are not

limited to: randomization bias, substitution bias, and treatment compliance; practical diﬃculties

associated with the randomization process; limitations on sample size; and selection issues.

Treatment Compliance, Substitution Bias, and Randomization Bias

For RCTs to provide unbiased estimates of the average treatment eﬀects, the following three

conditions must hold: (1) individuals in the treatment group must comply with their assignment

(treatment compliance); (2) randomization cannot alter participants’ behavior (no randomization

bias); and (3) the control group must comply with their assignment (no substitution bias). Violations

of these assumptions complicate causal inference from randomized control trials.

Let us ﬁ rst consider the issue of treatment compliance. While researchers clearly hope that

individuals assigned to the treatment group will comply with their assignment—namely parti-

cipate in the program or accept the experimental treatment—this cannot be taken for granted.

Returning to our PoU water technology example from the second section, consider a situation

in which a researcher is interested in identifying the impact of a PoU water ﬁlter on reducing

incidence of diarrhea among poor households. Ideally the researcher randomly assigns house-

holds in the study area to a treatment and a control group. The researcher then provides the

PoU water ﬁlter as well as a set of common user instructions to households in the treatment

group, and tracks the number of episodes of diarrhea in each household in both the treatment

and control groups over the duration of the study period.

There are numerous plausible reasons why households in the treatment group may not

comply with their assignment. First, some households may ﬁnd the PoU water ﬁlter incon-

venient to use and decide not to use the ﬁlter or use it only rarely, perhaps when guests visit the

house. Second, some households may not like the taste of the ﬁltered water and again decide

not to use the ﬁlter. Third, some of the households who like the taste of the ﬁltered water and

ﬁnd it easy or acceptable to use, may simply not operate the ﬁlter correctly and thus continue to

drink water that is eﬀectively untreated.

In each of the above examples, households will not be in compliance with their assignment

to the treatment group and unadjusted estimates of program impact will be biased. Thus

researchers must adjust their estimates using various statistical techniques to obtain unbiased

estimates of the program outcome. Here, researchers can no longer estimate the average

Randomized control trials

421

treatment eﬀect, but must estimate intent to treat (ITT), treatment on treated (TOT), and local

average treatment e ﬀects (LATE). As noted by Deaton (2010) above, in principle there is nothing

wrong with using econometric techniques to correct such biases. However, when there is

noncompliance among the treatment group, the experiment deviates from the ideal conditions

from which RCTs derive their comparative strength.

Similarly, in order to obtain unbiased estimates of average treatment eﬀects in an RCT, the

process of randomization cannot, in theory, aﬀect the behavior of individuals in the treatment

and control groups (randomization bias). With respect to the treatment and control groups, this

refers to the well-known threats of Hawthorne and John Henry eﬀects,

respectively. Medical

trials seek to eliminate the possibility of randomization bias through a double blind process

where neither the researcher nor participants know their treatment status (Deaton 2010). How-

ever, this is often not possible in the context of policy-relevant RCTs or social experiments.

Referring to Fischer’s (1935) seminal presentation of the randomized experimental design in

agricultural productivity research, Heckman (1992) quips that while plots of land do not respond

to their treatment status, people do (quoted in Deaton 2010). In our hypothetical PoU water

ﬁlter experiment, households who have consented to participate in the experiment will clearly

know whether or not they have received a water ﬁlter.

Closely related to John Henry eﬀects, substitution bias presents a threat to the internal

validity of RCTs. Substitution bias occurs when individuals assigned to the control group seek

close or exact substitutes for the intervention delivered in the experiment to the treatment

group. In the PoU example, households assigned to the control group may suspect that because

some households are receiving PoU water ﬁ lters, their water must be dirty. This may prompt

some households in the control group to either purchase some form of water puriﬁcation tech-

nology or begin to boil their water. If this occurs, the control group no longer reﬂects the

counterfactual situation required for causal inference. As with noncompliance among households

in the treatment group, substitute-seeking behavior by households in the control group will bias

estimates of a program’s average treatment eﬀects. In the case of noncompliance among the control

group, estimates of program eﬀects will be biased downward (Heckman and Smith 1995).

Treatment compliance and randomization bias are well known to scholars and practitioners of

program evaluation, and there are widely accepted means of addressing these issues. For example,

in the presence of potential endogeniety caused by treatment noncompliance or randomization

bias, the researcher can use the randomized assignment as an instrumental variable for treatment

status in a regression to obtain an estimate of the Local Average Treatment Eﬀect (LATE).

The estimate of the LATE provides a means of identifying a causal eﬀect of a given treatment.

However, when participants do not comply with their treatment assignments or the process of

randomization inﬂuences participants’ behavior, the experiment no longer complies with the ideal

conditions from which it derives its strong internal validity relative to other modes of program

evaluation. This presents challenges to both the parsimony and internal validity of RCTs. With

regard to parsimony, Heckman and Smith suggest, “[i]n the presence of randomization bias, the

meaning of an experimental impact estimate would be just as diﬃcult to interpret honestly in

front of a congressional committee as any non-experimental study” (1995: 93).

Selection issues in RCTs

Researchers who seek to identify a causal relationship between a program and an observed out-

come must also be aware of the threat from selection issues. In theory, the randomization pro-

cess ensures that any bias between the treatment and control groups is averaged out. While this

is technically correct, the randomization process removes selection bias among the individuals

Fuente and Whittington

422

(households) who participate in the study. In order for the results of a RCT to be applicable to

the broader population from which the study sample was drawn, the sample must be reﬂective

of the broader target population of an intervention. Researchers cannot assume a priori that the

sample of individuals who participate in a given study is representative of the broader target

population.

Standard research ethics dictate that researchers must obtain informed consent from participants

in a research study. Thus, it is plausible that some individuals in the broader target population

may not consent to participating in an RCT. In the PoU water ﬁlter example, some households

may be suspicious of a new technology or simply not want to be inconvenienced by repeated

visits from a research team. These households may not consent to participating in the study. If

this situation occurs, the researcher cannot guarantee that households that agree to participate in

the study are representative of the broader target population of interest.

Indeed, in the context

of medical trials, Kramer and Shapiro (1984) note that individuals are more likely to participate

in observational (quasi-experimental) studies than in RCTs (in Heckman and Smith 1995).

In general, the researcher can check whether participants are similar to the broader population

in terms of observable socioeconomic and demographic characteristics. However, the researcher

cannot rule out the possibility that the two populations diﬀer systematically in unobservable

ways. This raises many of the same challenges for RCTs inherent in quasi-experimental research

methods.

This type of selection issue will not necessarily be a concern for all RCTs. For example, policies

or programs may target individuals who want to participate in a program. This type of intervention

relies explicitly on voluntary participation and is quite common for a wide range of social

policies. In these instances, an oversubscription research design, in which willing participants are

assigned to the control group by a random denial or postponement of treatment, can yield

unbiased estimates of the treatment eﬀects on the target population of interest—namely those

who are inclined to participate in a given program or seek a given treatment.

There is a second type of selection eﬀect that has received relatively less attention in the litera-

ture: selection among the partner organization(s) who implement RCTs. Heckman and Smith

(1995) provide a detailed discussion of this issue in the context of the Job Training Partnership

Act (JTPA) in the United States, where there was a 90 percent refusal rate among potential

implementing agencies who feared that randomizing service delivery would impact their per-

formance benchmarks (see also Allcott and Mullainanthan (2011) and Deaton (2010) for a dis-

cussion of selection issues among implementing partners). In the context of social or development

policy, researchers often work with a partner organization such as a nongovernmental organi-

zation (NGO) who will implement and administer the program or treatment under considera-

tion. Potential implementation partners face a range of incentives that will inﬂuence whether or

not they agree to participate in an RCT.

Additionally, for obvious reasons researchers typically would like to partner with agencies or

organizations that are capable of delivering the program or treatment in a high quality manner

and closely following the research protocol. As a result, in developing country contexts researchers

often partner with prominent NGOs or government agencies under speciﬁc directive to parti-

cipate in the RCT. Clearly an NGO or government agency that values highly such academic

collaboration may not be representative of the broader population of agencies or organizations

who might be called upon to implement a particular program. Therefore, the researcher cannot

rule out the possibility that the estimated program impact is not related to particular characteristics

of the implementation partner or how the program was implemented.

Another concern is that staﬀ of an NGO that can deliver the intervention eﬃciently and

eﬀectively are likely to believe strongly in the eﬀectiveness of the treatment, and may

Randomized control trials

423

communicate this to participants. This type of implementation diﬀers markedly from a double-

blind design and threatens the internal validity of an experiment. In a recent working paper, Allcott

and Mullainanthan (2011) show evidence of partner selection bias in two sets of well-regarded

RCTs, one set examining the response of residential energy use to information on household

consumption and one set exploring the impact of microﬁnance in developing countries.

In some instances, selection eﬀects among implementation partners may not threaten the

internal validity of estimates derived from RCTs. However, they do raise concerns about the

external validity of some experimental ﬁndings—namely whether or not researchers or policy-

makers could expect roughly the same result if the program was scaled from a pilot program

implemented by a competent partner to a broader program implemented by partners of varying

ability or quality (Deaton 2010; Rodrik 2008).

Heterogeneity: parsimony reconsidered

Critics of RCTs also question the ability of policy-makers to understand the relative merits of

experimental and quasi-experimental techniques. Heckman and Smith (1995) point out that a

criterion of simplicity or parsimony would argue for simple before-and-after research designs

(without randomization).

Critics of RCTs also claim that their proponents mistake apparent simplicity for real simplicity

(Heckman and Smith 1995; Deaton 2010). In theory RCTs are based on a relatively simple set

of assumptions. However, in practice, addressing the challenges associated with implementing

RCTs and developing “robust” estimates of program impact are both complex and technical.

As a result of this complexity, critics of RCTs claim that there is no a priori reason to believe

that the assumptions and techniques employed to estimate program impacts from experimental

studies are more plausible or less complex than those used in quasi-experimental studies.

The issues of treatment compliance and randomization bias discussed above provide examples

of how practical challenges associated with implementing RCTs can compromise the types of

inference one can draw from them. There two other important issues where this becomes

apparent: (1) basic heteroscedasticity in the error term in the regression used to estimate pro-

gram eﬀects;

and (2) the use of baseline surveys and statistical controls to sharpen estimates of

program impact and reduce the required sample size.

Under ideal conditions, one can estimate the impact of the program intervention from

experimental data using the following simple model:

¼  þ T

þ "

ð6Þ

Where Ti is a dummy variable that represents whether individual i was randomly assigned to the

treatment group, randomization ensures that essential assumptions of the Gauss-Markov theorem

are met, so one can develop an unbiased estimate of β by ordinary least squares (OLS). The para-

meter estimate of β is the program impact we seek to estimate (Duﬂo et al. 2006; Deaton 2010).

Citing Fisher (1935), Deaton (2010) emphasizes that randomization plays two fundamental roles.

First, randomization ensures that the mechanism governing selection into the treatment group is

the same as the mechanism governing selection into the control group. This is what allows ran-

domization to eliminate selection bias. Second, randomization provides “a probability law that

enables us to judge whether the diﬀ erence between the two groups is signiﬁcant” (ibid. 442).

Thus, while randomization allows the researcher to obtain unbiased estimates of program impact

(the parameter estimate for β from equation 6) using OLS, this does not necessarily imply that

Fuente and Whittington

424

the parameter estimate is statistically signiﬁcant, namely that one would reject the null hypothesis

that there is no treatment eﬀect (Ho: β = 0). This can only be established by statistical inference,

which requires one to calculate the standard error of the parameter estimate of β.

In order for OLS to yield correct standard errors for parameter estimates, one must assume

that the error term is homoscedastic—that each observation has equal variance for all factors not

explicitly included in the model. As Deaton (2010) suggests, there is no a priori reason to believe

that the treatment and control groups in the sample have equal variance or that the act of ran-

domization itself did not aﬀect the variance of each group. Thus, in order to have conﬁdence

that the statistical inference from the OLS regression is correct, one must either invoke the

assumption of homoscedasticity or correct for heteroscedasticity. The latter can be accomplished

using standard econometric techniques. However, critics of the wholesale promotion of RCTs

claim that assuming homoscedasticity in the error term or correcting for heteroscedasticity using

econometric techniques compromises the claim that RCTs are based on more plausible

assumptions and are less technically complex than quasi-experimental methods.

Also, RCTs can require large sample sizes to obtain the statistical power needed to identify

program treatment eﬀects at a level convincing to policy-makers. This can be expensive. To

reduce the required sample size and hence the cost of a RCT, researchers often conduct base-

line surveys to obtain covariates to include in the regression equation used to estimate program

treatment eﬀects. The basic justiﬁcation for this practice is that including covariates that are

unrelated to the treatment but related to the outcome variable reduces the standard errors of the

parameter estimates and thus reduces the required sample size (Deaton 2010; Duﬂo et al. 2008).

The general form of the resulting regression equation is:

¼ α þ β1T

þ γX

þ υ

þ ω

ð7Þ

where X

is a vector of individual and/or group control variables over individuals i and groups

j, γ is a vector of parameters associated with X

, and υ

+ω

represents the compound error term

after controlling for X

Conducting baseline surveys and including covariates in the regression used to estimate average

treatment eﬀects raises two issues that complicate RCTs claim to parsimony. First, covariates must

be speciﬁed ex ante. Otherwise, including covariates in the regression equation opens up the

possibility of searching for a “good” speciﬁcation after the RCT is implemented (Deaton, 2010;

Duﬂo et al. 2008). This raises similar concerns as post-trial subgroup analysis.

Second, when treat-

ment eﬀects are heterogeneous across the study population, including covariates in the regression

equation will yield consistent, but not unbiased, parameter estimates of the average treatment

eﬀect (Deaton 2010).

Biased yet consistent parameter estimates are typically not a concern in large

samples. However, as Deaton notes, in small samples the bias may be substantial (ibid.: 444).

Scope of questions RCTs are well suited to answer

Up to this point in the chapter, we have discussed the basic rationale for RCTs, the core

arguments advanced by proponents of RCTs, and the various ways in which RCTs can deviate

from ideal conditions in practice. Each of these topics, while important, examines RCTs on

their own turf, that is in problem contexts that are conceivably well suited to them. As Imbens

notes “conditional on the question, the methodological case for randomized experiments seems

unassailable” (2010: 14). We now directly consider the scope of policy questions RCTs are well

suited to answer. This issue is well recognized in the program evaluation literature, but has

Randomized control trials

425

received relatively little attention in the current debate about the relative merits of RCTs or

from policy analysts.

Even the most ardent advocates of RCTs acknowledge that experimental methods can be

used to answer only a small subset of potentially important policy questions. Banerjee and Duﬂo

state that, “we [must] not forget that there are a lot of important questions that randomized experi-

ments cannot answer” (2009: 152). Critics of RCTs have decidedly stronger feelings about the

value of RCTs in scholarly and policy discourse. Indeed Heckman and Smith state that “experi-

mental data provide no answers to many of the questions of interest to program evaluators”

(1995: 86). Similarly, Deaton argues:

[t]he price for [the success of RCTs] is a focus that is too narrow to tell us “what works” in

development, to design policy, or to advance scientiﬁc knowledge about development

processes. Project evaluation using randomized controlled trials is unlikely to discover the

elusive keys to development, nor to be the basis for a cumulative research program that

might progressively lead to a better understanding of development.

(2010: 3)

Policymakers and policy analysts need a clear, nuanced understanding of the scope of questions

that RCTs are well suited to answer and the types of questions they cannot address. We highlight

below a number of considerations about the type of questions RCTs can address.

Average impact vs. distributional concerns

Policymakers are usually interested in the distribution of outcomes across the target population.

Distributional impacts that might be of interest to policy-makers include but are surely not

limited to: outcomes by income decile, percent of participants who experienced positive program

outcomes, the median (rather than mean) program impact, and so on.

In many circumstances,

a policymaker would not want to implement a program that bestowed large beneﬁts on a small

subset of the population while inﬂicting harm on the rest of the population. However, without

additional assumptions, statistical adjustments, or a research design speciﬁcally targeted toward

identifying program impacts on diﬀerent groups, RCTs typically cannot provide insight into the

distributional impacts of a policy or program (Heckman and Smith 1995; Heckman and Vytlacil

2001; Heckman 2010; Deaton 2010; and Rodrik 2008).

Causal mechanisms: why and how vs. whether and how much

Under the most favorable conditions RCTs provide an accurate, reliable means of identifying

the causal eﬀect of a policy or program. RCTs can identify whether or not a program “works”

(has positive outcomes for program participants) and provide an estimate of the magnitude of

the treatment eﬀect (how much). However, RCTs do not provide insight into how or why a

program “works.” That is, RCTs do not provide insight into the causal mechanisms that drive

program outcomes among a certain population.

This is particularly problematic for Deaton

who states:

RCTs of “what works,” even when done without error or contamination, are unlikely to be

helpful for policy, or to move beyond the local, unless they tell us something about why

the program worked, something to which they are often neither targeted nor well-suited.

(2010: 448).

Fuente and Whittington

426

Policy-makers need to understand why a program works in a speciﬁc context or location in order

to assess whether it is likely to work elsewhere.

In other words, they need some assurance of

not only a program’s internal validity, but also its external validity.

Similarly, RCTs do not provide information on why individuals might choose to participate

in a particular program or accept a particular treatment. For programs that rely upon voluntary

participation, policy-makers need to understand the factors that inﬂuence individuals’ decisions

about whether or not to opt into a voluntary program or use a particular technology. For example,

in our example of a PoU water ﬁlter, policy-makers or donors may want insight into why

households might agree, or continue, to use PoU water ﬁlters or be willing to pay for them.

RCTs cannot provide this information. However, these perspectives are essential for policy-

makers or donors to assess whether or not a program to market or distribute water ﬁlters is likely to

be eﬀective in a particular context and, perhaps more importantly, identify ways in which they

can increase the uptake and usage of the ﬁlter.

RCTs are only well suited to interventions that can be randomized

Policy researchers can only use RCTs if the interventions of interest can be randomized. There

are a number of types of development programs that are amenable to randomization. Examples

include insecticide-treated bed nets, point-of-use water ﬁlters, improved cookstoves, ﬁnancial

products among microﬁnance customers, and so on. However, many types of policy interven-

tions cannot be eﬀectively randomized for practical or ethical reasons. Indeed Banerjee states,

“there are times when randomized experiments are simply not feasible, such as in the case of

exchange rate policy or central bank independence: it clearly makes no sense to assign countries’

exchange rates at random as you might assign them ﬂip charts” (2006: 8).

Let us again consider PoU water ﬁlters and their impact on diarrheal disease reduction. As a

practical matter, researchers may be able to randomize the distribution of PoU water ﬁlters to

communities and particular households within communities. Indeed, a researcher could dis-

tribute multiple types of PoU ﬁlters randomly among sample households to determine which

type of ﬁlter is more eﬀective in reducing the episodes of diarrhea per household per month in

the study population. However, what if the researcher were interested in determining the

eﬀectiveness of large dams on economic growth? It is impossible to imagine a situation in which

randomizing such large-scale infrastructure projects would be politically feasible or ethically

appropriate.

This raises a broader issue. If RCTs are the gold standard for causal identiﬁcation, how does a

researcher or policy-maker compare two interventions when one intervention is relatively easy

to randomize (e.g. PoU water ﬁlters) and another is much more diﬃcult (e.g. piped and sewerage

networks)? From a policy perspective it would not make sense to limit the policy choice set to

interventions than can be randomized. However, it is not immediately clear on what basis one

can compare two interventions if the eﬀectiveness of the interventions cannot be determined by

comparable or epistemologically acceptable methods.

Value in reﬁning a single parameter

RCTs can provide insight into the eﬀectiveness of a particular intervention, i.e. whether or not

an intervention “works” and to what degree. However, this is only one of numerous pieces of

information a researcher needs to know to conduct meaningful policy analysis. For example,

suppose a researcher was interested in conducting a beneﬁt cost analysis of competing health

interventions, including PoU water ﬁ lters (e.g. Whittington et al. 2009 and Jeuland and

Randomized control trials

427

Whittington 2009). The researcher would need to calculate the economic value of the reduc-

tions in mortality and morbidity associated with the use of the water ﬁ lter. To estimate the

economic beneﬁts of mortality reduction associated with a reduction in diarrheal disease due

to the use of the PoU water ﬁlter the researcher could use the following equation from

Whittington et al. (2012):

Mortality Benefits ¼ Pop



Eff



CFR



VSL ð8Þ

Where Pop is the number of individuals in the target population, I is the incidence of diarrheal

disease in the target population, Eﬀ is the eﬀectiveness of the PoU ﬁlter in reducing diarrheal

disease, CFR is the case fatality rate for diarrheal disease in the target population, and VSL is the

value of statistical life in the target population. Often little is known about the exact values of

the following parameters for a given study location: I, Eﬀ, CFR, and VSL. RCTs can provide

insight into the eﬀectiveness of PoU ﬁlters (Eﬀ). However, in many instances researchers know

the most about the eﬀectiveness of a given intervention. They are far less certain about the

values of I, CFR, and the VSL for the study population (Whittington et al. 2012). Because the

mortality beneﬁts are the product of the ﬁve parameters, the uncertainty about parameter values

(or error) is multiplicative. Thus, there is limited value in spending time and resources reﬁning a

single parameter value—e.g. eﬀectiveness—when the values of the other parameters are less

well known or highly uncertain.

Categories of policy questions RCTs are well suited to answer

Despite that fact that RCTs can provide strong internal validity for the identiﬁcation of the

causal eﬀ ects of an intervention, they do not provide researchers information required to answer

a number of questions of potential interest to policy-makers. Heckman (2010) provides a useful

framework for thinking about the types of questions to which RCTs are well suited. Heckman

(ibid.) states that there are three types of policy questions in which policy-makers might be inter-

ested. First, policy-makers might want to evaluate the impact of interventions that have been or

will be implemented. Typically these types of interventions represent an incremental, or marginal,

change from existing policies. Second, policy-makers might be interested in predicting the impacts

of an intervention originally implemented in one context in another context.

Third, policy-

makers might be interested in predicting the impact of an intervention that has never been

experienced. This refers to policies that do not represent an incremental change from an existing

policy context, e.g. interventions to mitigate climate change.

In general, RCTs are well suited to answer the ﬁrst type of policy question—i.e. where the

intervention has been implemented or represents an incremental change from the existing policy

context. RCTs may also be useful in helping to predict the impact of policies implemented in

one context where such a policy has not been implemented in the past. However, because RCTs

do not prima facie have strong external validity, they do not have a comparative advantage over

other methods with respect to their ability to answer this type of policy question (Heckman

2010). Similarly, RCTs do not provide information that can help predict the consequences of

policies that that never been experienced.

We would add a fourth type of policy situation where RCTs provided limited value or

insuﬃcient information. RCTs do not provide insight into whether an intervention should be

implemented in a given context nor how to choose between multiple interventions ‘that work’.

This is well reﬂected in Das et al. who state:

Fuente and Whittington

428

[R]andomized control trials typically tell you what would happen if you were to intervene.

For instance, what would happen if the government were to provide free access to water.

[sic] Randomized control trials do not tell you whether you should intervene.

(2011)

Whether or not to implement a particular policy or program is a normative question that is

typically addressed through the lens of welfare economics and beneﬁt-cost analysis. Evidence

from RCTs can certainly contribute to such analysis but is not a substitute for it.

Summary and conclusions

Faced with very real monetary, political, and temporal constraints, researchers and policy-makers

should have a clear sense of the types of questions to which RCTs are well suited. In this

chapter, we discuss a set of fairly simple questions that policy professionals can use as a guide to

determine whether or not RCTs are well suited to their own research question or policy problem.

First, policy professionals must ask the basic question of whether they are interested in iden-

tifying the causal impact of a policy or program and estimating the magnitude of this impact. If

not, RCTs will not be of use to them. If so, RCTs may provide a useful means of obtaining

this information. Whether or not this is the case will depend on a number of other factors,

which are addressed in the questions that follow.

Second, policy professionals must ask whether or not they are interested in identifying why

or how a program works, namely the causal mechanism behind program outcomes. Under-

standing why or how a program works can provide policy professionals with a sense of whether

or not a policy or intervention is likely to work in a population or setting that diﬀers from the

particular context of the study. Guided by theory, RCTs can be designed to explore competing

hypotheses about causal mechanisms. However, they cannot provide this information without

an explicit research design to do so and do not necessarily have a comparative advantage over

other modes of social inquiry in this regard.

Third, the policy professional must ask whether or not the policy or intervention of interest

can be randomized. As a practical matter only certain types of policies or interventions can be

randomized. Recalling our water treatment example, it may be possible to randomize the distribu-

tion of PoU water ﬁlters but it is hard to imagine a situation where one could randomize the

provision of large water resources infrastructure such as dams. Similar concerns arise for large reg-

ulatory initiatives designed to improve environmental quality (e.g. air quality). This highlights

two important points. First, RCTs can only be used to evaluate interventions or policies that can

be randomized. Second, policy professionals must think carefully about whether and to what

degree their choice of research methodology constrains their policy or intervention choice set.

Fourth, the policy professional must determine when they need the information. To conduct

an RCT, a researcher must identify a partner organization—typically an NGO or less often a

government department—or build the programmatic infrastructure needed to deliver the inter-

vention of interest. This can take a considerable amount of time. More importantly, the potential

impacts of a social or environmental intervention are often not immediate. Thus, even if the

researcher has an implementation partner, the results of the RCT may not be available for

numerous years. If a policy professional needs information on program impacts immediately,

RCTs will be of limited use. Fifth, policy-makers must consider whether the randomization for

the treatment is ethically sound and political feasible.

This chapter discuses the traditional motivation for RCTs and provides policy professionals

with a more nuanced sense of the commonly cited strengths of RCTs as well as their limitations.

Randomized control trials

429

We devoted substantial attention to the limitations of RCTs and how the practical diﬃculties

associated with implementing them can cause RCTs to deviate from the ideal conditions from

which they draw their relative strength. We have done this not to advocate other modes of

program evaluation over RCTs, but to provide policy professionals with a richer sense of factors

that can complicate the ability of RCTs to provide internally valid estimates of program impact.

Perhaps more importantly, this chapter encourages policy professionals to view RCTs in the

broader context of the types of research that can inform public policy. Much of the current

debate about RCTs implicitly assumes that policy professionals seek to identify a causal eﬀect of

a policy intervention or program. While this is certainly an important goal for policy research, it

is not the only one. There are numerous other important questions that public policy research

must answer, including whether or not a particular policy or program should be implemented.

Acknowledgments

We would like to express our appreciation to Dr. Juan Robalino, Director of the Latin Amer-

ican and Caribbean Environmental Economics Program, for his comments and suggestions on a

previous draft.

Notes

1 http://www.povertyactionlab.org/

2 http://www.poverty-action.org/

3 The arguments in favor of RCTs are best reﬂected in Banerjee (2005), Banerjee and Duﬂo (2009),

Duﬂo (2004), Duﬂo et al. (2008), and Angrist and Pischke (2010).

4 Readers interested in a review of the recent books by Banerjee and Duﬂo (2011) and Karlan and

Appel (2011) should refer to Harrison (2011). The full discussion in the Boston Review can be found at:

http://www.bostonreview.net/BR36.2/ndf_behavioral_economics_global_development.php

5 See Handa and Maluccio (2010) for a more recent comparison between experimental and quasi-

experimental methods in the context of PROGRESSA.

6 A number of authors, including Duﬂo et al. (2008), Deaton (2010), and Heckman and Smith (1995)

present similar discussions of causal inference in randomized control trials. We follow the notation in

Duﬂo et al. (2008). Interested readers should refer to Morgan and Winship (2007) for a more detailed

overview of the potential outcomes framework.

7 For nontechnical readers the expectation operator E[ ] can be thought of as a simple mean.

8 This is often referred to as omitted variable bias in the econometrics literature.

9 In other words, it is not clear that: E[Y

T|T]- E[Y

C|C] = E[Y

T-Y

C|T] = E[Y

T-Y

C]. To arrive at

this general conclusion, one must invoke the Stable Unit Treatment Value Assumption developed in

Angrist et al. (1996) (see Duﬂo et al. 2008), which states that “the potential outcomes of an individual

are unrelated to the treatment status of any other individual” (p. 8). Interested readers should refer to

Morgan and Winship (2007) for a more detailed discussion of the Stable Unit Treatment Value

Assumption.

10 Cartwright (2010) calls into question whether any research method can be considered the gold standard

at all.

11 It is important to note that comparing means is one identiﬁcation strategy, which one could use with

both observational and experimental data. One could also use a diﬀerence in diﬀerence identiﬁcation

strategy with RCT data.

12 Readers interested in a more in-depth discussion of these as well as other issues associated with the

implementation of RCTs should refer to Deaton (2010), Rodrik (2008), Heckman and Smith (1995),

Heckman (2010), Imbens and Wooldridge (2009), Imbens (2010), and Duﬂo et al. (2008).

13 If conditions 1 and 3 are violated by the majority of study participants, the study cannot be considered an

RCT at all. However, “minor” deviations from conditions 1 and 3 are often experienced in real-life ﬁeld

situations. It is a matter of judgment as to how severe the violations need to be before one concluded

that the implementation of an RCT protocol failed.

Fuente and Whittington

430

14 These issues are well recognized by scholars who work on this issue. As a result, they pay considerable

attention to obtaining objective assessments of whether or not households are using the water ﬁlter, e.g.

conducting tests for residual chlorine in stored water.

15 Hawthorne and John Henry eﬀects threaten the internal validity of social experiments. Hawthorne

eﬀects refer to the fact that the act of participating in a study, or knowledge of being studied, often

inﬂuences the behavior of individuals in the treatment group. (See Gillespie (1991) for a history of the

term Hawthorne eﬀect.) For example, an individual who has been randomly assigned to the treat-

ment group in the hypothetical PoU water experiment scenario may feel fortunate to have been

selected to participate in the program and may wash their hands more often to ensure that they obtain

the maximum beneﬁt from the ﬁlter. Alternatively, the individual might engage in improved hygiene

practices to give the study team a good impression of them. In these instances, the treatment eﬀect

measured by researchers would include the co-beneﬁts of hand washing and the PoU ﬁlter, not simply

the PoU ﬁlter. Conversely, John Henry eﬀects refer to a situation in which individuals in the control

group change their behavior in response to the experiment or their treatment assignment (Zdep and

Irvine 1970).

16 Note that this would also be an issue with a quasi-experimental research design.

17 For a detailed discussion of LATE see Imbens and Angrist (1994), Heckman (1996), Heckman (1999),

Urzua and Heckman (2009), and Imbens (2010).

18 Recall that randomization addresses selection bias by ensuring that the treatment and control groups

are similar on average along both observable and unobservable dimensions. In this situation, even if the

researcher can conﬁrm that treatment and control groups are similar in observable ways, she cannot

rule out the possibility that those who agree to participate in the study are diﬀerent from those who refuse

to participate in unobservable ways.

19 This issue will also be present in quasi-experimental studies and thus should not be interpreted as a

problem with RCTs relative to other modes of research design. We raise the issue here simply to highlight

a level of complexity that is typically overlooked in discussion about RCTs.

20 Recognizing the threat to the internal validity, proponents of RCTs have openly supported calls for a

RCT registry where experiments would be registered ex ante along with proposed subgroups for analysis

and model speciﬁcation.

21 See Deaton (2010) for a summary of this issue and Freedman (2008) for a more detailed treatment.

22 See Deaton (2010) for an overview of distributional program impacts and Heckman and Smith (1995)

for a more extensive treatment of this issue.

23 Contrary to the common complaint that RCTs are atheoretical (e.g. Bardhan 2005; Basu 2005; Moo-

kherjee 2005) well-designed RCTs can be used to test hypotheses about competing causal mechanisms.

Deaton (2010) holds up Duﬂo et al. (2009), Bertrand et al. (2010), Giné et al. (2010), and Karlan and

Zinman (2008) as exemplary uses of RCTs to advance economic theory.

24 Pawson and Tilley (1997) present an interesting call for what they term “realistic evaluation” that

explicitly examines the interaction between context and mechanisms in program outcomes.

25 See Whittington et al. (2012) for a discussion of how uptake and usage can aﬀect the economic

attractiveness of a variety of interventions.

26 This is referred to as “beneﬁt transfer” in the environmental economics literature.

Bibliography

Ahuja, A., M. Kremer and A.P. Zwane. 2010. ‘Providing safe water: Evidence from randomized evaluations’,

unpublished working paper. Available at: http://www.economics.harvard.edu/faculty/kremer/ﬁles/

ARRE_CLEAN_2010_04_14.pdf

Allcott, H. and S. Mullainanthan. 2011. ‘External validity and partner selection’ , unpublished working paper.

Angrist, J. and J.S. Pischke. 2010. ‘The credibility revolution in empirical economics: How better research

design is taking the con out of econometrics’, Journal of Economic Perspectives 24(2): 3–30.

——, G. W. Imbens, and D.B. Rubin. 1996. ‘Identiﬁcation of causal eﬀects using instrumental variables’,

Journal of the American Statistical Association 91(434): 444–55.

Banerjee, A.V. 2005. ‘“New development economics” and the challenge to theory’, Economic and Political

Weekly: 4340–4.

—— 2006. ‘Making aid work: how to ﬁght global poverty – eﬀectively’, Boston Review 31(4): 7–9.

—— 2007. Making Aid Work. Cambridge, MA: MIT Press.

Randomized control trials

431

——and E. Duﬂo. 2009. ‘The experimental approach to development economics’, Annual Review of Economics

1: 151–78.

——and E. Duﬂo. 2011. Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty. Cambridge,

MA: Public Aﬀairs.

——,E.Duﬂo, R. Glennerster and C. Kinnan. 2010. ‘The miracle of microﬁnance? Evidence from a

randomized evaluation’, J-PAL Working Paper.

Bardhan, P. 2005. ‘Theory or empirics in development economics’, Economic and Political Weekly 40(40):

4333–5.

Basu, K. 2005. ‘New empirical development economics: Remarks on its philosophical foundations’, Economic

and Political Weekly 40(40): 4336–9.

Bertrand, M., D. Karlan, S. Mullainathan, E. Shaﬁr and J. Zinman. 2010. ‘What’s advertising content

worth? Evidence from a consumer credit marketing ﬁ eld experiment’, The Quarterly Journal of Economics

125(1): 263.

Bloom, H., L. Orr, G. Cave, S. Bell and F. Doolittle. 1993. ‘The National JTPA Study: Title IIA Impacts on

Earnings and Employment at 18 Months’, Bethesda, MD: Abt Associates.

Cartwright, N. 2010. ‘What are randomized trials good for?’, Philosophical Studies 147(1), 59–70.

Das, J., D. Shantayana and S. Hammer. 2011. ‘Lost in translation’, Boston Review. Available at: http://

www.bostonreview.net/BR36.2/devarajan_das_hammer_behavioral_economics_global_development.php

Deaton, A. 2010. ‘Instruments, randomization, and learning about development’. Journal of Economic Literature

48(2): 424–55.

Doolittle, F. and L. Traeger. 1990. Implementing the National JTPA Study. New York: Manpower

Demonstration Research Corporation.

Duﬂo,

E. 2004. ‘Scaling up

evaluation’, Annual World Bank Conference on Development Economics.

——, Pascaline Dupas, Michael Kremer, and Samuel Sinei. 2006. ‘Education and HIV/AIDS prevention:

Evidence from a randomized evaluation in western Kenya,” working paper, MIT.

——, R. Glennerster and M. Kremer. 2008. ‘Using randomization in development economics research: A

toolkit’, Handbook of Development Economics 4: 3895–962.

——, M. Greenstone and R. Hanna. 2008. ‘Cooking stoves, indoor air pollution and respiratory health in

rural Orissa’, Economic and Political Weekly 43(2): 71–6. Available at: http://web.mit.edu/ceepr/www/

publications/reprints/Reprint_205_WC.pdf

——, M. Kremer and J. Robinson. 2009. ‘Nudging farmers to use fertilizer: Evidence from Kenya’,

NBER Working Paper, 15131.

Fisher, Ronald A. 1935 (1960). The Design of Experiments. 8th edn. New York: Hafner.

Freedman, D.A. 2008. ‘On regression adjustments to experimental data’, Advances in Applied Mathematics 40

(2): 180–93.

Gillespie, Richard. 1991. Manufacturing Knowledge: A History of the Hawthorne Experiments. Cambridge:

Cambridge University Press.

Giné, X., T. Harigaya (IPA), D.S. Karlan and B. Nguyen. 2006. ‘Evaluating microﬁnance program innova-

tion with randomized control trials: an example from group versus individual lending’, Asian Development

Bank Economics and Research Department Technical Note Series 16.

——, D. Karlan and J. Zinman. 2010. ‘ Put your money where your butt is: A commitment contract for

smoking cessation’, American Economic Journal: Applied Economics 2(4): 213–35.

Glennerster, R. and M. Kremer. 2011. ‘Small changes, big results’, Boston Review. Available at: http://

bostonreview.net/BR36.2/ndf_behavioral_economics_global_development.php

Handa, S. and J. Maluccio. 2010. ‘Matching the gold standard: comparing experimental and non-experimental

evaluation techniques for a geographically targeted program’, Economic Development and Cultural Change

58(3): 415–47.

Harrison, G.W. 2011. ‘Randomization

and its discontents’, Journal

of African Economies 20(4): 626–52.

Heckman, J.J. 1992. ‘Randomization and social policy evaluation’, in C.F. Manski and I. Garﬁnkel (eds)

Evaluating Welfare and Training Programs. Cambridge, MA and London: Harvard University Press, 201–30.

—— 1996. ‘Randomization as an instrumental variable’, The Review of Economics and Statistics 78(2): 336–41.

—— 1999. ‘Instrumental variables: response to Angrist and Imbens’, The Journal of Human Resources 34(4):

828–37.

—— 2010. ‘Building bridges between structural and program evaluation approaches to evaluating policy’,

Journal of Economic Literature 48(2): 356–98.

——and J.A. Smith. 1995. ‘Assessing the case for social experiments’, The Journal of Economic Perspectives

9(2): 85–110.

Fuente and Whittington

432

——and E. Vytlacil. 2001. ‘Policy-relevant treatment eﬀects ’, American Economic Review 91(2): 107–11.

Imbens, G.W. 2010. ‘Better LATE than nothing: Some comments on Deaton (2009) and Heckman and

Urzua (2009)’, Journal of Economic Literature 48(2): 399–423.

——and J.D. Angrist. 1994. ‘Identiﬁcation and estimation of local average treatment eﬀects’, Econometrica

62(2): 467–75.

——and J.M. Wooldridge. 2009. ‘Recent developments in the econometrics of program evaluation’,

Journal of Economic Literature 47(1): 5–86.

Jeuland, M. and D. Whittington. 2009. ‘Cost-beneﬁt comparisons of investments in improved water

supply and cholera vaccination programs’, Vaccine 27(23): 3109–20.

Kanbur, R. 2005. ‘Goldilocks development economics: Not too theoretical, not too empirical, but watch

out for the bears!’, Economic and Political Weekly 40(40): 4344–6.

Karlan, D.S. and J. Zinman. 2008. ‘Credit elasticities in less-developed economies: Implications for

microﬁnance’, The American Economic Review 98(3): 1040–68.

——and J. Appel. 2011. More than Good Intentions: How a New Economics Is Helping to Solve Global Poverty.

New York: Penguin.

Keane, M.P. 2010. ‘A structural perspective on the experimentalist school’, The Journal of Economic Perspectives

24(2): 47–58.

Kramer, M.S. and S.H. Shapiro. 1984. ‘Scientiﬁc challenges in the application of randomized trials’,

JAMA: The Journal of the American Medical Association 252(19):

2739–45.

Kremer, M. 2003. ‘Randomized

evaluations of educational programs in developing countries: Some lessons ’,

The American Economic Review 93(2): 102– 6.

——, E. Miguel, S. Mullainathan, C. Null and Alix Peterson Zwane. 2009. ‘Making water safe: price,

persuasion, peers, promoters, or product design?’, unpublished working paper. Available at: http://ase.tufts.

edu/econ/events/neudcDocs/SaturdaySession/Session17/MKremerMakingWaterSafePricePersuasion.pdf

LaLonde, R.J. 1986. ‘Evaluating the econometric evaluations of training programs with experimental data’,

The American Economic Review 76(4): 604– 20.

Miguel, E. and M. Kremer. 2004. ‘Worms: Identifying impacts on education and health in the presence of

treatment externalities’, Econometrica 72(1): 159–217.

Mookherjee, D. 2005. ‘Is there too little theory in development economics today?’, Economic and Political

Weekly 40(40): 4328–33.

Morgan, S.L. and C. Winship. 2007. Counterfactuals and Causal Inference: Methods and Principles for Social

Research. Cambridge: Cambridge University Press.

Pawson, R. and N. Tilley. 1997. Realistic Evaluation, illustrated, reprint edn. London: Sage.

Rodrik, D. 2008. ‘The new development economics: We shall experiment, but how shall we learn?’

Faculty Research Working Papers Series, RWP08–055.

Todd, P.E., and K.I. Wolpin. 2006. ‘Assessing the impact of a school subsidy program in Mexico: Using a

social experiment to validate a dynamic behavioral model of child schooling and fertility’, The American

Economic Review 96(5): 1384–417.

Urzua, S. and J.J. Heckman. 2009. ‘Comparing IV with structural models: What simple IV can and cannot

identify’, NBER Working Paper 14706.

Whittington, D., W.M. Hanemann, C. Sadoﬀ and M. Jeuland. 2009. ‘The challenge of improving water and

sanitation services in less developed countries’, Foundations and Trends in Microeconomics 4(6–7), 469–609.

——, M. Jeuland, K. Barker and Y. Yuen. 2012. ‘Setting priorities, targeting subsidies among water,

sanitation, and health interventions in developing countries’, World

Development (under review).

Zdep, S.M.

and S.H. Irvine.1970. ‘A reverse Hawthorne eﬀect in educational evaluation’, Journal of School

Psychology 8: 89–95.

Randomized control trials

433

Policy evaluation and

public participation

Carolyn M. Hendriks

Introduction

This chapter is concerned with the methods and challenges of evaluating policies through the

participation of aﬀected publics. For this purpose, policy evaluation is understood as an act of

judgment about the performance of a particular policy process, department or program based on

its desirability, worth, or value. Evaluation is the phase of the policy process where there is

ideally an explicit opportunity for learning, reﬂection and improvement. It represents the moment

where the policy cycle ends but also restarts (Althaus et al. 2007). For the most part, this chapter

focuses on policy evaluation as a retrospective (ex post) exercise where past decisions and pro-

grams are formally assessed. Evaluations of this kind might be legally required, or they might be

stimulated by a budget or planning process. In some cases evaluations are triggered by

speciﬁc policy events, such as a perceived policy ﬁasco or change in political leadership (Bovens

et al. 2006).

The dominant approach to policy evaluation has been inspired by a rationalistic or traditional

model of policy analysis (Bovens et al. 2006). Evaluation of this kind typically sets out to

determine the facts at hand using predetermined evaluation criteria, and attempts to keep values

and ultimately politics out of the equation. In other words, the goal is to assess whether a

program or policy achieved its set objectives (is it eﬀective?), whether it is the best way to achieve

these objectives (is it eﬃcient?) and whether the policy or program is relevant under current

conditions (is it appropriate?) (Althaus et al. 2007). Formal evaluations of this kind often proceed

with an independent body (for example, an audit oﬃce, an ombudsman, a consultant and so on)

establishing the scope of the evaluation, determining the assessment criteria (for example, cost-

eﬀectiveness), then collecting relevant data (often focused on the operation and implementation

of the policy or program), and ﬁnally making an objective assessment and recommendations.

The inherent assumption with this more traditional approach to policy evaluation is that the

criteria used are appropriate and measurable, and that the conclusions drawn will be deﬁnitive

and uncontroversial.

However, in practice, this kind of systematic and rational assessment of a policy program is

rare (Bovens et al. 2006; O’Faircheallaigh 2002). Indeed, policy practice is littered with exam-

ples that demonstrate how policy evaluation is far from value neutral and apolitical. Diﬀerent

434

policy actors will almost always hold competing views on what constitutes policy success and

failure, how it might be judged and what lessons the results imply. This makes evaluation an

inevitably controversial and indeed political process, one that often generates more conﬂict and

contestation than any other stage of the policy process. Even the most low proﬁle and seemingly

uncontroversial policy programs can become political battle grounds when an evaluation process

commences because it triggers a host of questions about how funding is allocated and spent,

who wins and loses, who is accountable and responsible, and it also opens up possibilities for

reframing debates (Bovens et al. 2006).

Given the inherent political nature of policy evaluation in contemporary societies, its practice

is best understood as a process that seeks to reﬂect the plurality of perspectives on any given

policy or program. According to Majone (1989: 183) the imperative is thus “less to develop

‘objective’ measures of outcomes—the traditional aim of evaluation research—than to facilitate

a wide ranging dialogue among advocates of diﬀerent criteria.” In other words, the evaluative task

involves recognizing the multiplicity of criteria and perspectives on a given issue and helping to

“contribute to a shared understanding of the various critical perspectives and of their diﬀerent

functions in the process of public deliberation” (1989: 170). This more participatory approach to

policy evaluation requires that policy analysts (such as independent evaluators and public man-

agers) elicit multiple perspectives. In practice, this involves going well beyond engaging diverse

experts and elites, or conducting a simple client satisfaction survey. Ideally participatory policy

evaluation involves engaging a diversity of actors that aﬀect, and are aﬀected by, a policy or program

in the process of evaluation.

The perspective oﬀered in this chapter sits within this participatory or argumentative tradition

of policy evaluation. It views the assessment of any policy performance or program as something

shaped by the underlying assumptions and values of the evaluation process, and those interpreting

the evidence (Fischer and Forester 1993; Majone 1989). The argumentative approach to policy

evaluation takes seriously the role of values in shaping and determining the worth of a policy

program (Stewart 2009a). In other words, policy evaluation involves a process of judgment that

is inherently subjective; it is inﬂuenced by numerous factors such as who is doing the evaluation,

the scope of evaluation, the evaluative criteria used, and the way material is interpreted.

Participation is not a particularly well-understood or articulated theme in the literature on

policy evaluation. This is despite the growing interest in, and practice with, more participatory

and deliberative forms of governing. One reason for this could be that participation is typically

framed as an activity that occurs at the front end of the policy process, for example during agenda-

setting, policy design, and occasionally in decision-making. Yet participation is playing an

expanding role in policy evaluation particularly with the push to create “public value,” and the

broader trend toward more participatory forms of governing.

In this chapter I consider ex post policy evaluation in light of various themes emerging from

the practice and theory of participatory governance. I focus particularly on how multiple per-

spectives might be brought into the policy evaluation process, and the challenges and politics of

doing so. Consideration is given to both conventional methods of public involvement in policy

evaluation such as citizens’ surveys and public hearings, as well as more inclusive and delib-

erative forms of citizen engagement. The chapter concludes by discussing some of the particular

challenges of engaging the public in retrospective policy evaluation.

The role of participation in policy evaluation

The idea of participation is not new in discussions on policy evaluation. Indeed, some scholars

have identiﬁed participation as a central part of the evaluation process. For example, Guba and

Policy evaluation and public participation

435

Lincoln (1989), view evaluation as a process of encouraging consensus among various stake-

holders. In their methodology (referred to as “fourth generation evaluation”), stakeholders engage

in a collaborative process, and are empowered to develop and negotiate joint recommendations.

Participation also features as a theme in the work of post-positivist scholars interested in expanding

the argumentative basis for policy analysis (Fischer 1995; Forester 1999; Hajer and Wagenaar

2003b; Majone 1989). Here the emphasis is on policy analysts not so much “solving” a parti-

cular problem or ﬁnding its “truth,” but in creating interactive settings where policy arguments

and evidence can be exposed to public deliberation. More recently, scholars have argued that

interactive or deliberative forms of policy analysis are essential for making practical judgments in

contexts where governance increasingly involves a complex networks of actors, deep-value plur-

alism and radical uncertainty (Hajer and Wagenaar 2003a). Participation thus becomes central to

the tasks of evaluating how policies work, and how they might be improved. As Hajer and

Wagenaar put it: “Whatever knowledge we possess must be assessed for its relevance and usefulness

in interaction with the concrete situation at hand, and that this ongoing process of assessment

occurs in situations of intense social interaction” (2003a: 24, emphasis in the original).

Some themes on participatory evaluation have also emerged from the ﬁeld of public man-

agement. The push here has been to bring more “client focus” to the monitoring and assess-

ment of service delivery through consumer involvement. Participation in this context is about

improving knowledge on client needs, increased transparency and improved accountability for

services (Goetz and Gaventa 2001). Some of these imperatives have been promoted under the

banner of certain managerial discourses, such as New Public Management (NPM), which empha-

sizes the importance of service providers satisfying the needs and desires of their clients and con-

sumers (much like customers in the private sector). NPM has promoted passive and one-way

forms of participation, such as satisfaction surveys, that encourage respondents to think as consumers

rather than as citizens (Howard 2010; Parkinson 2004)—a theme I return to later.

“Participation” has also surfaced in recent academic and practitioner discussions on the impor-

tance of “public value” in public sector management. Public value is a concept that continues to

evolve (see Alford and O’Flynn 2009; Jørgensen and Bozeman 2007) but in broad terms it refers to

what the public values (Benington 2009; Horner et al. 2006). The idea of public value emerged

in the US as a construct to help public managers make sense of their policy tasks and to encourage

them to be more responsive to their public constituents (Moore, 1995). The notion of “public

value” has gained interest in other Western democracies, such as the United Kingdom, where it

is seen as a kind of corrective response to the managerial discourse of NPM (Horner et al. 2006). In

contrast to NPM, the concept of public value does not conﬂate the views and preferences of

consumers and citizens, and in doing so it encourages public managers to listen to the voices of

diﬀerent publics. While some of the literature on public value celebrates public participation,

little detail is given on who should participate in the policy process and how.

Participation has also surfaced in recent discussions on performance management in the public

sector. Measuring the performance of a program, department, or government is an important

component of policy evaluation—though the roots of evaluation research and performance man-

agement are diﬀerent (see Blalock 1999). Performance management is typically aimed at ensur-

ing that public agencies are doing their job in an eﬃcient and productive manner. This is as much

about promoting public accountability as it is about ensuring that governments are eﬀective

service deliverers. Typically performance measures are quantitative, for example, drawn from

budget data or from output measurements such as the units and standard of services provided

(Holzer and Kloby 2005b).

More recently, there has been a call to include citizens directly in the assessment and mea-

surement of government performance. Advocates argue that citizens can help to develop socially

Hendriks

436

relevant measures as well as provide assistance with data collection and service design (Callahan

2000; Ho 2005). Some go further and argue that citizen-led performance management can pro-

mote accountability, and strengthen democracy by fostering trust and social capital (Halachimi

and Holzer 2010). However, like discussions on public value, there has been limited discussion

on the practicalities and politics of engaging citizens in the policy process. This may change as

the practice of citizen engagement in performance management expands. Already in the United

States hundreds of performance projects have been conducted where citizens have been asked

for their feedback on services, or involved in developing performance measures. In most cases

participation appears to involve relatively non-interactive methods such as customer surveys (e.g.

Heikkila and Isett 2007; Holzer and Kloby 2005a, 2005b).

The discussion above reveals that the arguments for greater public participation in policy eva-

luation are mostly couched in epistemic terms. By engaging diﬀerent publics in the evaluation

process, we gain access to new information, alternative knowledge, and values. While this may be

the case, this particular framing of public involvement bypasses a host of democratic and managerial

reasons why participation is becoming such a central theme in contemporary governance.

Broader drivers for public participation in policy

One of the most signiﬁcant drivers for participation in public policy is a desire to democratize

the policy process and public services. On one level this is coming from increased demands from

diﬀerent publics to contribute to decisions, services, or policies that they use, fund or support.

Yet on another level there has been a push by some scholars and international bodies such as

the Organisation for Economic Co-operation and Development (OECD) to address some of

the democratic deﬁcits in the policy process by giving communities the opportunity to partici-

pate in decisions that aﬀect them (Dryzek 2000; Fung 2006b; OECD 2001). When commu-

nities are excluded from contributing to decision-making processes, they are left frustrated and

angry, and question the fairness and authority of the “imposed decision” (Renn et al. 1995). In

other words they may not accept the legitimacy of the decision, and consent to necessary beha-

vioral or funding changes (Parkinson 2006). Others point to a host of other moral reasons why

the public should have greater say in the policy process. For example, some argue that active citi-

zen involvement in public policy (especially over the long term) can promote more equitable

outcomes and address systemic inequalities in communities (Schneider and Ingram 1997). Par-

ticipation is also said to promote more transparent decision-making, and thereby improve trust

in public institutions and citizenship more broadly (Putnam 1993). These normative arguments

are all relevant to policy evaluation.

Participation has also become a key managerial and administrative theme for many govern-

ments around the world (see OECD 2009). At a minimum, participation may be a necessary

administrative requirement, for example, to fulﬁll certain legal requirements, such as an Envir-

onmental Impact Statement. Alternatively, a participatory process might be conducted to help

prevent or resolve a con ﬂict. For example, since the 1970s consultation exercises have been used

by governments as a means to prevent protest politics and restore “trust” in political institutions.

A related motive here is to use participation to help share the responsibility for policy outcomes

(including successes and failures) (Head 2007).

Public participation is increasingly advocated as a means to improve the managerial eﬃciency,

service delivery and accountability of government agencies—as discussed above. There has also

been increased recognition that the task of governing is not something that governments do

alone—it requires working with multiple interdependent actors (Wanna, 2009). Participation

promises to improve the problem-solving capacity of governments and other organizations.

Policy evaluation and public participation

437

When diﬀerent actors work together on problems, they have the opportunity to share knowledge

and resources, which can improve the eﬃciency of public programs and services (Kickert et al.

1997; Kooiman 1993). Participatory and collaborative approaches also present decision-makers

with an alternative approach to dealing with the complexities and uncertainties of modern

policy issues (Koppenjan and Klijn 2004).

More ambitious managers and administrators might utilize public participation to catalyze

change, or to shift the nature of the debate. For example, they might turn to more deliberative

processes in order to expand discussion on a controversial policy issue beyond symbols and

narrow interests (e.g. Carson et al. 2002; Einsiedel et al. 2001; Niemeyer 2004).

Public adminis-

trators might also seek to overcome some of the limitations of conventional modes of com-

munity involvement (such as opinion polls and pressure group politics), by experimenting with

new participatory processes (Hendriks 2002; Reddel and Woolcock 2003).

Who participates in policy evaluation, and how?

It would be a mistake to interpret participatory policy evaluation as a process in which relevant

experts work together to reach a consensus on the “eﬀectivenes” of a policy or program. What

this approach bypasses is the important democratic argument underlying participatory policy

evaluation; that the evaluative process (and its outcome) is legitimate to the extent that it

engages those who might be aﬀected by the policy in a process of informed public deliberation

(Dryzek 2000).

One common way to include potentially aﬀected populations in policy evaluation is to target

“stakeholder” groups (e.g. Guba and Lincoln 1989). This term is typically reserved for identi-

ﬁable groups or organizations that have expressed an interest in a policy issue, program, or

proposal. Stakeholders might also be entities that have useful knowledge or perspectives, or the

power and resources to block or promote policy change. Stakeholder engagement of this kind

carries many of the democratic dangers associated with interest group pluralism—most notably that

it tends to privilege institutionalized, well-organized and well-funded interests (Lowi 1969;

Olson 1965). In practice, stakeholder forms of participation tend to restrict debate to a select group

of elites and policy entrepreneurs (Curtain 2006; Hunold and Young 1998).

To address this mobilization bias, some scholars stress the importance of policy analysts lis-

tening to a diversity of voices including the under-represented and marginalized (Dryzek 2010;

Fischer 2000; Forester 1999). A similar idea is promoted by those working from the perspective

of public value who explicitly stress the need for broader public input beyond special interest

groups and experts. For example, Horner et al. (2006: 19) argue that the goal should be “about

placing individuals as citizens centre stage of decision-making so that public resources best serve

the public’s needs and not the self-interest of public managers, professionals or the interest of one

particular group of citizens.” Other scholars have been motivated to make emerging forms of

network governance (which typically involve public, private, and nongovernment representatives)

more inclusive of everyday citizens (Bingham et al. 2005; Hendriks 2008).

Clearly for any given policy issue or program there is a variety of relevant communities and

groups. These publics will be constituted by the particular focus of the evaluation process, as is

the case with all participatory projects (Barnes et al., 2003). It can be useful to think about how

diﬀerent categories of “publics” might be called upon to contribute to di ﬀerent stages of the

evaluation process. Such categories might include (adapted from Salter, 2007):

1 interest groups (including powerful and marginalized groups);

2 speciﬁc communities (particular socio-economic, user or disaﬀected groups);

Hendriks

438

3 elite stakeholders (businesses, experts, think tanks, consultants);

4 broader public (lay public, “ users ”, taxpayers); and

5 government oﬃcials and departments.

The question of how these various publics might practically engage in policy evaluation is less

straightforward. There is no shortage of participatory mechanisms available for engaging diﬀer-

ent kinds of publics in policy issues (Fung 2006a; Gastil and Levine 2005; Roberts 2004; Smith

2009). Participatory methods are often categorized in terms of a ladder or spectrum of public

participation where mechanisms vary in purpose ranging from informing (for example, open days,

websites and education campaigns) through to feedback and consultation (for example, surveys,

focus groups, and public hearings), through to more interactive and collaborative processes (for

example, advisory committees, citizens’ juries and partnerships) (see Head 2007; IAP2 2007). In

practice, the most common approach to public involvement in policy evaluation is to invite feed-

back or comments from the public, for example through surveys, focus groups, submissions, and

public hearings. In most of these processes, participation is not particularly interactive and the

scope for broader public input is limited. In some cases participatory processes are attached to

more formal evaluation processes involving experts, such as advisory committees and inquiries

(see Marier: Chapter 30).

In the following sections, I discuss two traditional modes of eliciting public input that are par-

ticularly prevalent in policy evaluation—satisfaction surveys and public meetings. I then take a

brief look at some of the more innovative approaches to public involvement that emphasize

inclusion and public deliberation. To be clear, my purpose here is not to provide a comprehensive

survey of diﬀerent participatory mechanisms, but rather to get a taste of their diversity, and

discuss some of the broader policy and political issues that their use in policy evaluation raises.

Citizen surveys

Citizen- or client-focused surveys constitute a common means to seek feedback on the performance

of policy program, department, or initiative. There are diﬀerent varieties but the most common

format is a satisfaction survey that uses standard survey research methodology where participants

are selected using random sampling and contacted either by mail or phone. In some cases,

citizen satisfaction might be explored using more interactive methods such as focus groups.

Satisfaction surveys are not particularly new to public administration but their popularity has

increased since the 1990s with the greater focus on customer-centered service delivery (Osborne

and Gaebler 1992). In this sense, surveys are seen as a way for public managers to gain insight

into how to improve their services to better meet the needs of their clients and customers (Kelly

2005). Satisfaction surveys also sit well with the discourse on more “evidence-based” policy-

making because they produce “evidence” from a technique largely viewed as methodologically

rigorous (Howard 2010).

There has also been a trend toward using citizen surveys as a benchmarking tool to track how

citizens’ satisfaction levels with public services change over time. In this respect Canada ’s Citi-

zens First survey has become a particularly inﬂuential model that has been replicated around the

world, providing public managers with data on program eﬀectiveness and citizen demands.

Since 1998 Canadian governments have sought the opinions of thousands of citizens on public

services in biennial surveys (for a critical perspective, see Howard 2010).

Satisfaction surveys are an attractive and popular tool for bringing the public into policy

evaluation. They provide a quick barometer of users and their expectations of public services.

However, such surveys represent shallow forms of citizen input, and are a long way from the

Policy evaluation and public participation

439

kind of participatory or deliberative evaluation that some advocate (e.g. Majone 1989). A

central problem is that citizens are given a narrow and passive role as a consumer; it also not

always clear who these consumers are, and how to ﬁnd them (Howard 2010). Evaluation sur-

veys of this kind also focus respondents on questions about whether they are satisﬁed with a

particular service, but not about broader issues such as their expectations for what the public

service should do, how services should be delivered/provided and who should use them

(Horner et al. 2006).

The data from citizen surveys has often been used to inform education campaigns rather than as

a useful external measure of performance. So, for example, where citizens’ level of satisfaction

with a service is much lower than the “objective” administrative performance data—a common

response by public managers is to better inform citizens on quality of services (Kelly and Swindell

2002). More problematic is that satisfaction surveys assume that “personal experience and

indeed self-interest is the sole determinant of a user’s evaluation of a service” (Horner et al.,

2006, p. 20). As a result they can unintentionally encourage particular public sector reforms that

are focused on customer service delivery, such as the expansion of consumer choice and the

increasing use of private sector providers (Howard 2010).

Public meetings and hearings

Another proliﬁc form of public involvement used in policy evaluation is the public meeting. In

a typical public meeting, citizens and aﬀected groups are invited to an open gathering to pro-

vide feedback on a speciﬁc policy issue, project, or proposal. Usually participation is open to

anyone to attend, and the entire proceedings are ostensibly “public”—though size typically limits

numbers (Snider 2003). Meetings vary in their formality, but the general structure involves a

public audience facing a panel of experts and oﬃcials, who may deliver a formal or technical

presentation. Following this, most public meetings allow members of the public to stand up and

brieﬂy voice their opinion. Dialogue is typically discouraged both between citizens, as well as

between citizens and oﬃcials (Adams 2004).

Some public meetings might be part of a formal policy evaluation process such as a public

inquiry. In such cases, public meetings that are legally required are referred to as public hearings.

In some countries such as Canada, public hearings are a regular part of formal public inquiries

(see Chapter 30). But as Salter (2007) points out hearings vary considerably depending on how

the public is constituted. Some public hearings call on stakeholder representatives to engage in a

process of interest negotiation, while others invite general members of the public to give personal

testimonies or oﬀer community expertise.

Public meetings remain a popular and often mandated form of public involvement in many

Western democracies (Baker et al. 2005; McComas 2001; Snider 2003). Their practice is also

increasing in more authoritarian contexts such as China (Zhong and Mol 2008). A recent vari-

ety in the UK and Australia is the use of community cabinets where ministers visit diﬀerent

locations in their jurisdictions to take questions from the community (Marsh et al. 2010; Reddel

and Woolcock 2003).

Despite their proli ﬁc use, public meetings do not enjoy a good reputation (Heberlein 1976;

Kathelene and Martin 1991). The primary criticism of public hearings is that citizens rarely inﬂu-

ence policy decisions. They are often performed as a minimalist mode of participation where

the focus is on transferring information to citizens rather than consideration of input from the

public. In many cases, meetings are timed too late in the decision-making process where there is

little room to take citizens’ concerns or ideas into account. For this reason, public meetings are often

experienced as forums for community venting, or opportunities for proponents to announce

Hendriks

440

and defend policies (King et al. 1998: 322). Some argue that public meetings are hollow democratic

rituals for manufacturing legitimacy (Topal 2009).

The participants of public meetings are also criticized on the grounds that they are unrepre-

sentative of the broader community (Roberts 2004). A common observation is that meetings

are typically occupied by people from organized interest groups; and the discussion dominated

by the articulate and those with the most intense feelings (for an overview, see Williamson and

Fung 2004).

Public meetings and hearings are also not particularly deliberative events. Much of their time

is dedicated to technical presentations or speeches. Public input is typically restricted to short

comment sessions in which there is limited or no opportunity for an exchange of arguments.

Agendas are typically set in advance and the scope of discussion tightly controlled by the com-

missioning body. The structure of public meetings also tends to foster adversarial conditions. For

example, contributors from the public typically stand at microphones at the front of the room

facing oﬃcials with their backs to the other public participants (Williamson and Fung 2004).

While public meetings may not be a useful way for citizens to deliberate or directly inﬂuence

policy decisions, some scholars argue that they provide an opportunity for citizens to perform a

number of other political functions. Empirical research ﬁnds that public meetings enable citizens:

(1) to send information and signals to decision-makers and the media; (2) to publically support

or shame public oﬃcials; (3) to set the policy agenda; (4) to delay decisions; and (5) to form

networks (Adams 2004). Unlike more structured consultative or deliberative forums (discussed

below), public meetings are open events for anyone to attend, and therefore carry opportunities

for citizens to employ advocacy tactics and political manoeuvring (Adams 2004). In this sense,

public meetings fulﬁll an important democratic role in the policy system by encouraging public

accountability and government responsiveness (Adams 2004; Snider 2003). Others speculate that

public meetings might promote positive rituals that reaﬃrm civic values and encourage social

cohesion (McComas et al. 2010).

Over the past decade, various improvements to public meetings have been proposed (e.g. see

Baker et al. 2005), including incorporating more deliberative elements (McComas et al. 2010),

linking them more closely to other forms of citizen engagement (Adams 2004), and utilizing

new forms of information technology to make them more accessible, transparent, and publicly

accountable (Snider 2003).

Innovative forms of public engagement

Many of the limitations identiﬁed above with conventional approaches to public engagement

have given rise to a host of innovative ways to involve diﬀerent publics more deeply and

meaningfully in policy issues. A particular emphasis of this innovation has been to diversify the

kinds of participants that engage and their level of participation. I focus here on procedures that

engage everyday citizens in a process of extensive public deliberation. Typically, the citizens’

deliberations are independently facilitated, and informed by input from experts and other policy

actors. The essence of the deliberative process is that the discussions and emerging recommen-

dations are based on a thorough consideration of all the arguments in view of the public interest

(Hendriks 2011).

The most popular deliberative process, particularly for local governments, is the citizens’ jury,

which typically brings together a panel of approximately 20 randomly selected citizens for sev-

eral days to deliberate on a particular policy issue (Crosby and Nethercut 2005). Bigger events

such as deliberative polls can engage hundreds of people in policy deliberations by combining

small and large group discussions (Fishkin 2009). Even larger is the 21st Century Town Meeting

Policy evaluation and public participation

441

model in which thousands of people can participate (often in multiple venues) with the aid of

networked computers and polling key pads (Lukensmeyer et al. 2005). Participants work in small

groups of around ten but are also collectively linked up to a central database for large-scale

dialogues. Some innovative deliberative procedures, such as citizens’ assemblies have been speciﬁ-

cally designed to connect to formal institutions of representative democracy, such as parliaments

and referenda (Warren and Pearse 2008).

Deliberative forms of public participation have been applied in most Western democracies on

a range of issues including urban planning, energy, gene technology, health care, housing, nuclear

waste, consumer protection, and indigenous aﬀairs (Gastil and Levine 2005; Goodin and Dryzek

2006; Hajer 2005; Hendriks 2005; Johnson 2009; Parkinson 2004).

Notwithstanding the growth of deliberative governance, its practice faces some signiﬁcant

challenges. One of the key limitations with deliberative procedures to date is their lack of

institutionalization. In practice surprisingly few deliberative projects are connected to, and have

an impact on, existing governance arrangements (e.g. Goodin and Dryzek 2006). The most well-

documented case is the British Columbia citizens’ assembly which considered electoral reform

(Warren and Pearse 2008). Limited institutionalization has implications for the sustainability of

public deliberation because it relies on actors outside the state to instigate and fund projects

(Williamson and Fung 2004). Indeed, the most innovative participatory projects (in terms of out-

reach and scale) in the United States, and in Australia have been instigated and funded by various

non-state actors such as foundations, academics, entrepreneurs, think tanks and nongovernmental

organizations (NGOs). But the promotion of public deliberation by non-state actors can be

viewed as an important part of democratizing the policy process. Indeed these participatory

projects represent “insisted spaces” used by non-state actors to build community capacity or

evaluate projects or their own internal decision-making processes (Carson 2008).

Another important limitation of deliberative governance is that its practice does not typically

extend into the realm of policy evaluation. To date most deliberative forms of public engage-

ment have been used in the early stages of policy development, for example, to assist citizens to

deﬁne policy problems and/or to develop future policy scenarios or prioritize policy options

(Hartz-Karp 2005). Typically citizens have limited opportunity to deliberate on the eﬀectiveness

of a particular policy program and its implementation. Indeed, there are only a few documented

cases where the deliberative process has given citizens the opportunity to not only set the policy

agenda, but also to monitor and evaluate progress on proposed strategies or goals. For example,

Fung (2003) describes a community policing project in Chicago in which there were extensive

ongoing deliberations with opportunities for participants to review policies and recommend

changes. A number of participatory budgeting projects have also provided opportunities for

citizens to evaluate policy programs in the context of allocating and monitoring funds for local

infrastructure and public services (Baiocchi 2005; Johnson 2009).

Innovative forms of public engagement have also taken place on the web in the form of

e-governance. Governments are increasingly placing services online and encouraging citizens to

provide feedback. Social media technologies such as Twitter and Facebook have opened up

communication channels (though mostly one-way) for agencies to communicate with their

constituents. Some departments now conduct web-based surveys, but these tend to replicate

many of the problems with traditional surveys as discussed above (Robbins et al. 2008). For this

reason there has been some experimentation with more interactive forms of online engagement

where discussion is facilitated (Boner et al. 2005).

A growing number of government agencies have also been championing the application of

collaborative web technologies (web 2.0) such as wikis and social networking platforms as a

means to promote more open, accountable and responsive government (e.g. Government 2.0

Hendriks

442

Taskforce 2009). While these “crowd sourcing” approaches oﬀer new interfaces for citizens to

learn more about, and possibly engage in, government activities, there is the danger that they oﬀer

only a limited form of “e-engagement” among groups of like-minded people (Lubensky 2009).

The Internet oﬀers a number of opportunities for expanding who engages in participatory

processes, and how. Web technology has also been used eﬀectively to complement face-to-face

deliberative processes (Hartz-Karp 2005). Notwithstanding the potential of the Internet for

enhancing public participation, its use to date in public agencies has been mostly about enhan-

cing customer service as opposed to consultation and engagement (Dutil et al. 2007). Indeed the

long-term danger, particularly for policy evaluation is that opportunities for public engagement

are ﬁlled with shallow forms of e-governance that seek limited customer and client feedback

rather than considered input from citizens.

Pursuing participatory policy evaluation

The expansion of participatory governance over the past decade has gone largely unnoticed by the

literature on policy evaluation. Instead the role of participation in evaluation has been construed

in narrow terms through the lens of performance management where public engagement is

limited to one-way modes of involvement, such as citizen satisfaction surveys.

Yet as this chapter has shown ex post policy evaluation can be greatly enriched by the appli-

cation of more innovative forms of public participation. Processes that emphasize inclusion and

deliberation can help to elicit important perspectives on policy programs and agencies, and expose

relevant arguments, perspectives, and values to public scrutiny. Participation can also help public

managers to negotiate the increasing complexity of policy issues by bringing those aﬀected by

policies into the evaluation process. Most importantly the inclusion of aﬀected publics helps to

secure the democratic legitimacy of an evaluative process and its outcomes.

Participation, however, poses a number of challenges for policy evaluation. First, drawing more

voices into the evaluation process can intensify the politics of an evaluation program. While all

forms of policy evaluation involve competing actors “framing, blaming and credit claiming”

(Bovens et al. 2006: 323), opening the process up to more inclusive and deliberative forms of

participation risks intensifying contestation. This is especially the case when participation is

extended to everyday citizens, whose perspectives may not be considered valid or legitimate.

Greater deliberation in policy evaluation is likely to be resisted by powerful actors who beneﬁt

from the status quo or those who are keen to keep issues out of the public spotlight (Hendriks

2002, 2011). It also needs to be recognized that not all services and issues are amenable to direct

citizen engagement and participation. Fair and open dialogue will be diﬃcult in those contexts

where there is a sense of policy failure, crisis or declining trust (Bovens et al. 2006).

The second challenge concerns the common criticism that public participation is too often

tokenistic and unsustainable. Many participatory eﬀorts represent one-oﬀ projects that are

poorly integrated into existing governance arrangements and institutions (Williamson and Fung

2004). Evaluation processes are especially vulnerable in this respect because they are often out-

sourced to independent bodies. Moreover, it is not uncommon for the one agency or policy to

be evaluated multiple times resulting in piecemeal and disconnected recommendations (Bovens

et al. 2006). The challenge with any participatory evaluation is to ensure that it is connected to

other forms of evaluation and the broader institutions of governance. It is also important that

any participatory evaluation is situated in an overall strategy for public engagement that targets a

range of policy actors using diﬀerent processes.

Third, participation can generate tensions and complexities for public managers. One core

challenge is trying to reconcile the public’s desire for greater involvement while at the same

Policy evaluation and public participation

443

time attempting to be an authoritative voice or decision-maker (Yang and Callahan 2007). Public

managers also face constraints within their organizations, for example they are often unable to

choose the way they engage with the community, since “the mandate and powers of their agency

shape purpose and practice” (Stewart 2009b: 49). As a result, many participatory activities are

outsourced to private consultants, which has implications for public sector accountability (see

Speers 2007) and the sustainability of public engagement more generally (see Hendriks and Carson

2008; Williamson and Fung 2004). The broader challenge here is that participatory policy evalua-

tion requires a shift in the way bureaucracies authorize, manage and demonstrate accountability

(Goetz and Gaventa 2001).

Participation also requires public managers to accommodate alternative forms of knowledge

that typically do not sit well alongside more traditional evaluation approaches which produce

quantitative data with seemingly “deﬁnitive” results. Ultimately, the outputs from any evalua-

tion process involving participation must compete with the ﬁndings emerging from other more

traditional forms of policy evaluation. In the process of judging di ﬀerent arguments, the validity

of citizens’ preferences can often be called into question, and labeled as too impressionistic,

irrational, or idiosyncratic (Hendriks 2011; Yang and Callahan 2007). The challenge for public

managers is to ensure that citizens’ input into an evaluation process receives equal consideration

alongside the perspectives of elites and lobby groups.

In order to tackle these and other challenges, the ﬁeld of public administration needs to better

equip its practitioners with skills for undertaking participatory forms of governance (Bingham et

al. 2005; OECD 2001). New skills and incentive structures are required to help practitioners

expand their participatory eﬀorts in evaluation beyond citizen surveys towards more inclusive

and deliberative approaches. Skills are also needed to not only manage and conduct participation,

but also to eﬀectively communicate its limits and constraints. As Stewart suggests (2009b: 49):

“Engagement is about understanding where people are coming from and making sure they know

which of their concerns might (or might not) be aﬀected as a result of their participation.”

Deepening the role of participation in policy evaluation oﬀers an important opportunity to

democratize the policy process. After all evaluation is the stage where the eﬀectiveness of policy

programs are assessed, and the site where new ideas and agendas are formed. Thus widening and

deepening public involvement in the policy evaluation process will enrich the conclusions

reached, and ultimately their legitimacy.

Notes

1By“deliberative” here I mean an interactive form of communication in which participants provide

reasons for their positions, listen to the arguments of others, and are open to changing their position in

view of the better argument (Dryzek 2000).

2 For more, see Institute for Citizen Centre Service (ICCS), available at: http://www.iccs-isac.org/en/cf/

Bibliography

Adams, B. 2004. ‘Public meetings and the democratic process’, Public Administration Review 64(1): 43–54.

Alford, J. and J. O’Flynn. 2009. ‘Making sense of public value: Concepts, critques and emergent meanings’,

International Journal of Public Administration 32(3): 171–91.

Althaus, C., P. Bridgman and G. Davis. 2007. The Australian Policy Handbook. Crows Nest: Allen and

Unwin.

Baiocchi, G. 2005. Militants and Citizens: The Politics of Participatory Democracy in Porto Alegre. Stanford, CA:

Stanford University Press.

Baker, W.H., H.L. Addams and B. Davis. 2005. ‘Critical factors for enhancing municipal public hearings’,

Public Administration Review 65(4): 490–9.

Hendriks

444

Barnes, M., J. Newman, A. Knops and H. Sullivan. 2003. ‘Constituting “the public” in public participation’,

Public Administration 81(2): 379–99.

Benington, J. 2009. ‘Creating the public in order to create public value?’, International Journal of Public

Administration 32(2): 232–49.

Bingham, L.B., T. Nabatchi and R.O. Leary. 2005. ‘The new governance: Practices and processes for

stakeholder and citizen participation in the work of government’, Public Administration Review 65(5): 547.

Blalock, A.B. 1999. ‘Evaluation reseach and the performance management movement: From estrangement

to useful integration?’, Evaluation 5(2): 117–49.

Boner, P.A., R. Carlitz, R. Gunn, L.E. Maak and C.A. Ratliﬀ. 2005. ‘Bringing the public and government

together through online dialogues’, in J. Gastil and P. Levine (eds) The Deliberative Democracy Handbook:

Strategies for Eﬀective Civic Engagement in the 21st Century. San Francisco: Jossey-Bass, 141–53.

Bovens, M., P. ’t Hart and S. Kuipers. 2006. ‘The politics of policy evaluation’, in M. Moran, M. Rein

and R.E. Goodin (eds) The Oxford Handbook of Public Policy. Oxford: Oxford University Press, 319 – 35.

Callahan, K. 2000. ‘Performance measurement and citizen participation’, in M. Holzer and S. Lee (eds)

Public Productivity Handbook, 2nd edn. New York: Marcel Dekker.

Carson, L. 2008. ‘Creating democratic surplus through citizens’ assemblies’, Journal of Public Deliberation

4(1): Article 5. Available at: http://services.bepress.com/jpd/vol4/iss1/art5/

——, S. White, C. Hendriks and J. Palmer. 2002. ‘Community consultation in environmental policy

making’, The Drawing Board: An Australian Review of Public Aﬀairs 3(1): 1–13.

Crosby, N. and D. Nethercut. 2005. ‘Citizens juries: Creating a trustworthy voice of the people’,in

Gastil and P.

Levine (eds) The Deliberative Democracy Handbook: Strategies for Eﬀective Civic Engagement in

the Twenty-First Century. San Francisco: Jossey Bass, 111–19.

Curtain, R. 2006. ‘Engaging citizens to solve major public challenges’, in H.K. Colebatch (ed.) Beyond the

Policy Cycle: The Policy Process in Australia. Crows Nest: Allen and Unwin, 121–42.

Dryzek, J.S. 2000. Deliberative Democracy and Beyond: Liberals, Critics, Contestations. Oxford: Oxford University

Press.

—— 2010. Foundations and Frontiers of Deliberative Governance. Oxford: Oxford University Press.

Dutil, P.A., C. Howard, J. Langford and J. Roy. 2007. ‘Rethinking government-public relationships in a

digital world’, Journal of Information Technology and Politics 4(1): 77–90.

Einsiedel, E.F., E. Jelsøe and T. Breck, T. 2001. ‘Publics at the technology table: The Australian, Canadian and

Danish Consensus Conferences on Food Biotechnology’ , Public Understanding of Science 10(1): 83–98.

Fischer, F. 1995. Evaluating Public Policy. Chicago: Nelson Hall.

—— 2000. Citizens, Experts, and the Environment. Durham, NC: Duke University.

—— and J. Forester (eds). 1993. The Argumentative Turn in Policy Analysis and Planning. Durham, NC:

Duke University Press.

Fishkin, J. 2009. When the People Speak: Deliberative Democracy and Public Consultation. Oxford: Oxford

University Press.

Forester, J. 1999. The Deliberative Practitioner: Encouraging Participatory Planning Processes. Cambridge, MA:

The MIT Press.

Fung, A. 2003. ‘Deliberative democracy, Chicago style: Grass-roots governance in policing and public

education’, in A. Fung and E.O. Wright (eds) Deepening Democracy: Institutional Innovation in Empowered

Participatory Governance. London: Verso, 111 – 43.

—— 2006a. ‘Varieties of participation in complex governance’, Public Administration Review 66(1)

(supplementary): 66–75.

—— 2006b. ‘Democratising the policy process’, in M. Moran, M. Rein and R.E. Goodin (eds) The

Oxford Handbook of Public Policy. Oxford: Oxford University Press, 669–85.

Gastil, J. and P. Levine. (eds). 2005. The Deliberative Democracy Handbook: Strategies for Eﬀective

Civic

Engagement in the

Twenty-First Century. San Francisco: Jossey-Bass.

Goetz, A.M. and J. Gaventa. 2001. Bringing Citizen Voice and Client Focus into Service Delivery, IDS Working

Paper 138. Brighton: Institute of Development Studies.

Goodin, R.E. and J. Dryzek. 2006. ‘Deliberative impacts: The macro-political uptake of mini-publics’,

Politics and Society 34(2): 219–44.

Government 2.0 Taskforce. 2009. Engage: Getting on with Government 2.0. Canberra: Commonwealth of

Australia.

Guba, E.G. and V.S. Lincoln. 1989. Fourth Generation Policy Evaluation. Newbury Park, CA: Sage.

Hajer, M. 2005. ‘Rebuilding Ground Zero. The politics of performance’, Planning Theory and Practice 6(4):

445–64.

Policy evaluation and public participation

445

—— and H. Wagenaar. 2003a. ‘Introduction’, in M. Hajer and H. Wagenaar (eds) Deliberative Policy

Analysis: Understanding Governance in the Network Society. Cambridge: Cambridge University Press, pp. 1–30.

—— and H. Wagenaar. (eds). 2003b. Deliberative Policy Analysis: Understanding Governance in the Network

Society. Cambridge: Cambridge University Press.

Halachimi, A. and M. Holzer. 2010. ‘Citizen participation and performance measurement: Operationalising

democracy through better accountability’ , Public Administration Quarterly 34(3): 378–99.

Hartz-Karp, J. 2005. ‘A case study in deliberative democracy: Dialogue with the city’, Journal of Public

Deliberation 1(1): Article 6. Available at: http://services.bepress.com/jpd/vol1/iss1/art6.

Head, B. 2007. ‘Community engagement: Participation on whose terms?’, Australian Journal of Political

Science 42(3): 441–54.

Heberlein, T.A. 1976. ‘Some observations on alternative mechanisms for public involvement: The hearing,

the public opinion poll, the workshop and the quasi-experiment’, Natural Resources Journal 16: 197– 212.

Heikkila, T. and K.R. Isett. 2007. ‘Citizen involvement and performance management in special-purpose

governments’, Public Administration Review 67(2): 238.

Hendriks, C.M. 2002. ‘Institutions of deliberative democratic processes and interest groups: Roles, tensions

and incentives’, Australian Journal of Public Administration 61(1): 64–75.

—— 2005. ‘Participatory storylines and their impact on deliberative forums’, Policy Sciences 38(4): 1–20.

—— 2008. ‘On inclusion and network governance: the democratic disconnect of Dutch energy transitions’,

Public Administration 86(4): 1009–31.

—— 2011. The Politics of Public Deliberation: Citizen Engagement and Interest Advocacy. London: Palgrave

Macmillan.

—— and L. Carson. 2008. ‘Can the market help the forum? Negotiating the commercialization of deliberative

democracy’, Policy

Sciences 41(4): 293–313.

Ho, A.T.

2005. ‘Citizen participation in performance management’, in R.C. Box (ed.) Democracy and Public

Administration. Armonk, NY: M.E. Sharpe.

Holzer, M. and K. Kloby. 2005a. ‘Sustaining citizen-driven performance improvement: models for adoption

and issues of sustainability’, The Innovation Journal: The Public Sector Innovation Journal 10: 1–20.

—— and K. Kloby. 2005b. ‘Public performance measurement: An assessment of the state-of-the-art and

models for citizen participation’, International Journal of Productivity and Performnce Management 54(7): 517–32.

Horner, L., R. Lekhi and R. Blaug. 2006. Deliberative Democracy and the Role of Public Managers. London:

The Work Foundation.

Howard, C. 2010. ‘Are we being served? A critical perspective on Canada’s Citizens First satisfaction surveys’,

International Review of Administrative Sciences 76: 65–83.

Hunold, C. and I.M. Young. 1998. ‘Justice, democracy and hazardous siting’, Political Studies, 46(1): 82– 95.

IAP2. 2007. ‘IAP2 Spectrum of public participaton’ (International Association for Public Participation

(IAP2). Available at: http://iap2.org/associations/4748/ﬁles/IAP2 Spectrum_vertical.pdf (accessed 6

January 2009).

Johnson, G.F. 2009. ‘Deliberative democratic practices in Canada: An analysis of institutional empowerment

in three cases’, Canadian Journal of Political Science 42(3): 679–703.

Jørgensen, T.B. and B. Bozeman. 2007. ‘Public values: An inventory’, Administration and Society 39(3):

354–81.

Kathelene, L. and J. Martin. 1991. ‘Enhancing citizen participation: Panel designs, perspectives and policy

formulation’, Policy Analysis and Management 10(1): 46– 63.

Kelly, J.M. 2005. ‘The dilemma of the unsatisﬁed

customer in a

market model of public administration’,

Public Administration Review 65(1): 76.

—— and D. Swindell. 2002. ‘A multiple-indicator approach to municipal service evaluation: Correlating

performance measurement and citizen satisfaction across jurisdictions’, Public Administration Review

62(5): 610.

Kickert, W.J.M., E.-H. Klijn and J.F.M. Koppenjan. (eds). 1997. Managing Complex Networks: Strategies for

the Public Sector. London: Sage.

King, C.S., K.M. Feltey and B.O.N. Susel. 1998. ‘The question of participation: Toward authentic public

participation in public administration’, Public Administration Review 58(4): 317–26.

Kooiman, J. (ed.). 1993. Modern Governance: New Government: Society Interactions. London: Sage.

Koppenjan, J. and E.-H. Klijn. 2004. Managing Uncertainties in Networks: A Network Approach to Problem

Solving and Decision Making. London: Routledge.

Lowi, T.J. 1969. The End of Liberalism: The Second Republic of the United States, 2nd edn. New York:

Norton.

Hendriks

446

Lubensky, R. 2009. Submission to Government 2.0 Taskforce (Canberra). Available at: http://gov2.net.

au/submissions/ron-lubensky/.

Lukensmeyer, C.J., J. Goldman and S. Brigham. 2005. ‘A town meeting for the twenty-ﬁrst century’,in

J. Gastil and P. Levine (eds) The Deliberative Democracy Handbook: Strategies for Eﬀective Civic Engagement in

the 21st Century. San Francisco: Jossey-Bass, 154–63.

McComas, K.A. 2001. ‘Theory and practice of public meetings’, Communication Theory 11(1): 36–55.

——, J.C. Besley and L.W. Black. 2010. ‘The rituals of public meetings’, Public Administration Review

70(1): 122–30.

Majone, G. 1989. Evidence, Argument and Persuassion in the Policy Process. New Haven, CT: Yale University

Press.

Marsh, D., C. Lewis and P. Fawcett. 2010. ‘Citizen-centred policy making under Rudd: Network gov-

ernance in the shadow of heirarchy?’, in C. Aulich and M. Evans (eds) The Rudd Government: Australian

Commonwealth Government 2007–2010. Canberra: ANU E-press, 143–60.

Moore, M. 1995. Creating Public Value: Strategic Management in Government. Cambridge, MA: Harvard

University Press.

Niemeyer, S. 2004. ‘Deliberation in the wilderness: Displacing symbolic politics’, Environmental Politics

13(2): 347–72.

O’Faircheallaigh, C. 2002. A New Approach to Policy Evaluation: Mining and Indigenous People. Aldershot:

Ashgate.

OECD. 2001. Citizens as Partners: Information, Consultation and Public Participation. Paris: OECD.

—— 2009. Focus on Citizens: Public Engagement for Better Policy and Services. Paris: OECD.

Olson, M. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA:

Harvard University Press.

Osborne, D. and T. Gaebler. 1992. Reinventing Government. New York: Plume.

Parkinson, J. 2004. ‘Why deliberate? The encounter between deliberation and new public managers’,

Public Administration 82(2): 377–95.

—— 2006. Deliberating in the Real World: Problems of Legitimacy in Deliberative Democracy. Oxford Oxford

University Press.

Putnam, R.D. 1993. Making Democracy Work: Civic Traditions in Modern Italy.

Princeton, NJ: Princeton

University

Press.

Reddel, T. and G. Woolcock. 2003. ‘From consultation to participatory governance? A critical review of

citizen engagement strategies in Queensland’, Australian Journal of Public Administration 63(3): 75–87.

Renn, O., T. Webler and P. Wiedemann (eds). 1995. Fairness and Competence in Citizen Participation.

Dordrecht: Kluwer.

Robbins, M.D., B. Simonsen and B. Feldman. 2008. ‘Citizens and resource allocation: Improving decision

making with interactive web-based citizen participation’, Public Administration Review 68(3): 564.

Roberts, N. 2004. ‘Public deliberation in an age of direct citizen participation’, American Review of Public

Administration 34(4): 315–53.

Salter, L. 2007. ‘The public of public inquiries’, in L. Dobuzinskis, M. Howlett and D. Laycock (eds) Policy

Analysis in Canada. Toronto: University of Toronto, 291–314.

Schneider, A.L. and H. Ingram. 1997. Policy Design for Democracy. Lawrence, KS: University of Kansas.

Smith, G. 2009. Democratic Innovations: Designing Institutions for Citizen Participation. Cambridge: Cambridge

University Press.

Snider, J.H. 2003. ‘Should the public meeting enter the information age?’, National Civic Review, 92(3):

20–9.

Speers, K. 2007. ‘The invisible private sector: Consultants and public policy in Canada’, in L. Dobu-

zinskis, M. Howlett and D. Laycock (eds) Policy Analysis in Canada. Toronto: University of Toronto,

399–422.

Stewart, J. 2009a. Public Policy Values. London: Palgrave Macmillan.

—— 2009b. The Dilemmas of Engagement: The Role of Consultation in Governance. Canberra: Australian New

Zealand School of Government and ANU E-Press. Available at: http://epress.anu.edu.au/dilemmas_cit

ation.html.

Topal, C. 2009. ‘The construction of general public interest: Risk, legitimacy and power in a public

hearing’, Organization Studies 30(2–3): 277–300.

Wanna, J. 2009. ‘Collaborative government: Meanings, dimension, drivers and outcomes’,inJ.O’Flynn

and J. Wanna (eds) Collaborative

Governance: A New

Era of Public Policy in Australia? Canberra: ANU

EPress, 3–12. Available at http://epress.anu.edu.au/anzsog/collab_gov/pdf/whole_book.pdf

Policy evaluation and public participation

447

Warren, M. and H. Pearse. (eds). 2008. Designing Deliberative Democracy: The British Columbia Citizens’

Assembly. Cambridge: Cambridge University Press.

Williamson, A. and A. Fung. 2004. ‘Public deliberation: Where we are and where can we go?’, National

Civic Review, winter.

Yang, K. and K. Callahan. 2007. ‘Citizen involvement eﬀorts and bureaucratic responsiveness: Participa-

tory values, stakeholder pressures, and administrative practicality’, Public Administrative Review March/

April: 249–64.

Zhong, L.J. and A.P.J. Mol. 2008. ‘Participatory environmental governance in China: Public hearings on

urban water tariﬀ setting’, Journal of Environmental Management 88(4): 899–913.

Hendriks

448

Part IX

Policy dynamics

Patterns of stability and change

Policy dynamics and change

The never-ending puzzle

Giliberto Capano

Understanding policy change in order to grasp policy dynamics

Public policy is a dynamic phenomenon. This means not only that policies are intrinsically

processual, but also that as a whole they are not the simple sum of their individual parts (stages,

actors, institutions, time, instruments, etc); at the same time their parts are variously interlinked

via a dense web of interconnections, the nature, dynamics and eﬀects of which should not be

taken for granted. The world of policy dynamics is an incredibly fascinating puzzle of which

scholars are constantly striving to discover the basic elements and reliable, regular patterns. It is

precisely because of their dynamic nature that policies may appear at one and the same time

both stable and changing. However, stability and change are a product of observers’ viewpoints.

They depend on which analytical categories are used, and on what observers are actually looking

for. The same element of dynamics, or the same time span, in the historical development of a

policy may appear either stable or characterized by signiﬁcant changes. The diﬀerent theoretical

perspectives cannot, however, hide the dynamic nature of public policy, and the diﬃculties

encountered in trying to explain its underlying features. Moreover, those scholars who assume

that policies are characterized by long periods of stability have to admit that in the end these

same policies change, possibly in a dramatic, radical manner, and they often struggle to explain

such structural movements. On the other hand, those who assume that the dynamic character of

policy means that it continuously changes, albeit from another point of view, have to deal with

a complex, multifaceted, evasive set of questions.

The truth is that in order to understand policy dynamics, the analytical focus has to be directed

towards change. As Heraclitus stated, “everything ﬂows”: change is a feature of human devel-

opment, and also of social and political phenomena. The same is true of public policies: they are

“moving events, routines, strategies, and adaptations” (Heclo 1972: 83). In the policy ﬁeld, as in

politics, change is the norm (Lewis and Steinmo 2010), so to say that “all policy is policy

change” (Hogwood and Peters 1983) is not a provocation or a symbolic metaphor, but an

inexorable fact.

Change is the normal status of policy-making, whereas stability is simply contingent upon

human agency (through rules, institutions, social norms, ideas, etc.) striving to control envir-

onmental uncertainty (Blyth 2011). However, this collective eﬀort is not always coherent, and

451

is very often conﬂictual (because diﬀerent actors have diﬀerent ideas about what constitutes stabi-

lity), and thus the stability of public policy is simple a partial, static point of equilibrium in policy

dynamics. From this point of view, studying policy dynamics means developing, from both

theoretical and empirical points of view, patterns of policy change over time.

All theoretical attempts made to explain the puzzle of policy change have to deal with certain

fundamental issues which may be summarized as follows.

First of all, there is a set of basic questions. What has really changed? How do policies change?

When do they change? Why do they change? (Capano 2009). These questions focus on the

fundamental elements that every scholar interested in solving the puzzle of policy dynamics has

to face up to, namely: (1) the deﬁnition of what change is at stake (the dependent variable

problem); (2) an understanding of how policy change develops (the logic of change direction);

(3) the explanation of why policy change occurs (the causality problem); and (4) the reasons for

the timing of change.

Second, all eﬀorts to explain policy have to identify its fundamental causal mechanisms (also

perceived as the drivers, generators, motors of change).

Third, all theoretical frameworks have to deal with the problem of the role of agency in

policy change (are actors really agents of change, or are they simply tools in the hands of structural

factors?).

The most important frameworks present in studies of policy dynamics—including the Advo-

cacy Coalition Framework (ACF) (Sabatier and Jenkins-Smith 1993; Sabatier and Weible 2007),

the Punctuated Equilibrium Framework (PEF) (Baumgartner and Jones 2002, 2009; Baum-

gartner et al. 2009), the Multiple Stream Approach (MSA) (Kingdon 1984; Zahariadis 2003),

the Path Dependence Framework (PDF) (Mahoney 2000; Pierson 2000, 2004), the Policy

Generations Framework (PGF) (de Vries 2005, 2010), as well as neoinstitutionalist theories (both

from the historical perspective (Streeck and Thelen 2005; Mahoney and Thelen 2010) and from

the rational perspective (Ostrom 1990, 2005), have all proﬀered diﬀerent sets of answers to the

above-mentioned issues.

The basic questions

The dependent variable problem

The study of the content of policy change is a puzzle within a puzzle, since very often the object of

analysis is presented in an ambiguous, unclear manner. Policies are not only dynamic, but are

also composed of diﬀerent elements (actors, goals, instruments, strategies, laws, plans, paradigms,

preferences); furthermore, from a structural point of view, policies are multilevel processes (health

policy ranges from the state level to the individual level, via an intermediate political-institutional

level). So when policy change is the dependent variable, we need to clarify exactly “what”

dimension of policy is under scrutiny, and “where” this is situated in the institutional arrange-

ment. The “what” question represents the real cornerstone of all attempts to analyze policy change

(although the “where” question may still be important). The “what” question is not a problem

for mainstream political science, whereas policy change is simply the output of the political process.

However, this perspective (which can help us to understand, for example, the content of policy

decisions at the legislative level) is somewhat limited, and is not useful to those who believe in

the dynamic, processual, and multilevel nature of policy-making. For policy scholars studying

policy as political scientists, policy is not simply an output of the political process but is the

political process itself. So, in order to resolve the question, the complex composition of policies

should be disaggregated for analytical purposes (Howlett and Cashore 2009).

Capano

452

It is clear that this analytical disaggregation is a researcher’s decision. It depends on the deﬁ-

nition attributed to policy, and on the scientiﬁc goals of the investigation. Policy literature since

the end of the 1980s has witnessed various deﬁnitions of the diﬀerent components of policy,

and thus of policy change, starting with the idea of tripartition, submitted by Sabatier (1988), based

on the distinction between technical aspects, policy strategies, and deep policies values. Peter

Hall (1993) echoed this tripartition idea, albeit from a diﬀerent theoretical perspective. Cashore

and Howlett (2007) proposed a sixfold taxonomy of policy elements, based on the separation

between ends and goals, and the tri-cotomization of the level of abstraction through which ends

and means can be dealt with (high level abstraction, program level operationalization and spe-

ciﬁc on-the-ground measures). However, none of these examples is exhaustive, since the dis-

aggregation of the components of policies, and thus of the possible dependent variable in the

case of the analysis of policy change, is based on the speciﬁcdeﬁnition of what policy is, as there

could be several potential lists of the constitutive components of policies. What is important

here, nevertheless, is that research attempts to clearly deﬁne which dimension, aspect or element

of policy is the real object of its analysis.

The game being played out here is not a simple one: due to the constitutive multidimensional

nature of policies, their dynamic nature, and the unclear character of their horizontal and ver-

tical borders, the problem of identifying the dependent variable in policy change is highly

demanding, and one that needs to be dealt with while avoiding any form of reductionism. You

can choose to study policy change in a speciﬁc sector by focusing on changes in instruments, or

in the legislative content of policies, or in the role of policy actors; however, it is theoretically

and methodologically wrong to assume that the chosen aspect perfectly represents the overall trend

in that policy sector. Thus a strictly epistemological consciousness is required in order to accept

that the choice of dependent variable—the way in which policy change is operationalized—

does not perfectly ﬁt with the observed general developments of policy dynamics, whereas it

can help to explain a part thereof. The possible implications of the ways in which these partially

explained changes can inﬂuence the rest of the dynamics in question, are a matter for further

research.

The logic of direction problem

The “how” problem focuses on the directionality of policy change (Nisbet 1972; Baumgartner

and Jones 2002; Cashore and Howlett 2007). Policy change can be incremental or radical; linear

(a coherent development of the past) or nonlinear (when a change represents a break with the

past, creating a completely new situation); reversible or irreversible (Capano 2009). Obviously,

these dimensions of the “how” problem are not objective, but are the researcher’s intellectual

construct. “How” policies change is another puzzle within a puzzle, the solution of which lies

with the researcher, that is, it depends on his/her epistemological and theoretical choices.

However, what needs to be underlined here is that there are no strict relations between these

three dimensions of the question of “how” policy changes. For example, even though one would

think of linearity as being linked with incremental dynamics, studies show that incremental pro-

cesses can lead to radical, apparently irreversible, changes in the medium-/long-term nonlinear

output (as all historical-institutionalist studies have shown; see Mahoney and Thelen 2010 for all).

However, at the same time, linearity and incrementalism are not logically connected with

irreversible outputs (as the renewed role of governments in addressing the economy following

the ﬁnancial crisis of 2008 clearly shows). The same reasoning can be adopted with regard to the

possible links between radical change and its irreversibility. As the PEF and PGF have shown,

the punctuated developments of policy dynamics (radical changes following long periods of

Policy dynamics and change

453

stability) are not necessarily irreversible, and are very often characterized by oscillation between

policy choices (but against this conclusion are those historical neoinstitutionalists who adopt a

path-dependent perspective). It is clear here that the temporal dimension is essential if we are to

establish whether a change is linear, radical, or irreversible. However, as we shall see “tempo” is

a key factor in policy dynamics and change. It is clear that a change may seem incremental if the

period of time in question is very brief, whereas it can appear radical if the time span is increased.

The temporal dimension makes the diﬀerence in the perception of “how” policies change. How-

ever, once again what matters is the researcher’s choice: the temporal aspect of the considered

change is clearly at his/her discretion. From this point of view, all debate between the incre-

mentalist and radical perspectives, vis-à-vis policy change (as well as political change), yields

very little. Given that “everything ﬂows” in policy dynamics, in the long run everything changes

radically, albeit through an incremental process. At the same time, a radical change in time t1

can open the door to the reestablishment of old policy features in time t2. From this point of

view, the “how” of policy change should be contextualized from both the structural and the

diachronic perspective. It is intrinsically contingent and related to the speciﬁc context in which

change occurs.

The causality problem

Why do policies change, and why do they change in a given manner? In order to answer the

third fundamental question when dealing with the puzzle of policy dynamics, account must be

taken of the structural complexity of policy. In fact, policy is a process, and thus the dependant

variable chosen to operationalize policy change, is embedded in a ﬂ ux of interrelated factors

(actors, institutions, interests, routines) which inﬂuence each other; since it is a process, policy

develops over time, so the same dependent variable may be aﬀected by diﬀerent factors at diﬀerent

times; since policy does not exist in a vacuum, but is ﬁrmly situated within a speciﬁc environ-

ment (with speci ﬁc political, social, economic characteristics) there are many potential inde-

pendent and intervenient variables which could inﬂuence policy change (Capano and Howlett

2009b). This structural complexity of policy dynamics cannot be understood by pursuing over-

simpliﬁcations or general theories. In fact there is not only one possible way by which policies

change, and there are several diﬀerent possible explanations of such change. Nevertheless, recent

research may help to clarify our explanations of policy change.

First of all, the complexity of policy dynamics (and the highly intricate interdependence of its

components) is a structural factor inviting scholars to abandon their exogenous-driven expla-

nations. External factors matter, obviously, but they do not directly determine policy change.

Environmental changes can be facilitators or triggers of change, but the “what change ” and “how

policy changes” conundrums depend on the internal characteristics of the policy ﬁeld aﬀected by

the external stimuli. So all the “external shock” hypotheses should be put aside when studying

policy change. This is a real problem for the more commonly adopted explanatory frameworks:

it is particularly true for the ACF and for the PEF, which assume the fundamental role of external

perturbations, events, and changes in determining policy change. However, the relevance of internal

dynamics as a reaction (not only in an adaptive way) to environmental pressures, has been pointed

out on several occasions in the past (Mahwinney 1993; Capano 1996), and empirically proven in

recent years by research in various diﬀerent policy ﬁelds (Capano and Howlett 2009a; Williams

2009), while also being admitted by those scholars who have done relevant research following the

ACF (Weible et al. 2009). So the explanation of policy change should be integrative, by ordering

relations between external and internal dynamics. Thus the “why” of policy change can be

externally inﬂuenced, although the ﬁnal word rests with internal characteristics and developments.

Capano

454

Second, due to the complexity of policy dynamics, the relationship between cause and eﬀect

is extremely problematic, and strictly linked to context, timing, and contingency. So policy change

is unforeseeable: small movements can lead to large-scale change, while important events may

have extremely limited eﬀects.

Third, and most importantly, due to its complexity, the “why” question cannot be answered

from a linear, causal perspective. Explaining policy change cannot be done by searching for net

causal eﬀects, that is, by assuming simple and linear causal relations between independent and

dependent variables. Policy change is a matter of combinatorial causes (Ragin 2006), and due to

its complexity, the same policy change can be caused by diﬀerent combinations of causes or of

causal mechanisms present at any given moment (Jervis 1997). It would seem that Aaron Wild-

avsky (1979) was right when he invited scholars to consider policy as its own cause, because its

complexity means that its speciﬁc conﬁguration (its contextual pattern of behavior and interac-

tions) matters in the policy change process. Furthermore, policy conﬁguration is stable but not

static, due to its continual adjustments caused by both its internal and external interactions.

So in order to explain “why” policies change, the researcher has to resolve a moving puzzle:

the dependent variable is constantly moving within a context (the conﬁguration of the policy)

that is stable, but not static.

The temporal problem

For too long, policies have been considered as historical phenomena. However, policy dynamics

represent a process, meaning a sequence of events, actions, interactions, outputs and outcomes

over the course of time. The temporal dimension of policy change has been rediscovered thanks to

the historical-neoinstitutionalism and path-dependent perspectives, although both approaches are

based on a traditional perception of the temporal dimension of policies as another way of

representing historical time.

However, there are other dimensions of “time” involving policy dynamics and change. Fol-

lowing the theoretical proposals of Barbara Adam (2004), these diﬀerent dimensions of temporality

constitute the timescape of policy change. Time frame, timing, temporality, tempo, duration, and

sequence are the most relevant dimensions of the intrinsic temporality of policy dynamics and

change. The time frame is the period of time of policy dynamics that is analyzed when studying

changes in said policy dynamics. The ACF, for example, suggests that the study of policy change

should focus on a period of at least ten years of the policy dynamics, because changes need time

to develop and emerge. The time frame chosen by the researcher has a deep impact on the expla-

natory eﬀort. In fact, the longer the period analyzed, the greater the likelihood that the status of

the dependant variable changes, becoming an independent or intervenient variable Further-

more, it is clear that the time frame also aﬀects the evaluation of “how” policy has changed:

what seems to be incremental in the short run could become radical in the long run, and vice

versa. The timing of policy dynamics and change focuses on the synchronization between policy

dynamics and the external environment. For example, there could be a temporal misalignment

between the debate over possible changes inside and outside the policy domain (i.e. in the

public’s opinion or in the political system). From this point of view, timing is also the temporal

dimension of the “policy windows” (Kingdon 1984), which can be conceived as the perfect over-

lapping of the timing of diﬀerent dynamics (the political and policy ones). Temporality represents

the directionality of time: it can be linear or cyclical. It is linear when “that time returns no

more,” indicating the irreversibility of the change in question. It is cyclical when “that time

ﬁnally returns,” thus representing the reversibility of change. The tempo dimension accounts for

the intensity and the density, and thus the speed, of the dynamics, and therefore of the achieved

Policy dynamics and change

455

change. A high tempo means that the change in a speciﬁc policy time frame has been very fast,

and that time has been utilized intensely. Duration is the dimension through which policy

dynamics develop, and the time required to achieve policy change.

The sequence is the order of succession through which policy dynamics develops, and policy

scholars have speciﬁcally focused on this temporal dimension of public policy in their analytical

eﬀorts to take account of the path dependency inﬂuence on policy trajectories (see Adrian Kay

in this volume) and to explain the policy sequence through the process tracing (see Carsten

Daugbjerg in this volume). There are two ways of conceiving the temporal sequence of the policy

dynamics and change (Howlett 2009). The ﬁrst one, following the path dependence approach,

is characterized by turning points, that is, critical junctures at which policy change occurs—very

often due to contingent events—and long periods of inertia. This type of sequence represents

the historical trajectory of policy dynamics, characterized by periods of momentum in which

changes occur. It is a kind of linear succession between changes which are intended as irrever-

sible, and produced substantially by external shocks or by the incapacity of current policy (and

its inertial character) to adapt to external transformations. In this version, the sequence seems to be

external to policy dynamics and change. The second type of sequence, suggested by Haydu

(1998), is conceived as a kind of “dialectical” confrontation over time between contrasting solu-

tions to recurring problems. From this point of view, the temporal sequence of policy dynamics

and change is an ongoing battle between solutions to the same problems, in which the temporary

winner is the result of contingent factors which are not random, as in the path-dependent

sequence, but the product of agency (actors pursuing their own interests or ideas). So the logic

of this sequence is based on reversible changes. This way of deﬁning the sequence considers the

succession of events as an “internal, ” intrinsic process of policy dynamics and change. Conceived

in this way, the temporal sequence is another way of emphasizing the potential that endogenous

factors have to forcibly inﬂuence policy change.

This deconstruction of time in policy dynamics and change considerably helps our under-

standing of how important the choice of time dimension is when analyzing and accounting for

policy change. For example, were one to decide to focus on a time frame of one year, during

which a new legislative reform has been adopted, the result would be a change characterized by

high tempo (high density and speed), right timing and linear temporality. However, were this

time frame extended to 20 years, the researcher may discover, for example, that ten years later

the new law has been reversed (cyclical temporality) or its implementation has been slow and

ineﬀective (low tempo) and characterized by a contingent, and so nondecisive, overlapping between

the policy timing and the external one (Loomis 1994; Pollit 2008). The concept of time, then,

should be taken seriously by policy scholars, especially by those involved both in better devel-

oping the sequential nature of policy dynamics and in designing ﬁne-grained frameworks of

policy trajectories, while it should be pointed out here that the most highly reputed theoretical

frameworks give only limited importance to it. The PDF simply adopts the ﬁrst type of tem-

poral sequence, while the ACF takes the time frame dimension into consideration, since long-

itudinal analysis is at the basis of the analytical logic of those advancing the latter. The PGF assumes

the second type of sequence. In the case of the MSA, only timing is a relevant dimension of tem-

porality. Finally the IAD seems to be uninterested in the temporal dimension of policy-making

altogether.

Causal mechanisms of policy change

Explaining policy change is a frustrating activity for those who do not adopt a covering-law legal

perspective of explanation. If the search for linear causality is abandoned, as I have proposed

Capano

456

above, the explanatory enterprise becomes very complex, and basically never-ending. It is no

coincidence that many theoretical frameworks have been proposed in the policy literature to

explain policy change, and that they are all complex designs of multiple interconnected factors.

Combinatory causality, that is, the search for the interconnected mechanisms whereby change

occurs, characterize all of them. There is not suﬃcient space here to portray their fundamental

characteristics, and as such they are well-known to the policy scholars. Multiple Stream Approach,

Punctuated Equilibrium Framework, Advocacy Coalition Framework, Policy Generation Fra-

mework, Path Dependence approaches and Institutional Analysis and Development Framework

have for a long time been part of policy scholars’ toolkit. They are all diﬀerent from each other,

although they have certain characteristics in common: they all assume the processual nature of

policy dynamics, that is, that policy change occurs through a speciﬁc process; they analyze epi-

sodes that are continuous streams of social and political life (Tilly 2001), and they try to grasp

the most relevant factors which inﬂuence change. To be more precise, all these frameworks try

to deﬁne the most relevant causal mechanisms of policy change.

Causal mechanisms are “portable concepts that describe how causation occur” (Falleti and Lynch

2009: 1148). As Tilly points out, mechanisms “form a delimited class of events that change relations

among speciﬁed sets of elements in identical or closely similar ways over a variety of situations”

(2001: 25–6). Mechanisms constitute the links between input and outputs, between external shocks

and policy change, between internal policy development and subsequent changes. As Falleti and

Lynch point out: “Mechanisms tell us how things happen: how actors relate, how individuals

come to believe what they do or what the draw from past experience, how policies and institutions

endure or change.” (2009: 1147). So, focusing on mechanisms means searching for the “the pathway

or process by which an eﬀ ect is produced or a purpose is accomplished” (Gerring 2008: 178).

The concept of causal mechanisms is more precise than that of drivers or factors. It assumes

that a potential driver or factor of change is exactly the means by which causes produce eﬀects.

A causal mechanism means that a factor or a driver, thanks to its action, produces certain out-

comes. Another property of mechanisms is that they do not produce eﬀects in a determinist

way; their outcomes are inﬂuenced by the context. This is because the same causal mechanism

can produce diﬀerent eﬀects in diﬀerent contexts, and the same eﬀects can be produced by diﬀerent

causal mechanisms.

Assuming a mechanismic perspective leads scholars to choose the most relevant causal mechanisms

from among all potential factors, thus simplifying the possible theoretical framework.

Seen from this point of view, all of the most commonly used theoretical frameworks for

explaining policy change have clearly some speciﬁc factors handled as causal mechanisms. The

ACF, for example, has particularly emphasized actors’ beliefs and learning; the path dependence

approaches and certain versions of historical-institutionalism have focused on critical junctures

and increasing returns; other streams of neoinstitutionalist thought have placed the emphasis on

policy drift, conversion, layering, and replacement; the PEF and the PGF have given impor-

tance to the shifting of attention and on the ﬂux of new information. The PEF (in an initial

version) also emphasized the role of multiple institutional venues, which were later perceived as

‘institutional frictions’. The IAD approach has focused on institutional arrangements. The MSA

has emphasized the relevance of policy windows and policy entrepreneurs. Furthermore, all of

the aforementioned frameworks and theories pay attention, albeit to diﬀerent degrees, to the

role of ideas and institutions as signiﬁcant causal mechanisms.

So from this point of view, the variety of theoretical frameworks designed to explain policy

change re ﬂects diﬀerent

choices with regard to what are considered the most signiﬁcant causal

mechanisms in

play within policy-making. They vary also regarding another causal mechanism,

time, as I have shown above.

Policy dynamics and change

457

There are some positives to reasoning in terms of a mechanismic perspective when trying to

explain policy change. The ﬁrst one is that this approach takes due account of a chosen set of

potential causal mechanisms, handling them in a more comfortable way, thus avoiding the risk of

overestimation (for example, globalization is clearly a causal mechanism inﬂuencing many national

policy ﬁelds, but should be considered together with other speciﬁc causal mechanisms which

vary from one country to the next, according to the speciﬁc national context).

The second one is that the mechanismic approach to explanation perfectly ﬁts with the processual

nature of change: ﬁnding causal mechanisms means, in fact, reconstructing the diachronic

dynamics leading to change.

The third one is that, through a comparative analysis, this approach provides us with a better

understanding of which contextual factors favor the eﬀectiveness of speciﬁc causal mechanisms

(policy learning does not work in the same way in diﬀerent countries and in diﬀerent policy ﬁelds;

new information do not have the same eﬀects in diﬀerent countries or in diﬀerent policy ﬁelds).

Fourth, the search for causal mechanisms could be the best way of dealing with the structural,

multilevel nature of policy dynamics and change. In fact the multi-institutional arrangements of

public policy means that diﬀerent causal mechanisms can work in the same policy ﬁeld, but because

every institutional layer represents a diﬀerent context, changes are not only of diﬀerent types,

according to the diversity of the layer, but are also based on diﬀerent sets of causal mechanisms.

The same mechanism may work in diﬀerent ways, depending on the speciﬁc level at which the

change is occurring (learning at the legislative stage is somewhat diﬀerent from, and works in a

diﬀerent way from, learning at the micro level, where the change has not only to be approved

but also to be implemented).

Fifth, causal mechanisms can also help us to understand how agency and structure actually interact

and matter in determining policy change (for example, whether the eﬀect of new information

alters the behavior of actors because of structural conditions, or because of the rational will of

policy actors).

Sixth, focusing on causal mechanisms could be very fruitful when performing comparative

processual analyses, by reconstructing policy trajectories and sequences. In fact, the majority of

current theoretical frameworks are not very helpful when it comes to comparing policies or

countries, since they contain too many factors and elements (especially the ACF and IAD, whereas

PEF is more interested in analyzing policy change as a sequence of discrete decisions). This is

not a call for theoretical parsimony, which would be an impossible mission when analyzing policy

change, but simply for more aﬀordable theoretically oriented research, in which the comparative

perspective constitutes an ineludible methodological tool.

So, by focusing on sets of causal mechanisms, the study of policy change could be more man-

ageable and better suited to the accumulation of knowledge. At the same time, the adoption of

a mechanismic perspective seems to be very much in keeping both with the trajectory and the

sequence of policy dynamics and with the intrinsic processual nature of policy change. By

working on sets of causal mechanisms, the interaction and eﬀects of which develop over time, it

is possible to investigate and order the possible patterns (that is, trajectories and sequences) of

policy dynamics and change.

One ﬁnal point which deserves to be mentioned here is that the present plethora of studies of

policy change by policy scholars contains certain potential causal mechanisms which have not

been taken into due consideration, or which have been aﬀorded marginal importance. In fact,

the theoretical frameworks proposed in an attempt to explain policy change, have paid con-

siderable attention to causal mechanisms such as learning, policy legacy, lock-in eﬀects,

feed-

back, information, rules,

ideas, and institutional arrangements, and so on, but have given little

importance to other possible factors which could play a relevant role as causal mechanisms in

Capano

458

policy change, such as leadership, the features of political systems, and interest groups. These factors

are basically absent from policy change research, perhaps because they belong to the mainstream

of political science, and are thus considered to be the conceptual tools of policy scholars’

“rivals.” I think that this gap should be ﬁlled, because policy change is replete with leaders

trying to lead processes (that is, to resist or to address change) at all institutional levels; parties,

politicians, bureaucrats and governments which are all constantly involved in policy dynamics;

interests groups which, day by day, attempt to maintain or change their status and rewards

within the dynamics of policy change (for example, how much of the comparative diﬀerence in

welfare reforms is due to the national structure of interests groups?).

Thus, broadening the perspective of the potential causal mechanisms, by taking in the theore-

tical toolbox of policy scholars and the abovementioned three causal mechanisms, is a necessary

step toward improving the quality of research into policy change.

The puzzle and the agents

The dynamics and characteristics of change may vary, depending on the country in question,

the chosen policy ﬁeld, the content at stake, the institutional level, and the historical period.

What deﬁnitely does not vary is the central role of agency in inﬂuencing such change. Policies

are made by actors, not by structural conditions, ideas, institutions, political parties, or interest

groups. (Lundquist 1980). Policy actors are those who act in the policy arena, during all phases of

the policy-making process. They are the ones who learn, who receive new information on

policy problems, and who interpret such problems. They are the ones whose attention shifts dia-

chronically. They are the ones who pursue their own interests and ideas. They are the ones who

determine the internal temporal dimensions of policy change and dynamics. They are the ones

who, through their daily behavior (consisting of decisions and nondecisions, interactions, learning,

search for solutions) determine the persistence of, or change in, policy dynamics. Obviously

they are strongly inﬂuenced and guided by the context, that is, by institutions, social norms, and

patterns of institutionalized roles. However, in the end, actual decisions are the results of their

own will—their constrained will—but nevertheless their will. Policy actors are the ones who,

take causal mechanisms and make them work within the reality of policy-making. Policy actors

are those who determine how policy changes. They do so both for incremental and radical

changes (Wilsford 2010).

In fact, even in an apparently static situation, the actors’ decision to choose Y instead of X

(for example, to eliminate a speciﬁc policy instrument, or to adopt another in its place) can make a

diﬀerence because this incremental choice could alter the policy conﬁgurations, inducing other

actors to change their behavior by creating a slow, but clear, cascade of eﬀects. Although Y or

X are limited alternatives permitted by the present equilibrium, and by the strong institutiona-

lized constraints on actors, their choice is not determined by the context but depends on the

will and the actions of policy-makers. Why do policy actors opt for Y rather than X in a similar

context, given their similar preferences and boundaries? The question may be a small one, but

small decisions in similar contexts may lead to changes in policy trajectory and sequence in the

medium and long term.

Policy actors are also those responsible for deciding what can be done at critical junctures in

time and at turning points, precisely when existing boundaries and contextualized constraints

become minimized. In such contingent situations, when a new trajectory can be chosen or the

policy sequence can be redirected, it is precisely the policy actors and their coalitions who decide

what the new path to be taken is going to be, and what the contents of radical change shall consist

of. Policy actors are the ones who constantly interpret contingencies, and take responsibility for

Policy dynamics and change

459

deciding whether there are any opportunities for change, and they sometimes act so as to create

the conditions for change.

Policy dynamics is an abstract concept constantly fed by the real actions of policy actors. Such

policy actors may behave as leaders, entrepreneurs, interest-driven politicians, old-fashioned

bureaucrats, and so on, but they nevertheless constitute the core of policy dynamics and its

sequence, and thanks to their actions and interactions, policy dynamics survive and policy

change occurs.

It is precisely due to this agency role that the theoretical eﬀorts made to order policy dynamics

and to understand policy change constitute a never-ending puzzle. We constantly try to conﬁne

agency within patterns of regular behavior (institutional arrangements, arenas, networks, etc.),

but very often it evades our theoretical restraints, and thus the search for the solution to the

puzzle continues.

Bibliography

Adam, Barbara. 2004. Time. Cambridge: Polity Press.

Baumgartner, Frank R. and Bryan D. Jones (eds). 2002. Policy Dynamics. Chicago: University of Chicago

Press.

——and Bryan D. Jones. 2009. Agendas and Instability in American Politics, 2nd edn. Chicago: University of

Chicago Press.

——, Christian Breunig, Christoﬀer Green-Pedersen, Bryan D. Jones, Peter B. Mortensen, Michiel

Nuytemans and Stefaan Walgrave. 2009. ‘Punctuated equilibrium in comparative perspective’, American

Journal of Political Science 53(3): 603–20.

Blyth, Mark. 2011. ‘Ideas, uncertainty, and evolution’, in Daniel Beland and Robert Henry Cox (eds) Ideas

and Politics in Social Science Research. Oxford: Oxford University Press, 83–101.

Capano, Giliberto. 1996. ‘Political science and the comparative study of policy change in higher education:

Theorico-methodological notes from a policy perspective’, Higher Education 31(3): 263–82.

—— 2009. ‘Understanding policy change as an epistemological and theoretical problem’, Journal of Comparative

Policy Analysis 1(1): 7–31.

——and Michael Howlett (eds). 2009a. European and North American Policy Change. Drivers and Dynamics.

London: Routledge.

——and Michael Howlett. 2009b. ‘Conclusion: A research agenda for policy dynamics’, in Giliberto

Capano and Michael Howlett (eds) European and North American Policy Change. Drivers and Dynamics.

London: Routledge, 215–31.

Cashore, Benjamin and Michael Howlett. 2007. ‘Punctuating which equilibrium? Understanding thermostatic

policy dynamics in Paciﬁc Northwest forestry’, American Journal of Political Science 51(3): 532–51.

de Vries, Michiel S. 2005. ‘Generations of interactive policy-making in the Netherlands’, International

Review of Administrative Sciences 71(4): 577–91.

—— 2010. The Importance of Neglect in Policy-Making. London: Macmillan.

Falleti, Tulia and Julia Lynch. 2009. ‘Context

and causal mechanisms in political analysis’, Comparative Political

Studies 42(9): 1143–66.

Gerring,

John. 2008. ‘The mechanismic worldview: Thinking inside the box’, British Journal of Political

Science 38(1): 161–79.

Hall, Peter. 1993. ‘Policy paradigms, social learning, and the state: The case of economic policymaking in

Britain’, Comparative Politics 25(3): 275–96.

Haydu, Jeﬀrey. 1988. ‘Making use of the past: Time periods as cases to compare and as sequences of

problem solving’, The American Journal of Sociology 104(2): 339–71.

Heclo, Hugh. 1972. ‘Review article: Policy analysis’, British Journal of Political Science 2(1): 83–108.

Hogwood, Peter W. and Guy B. Peters. 1983. Policy Dynamics. New York: St. Martin’s Press.

Howlett, Michael. 2009. ‘Process sequencing policy dynamics: Beyond homeostasis and path dependency’,

Journal of Public Policy 29(3): 241–62.

——and Ben Cashore. 2009. ‘The dependent variable problem in the study of policy change: Understanding

policy change as a methodological problem’, Journal of Comparative Policy Analysis 1(1): 33–46.

Kingdon, John W. 1984. Agendas, Alternatives, and Public Policies. Boston: Little, Brown.

Capano

460

Jervis, Robert. 1997. System Eﬀects: Complexity in Political and Social Life. Princeton, NJ: Princeton University

Press.

Lewis, Orion and Sven Steinmo. 2010. ‘Taking evolution seriously in political science’, Theory in Biosciences

129 (2–3): 235–45.

Loomis, Burdett A. 1994. Time, Politics, and Policies: A Legislative Year. Lawrence, KS: University Press of

Kansas.

Lundquist, Lennart. 1980. The Hare and the Tortoise: Clean Air Policies in the United States and Sweden. Ann

Arbor, MI: University of Michigan Press.

Mahoney, James. 2000. ‘Path dependence in historical sociology’, Theory and Society 29(4): 507–48.

——and Kathleen Thelen (eds). 2010. Explaining Institutional Change. Cambridge: Cambridge University

Press.

Mawhinney, Hanne B. 1993. ‘An advocacy coalition approach to change in Canadian education’, in Paul

Sabatier and Henk Jenkins-Smith (eds) Policy Change and Learning. Boulder, CO: Westview Press, 59–81.

Nisbet, Robert. 1972. ‘Introduction: The problem of social change’, in Robert Nisbet (ed.) Social Change.

New York: Harper and Row, 1–45.

Ostrom, Elinor. 1990. Governing the Commons. Cambridge: Cambridge University Press.

—— 2005. Understanding Institutional Diversity. Princeton, NJ: Princeton University Press.

Pierson, Paul. 2000. ‘Increasing returns, path dependence, and the study of politics’, American Political Science

Review 94(2): 251–67.

—— 2004. Politics in Time: History, Institutions, and Social Analysis. Princeton, NJ: Princeton University

Press.

Pollitt, Christopher. 2008. Time, Policy, Management. Oxford, Oxford University Press.

Ragin, Charles. 2006. ‘How to lure analytic social science out of the doldrums: Some lessons from comparative

research’, International Sociology 21(5): 633–46.

Sabatier, Paul. 1988. ‘An advocacy coalition framework of policy change and the role of policy oriented

learning therein’, Policy Science 21(2–3):

129–68.

——and Henk Jenkins-Smith

(eds) 1993. Policy Change and Learning. An Advocacy Coalition Approach.

Boulder, CO: Westview Press.

——and Christopher Weible. 2007. ‘The advocacy coalition framework: Innovations and clariﬁcations’,in

Paul A. Sabatier (ed.), Theories of the Policy Process, 2nd edn. Boulder, CO: Westview Press, 189–220.

Streeck, Wolfgang, and Kathleen Thelen (eds). 2005. Beyond Continuity: Institutional Change in Advanced

Political Economies. Oxford: Oxford University Press.

Tilly, Charles. 2001. ‘Mechanisms in political processes’, Annual Review of Political Science 4: 21–41.

Weible, Christopher, Paul Sabatier and Kelly McQueen. 2009. ‘Themes and variations: Taking stock of

the advocacy coalition framework’, Policy Studies Journal 37(1): 121–40.

Wildavsky, Aaron. 1979. Speaking Truth to Power: The Art and Craft of Policy Analysis. Boston: Little, Brown.

Williams, Russell A. 2009. ‘Exogenous shocks in subsystem adjustment and policy change: The credit

crunch and Canadian banking regulation’, Journal of Public Policy 29(1): 29 – 53.

Wilsford, David. 2010. ‘The logic of policy change: Structure and agency in political life’, Journal of Health

Politics, Policy and Law 35(4): 663–80.

Zahariadis, Nikolaos. 2003. Ambiguity and Choice in Public Policy. Washington, DC: Georgetown University

Press.

Policy dynamics and change

461

Policy trajectories and legacies

Path dependency revisited

Adrian Kay

The emergence of a ‘new’ institutionalism across the social sciences has coincided with the

increased interest in temporality, change and history in social and political analysis. Institutions are

structures that endure, have a history and can be used to link temporally events and processes.

The concept of path dependency has been used within policy studies and political science almost

exclusively within a broad institutionalist framework. It is institutions that are path dependent;

as Raadschelders (1998: 569) states: ‘whatever the discipline … contemporary neo-institutional

analysis has one feature in common: the notion of path dependency’. The widespread and

cross-disciplinary use of path dependency for the analysis of institutional ‘stickiness’ makes the

concept an obvious starting point for the examination of policy trajectories and legacies. Indeed,

the concept appeals as a label for the simplest of policy dynamics: that past policy decisions act as

an institution-like constraint on the options available to current policy-makers; or to use the

language of dynamics, past policy decisions act to circumscribe or foreclose parts of policy space.

This chapter revisits path dependency in light of its appeal as a means of understanding the

process and logic of policy trajectories and legacies, with a particular focus on the reﬁnement of

the concept in response to complaints of determinism and an inability to accommodate change.

A process is path dependent if initial moves in one direction elicit further moves in that same

direction; in other words the order in which things happen aﬀects how they happen; the tra-

jectory of change up to a certain point constrains the trajectory after that point. As Douglass North

puts it, path dependency is a process that constrains future choice sets:

At every step along the way there are choices – political and economic – that provide …

real alternatives. Path dependence is a way to narrow conceptually the choice set and link

decision-making through time. It is not a story of inevitability in which the past neatly

predicts the future.

(1990: 98–9)

The reference to choice sets and decision-making reveals the origins of the concept in economics.

Indeed, path dependency is problematic for that discipline because it implies that decentralised

interactions between economically rational actors do not necessarily lead to eﬃcient outcomes;

indeed, ineﬃcient equilibria may be recognised as such but still persist.

462

The concept of path dependency is not a framework or theory or model in the terms of

Ostrom (1999: 39–41): it does not provide a general list of variables that can be used to organise

‘diagnostic and prescriptive inquiry’ nor does it provide hypotheses about speciﬁc links between

variables or particular parameters of those links. Instead, path dependency is an empirical category,

an organising concept or metaphor which can be used to label a certain type of temporal process.

As Hall and Taylor put it:

they [historical institutionalists] have been strong proponents of an image of social causation

that is path dependent in the sense that it rejects the traditional postulate that the same

operative forces will generate the same results everywhere in favour of the view that the

eﬀect of such forces will be mediated by the contextual features of a given situation often

inherited from the past.

(1996: 941)

The application of this organising concept or metaphor to a phenomenon is only the beginning

of a form of explanation. Although it asserts a relationship between the sequence of early events

and the probability of later events, the concept of path dependency does not per se provide

necessary or suﬃcient conditions to understand or explain that which it labels: path dependent

processes, even when identiﬁ ed, require theorising; it is the mechanisms that connect decisions

or actions across time that explain a path-dependent process.

Although both refer to mid-range phenomena, ‘policy’ and ‘institution’ are not synonyms.

Within the policy system, there are various structures at diﬀerent scales which act as institutions

in shaping agents’ decision-making in the formulation, enactment and implementation of policy.

These are not reducible to individual level agents or single elements in the policy process.

Examples of policy institutions are budget rules, policy networks, shared mental maps as well as

the standard operating procedures for policy-making in government departments and agencies.

Most importantly in terms of understanding policy development as path dependent, past policy

decisions are institutions in terms of current policy decisions: they can act as structures that can

limit or shape current policy options.

The question of what about a policy is path dependent does not admit a single, conclusive

answer; rather it remains an open question for scholars applying the concept with theoretical

and empirical corollaries. If the policy whole or system is path dependent, there may be several

potential underlying mechanisms operating, independently or in combinations. This property of

‘multiple realisability’ has theoretical implications in terms of the micro foundations of path depen-

dency and the spatial and temporal scales of policy analysis. It is necessary when using the concept,

either theoretically or empirically, to be clear about the perspective being adopted. The devel-

opment of a policy may be labelled path dependent over some period, but the various mechanisms

that underlie that process remain unclear unless a more ﬁne-grained perspective is adopted.

Without micro foundations, the value of the concept in making sense of policy trajectories is

doubtful. Indeed, path-dependent processes may coexist with other types of processes within

policy systems. This is important in order to avoid the imputation that path dependency is

simply a fashionable neologism for the notion that ‘history matters’. An adequately ﬁne-grained

perspective is essential in using path dependency analytically; when policies or elements of

policies are seen as strongly interrelated or where our analytical lens shows policy institutions as

deeply interwoven then a much clearer sense of the mechanisms that underlie path-dependent

processes is gained.

The ﬁrst section of the chapter considers the application of path dependency to the analysis

of policy development and its potential advantages in understanding the dynamics of that

Policy trajectories and legacies

463

development. The next section considers several criticisms of the concept: that it lacks a convin-

cing account of decision-making over time, both of the accumulation of constraints and context

bound rationality; it is incapable of dealing with policy change; and it lacks a clear normative

focus. In the ﬁnal section, I argue that despite its theoretical underdevelopment and relatively

limited number of successful empirical applications, the concept of path dependency does have

potential utility in the ﬁeld of policy studies in terms of understanding why policies might be

diﬃcult to reform and also why they may tend to become more complex over time.

Beneﬁts of path dependency for understanding policy trajectories

and legacies

Path dependency is an appealing concept for understanding public policy development; it pro-

vides a label for the observations and intuitions that policies, once established, can be diﬃcult to

change or reform. Early and inﬂuential examples of the use of path dependency for understanding

policy development include health care policy in the US (Hacker 1998, 2002; Wilsford 1994)

and the UK (Greener 2002); the reform of housing beneﬁt in the UK (Kemp 2001); to UK

pension policy (Pemberton 2003); and the Common Agricultural Policy (CAP) of the European

Union (Kay 2003).

Path dependency encapsulates the insight that policy decisions accumulate over time; a pro-

cess of accretion can occur in a policy area that restricts options for future policy-makers. In this

sense, path dependency arguments can ‘provide an important caution against a too easy con-

clusion of the inevitability, ‘ naturalness’, or functionality of observed outcomes’ (Pierson 2000a:

252). For example, Pemberton (2003) argues that the pensions ‘crisis’ in the UK is not primarily

demographic but rather due to a low savings rate; further, this low rate is a function of the path

dependency and increasing complexity of pension policy. The system of pension provision in the

UK has shifted over the last 25 years from one dominated by state provision to one in which

the state pension plays a residual welfare role. Despite this large change at the policy system level,

there is evidence of path dependency in particular policy subsystems. In the case of UK pen-

sions, policy subsystems exist around speciﬁc pension schemes. An individual contract established

under a particular pension scheme at a particular time is costly to change: there are large sunk

costs; increasing returns associated with rising numbers of contributors and pensioners in a par-

ticular scheme; further, there may be signiﬁcant learning eﬀects. All of these factors contribute

to signiﬁcant switching costs for the abolition of one scheme and the transferring of that set of

individual contracts into a superseding scheme. Particular schemes are ‘locked in’ for particular

individuals. Nonetheless, pension reform has been possible but change has come in the form of

the addition of new schemes or elements to the system. This amounts not to a single, path dependent

policy trajectory but rather to a widening array of ‘locked in’ subsystems over time. This accretion

of new subsystems or schemes has led to the increasing complexity of the overall system of pen-

sions and raised questions of eﬀectiveness at a policy ‘whole’ level. Kemp (2001) reports similar

dynamics with regard to housing beneﬁt policy in the UK.

Path dependency can help separate not just diﬀerent orders of policy change as in Hall (1993),

but diﬀerent rates of policy change. One of the foundations of dynamic policy analysis is the

assumption that there are a multitude of temporal scales immanent in any system. Therefore, within

a policy system, there may be some elements that are path dependent, and others that are not.

Further, there may be a relationship between the diﬀerent processes at diﬀerent speeds. Later in

the chapter, I develop the point that it is the combinations of institutions and policies that

provide the important mechanisms underlying path dependency but also create the potential for

strategic action and policy innovation by agents.

Kay

464

The concept of path dependency has the additional advantage of ﬂexibility for policy scholars.

In particular, because the concept does not contain within it a ﬁxed temporal or spatial scale of

analysis, the insights of path dependency can often complement rather than rival other accounts

of policy change. Social housing in the UK is an example of where path dependency can hold

at the subsystem level with interesting consequences, but where the policy system as a whole

has changed profoundly. Since the mid-1970s there has been a series of failed initiatives by

central government to directly control the rents charged in the social housing sector and ensure

equity between local authorities and housing associations. This particular element of social

housing policy is path dependent; each local authority has an established policy for rent calcu-

lation, often determined by initial decisions made in the immediate post-war era. That these

have proved resistant to central control or inﬂuence signiﬁcantly constrains the ability of central

government to pursue some of its objectives for social housing. However, during the same

period as this path-dependent process there has been a clear shift in the social housing policy

paradigm as home ownership increased substantially, notably through the central government’s

‘right to buy’ scheme introduced in the early 1980s that allowed local authority tenants to buy

their own home.

Problems of path dependency for understanding policy trajectories

and legacies

Accounts of decision-making over time

The criticism that the concept of path dependency lacks explanatory power is well expressed by

Raadschelders:

it is only by virtue of retrospect that we are aware of stages or paths of development. ‘Path

dependency’ refers to a string of related events: causality in retrospect. The concept does

not come even close to pinpointing a mechanism or the mechanisms that propel social

change.

(1998: 576)

The quotation contains two criticisms. The ﬁrst is that the concept cannot be used for current or

future phenomena. This is, of course, not a singular feature of path dependency but common to

all concepts that are useful for structuring retrospective, ‘ thick’ historical descriptions to support

narrative explanations in the social sciences. Indeed, as discussed in the previous two chapters, a

dynamic perspective reveals that this is not an appropriate standard for considering the utility of

diﬀerent concepts, theories or metaphors.

The more important criticism is that even if one accepts path dependency as intuitively appeal-

ing, it is unlikely to be convincing analytically because the notion does not provide any ﬁne-

grained mechanisms that might provide necessary and suﬃcient conditions for the process observed.

The challenge for the use of path dependency in the study of policy trajectories is the unco-

vering of mechanisms that can help make sense of a path dependent process. One inﬂuential strand

of the literature on path dependency has worked on the micro foundations of the concept using

insights from new institutional economics. Much of this literature proceeds by analogy from

technological development to institutional development. In simple terms, imagine two tech-

nologies, A and B, both of which are subject to increasing returns but there is uncertainty over

the rate of increasing returns. Initial adoptions of one technology, say B, which may occur for a

number of small or chance reasons, beget further adoptions of B in the market because of

Policy trajectories and legacies

465

increasing returns namely it becomes cheaper for future ﬁrms to adopt technology B rather than

A. The interesting results from the models built on these assumptions (for economists at least)

are that you might get ineﬃcient technologies adopted by markets. The normative implications

of this borrowing from economics are considered in more detail later.

Arthur (1994) states the circumstances in which path dependence as an increasing returns

process is likely: the presence of large ﬁxed (and sunk) costs; network eﬀects; learning eﬀects;

and adaptive expectations. As noted, these factors have been used at a macro, constitutional level

to make arguments about path dependency in institutional development (North 1990; Pierson

2000a, 2000b, 2000c). Within this list of sources of increasing returns, it is useful to distinguish

between those factors that relate to the internal eﬃciency of ﬁrms – large ﬁxed costs leading to

declining average costs as production increases and learning eﬀects – from those that are external

to the ﬁrm, in particular, network eﬀects. The distinction is important because later in the chapter,

path dependency is discussed in terms of the increasing returns involved in combinations of

institutions and policies rather than increasing returns as a property of the internal operations of

ﬁrms. This complements the insistence on a ﬁne-grained perspective of policy systems developed

in this chapter.

A focus on increasing returns is only a partial interpretation of the economics of path depen-

dency. Increasing returns are suﬃcient but not necessary for path dependency. As Arrow (2000)

points out, the existence of signiﬁcant sunk costs along with sequencing arguments can support

many of the path dependency narratives of technological change. Although he does not use the

concept of path dependency, Arthur Stinchcombe’s (1968: Chapter 5) celebrated work on

constructing theories of historical causation emphasises the central importance of sunk costs. I

submit that any decision that is diﬃcult to reverse and which has enduring and ongoing eﬀects

can be said to have initiated a path-dependent process; and work on path-dependent processes

should not focus exclusively on increasing returns processes and it is moot whether they should

privilege this mechanism-type over other sources of path dependency.

Contrarily, Schwarz (2004) argues that the combination of initial, small and contingent steps

with increasing returns deﬁnes a path dependent process; if there was an large initial cause that

had signiﬁcant and enduring eﬀect on the subsequent process then this historical cause would be

salient in any explanation of the process. The process would no longer be path dependent but

rather, as a policy dynamic, be better characterised as the temporal unravelling of the consequences

of some initial event. In a similar vein, Page (2006) introduces the idea of phat dependence to

refer to cases where it is the distribution rather than the precise sequence of early events that is

important in shaping later events. Both these insights are noteworthy in highlighting the need

for future scholarly work in cataloguing and explaining diﬀerent types of path dependency and

its relationship to other policy dynamics such as process sequencing (see Daugbjerg in this volume).

However, for the sake of expositional clarity this chapter remains with the deﬁnition of path

dependency as a general metaphor for sticky policy or institutional processes.

A number of non-increasing returns mechanisms have been suggested as underlying path

dependency in policy development: the eﬀect of policy on interest groups as when policies

constrain some groups and enable others (Pierson 2000a); self-reinforcing mechanisms where

policies involve investment or disinvestments in administrative infrastructure which transforms gov-

ernmental capacity and the set of possible future policies that may be enacted (Skocpol 1992);

policies that involve the establishment of formal or informal contracts with individuals (Pem-

berton 2003; Kay 2003; Kemp 2001) which are costly to change. Further, there are network

eﬀects to types of contracts rather than the number of signatories. Once a contract is established,

the transaction costs of agreeing another contract of that type in that area of public policy will

be considerably lower than any alternative contract.

Kay

466

All these policy-speciﬁc mechanisms are based on deﬁnite, conscious choices, which have the

foreseeable consequence of high future switching costs; none relies on an increasing returns pro-

cess. Nevertheless, there are examples that suggest that increasing returns processes can occur in

policy development. In the structural reforms of the primary care sector in the UK after 1997, a

series of primary care models were piloted. By a series of chance factors, a particular primary care

trust model quickly became popular. This model subsequently became the government’s tem-

plate for all future combinations of primary care organisations. There was no particular feature

to this model to recommend it over any of the others that were piloted between 1997 and 2000;

instead it was the case that this model was adopted early in the government’s reform process,

which made it considerably easier (or cheaper) for subsequent primary care groups to use it, and

with such a momentum became the template adopted by the government for all primary care

agglomerations. At a general level, all metaphors of policy ideas or proposals emerging from the

‘policy soup’ or ‘garbage can’ share a notion of a market place of ideas; analogous to a market

system, ideas compete for attention and inﬂuence. Where the market structure produces increasing

returns, policy ideas can succeed into proposals and eventual enactment by an initial series of

small, contingent steps as early adopters of an idea increase the return to future adopters.

Any borrowing from microeconomics, including the idea of increasing returns, inevitably situ-

ates the agent in terms of responding to the costs and beneﬁts of diﬀerent options in a manner

consistent with straightforward parametric rationality. The assumption of this type of rationality

serves certain purposes in formal economic modelling, but to use it in more informal, intuitive

and post-positivist accounts of path dependency in public policy is problematic (Hay 2004). One

response is to use the notion of context-bound rationality in an account of decision-making in a

path dependent process. Nooteboom (1997) describes the manner in which markets lock in to

certain technologies in similar terms to how philosophers of science characterise the entrench-

ment of scientiﬁc theories. Both can be path dependent. He cites Kuhn’s famous account of

how scientiﬁc theories develop according to paradigms, a set of tacit and unarticulated guiding

assumptions, rather than the standard conceptions of ‘pure’ rationality at the heart of a scientiﬁc

approach. Further, just as the Kuhnian model challenges the conception of scientiﬁc activity approx-

imating to certain canons’ rationality, so will any parallel model applied to the economic case.

Hall’s (1993) widely cited notion of a policy paradigm exempliﬁes this insight in policy studies:

a policy paradigm is an interpretative framework that operates in the policy-making process;

speciﬁcally, it refers to the ideas and standards that specify the goals, instruments and the very

nature of a policy issue.

In these terms, the mechanism that underlies path dependency in the policy process is a form

of context-bound rationality among policy actors. The current path dependency literature is

mostly developed around the following two claims: (1) that the analogy from economics to

institutions can be extended to policy, and (2) that microeconomics can be borrowed as the

micro foundations of path dependency in policy development. Importantly, these claims require

a rational choice actor making the decisions. The assumption of this type of rationality is a strict

corollary of claims (1) and (2). This is potentially problematic for policy studies; in particular,

those parts of contemporary public policy theory that have begun to move away from a reliance

on the simple postulates of instrumental rationality to a more nuanced and contextualised views

of rationality.

Policy change and stability

At the heart of any account of path dependency is stability: observations of change challenge the

notion. This is a common criticism of the historical institutionalist school, ‘in its emphasis upon

Policy trajectories and legacies

467

path dependence and historical legacies it is rather better at explaining stability than change’ (Hay

2002:15). Thelen (1999) argues that path dependency is simultaneously overly sensitive to initial

conditions and too deterministic and mechanical with respect to subsequent policy development.

In terms of policy studies, one possible counterargument is based on the interpretation of

stability in path dependency. Speciﬁcally, the notion does allow policy change; policy legacies

constrain rather than determine current policy. Policy does change but within a particular set of

possibilities, and thus the policy may be said to exhibit stability. There are two main implica-

tions of the constrained change argument. The ﬁrst is that these bounded possibility sets may be

large or paths wide; and, the wider they are, the less the notion of path dependency can

account for current policy development. In alternative terms, the weaker is the ‘echo’ of past

policy developments in the present and the more other concepts, framework and theories are

required.

The notion of policy direction may assist constrained change accounts of policy development. A

stable policy path when projected into policy space may well imply signiﬁcant cumulative policy

change over time or in other terms, a signiﬁcant distance from the initial position in policy

space and time. Rose and Davies (1994) show the importance of compounding eﬀects as small,

incremental and constrained changes in annual budget allocations can accumulate to signiﬁcant

policy shifts over a period of a decade or more. Further, a change in direction may appear at

one distance a small perturbation but by shifting the direction of the policy may turn out in

retrospect to have been a critical juncture and a problem for the validity of path dependency as

a description.

Nonetheless, the limitations of path dependency as a conception of policy change have been

highlighted in empirical applications in public policy: for example, Kemp (2001) with respect to

housing beneﬁt reform in the UK; Pemberton (2003) with regard to pensions; Greener (2002)

on the NHS. Each of these studies ﬁnds path dependency in policy development alongside

some policy change. They consider change as a reaction to the unintended consequences or side

eﬀects of policy, or from pressure for reform due to exogenous shifts in the wider policy envir-

onment, for example where the distribution of power between interested groups has changed.

Once a dualism between policy stability and policy change is established, the notion of path

dependency is only useful for accounting for the former; indeed the purpose of the concept is

to aid understanding of policy stickiness and why actors do not change policy across time.

However, the dualism between stability and change can be avoided by considering the

sedimentation of policy decisions or the growing complexity of policy space that is implied by

the notion of path dependency. The dynamics of policy subsystem accumulation are theoreti-

cally underdeveloped but are important for the use of path dependency in policy narratives.

The development of UK pharmaceutical policy since the 1980s is an example of new policies

being added on as a ‘patch’ or ‘ﬁx’ to satisfy pressure to mitigate the consequences of the ori-

ginal policy. Relatively high prices for medicines were agreed by the government to reward

innovation by the industry under the Pharmaceutical Price Regulation Scheme (PPRS), and

this contributed to the rapid increases in public expenditure on medicines observed since the late

1980s. The PPRS remains unchanged and potentially path dependent but its budgetary con-

sequences have precipitated a series of new policies aimed at controlling the demand for medicines

in the NHS e.g. cash-limited prescribing budgets. The path dependency of particular policy

subsystems is a contributory factor in the explanation of the accumulation of these policy patches

and the growing complexity of the policy system, with potential consequences for the overall

coherence and eﬀectiveness of policy.

To reprise, a key issue when using the concept of path dependency is the granularity of the

perspective. Much of the work within the historical institutionalist literature uses the concept at

Kay

468

a macro perspective in which there is a single whole, that allows for discussion of an ‘institutional

setting’ or a ‘policy’. The path then refers to the trajectory for that composite variable, the direc-

tion of which is reinforced after early moves in the sequence. While this is valid for some nar-

ratives, from a more ﬁne-grained perspective the issue is which elements of that composite

system are ﬁxed or locked-in, and which are capable of being reformed. Further, within the policy

space occupied by the composite whole there may be potential for the introduction of new

institutions or policy subsystems.

In more ﬁne-grained analysis, increasing returns processes operate at the level of sequences of

institutional or policy choices: once an initial policy framework is established, there are strong

increasing returns involved in the choice of new, supplementary policies within that framework.

That is, an increasing returns process explains policy change qua the introduction of new, sup-

plementary policies. As North (1990: 95) states, it is ‘the interdependent web of an institutional

matrix that produces massive increasing returns’. This view of increasing returns helps to avoid

too sharp a distinction between stability and change, as seen in on-path versus oﬀ-path change

or where stability is followed by a path-breaking juncture, and the introduction of a new institu-

tional or policy setting. At a more ﬁne-grained perspective, institutions exist in combinations:

they are interdependent, with necessary and contingent relationships. Thelen describes examples

where institutional lock-in is combined with elements of institutional innovation that can push

the overall trajectory of policy and politics in a diﬀerent direction; indeed ‘to understand how

institutions evolve, it may be more fruitful to aim for a more ﬁ ne-grained analysis that seeks to

identify what aspects of a speciﬁc institutional conﬁguration are (or are not) negotiable and

under what conditions’ (2003: 233).

In an important recent work pushing the concept path dependency beyond simply the under-

standing of continuity, Crouch and Farrell (2004) consider how actors cope with exogenous

changes in their environment. At some point, the once reliably successful path no longer works,

and even though policy actors know this, they ﬁnd it extremely diﬃcult to change. Simple path

dependency has the actor trapped—in a strict sense the concept does not admit any other possibility;

however, a more nuanced account would look at how the perceived failure of a habitual path

may lead to the search for alternatives, but where that search process is itself path dependent.

The relatively informal models oﬀered by Crouch and Farrell (2004) are designed to address

the apparent determinism of path dependency once a path is selected. Change is explicitly mod-

elled as the intentional adaptation of agents to exogenous, environmental shifts. The emphasis is

on the ability of agents to reactivate redundant institutions, or convert existing institutions to

diﬀerent purposes, or borrow wholly new institutions from elsewhere to tackle exigencies.

This opens up an alternative theoretical line on the problem of policy change in historical

institutionalist theory: instead of insisting on a strict dualism between stability and change, policy

systems may be adapted to new circumstances through a gradual process of layering and con-

version rather than through periods of drastic and rapid change (Streeck and Thelen 2005; Ackrill

and Kay 2006). Layering and conversion are important causal mechanisms in a nascent non-

determinist institutional analysis of policy change (Campbell 2004; Crouch 2005, 2007). Recog-

nising that institutions shape the desires, motives and preferences of individual and group actors,

layering refers to the adding of new preferences onto an existing set of institutionally shaped

policy preferences. Such appending of new elements can lead to dysfunctionality and incoherence

in the policy system and set oﬀ path-changing dynamics. In the conversion mechanism, agents

develop new policy preferences from outside the institutionalised policy system, and convert

inherited institutions toward these novel goals and functions. There is a signiﬁcant empirical gap

in the comparative public policy literature that explores the explanatory power of the policy

layering and conversion mechanisms.

Policy trajectories and legacies

469

One implication of conversion and layering is that at the level of the composite whole, policy

systems cease to embody a simple unique logic, but rather a complex bundle of diﬀerent policy logics,

ideas and interests. Some of these may be dormant, unused or ‘forgotten’ for periods but are capable

of being reactivated by strategic action by agent in response to exogenous environmental shifts.

The reassertion of the capacity of agents situated within path dependent processes with increasing

returns to act to change the direction of the path in response to shifts in their environment is

important. Inheritance and policy legacies are not as hard or ﬁxed or as determined as some of

the simple path dependency analysis may suggest; or more accurately, increasing returns pro-

cesses in economics typically assume a static environment whereas changes in that environment

can attenuate (or amplify) feedback processes. In providing a set of mechanisms that may help to

structure a narrative in terms of transitions between paths, Crouch and Farrell (2004) provide a service

to the analysis of policy dynamics. This can complement the emphasis on inter-policy and inter-

institutional relationships, in particular combinational eﬀects, which imply that policy development

proceeds in a more subtle way than the two-speed view of policy development.

Normative aspects of the term

One of the consequences of constructing the explanatory foundations of path dependency in

public policy by analogy from the economics of technological development is to ‘import’ the

normative result that ineﬃciencies can persist in path dependent processes. This is a powerful

result for neoclassical economics: certain historical factors can ensure that ineﬃciencies occur

and markets do not eliminate these over time. Eﬃciency is understood here as social eﬃciency

namely a situation where both technical and allocative eﬃciency hold. The strength of this nor-

mative result depends on the judgement as to whether the ineﬃciency could have been foreseen at

some point in the initial stages in the path dependency process and corrected; and second,

whether the ineﬃciency remains remediable, namely that a Pareto improvement can be identiﬁed

and is achievable from the current situation.

For some writers the question of whether path dependency implies the persistence of ineﬃcient

institutions is an open and empirical question (e.g. Hay and Wincott 1998); for others path

dependency is more clearly something which inhibits the introduction of ‘better’ or perhaps

more rational policy or organisational form (e.g. Greener 2002). Overall, the normative impli-

cations of path dependency are less pressing for scholars outside the boundaries of neo-classical

economics; it is generally accepted that ine ﬃcient policies or institutions may persist. However,

it is a much stronger claim that policies in a path dependent process are necessarily ineﬃcient, or

alternatively, contained within the concept is the imputation of ineﬃciency. The claim is strong

at a theoretical level as the concept would require signiﬁcant elaboration in terms of both policy

design and the pressures that sustain path dependent and ineﬃcient policy. There is also the pro-

blem of indeterminism: path dependency emphasises that policy paths are unique and arrived at by

a series of small and contingent moves. As such, it is diﬃcult to say that there exists another path

that could have been arrived at which is more eﬃcient and without such a relevant counterfactual

it is diﬃcult to accept the imputation of ineﬃciency.

At an empirical level, the claim that policies are necessarily ineﬃcient is also strong. Devel-

opments in performance measurement in the public sector that might allow arguments that, for

example, health care or education policy are better in one system than another. Despite this, it

is diﬃcult to assert, within a particular political system, that there exist policy options which repre-

sent a welfare improvement over the current policy (net of switching costs and increased

transaction costs) and there is widespread recognition of this by policy actors. Without these

two conditions holding, the normative implications of path dependency in terms of public

Kay

470

policy are attenuated. Nonetheless empirical works on path dependency in policy development

seem willing to impute ineﬃciency to some degree (Pemberton 2003; Kay 2003; Greener

2002; Kemp 2001; Wilsford 1994). This often not so stark as labelling policies in terms of eﬃ-

ciency and certainly involves no quantitative analysis; however, it is not an over-interpretation

of these works to tease out the implicit assumption that a policy would be ‘better ’ without

path-dependent processes acting as a barrier to eﬀective reform.

Conclusion

The rise of the broad school of historical institutionalism has been a salient trend in the social

sciences over the last 20 years and its critical insights have recently been adapted for the more

ﬁne-grained concept of policy by recognising that clusters of governmental decisions, actions,

and norms can – over time – form policy systems, reinforced by feedback mechanisms, which

function as institutions. Historical institutional theory is becoming increasingly reﬁned in its atten-

tion to historical causality and aims to explain diﬀerent types of path dependency beyond the

simple version where the inheritance of accumulated prior commitments limits substantially

current options for policy-makers (Page 2006).

Historical institutionalism has often been judged as over-emphasising positive feedback pro-

cesses and the sensitivity of institutional development to small eﬀects at origin but under-

emphasising the subsequent opportunity for endogenous change in the process of institutional

reproduction over time (e.g. Hay 2002). It is perturbations occurring outside the institutiona-

lised policy system, often characterised as societal or political upheaval or learning, that explain

policy changes; without such exogenous shocks institutions tend to be stable.

In revisiting the concept of path dependency, this chapter has cast some light on a recent

seam of work that provides a particular focus on the capacity of policy-makers, in the

context of strong path dependent policy legacies, to innovate in response to novel demands

from the external environment. However, there are two problems for the analysis of change

within path dependent trajectories: (1) a theoretical one: it is unclear how much or which

elements of a policy system must change to be regarded as a structural break in policy devel-

opment, not just an insigniﬁcant, within path, ﬂuctuation in policy history; and (2) an empirical

one: comparative policy scholars have produced signiﬁcant and important works on policy

reform in response to various pressures, but there are few cases where diﬀerent periods of reform

have been matched neatly to salient structural breaks or punctuated equilibria in path-dependent

trajectories.

Bibliography

Ackrill, R.W. and A. Kay 2006. ‘Historical institutionalism and the EU budget’, Journal of European Public

Policy 13(1): 113–33.

Arthur, W.B. 1994. Increasing Returns and Path Dependence in the Economy. Ann Arbor, MI: University of

Michigan Press.

Arrow, K.J. 2000. ‘Increasing returns: Historiographic issues and path dependence’, The European Journal of

the History of Economic Thought 7(1): 171–80.

Campbell, J.L. 2004. Institutional Change and Globalization. Princeton NJ: Princeton University Press.

Crouch, C. 2005. Capitalist Diversity and Change. New York: Oxford University Press.

—— 2007. ‘How to “do” post-determinist institutional analysis’, Socio-economic Review 5(3): 527–537.

—— and H. Farrell. 2004. ‘Breaking the path of institutional development? Alternatives to the new

determinism’, Rationality and Society 16(1): 5–43.

Greener, I. 2002. ‘Understanding NHS reform: The policy-transfer, social learning and path dependency

perspectives’, Governance 15(2): 161–83.

Policy trajectories and legacies

471

Hacker, J.S. 1998. ‘The historical logic of National Health Insurance: Structure and sequence in the development

of British, Canadian, and US medical policy’, Studies in American Political Development 12: 57–130.

—— 2002. The Divided Welfare State. Cambridge: Cambridge University Press.

Hall, P.A. 1993. ‘Policy paradigms, social learning, and the state’, Comparative Politics 25(3): 275–96.

——and R.C.R. Taylor. 1996. ‘Political science and the three new institutionalisms’, Political Studies 44(4):

936–57.

Hay, C. 2002. Political Analysis: A Critical Introduction. Basingstoke: Palgrave Macmillan.

—— 2004. ‘Theory, stylized heuristic or self-fulﬁlling prophecy? The status of rational choice theory in

public administration’, Public Administration 82(1): 39– 62.

——and D. Wincott. 1998. ‘Structure, agency and historical institutionalism’, Political Studies 46(5): 951–7.

Kay, A. 2003. ‘Path dependency and the CAP’, Journal of European Public Policy 10(3): 405–21.

Kemp, P. 2001. ‘Housing beneﬁt and welfare state retrenchment in Britain’, Journal of Social Policy 29(2):

263–79.

North, D.C. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University

Press.

Nooteboom, B. 1997. ‘Path dependence of knowledge: Implications for the theory of the ﬁrm’, in Lars

Magnusson and Jan Ottosson (eds) Evolutionary Economics and Path Dependence. Cheltenham and Lyme,

NH: Edward Elgar.

Ostrom, E. 1999. ‘Institutional rational choice: An assessment of the institutional analysis and development

framework’,

in Paul Sabatier (ed.) Theories of the

Policy Process. Boulder, CO: Westview, 35–72.

Page, S. 2006. ‘Path dependence’, Quarterly Journal of Political Science 1(1): 87–115.

Pemberton, H. 2003. ‘Lock-in or lock-out? Path dependency and British pensions’, Annual Conference of

the Political Studies Association, 17 April.

Pierson, P. 2000a. ‘Increasing returns, path dependence, and the study of politics’, American Political Science

Review 94(2): 251–67.

—— 2000b. ‘Not just what but when: Timing and sequencing in political processes’, Studies in American

Political Development 14(1): 72–92.

—— 2000c. ‘The limits of design: Explaining institutional origins and change’, Governance 13(4): 474–99.

Raadschelders, J.C.N. 1998. ‘Evolution, institutional analysis and path dependency: An administrative-history

perspective on fashionable approaches and concepts’, International Review of Administrative Sciences 64(4):

565–82.

Rose, R. and P.L. Davies 1994. Inheritance in Public Policy: Change Without Choice in Britain. London: Yale

University Press.

Schwarz, Herman 2004. ‘Down the wrong path: Path dependence, increasing returns, and historical

institutionalism’.Available at: http://www.people.virginia.edu/~hms2f/Path.pdf (accessed August 2011).

Skocpol, T. 1992. Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States.

Cambridge: Belknap Press of Havard University Press.

Stinchcome, A.L. 1968. Constructing Social Theories. London: University of Chicago Press.

Streek, W. and K. Thelen 2005. ‘Introduction: Institutional change in advanced political economies’,in

Wolfgang Streek and Kathleen Thelen (eds) Beyond Continuity. Oxford: Oxford University Press.

Thelen, K. 1999. ‘Historical institutionalism in comparative politics’, Annual Review of Political Science 2:

369–404.

—— 2003. ‘How institutions evolve’,

in James Mahoney

and Dietrich Rueschmeyer (eds) Comparative

Historical Analysis in the Social Sciences. Cambridge: Cambridge University Press, 208–41.

Wilsford, D. 1994. ‘Path dependency, or why history makes it diﬃcult but not impossible to reform health

care services in a big way’, Journal of Public Policy 14: 251–83.

Kay

472

Process sequencing

Carsten Daugbjerg

Introduction

Much policy research is undertaken within a restricted time perspective. In explaining policy deci-

sions, many studies focus on the immediate, assuming that ‘causes and outcomes … are both tem-

porally contiguous and rapidly unfolding ’ (Pierson 2003: 178, see also Bulmer 2009). As Pierson

(2003: 180) argues: ‘There is … no reason to think that most political processes, or the most inter-

esting ones, are necessarily best understood by invoking accounts with this kind of temporal struc-

ture’. A research strategy in public policy analysis focusing on the immediate may at best provide

partial understanding of policy evolution since important dynamics which are rooted in past

events may be missed. Hence, in the recent two decades or so, there has been a growing interest

among policy analysts in policy development over time, in particular among historical institution-

alists who have attempted to establish the reproduction mechanisms maintaining path dependent

policy development (see Kay, this volume). An alternative, but related, approach to explaining policy

evolution over time is process sequencing. It focuses on the temporal connections between

policy events and attempts to establish how previous policy change enables and shapes subsequent

policy changes. While analysis on path dependency focuses on the mechanisms retaining policy

development within a particular path, allowing for bounded policy change, the sequencing approach

concentrates on the way in which policy evolves without assuming that the policy direction

resulting from a series of policy reforms would be predetermined by path dependency. Rather,

it is an analytical approach to policy evolution in which ﬁne-grained temporal analysis of the causal

relations between policy events is the key to explaining policy evolution. The basic assumption in

process sequencing is that an event in a policy sequence is both a reaction to an antecedent event

and a cause of a subsequent one. Thus, policy outcomes feed back and become inputs in the policy

process. In process sequencing, the analytical focus is on the structural constraints and opportunities

embedded in previous policy events and the ability of reform advocates to overcome the constraints

and utilise the opportunities in order to respond to challenges emerging from the broader policy con-

text or from within the policy. Though each policy adjustment is unlikely to be radical, the sequence

of adjustments may over time cumulatively amount to substantial policy change (Rose 1990: 264).

Process sequencing has only recently been suggested as an approach to analyse public policy

evolution (Howlett and Rayner 2006). It is under-theorised and therefore the purpose of this

473

chapter is to outline the approach and its basic concepts and an attempt is made to further develop

the approach by identifying some of the causal mechanisms which link events in a policy sequence.

It is suggested how learning and precedent are potentially important causal mechanisms con-

necting events in a policy sequence. Examples from the sequence of agricultural policy reforms

in the European Union (EU) over two decades are used to highlight the analytical advantages

of applying process sequencing in the study of policy evolution and reform.

Path dependency and process sequencing: two distinct approaches to

temporal analysis

Most recent dynamic analyses on policy change are based on the notion of punctuated equili-

brium which describes policy development as characterised by relatively short moments of

innovative change followed by extended periods of stability (Krasner 1984: 240–4). The notion

of punctuated equilibrium has been widely accepted as a plausible model of policy evolution.

For instance, though not explicitly positioning themselves in the historical institutionalist litera-

ture, but within the agenda-setting literature, Baumgartner and Jones (1993) apply the notion of

punctuated equilibrium in explaining agenda stability over long periods of time and how

punctuations interrupt and subsequently change the policy agenda. Similarly, the advocacy coali-

tion framework developed by Sabatier (1988) and the policy network analysis school (Rhodes

and Marsh 1992) also focus on explaining policy change from a not dissimilar perspective

though the notion of punctuated equilibrium is not explicitly applied.

From a historical institutionalist perspective, radical institutional change is caused by exogenous

shocks, creating a critical juncture which is generally understood as a ‘contingent event … that

was not expected to take place, given certain theoretical understandings of how causal processes

work’ (Mahoney 2000: 513). The concept of path dependency was used to explain the periods

of stasis though it was not until a decade ago that more analytical emphasis was put on identi-

fying the self-reinforcing reproduction mechanisms maintaining a particular path. Clearly, this

improved our understanding of the phenomenon of path dependency (see Thelen 1999; Pier-

son 2000a; Mahoney 2000). This theoretical preoccupation with path dependency within the

historical institutionalist literature left unanswered the question of what explains institutional/

policy change. The notion of path dependency does not enable policy changes which

amounts to more than incrementalism to be explained. The statement by Thelen and Steinmo

(1992: 15) that ‘institutions explain everything until they explain nothing’ still seems to apply.

Therefore, the notion of path dependency has limitations in explaining policy evolution because

policies are adjusted occasionally and such incremental adjustments may amount to substantial

change over time. Further, some adjustments appearing insigniﬁ cant at the time of their adop-

tion may turn out to be an important catalyst for more comprehensive changes at a later stage.

Thus, an important theoretical question in relation to policy change is whether a sequence of

cumulative incremental policy changes can gradually lead to paradigm shift within a policy. There

is a growing literature arguing that this may be possible (Coleman et al. 1997; Cashore and

Howlett 2007; Daugbjerg and Swinbank 2009), in eﬀect questioning the assumption of the

punctuated equilibrium model that radical change occurs in critical junctures over relatively

short time periods.

Indeed, some of the critique of path dependency originates from within the historical institu-

tionalist literature. It has been questioned whether the notion of punctuated equilibrium is a

valid metaphor for institutional and policy development, suggesting that ‘there often seems to be

too much continuity through putative breakpoints in history, but also often too much change

beneath the surface of apparently stable formal institutional arrangements’ (Thelen 2003: 211).

Daugbjerg

474

Among historical institutionalists, this scepticism towards the notion of punctuated equilibrium

as the only model of institutional development has triggered a theoretical discussion on gradual

institutional change which amounts to more than just minor adjustments. Thelen (2003) and

her associates (Streeck and Thelen 2005; Mahoney and Thelen 2009) have engaged in eﬀorts to

develop analytical devices which can help identify various types of gradual institutional transfor-

mation and understand their causes and how the transformation process unfolds. Recent theo-

retical developments suggest that much gradual institutional transformation is driven from within

the institution rather than by forces exogenous to the institution; however, institutional theory

has had little to say on this thus far (Streeck and Thelen 2005: 19; Mahoney and Thelen 2010:

6–7). Though this discussion focuses on institutional development, it is also highly relevant for

public policy evolution. Public policies can be considered institutions since they consist of a set

of rules which facilitate or dictate certain actions and constrain or preclude others and inﬂuence

‘the allocation of economic and political resources, modifying the costs and beneﬁts associated

with alternative political strategies, and consequently altering ensuing political development’

(Pierson 1993: 596).

Reactive sequencing and path dependency

A promising starting point for the development of analytical models enabling us to explain and

understand gradual institutional and policy change is the notion of reactive sequencing. Reactive

sequences ‘are chains of temporally ordered and causally connected events’ (Mahoney 2000:

509) and

in a reactive sequence, each event in the sequence is both a reaction to antecedent

events and a cause of subsequent events … In a reactive sequence, early events trigger

subsequent development … by setting in motion a chain of tightly linked reactions and

counterreactions.

(Mahoney 2000: 526, see also Pierson 2000b: 84)

Earlier events cause the subsequent events, ‘because they trigger a powerful response ’ (Pierson

2000: 85), thus producing an inherent logic in the chain of events. In other words, ‘events in

one period [are used] to explain outcomes in another’ (Haydu 1998: 353). This type of sequencing

has also been dubbed reiterated problem solving (Haydu 1998, 2010) or process sequencing

(Howlett and Rayner 2006). While Haydu’s discussion of reiterated problem solving tends to

be more general and focusing on macro-political developments and the big steps in a historical

trajectory, Howlett and Rayner’s discussion is focused on its relevance for the policy sciences.

Process sequencing focuses on the way in which policy events are causally connected in a policy

sequence. Some of the events may seem insigniﬁcant at the time they happen, but may turn out

to be signiﬁcant events changing direction of policy or even reversing it. Therefore process sequen-

cing requires ﬁne-grained empirical analysis and a detailed understanding of the case under

scrutiny. Since this chapter is focused on public policy issues, I shall refer to process sequencing

in the balance of the chapter.

Mahoney (2000: 509) considers reactive sequencing as a type of path dependency because

‘each step in the chain [of events] is “dependent” on prior steps’. This labelling of reactive

sequencing as a type of path dependency is based on his extraction of three basic deﬁning fea-

tures of path dependent analysis. According to Mahoney (2000: 510–11), the analysis of path

dependency is characterised, ﬁrst, by the ‘study of causal processes that are highly sensitive to

events that take place in the early stages of an overall historical sequence’. Second, he argues

Process sequencing

475

that ‘in a path dependent sequence, early historical events are contingent occurrences that cannot

be explained on the basis of prior events or “initial conditions”’. Third, after the initial step, ‘path

dependent sequences are marked by relatively deterministic causal patterns’.

From a diﬀerent perspective, Haydu (1998, 2010) and Howlett and Rayner (2006) make a

clear distinction between path-dependent sequencing and process (or problem solving) sequen-

cing. While the ﬁrst deﬁning feature of path dependency, as outlined by Mahoney, is unpro-

blematic for these writings, the second and the third are not deﬁning features for process (or

problem solving) sequencing. As to the claim that the initial event in a sequence is contingent,

scholars distinguishing sharply between the two approaches to temporal analysis have argued

that ‘outcomes at a given switch point are themselves products of the past rather than historical

accidents’ (Haydu 1998: 354). In their review of the public policy literature applying the punc-

tuated equilibrium model, Howlett and Rayner (2006: 7) also reach the conclusion that the

outcomes of policy switch points are ‘outgrowths of earlier trajectories’.

The third feature of path dependency, deterministic causal patterns, is also a feature diﬃcult

to recognise in process sequencing and perhaps even in Mahoney’s own portrayal of reactive

sequencing. While there is agreement that events in a temporal sequence are causally connected

and can be traced back to a historical juncture, process sequencing analysts do not recognise this

historical legacy as deterministic for future developments (Howlett and Rayner 2006: 7). In path

dependency, choices made at a critical juncture to a considerable extent lock in the trajectory of

future choices (Pierson 2000a). In process sequencing, the overall direction of the trajectory is

not conceived oﬀ as uni-directional (Howlett and Rayner 2006: 7–8, see also Sewell 1996: 263).

Events are causally linked through reiterated problem solving and this may involve a change of

direction. A particular trajectory is strengthened where the feedbacks from previous choices are

positive. Where feedbacks are negative, the trajectory may change direction but not necessarily

in the dramatic manner envisaged for critical junctures. Thus, as Jervis (1997: 129) argues in rela-

tion to policy evolution: ‘although some … feedbacks are amplifying or dampening, in many

other cases they force the policy in a diﬀerent direction’. The possibility of counter-reactions

caused by negative feedbacks weakens the argument, put forward by Mahoney, that the trajectory

laid down by the initial events in a sequence determines future choices and thus maintain the

path. A trajectory can only be retained in a particular direction when there are self-reinforcing

reproduction mechanisms at work in a sequence. Such mechanisms are essential to the path

dependency argument, but not to process sequencing. Though reactive sequencing, as outlined

by Mahoney, ‘involves reaction and counterreaction mechanisms that give an event chain an

“inherent logic” in which one event “naturally” leads to another event’ (Mahoney 2000: 511),

the resultant sequence is not path dependent because there are no reproduction mechanisms to

sustain the sequence on a particular path.

Process sequencing in public policy

Process sequencing is particularly useful in public policy analysis (Howlett and Rayner 2006:

13–14). To understand policy development, the approach emphasises ﬁne-grained temporal

analysis of reactions and counter-reactions in a process of reiterated problem solving. The downside

of process sequencing is its limited potential to generate theoretical statements on the way in

which events in sequences are connected (Pierson 2000b: 84). In this section, the under-theorised

nature of this approach to temporal analysis of policy is addressed and it is suggested how

learning and precedent can become causal mechanisms connecting events in a policy sequence.

Connecting events causally in a sequence is an important challenge facing process sequencing

analysis. In sequential analysis

Daugbjerg

476

explanations should respect historical time by casting causal analysis in the form of sequenced

events, with earlier happenings leading to and accounting for later ones. Explanations of

this kind should, moreover, carefully specify the mechanisms through which causal inﬂuence

is conveyed through time.

(Haydu 1998: 354)

The theoretical challenge is to consider what connects policy events without violating the ability

of the analytical approach to identify case-speciﬁc empirical details which are important for

understanding the sequence of events under scrutiny.

Policy problems and policy actors

In process sequencing, events are connected through a reiterated process of problem solving

unfolding over longer periods of time (Haydu 1998: 349). Since most public policy formulation

and implementation take place in sub-systems in which a limited number of actors are continuously

and actively involved in decision-making and often in many day-to-day administrative decisions

(e.g. Richardson and Jordan 1979; Rhodes and Marsh 1992; Baumgartner and Jones 1993;

Sabatier 1988) it is analytically useful to distinguish between endogenously or exogenously problem

dynamics. Problems are endogenously generated when they originate from within the subsystem.

These may be unintended policy eﬀects which threaten the core functions and thus viability of

the policy. They need not necessarily aﬀect political actors outside the subsystem but may often

do so. An example of endogenously generated policy eﬀects is recurring budgetary diﬃculties caused

by the cost of implementing certain policy instruments. Exogenously generated problems emerge

as a result of evolving broader policy contexts beyond the control of the policy subsystem. Devel-

opments in the broader policy context may challenge the policy and thus put pressure on the

members or the subsystem to respond with policy changes. Exogenous problems can, for instance,

be caused by the growing level of economic, cultural and political globalisation. An instance of the

latter is the inclusion of agricultural trade on the negotiating agenda of Uruguay Round of the

General Agreement on Tariﬀs and Trade (GATT) in the mid-1980s. As it became evident in 1990

that agricultural support and protection could not be sidelined in the Round, as had happened in

earlier rounds, the trade impacts of the Common Agricultural Policy (CAP) became a new key

policy problem which EU agricultural policy-makers had to address (Daugbjerg and Swinbank 2009).

An important theoretical challenge is to establish the conditions under which individual events in a

policy sequence become both an outcome of previous policy decisions and developments and a

cause of subsequent policy changes. Mahoney (2000: 511) refers to an ‘inherent logic’ and a ‘natural’

connection between events and thus puts forward a somewhat deterministic view on the driving

forces behind a sequence. He tends to emphasise structure and devotes limited attention to the role

of actors in connecting events. To establish the causal link between events in a policy sequence

more attention must be paid to the way in which policy actors act within a policy structure and how

they change it. Policy sequences ‘are continually shaped and reshaped by the creativity and stub-

bornness of their human creators’ (Sewell 1996: 272; see also Haydu 1998: 357, 2010: 32–3, 39–40).

Therefore, on the one hand, theory on sequencing must leave room for agency when theorising

about the causal connection between individual events in a sequence. On the other hand, since

public policies are institutions, they constrain and facilitate certain actions over others. To have some

independent impact on a policy sequence, actors must have capacity to reﬂect over the opportunities

and constraints embodied in previous policy events when responding to policy problems.

Hay’s (1995, 2002) concept of strategic learning spells out the potentials and limitations in

actors’ ability to reﬂect over the policy processes in which they are involved. While lacking

Process sequencing

477

complete information, political actors have some capacity to learn. As Hay (2002: 210) points

out: ‘given that actors are reﬂective, routinely monitoring the consequences of their action, we

might expect their knowledge of the context to evolve over time’. However, since the broader

and policy-speciﬁc context is complex and evolving, they are unlikely to reach a state of com-

plete knowledge. Through their actions and the consequences of these, policy-makers learn,

‘enhancing awareness of structures and the constraints/opportunities they impose, providing the

basis from which subsequent strategy might be formulated and perhaps prove more successful’

(1995: 201, see also 1998: 43). Learning is based on selective knowledge on the context within

which strategic action is conducted. Since agents have incomplete information on the constraints

and opportunities embodied in the settings within which they operate, they use ‘more or less

informed projections regarding the strategic motivations, intentions and likely actions of other

signiﬁcant players’ (ibid.: 44). However, since the context is ‘constantly evolving through the

consequences (both intended and unintended) of strategic action’ (ibid.: 43), cumulative learn-

ing may not necessarily result. There is no guarantee that actors draw the ‘right’ lessons from

past experience; indeed, they may not even have the insight to judge which lessons are right

and which ones are wrong (2002: 211). Thus, it is not the context as such which is the source

of learning but actors’ perceptions of the intended and unintended consequences of previous

policy decisions which connect events in a sequence.

Policy feedback: learning and precedent

The feedback from previous policy decisions is the key to establish how the actions of reﬂective

policy actors create a link between antecedent and subsequent events in a policy sequence.

Public policies produce various types of outcomes which feed back into the policy process and

thus they become inputs into the policy process (Pierson 1993). Feedbacks aﬀect the perceptions

and actions of government oﬃcials and various types of stakeholders continuously and actively

involved in policy formulation and implementation. These reﬂect over the feedback from policy

implementation when attempting to understand whether the policy works as intended or

whether it produces undesired unintended consequences (Lindblom 1959: 86).

As Heclo (1974: 305) pointed almost four decades ago, ‘Governments not only “power”…

they also puzzle’. What he meant was that policy development was not only driven by powering,

but also by attempts to learn from experience and adopt the best available solutions to policy

problems. However, the two processes of powering and puzzling cannot easily be separated. Policy

actors act within a structure which constrains certain actions and facilitates others. This structure

may also facilitate and constrain which lessons are learned from policy experience and which are

ignored. Since policy, as an institution, also reﬂects the outcome of past power struggles, learning

may be constrained by powering (Roederer-Rynning and Daugbjerg 2010: 316–17).

Policy-makers have to respond to policy problems emerging from an evolving broader policy

context and to policy problems generated from within the policy itself. Though some policy

problems can be ignored, at least for some time, others require action to maintain output legiti-

macy; that is, demonstrating to a broader universe of policy participants that policy can eﬀec-

tively address the problems with which it is confronted. The extent to which policy-makers are

able to respond to the policy problems depends on their ability to reﬂect continuously over the

opportunities and constraints embodied in the policy structure and develop strategies to address

the challenges.

Each event in a policy sequence embodies a number of constraints and opportunities which

policy actors can utilise to prevent or to facilitate further evolution of the policy; while previous

events constrain certain policy responses, they provide opportunities for others. Prior events

Daugbjerg

478

‘foster and shape the crises that prepare the ground for new solutions, and they inﬂuence the choices

made by actors in response to those crises’ (Haydu 2010: 36). Feedback process may reveal these

constraints and opportunities to reﬂective and skilful reform advocates who want to change

policy in particular directions and to those who desire the status quo and at the most willing to

accept limited adjustments to address the most immediate policy problems. These perceived oppor-

tunities and constraints inﬂuence the power balance within the policy process by empowering

some actors with particular interests and weaken others who hold diﬀering views on the desired

direction of policy evolution regardless of whether or not these views are motivated by power

concerns or are outcomes of genuine learning processes.

As argued above, feedback involves an element of learning albeit constrained by the policy

structure. In a larger policy complex, small-scale experiments with new policy solutions may be

undertaken for more limited parts of the policy. Positive experiences with these can produce

positive learning eﬀects modifying entrenched perceptions of certain types of policy instruments

among some policy-makers and thus create opportunities for extended and more intensive use

of these. For instance, the 1992 reform of the CAP shows how positive experiences with new

policy instruments can change preferences. The reform lowered the guaranteed minimum prices

by a third and introduced annual direct compensatory payments. This switched a substantial share

of farm support from indirect price support, in which subsidies to farmers were paid through

artiﬁcially high consumer prices, to direct payments. The farm unions were strongly opposed to this

change because it would make farm support look like social beneﬁts (Daugbjerg 2009). However,

‘the compensatory payments proved much more popular than anticipated in the farm commu-

nity’ (Moyer and Josling 2002: 194). Farmers learned that direct payments increased their income

security and safeguarded them against drastic income losses from, for example, poor harvests

(ibid.). Since the early 1990s, direct payments to farmers have been unquestioned and an integral

and growing component of EU farm policy.

On the contrary, negative policy experiences may produce counter-reactions precluding the

increased use of such instruments and perhaps even policy reversal. However, such learning eﬀects

may not materialise in policy change until the power balance shifts in favour of those interests

desiring an eﬀective response to the policy problems. For instance, surplus production emerged

as a major problem in the EUs agricultural policy in the late 1970s and early 1980s. In the dairy

sector, surplus production was eﬀectively dealt with by introducing dairy quotas in 1984. In the

arable sectors, the problem was not eﬀectively addressed until 1992 when compulsory set-aside

of arable farm land was introduced (Daugbjerg 2012). The 1992 reform of the main arable policy

sectors demonstrates that negative feedback can change the power balance within a policy sub-

system. Subsidised exports of EU surplus production had triggered the anger of the United

States and other cereals exporting countries and therefore the EU was unable to avoid farm trade

became an integral part of the GATT negotiating agenda (Daugbjerg and Swinbank 2009). This

led to a shift in the power balance between the EU Commission and the Council of Agri-

culture Minsters in favour of former and eventually enabled substantial reform of the CAP

(Coleman and Tangermann 1999).

Feedbacks may reveal policy inconsistencies which are counterproductive in relation to the

objectives to be achieved within the policy ﬁeld. Speciﬁc measures may pull policy in diﬀerent

directions and in some instances even neutralise the eﬀects of each other. Such eﬀects may be

ignored, particularly in situations in which contradictory measures have been deliberately

introduced to create or maintain a delicate balance between diﬀerent groups of stakeholder with

opposing interests. However, when these inconsistent measures start generating eﬀects that

threaten the core functions of the policy, and particularly when this amounts to a crisis, policy

change becomes inevitable.

Process sequencing

479

The most likely policy responses to learning feedbacks are those which are seen as repre-

senting the logical continuation of the direction set by the previous event in the policy sequence.

In situations in which the distance between the existing policy design and the policy changes

which from a rational perspective would be required to solve the problems is too large to be

undertaken in one step for political reasons, policy change will tend to be gradual. In some, but

more rare, situations the change of policy direction may be more abrupt and substantial. This

tends to happen when contextual changes provoke a policy crisis and immediate response is needed.

But even in these situations the policy response may grow out from the previous events in the

policy sequence (Howlett and Rayner 2006: 7). Learning feedbacks from previous events in the policy

sequence are most likely to impact on policy responses in situations in which the level of poli-

ticisation is low. As the level of politicisation increases, the less likely it is that learning will

inﬂuence policy responses. Rather, policy-makers will exercise power to attempt moving policy

in a particular direction when responding to policy problems and as a result learning processes

might be sidelined (Pierson 1993: 617–18).

The causal connection to previous policy events in such politicised situations may be a new

measure introduced in a previous reform to address a particular problem. This can be utilised to

legitimise a full-scale use of the new measures within the policy even when it was originally aimed

at a more marginal problem or was limited to solve a temporary problem. Policy-makers use

precedent to legitimise policy change by claiming that the ‘new’ measures have indeed already

been applied within the policy and that its extended use is a logical continuation of the direction of

policy evolution previously set out. Reform advocates can, thus, construct precedent to support

their interests in moving policy in a particular direction. Precedent can be a quite powerful political

tool. Even though a new measure may be insigniﬁcant when introduced, or appear so, it may

sow the seeds for a sequence of policy adjustment amounting to substantial change over time.

For instance, the introduction of the so-called Small Farmers Scheme in 2001 to simplify

administrative procedures in EU agricultural policy demonstrates that a policy change aimed at

addressing a relatively limited problem can form a powerful precedent for more comprehensive

reform later on. The scheme oﬀered farmers receiving less than €1,250 in direct aid payments

annually to transform these payments into a single ﬂat rate payment based on a historical

reference period. The introduction of this scheme seemed insigniﬁcant at the time. However, it

introduced a new form of direct aid payments to farmers. The direct payments introduced in

1992 required farmers to produce certain crops of keep particular types of livestock. The pay-

ments of the Small Farmers Scheme were not linked to production requirement and thus they

were what is known in agricultural policy terminology as decoupled payments. They eventually

spread to most of the commodity support schemes under the CAP in the 2003 policy reform.

Simpliﬁcation served as the precedent legitimising the proposal to apply the decoupled farm

payment model, introduced in the Small Farmers Scheme, as the general support model in the

CAP. As stated by the Commission: ‘The Small Farmers Scheme represents an important pre-

cedent for reducing the administrative burden’ (author’s italics, Commission of the European

Communities 2002: 10).

Conclusion

In this chapter, process sequencing has been outlined and it has been argued that it is a distinct

analytical approach to studying policy evolution. Though it shares a temporal approach to the

study of policy-making with the path dependency approach, it is not based on the assumption

that public policy evolution is predetermined by a particular path sustained by a set of self-

reinforcing reproduction mechanisms. Rather, events in a policy sequence are connected by the

Daugbjerg

480

attempts of policy actors to solve policy problems. As Haydu (2010: 36) points out: ‘Thinking

in terms of reiterated problem solving thus corrects the tendency of path dependency to make

historical trajectories overdetermined, with outcomes increasingly locked in over time’. Policy

events in a temporal sequence are casually linked by the feedbacks from previous events. These

feedbacks disclose the constraints and opportunities for further reform to policy-makers. Though

constrained by the policy structure, these actors utilise learning eﬀects and precedent to generate

support for policy changes. Thus, learning eﬀects and precedent can provide causal link between

antecedent and subsequent policy events in a sequence.

Process sequencing is theoretically underdeveloped in terms of specifying the factors linking

antecedent and subsequent events in a policy sequence. In this chapter, some initial theoretical

considerations have been presented, focusing on the role of learning and precedent as casual links

between individual events in a policy sequence. It has also been emphasised that policy sequences

are the outcomes of human action and therefore analysis of sequencing must focus on the

interaction between structure and agency as policy-makers respond to policy problems.

Future research should concentrate on identifying other factors which potentially provide the

causal connection between antecedent and subsequent policy event, specify how they privilege

certain policy decisions and constrain others and explain how policy actors identify and conceive

of the opportunities and constraints embodied in previous policy events. However, it is important

not to over-theorise process sequencing. There is a risk that too much reliance on theory in iden-

tifying the causal mechanisms linking events in a sequence tempts analysts to downgrade ﬁne-

grained empirical analysis of the way in which events are actually connected. As Bearman et al.

(1999, 508) caution: ‘Theory involves denying data. Thin narrative accounts are the product of

speciﬁc theories that direct the [analyst] to identify some events as salient and to deny others as

not salient’. Thus, the further development of process sequencing as an analytical framework to

studying policy evolution must balance between theory as an analytical tool to guide the search for

the causal mechanisms on the one hand and an open-minded empirical approach to how events

are casually connected in the speciﬁc case study on the other hand. Another but related chal-

lenge, not addressed in this chapter, is methodological. The challenge is to apply research methods

which enable the scholar to establish the causal connection between policy events and avoid

substituting causal analysis for accounts of a series of events which are temporally ordered but not

explained. Process tracing is the most obvious research method to apply in process sequencing.

The key focus of this research method is to develop tools which can help the scholar to establish

whether events in a sequence are causally connected (Beach and Pedersen 2012).

Notes

1 I would like to thank Darren Halpin, Adrian Kay and Rasmus Brun Pedersen for helpful comments on

earlier versions.

2 See Pierson (1993: 614–19) and Bennett and Howlett (1992) for a discussion of the diﬃculties of

applying the learning concept in empirical research.

Bibliography

Baumgartner, Frank R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics. Chicago:

University of Chicago Press, 1–102.

Beach, Derek and Rasmus Brun Pedersen. 2012. Process Tracing Methods: Guidelines and Foundations, Ann

Arbor: University of Michigan Press.

Bearman, Peter, Robert Faris and James Moody. 1999. ‘Blocking the future: New solutions for old problems

in historical science’, Social Science History 23(4): 501–33.

Process sequencing

481

Bennett, Colin J. and Michael Howlett. 1992. ‘The lessons of learning: Reconciling theories of policy

learning and policy change’, Policy Science 25(3): 275–92.

Bulmer, Simon. 2009. ‘Politics in Time meets the politics of time: Historical institutionalism and EU timescape’,

Journal of European Public Policy 16(2): 307–24.

Cashore, B. and M. Howlett. 2007. ‘Punctuating which equilibrium? Understanding thermostatic policy

dynamics in Paciﬁc Northwest forestry’, American Journal of Political Science 51(3): 532–51.

Coleman, W.D. and S. Tangermann. 1999. ‘The 1992 CAP reform, the Uruguay round and the commission’,

Journal of Common Market Studies: 37(3): 385–405.

——, G.D. Skogstad and M.M. Atkinson. 1997. ‘Paradigm shifts and policy networks: Cumulative change

in agriculture’, Journal of Public Policy 16(3): 273 – 301.

Commission of the European Communities. 2002. Communication from the Commission to the Council and the

European Parliament. Mid-Term Review of the Common Agricultural Policy, COM(2002)394, Brussels: CEC.

Daugbjerg, Carsten. 2009. ‘Sequencing in public policy: The evolution of the CAP over a decade’, Journal

of European Public Policy 16(3): 395–411.

—— 2012. ‘Globalization and internal policy dynamics in the reform of the Common Agricultural

Policy’, in Jeremy Richardson (ed.) Constructing a Policy-making State? Policy Dynamics in the EU, Oxford:

Oxford University Press.

——and Alan Swinbank. 2009. Ideas, Institutions and Trade: The WTO and the Curious Role of EU Farm

Policy in Trade Liberalization. Oxford: Oxford University Press.

Hay, Colin. 1995. ‘Structure and agency’, in D. Marsh and G. Stoker (eds) and Methods of Political Science.

Basingstoke: Palgrave Macmillan.

—— 1998. ‘The tangled webs we weave: The discourse, strategy and practice of networking’, in David

Marsh (ed.) Comparing Policy Networks. Buckingham: Open University Press, 33–51.

—— 2002. Political Analysis: A Critical Introduction.

Basingstoke: Palgrave Macmillan.

Haydu,

Jeﬀrey. 1998. ‘Making use of the past: Time periods as cases to compare and as sequences of

problem solving, American Journal of Sociology 1(2): 339–71.

—— 2010. ‘Reversal of fortune: path dependency, problem solving, and temporal analysis’, Theory and

Society 39(1): 39–48.

Heclo, Hugh. 1974. Modern Social Policy in Britain and Sweden. New Haven, CT: Yale University Press.

Howlett, Michael and Jeremy Rayner. 2006. ‘Understanding the historical turn in the policy sciences. A

critique of stochastic, narrative, path dependency and process-sequencing models of policy-making over

time’, Policy Sciences 39(1): 1–18.

Jervis, R. 1997. System Eﬀects: Complexity in Political and Social Life. Princeton, NJ: Princeton University Press.

Krasner, S.D. 1984. ‘Approaches to the state: Alternative conceptions and historical dynamics’, Comparative

Politics 16(2): 223–46.

Lindblom, Charles E. 1959. ‘The science of “muddling through”’, Public Administration Review 19(2): 79–88.

Mahoney, J. 2000. ‘Path dependence in historical sociology ’, Theory and Society 29(4): 507–48.

——and Kathleen Thelen. 2010. ‘A theory of gradual institutional changed’, in James Mahoney and

Kathleen Thelen (eds) Explaining Institutional Change: Ambiguity, Agency, and Power. New York: Cambridge

University Press.

Moyer, H.W. and T.E. Josling. 2002. Agricultural Policy Reform: Politics and Policy Process in the EC and the

US in the 1990s. Aldershot: Ashgate.

Pierson, P. 1993. ‘When eﬀect becomes cause: Policy feedback and political change’, World Politics 45(4):

595–628.

—— 2000a. ‘Increasing returns, path dependence and the study of politics’, American

Political Science Review

94(2): 251–67.

—— 2000b. ‘Not just

what, but when: Timing and sequence in political processes’, Studies in American

Political Development 14 (spring): 72–92.

—— 2003. ‘Big, slow moving, and … invisible: Macrosocial processes in the study of comparative poli-

tics’, in J. Mahoney and D. Rueschemeyer (eds) Comparative Historical Analysis in the Social Sciences.

Cambridge: Cambridge University Press, 177–207.

Rhodes, R.A.W and David Marsh. 1992. ‘New directions in the study of policy networks’, European

Journal of Political Research 21(1–2): 181–202.

Richardson, J.J and A.G. Jordan. 1979. Governing under Pressure: The Policy Process in a Post-Parliamentary

Democracy. Oxford: Robertson.

Roederer-Rynning, Christilla and Carsten Daugbjerg. 2010. ‘Power, learning or path dependency?

Investigating the roots of the European Food Safety Authority’, Public Administration 88(2): 315–30.

Daugbjerg

482

Rose, R. 1990. ‘Inheritance before choice in public policy’, Journal of Theoretical Politics 2(3): 263–91.

Sabatier, Paul. 1988. ‘An advocacy coalition framework of policy change and the role of policy-oriented

learning therein’, Policy Sciences 21(2): 129–68.

Sewell, William H. 1996. ‘Three temporalities: Toward an eventful sociology’, in Terence J. McDonald

(ed.) The Historic Turn in the Human Sciences. Ann Arbor, MI: University of Michigan Press, 245–80.

Streeck, Wolfgang and Kathleen Thelen. 2005. ‘Introduction: Institutional change in advanced political

economies’, in Wolfgang Streeck and Kathleen Thelen (eds) Beyond Continuity: Institutional Change in

Advanced Political Economies. Oxford: Oxford University Press, 1–39.

Thelen, Kathleen. 1999. ‘Historical institutionalism in comparative politics’, Annual Review of Political Science

2: 369–404.

—— 2003. ‘How institutionalism evolves: Insights from comparative historical analysis’, in J. Mahoney and

D. Rueschemeyer (eds) Comparative Historical Analysis in the Social Sciences, Cambridge: Cambridge

University Press, 208–40.

—— and S. Steinmo. 1992. ‘Historical institutionalism in comparative politics’, in S. Steinmo, K. Thelen

and F. Longstreth (eds) Structuring Politics: Historical Institutionalism in Comparative Analysis. Cambridge:

Cambridge University Press, 1–32.

Process sequencing

483

Learning from success and failure?

Allan McConnell

Introduction

A common sense view of the political world might expect policy-makers to seek success and avoid

failure. Such assumptions are echoed in the public language of policy and political elites, from pre-

sidents and prime ministers to government ministers and senior oﬃcials. Stated success goals include

building on what ‘already works’, as well as learning from mistakes when policies fail or disaster strikes.

To what extent do such assumptions and ‘oﬃcial’ rhetoric accord with the realities of public

policy? Perhaps surprisingly, the phenomenon of policy success has been given scant attention in

the public policy literature (for exceptions, see Kerr 1976; Marsh and McConnell 2010; McConnell

2010a, 2010b). More has been written on aspects of failure, including policy ﬁascoes, crises and

disasters, although the literature tends to concentrate on the management of acute episodes, rather

than on long-term repercussions (see Hogwood and Peters 1985; Dunleavy 1995; Bovens and

’t Hart 1996; Bovens et al. 2001; Birkland 2007; Boin et al. 2005). The purpose of this chapter is to

build on the insights and implications of this relatively small group of literature (particularly that

of McConnell 2010a, 2010b) in order to oﬀer a more nuanced analysis of the simple assumption

that policy-makers seek to cultivate success and avoid failure.

The chapter is structured as follows. First, it examines the innumerable methodological dif-

ﬁculties in conceiving of success and failure. Second, it deals with the issue of ‘success as fact’ vs.

‘success as interpretation’, and produces a working deﬁnition which allows us to conceive of success

(and failure) as occurring in three realms: processes, programmes and politics. Third, it argues

that by understanding the malleable, contested and multidimensional aspects of success, we are

better placed to understand why the common sense logic of learning from success and failure has

some validity, but not nearly as much as we might assume. Policymakers can certainly build on

existing successes and learn from failures, but they may also tolerate, mask, risk, exploit and

even cultivate failure.

The methodological difﬁculties of ascertaining ‘success’ and ‘failure’

Broadly speaking, public policy is whatever governments choose to do or choose not to do. We

know from the literature on policy instruments (see e.g. Hood and Margetts 2007; Howlett

484

2010) that governments have many tools at their disposal, from laissez-faire to active provision

of public services. They can refuse to intervene in industrial disputes, initiate anti-drink-driving

campaigns, set regulatory standards for food hygiene, impose new taxes, build public hospitals,

establish customs and immigration controls and send troops to foreign lands. What constitutes

‘success’ in making such policy choices? The issue is predictably contested and methodologically

diﬃcult, as it is with many phenomena within the political and social sciences.

Diﬀering perceptions: put simply, is success a matter of fact or a matter of interpretation? Broadly

speaking, there are two diﬀerent tendencies – often implicit in the literature and reﬂecting

broader trends within the policy sciences.

One tendency is interpretive or discursive, leaning heavily towards the assumption that success

is in the ‘eye of beholder’ (see e.g. Bovens and ’t Hart 1996; Fischer 1995, 2003). It is easy to see

why such a position has some analytical validity. For example, in relation to the coalition inter-

ventions in Iraq or Afghanistan, it is likely that no amount of argument or evidence would per-

suade anti-war campaigners that the incursions were successful in bringing peace and stability to

authoritarian regimes. It is also highly unlikely that animal rights activists would be persuaded that

legislation permitting drug testing on animals was successful because of the beneﬁts it provided in

developing vaccines, drugs and treatments. The corollary of this tradition is that ‘success’ and ‘failure’

are little more than labels used by governments, commentators, the media and others, depending

on their values, acceptance of government goals and the means of achieving them.

A counter-tendency is in the rational-scientiﬁc tradition, which leans heavily towards the

assumption that policy outcomes are objective facts. Therefore, armed with the knowledge of

government goals, we simply gather evidence which can be matched against these goals in order

to ascertain an outcome of ‘success’ or ‘failure’ (see e.g. Gupta 2001; Davidson 2005). Again, it

is easy to see some analytical worth in this tradition, because government can achieve broadly

what it sets out to achieve, regardless of whether the goals and the outcomes are supported. No

amount of diﬀering perceptions, for example, can refute the fact that a public school was built,

or that income taxes were levied and collected, or that NASA launched space shuttles into

space, or that Saddam Hussein was deposed as president of Iraq. One implication of the rational-

scientiﬁc tradition, therefore, is that, success and failure are matters of objectivity, with little or

no room for perceptions of the merits or otherwise of a policy.

Overall, we are confronted with the real diﬃculty of how to conceive of success, when two

such apparently diametrically opposed positions are evident. We will return to this issue shortly.

Multiple benchmarks: no deﬁnitive standards exist, by which we can judge success or failure.

An examination of discourse surrounding policy success reveals a number of potential success

measures. They are not mutually exclusive and they may complement or conﬂict, but they indicate

substantial diversity.

 Implemented as intended e.g. public servants were eﬃcient and eﬀective in carrying out

instructions.

 Original objectives were met e.g. target of 80 per cent of local property taxpayers, paying

their bills online.

 Beneﬁt was achieved for the target group(s) e.g. welfare payments to farmers.

 An improvement on a prior state of aﬀairs e.g. unemployment has fallen by 2 per cent.

 Criteria have been met which have wide value in that policy domain e.g. eﬃciency in budgeting,

adherence to risk management standards in continuity planning, secrecy in intelligence.

 The policy is supported by key interests e.g. citizens, key stakeholders, media.

 The

policy is achieving more when compared to another jurisdiction e.g. one country doing

better than another

in dealing with the global ﬁnancial crisis.

Learning from success and failure

485

 Beneﬁts outweigh the costs e.g. beneﬁts of waging war on another regime, outweigh the human

and ﬁnancial costs.

 The policy is innovative e.g. using new technology to tackle online fraud.

 The policy complies with moral, ethical or legal standards e.g. adherence to principles of

fairness; operates within the parameters of existing laws.

It seems, therefore, that success standards are malleable, to the point for example that a government

and its supporters, can claim policy success, while its opponents can claim failure. A June 2010

Gallup poll on President Obama’s management of the BP Gulf of Mexico oil spill, revealed a

56 per cent approval rating from Democrats and Democrat–leaners, but only 11 per cent

approval from Republicans and Republican-leaners.

Dealing with shortfalls, ambiguities and con ﬂicts: success isn’t ‘all or nothing’ based on the achieve-

ment (or not) of a singular goal. Policy outcomes typically fall short of aspirations to varying

degrees, If a government seeks to eliminate a 30 billion dollar budget de ﬁcit, many outcomes

are possible, such as reductions to 28 billion, 20 billion, 10 billion and 5 billion. Yet there is no

scientiﬁc line that can be drawn to demarcate success from failure. In all likelihood, government

and its supporters will emphasise the importance of what has been achieved, while opponents

and critics will highlight the shortfalls.

Furthermore, it is by no means certain that outcomes will be clear and unambiguous. Evi-

dence may be lacking, stated goals may mask hidden agendas and success may produce unin-

tended consequences. Indeed, one of the most common diﬃculties relates to the existence of

multiple and conﬂicting goals where some are achieved and some not (indeed the achievement

of one may be at the expense of another). For example, police resources focused heavily on

crime investigation, may be at the expense of crime prevention. Similarly, a drive by schools to

increase educational achievement, may conﬂict with a goal of increasing the number of pupils/

students from deprived backgrounds. In reality, therefore, policy outcomes are often messy,

requiring judgement on the signiﬁcance given to what has been achieved, weighed up against

the failures and the grey areas in between.

Success for whom? Public policy, by its very nature of intervening (or not) to persuade, ﬁnancially

reward/penalise, regulate and provide, typically beneﬁts some groups or interests by comparison

with others, albeit that the broader rationale is articulated as being in the public interest. So, for

example, the rights of smokers to smoke cigarettes, can be restricted in public places on the basis that

doing so limits passive smoking and improves public health. Similarly, income tax payers who have

no children, would be required to pay an increase in income tax that was levied speciﬁcally for the

purpose of generating additional resources for schools. It is hard to escape the fact that a policy may

be successful for those to whom it brings rights and rewards, but may not be conceived as such by

those whose freedoms or resources are curtailed/reduced in order to achieve the broader goal.

Variance over time: it may be tempting to view policies as succeeding or failing, as though a

snapshot is able to capture policy outcomes at an appropriate point in time. Yet, to continue the

analogy, snapshots at diﬀerent points in time might lead to diﬀerent assessments. Apparent short-

term failure may produce long-term success, such as the ‘great planning disaster’ of the Sydney

Opera house (see Hall 1982) and its transformation into an iconic, revenue-raising tourist attraction.

The opposite may be the case, where short-term success fades away to be replaced by condi-

tions more likely to be considered as failure. The 2008–9 global ﬁnancial crisis is a case in point,

where light touch regulation and faith in the eﬃciency and self-corrective capacities of markets,

has been viewed widely in hindsight as producing unsustainable levels of proﬁtability. Analysing

policy success needs to contend with the existence of multiple and perhaps contradictory outcomes

over time.

McConnell

486

Isolating the policy eﬀect: political science is home to multiple methods and approaches (Hay 2002)

and policy analysis is no diﬀerent, including widely diﬀering assumptions about how to evaluate

outcomes (for contrasting positions see e.g. Fischer 1995; Davidson 2005). The very fact that

multiple and conﬂicting assumptions exist, is a clue to the diﬃculties of explaining and assessing

policies and their outcomes. If, for example, we wanted to assess the success of a government’s

anti-drugs advertisements, we would confront the fact that changes in drugs habits can be the

product of many factors such as availability of supply, family pressures, peer inﬂuences, mass media

stories and individual psychologies. We would either need to operationalise all the main variables

and engage in rigorous statistical examination, or perhaps make a series of informed judgements,

based on imperfect and perhaps conﬂicting evidence. In the policy sciences, it would be very

diﬃcult to reach a universal consensus on our ability to isolate the outcome of a policy.

There are no magical answers to these questions, or solutions to any of the other methodo-

logical diﬃculties. However, I would argue that rather than being deterred by such matters,

there is an opportunity to gain insight from them.

Moving forwards: the nature of policy success and policy failure

For heuristic purposes, we need to confront the issue of whether success is a ‘fact’ or an ‘inter-

pretation’. I would argue that to side purely with one or the other is to be analytically blind to

important aspects of policy phenomena. A rational-scientiﬁc approach can never escape the exis-

tence of multiple interpretations and diﬀering levels of support in plural societies. Equally, an inter-

pretive approach can never escape the reality that governments can do broadly what they set out to

do, regardless of whether the outcomes are perceived as successful. We need a conception of

success that recognises and embraces both. Accordingly, I propose the following deﬁnition:

A policy is successful if it achieves the goals that proponents set out to achieve and attracts

no criticism of any signiﬁcance and/or support is virtually universal.

(McConnell 2010b: 351)

The key advantage of this deﬁnition is that it manages to combine more objective-oriented

achievements, with perceptions of whether such achievements deserve the label of ‘success’,

depending on popular and stakeholder support (explicit or implicit) for the means and ends. An

example of success is the system of dams and dykes in the Netherlands, aimed at preventing about a

quarter of the country’s land mass being submerged by water, and universally supported among the

populace. Many matters of low politics on non-controversial issues may also fall into this category,

where government goals are achieved and there is no objection or criticism of any signiﬁcance.

It is also possible to conceive of a spectrum of outcomes (see McConnell 2010a, 2010b)

moving from durable or resilient success (where shortfalls in goal achievement are second best

but tolerable and sustainable), to conﬂicted success (where achievements and support are ﬁnely

balanced with failures and criticism), through to precarious successes (where failures and criticism

outweigh some achievements and support) and ﬁnally to outright policy failure.

A policy fails if it does not achieve the goals that proponents set out to achieve, and

opposition is great and/or support is virtually non-existent.

(McConnell 2010: 357)

Failures may generate high levels of media interest (such as the poll tax in the UK) or they may

be contained within the realms of small policy subsystems, away from the glare of the media.

Learning from success and failure

487

Three realms of policy success

If we want to understand the dynamics of success and failure, a crucial step now needs to be taken in

developing our understanding. What ‘governments do’ can usefully be split into three realms.

Table 36.1 provides ﬁne detail and hence the discussion here is conﬁned to making the general

point that governments can succeed and/or fail in each of these realms.

Process: among other things, governments deﬁne problems, appraise options, design instruments,

engage with stakeholders and take decisions. Processes of policy formation involve strategies which,

if they succeed, allow governments to preserve their goals and policy instruments, produce

policy in a legitimate fashion, build sustainable coalitions, appear innovative and attract little or

no criticism for the process of policy formation. For example, a range of processes from adherence

to constitutional rules or norms, to utilising good practice in stakeholder engagement can form the

basis of ‘successful’ and legitimate, coalition-building processes. If policy is about what governments

do, they ‘do process’ and they may succeed and/or fail in this regard

Programmes: As we know, governments produce programmes, involving varying mixes of policy

instruments. Regardless of the process of policy formation, we may judge programmes separately,

to the extent of matching them against a number of possible indicators. They may be imple-

mented in line with objectives, achieve desired outcomes (such as the switchover from analogue

to digital television in Luxembourg, Finland and Sweden) create beneﬁt for the target group(s),

satisfy criteria valued highly in that policy domain (such as secrecy in the intelligence sector) and

attract widespread support or at least no criticism/opposition of any signiﬁcance. Governments

‘do programmes’ and they may succeed/and or fail in this realm.

Politics: the processes and programmes of governments also have political repercussions. They

can boost political reputations and electoral prospects, ease the business of governing through agenda

containment, contribute to desired policy trajectories, maintain broader governance agendas and

help to galvanise political support behind these moves. An example is Australian Prime Minister

Kevin Rudd’s 2007 historic ‘sorry’ speech to aboriginal people for the forced removal of a

Table 36.1 The three realms of policy success

Process success Programme success Political success

Preserving government policy

goals and instruments

Implementation in line with

objectives

Enhancing electoral prospects or

reputation of governments and

leaders

Conferring legitimacy on the

policy

Achievement of desired

outcomes

Controlling policy agenda and

easing the business of governing

Building a sustainable coalition Creating beneﬁt for a target

group

Sustaining the broad values and

direction of government

Symbolizing innovation and

inﬂuence

Meets policy domain criteria Opposition to political beneﬁts for

government is virtually

non-existent and/or support is

virtually universal

Opposition to process is virtually

non-existent and/or support is

virtually universal

Opposition to programme

aims, values, and means of

achieving them is virtually

non-existent, and/or support

is virtually universal

Source: Adapted from McConnell (2010b: Tables 1, 2 and 3)

McConnell

488

‘stolen generation’ of children from their families. Governments ‘do politics’ and they may be

more, or less, successful in doing so.

Policy-making as the juggling of success and failure

There is no doubt that governments can learn from both successes and failures. Learning from

success involves a form of positive feedback, where policies deemed successful are reproduced

and expanded, such as the waves of privatisation of the late 1980s and 1990s. Learning from

failure can be described as negative feedback, where a form of self-correction ensures that fail-

ures are addressed. Corrections may involve the ﬁne-tuning of existing policies (through, for

example, amended guidelines for front-line public housing oﬃcials), the replacement of existing

policies with new ones (Australian gun law reforms after the 1996 Port Arthur massacre) or

even a paradigm realignment which sets the agenda for reform (arguably, pre- and post-global

ﬁnancial crisis shifts in political elite attitudes to debt, regulation and the free market).

Yet an implication of the foregoing discussion of success and failure, is that governments may

aspire to diﬀerent ‘successes’. Indeed, doing so may involve prioritising some over others, and

using the malleable label of ‘success’ to help legitimise their choices and direction. Failure in

some realms is often a precondition for success in others. Failures in some areas can be tolerated,

masked, risked, exploited and even cultivated in the pursuit of success elsewhere.

Tolerating failure

Public policy failures are everywhere. Judging how signiﬁcant they are, may involve Wildavskian

art rather than science. Nevertheless, some simple examples help to make broader points. Low

level failures involve small shortfalls, for example 97 per cent of welfare claims processed within

two weeks as opposed to a target of 98 per cent. There is no incentive to address such minor

failures if broader programme successes are evident and not compromised by the failures.

Moderate or medium-level failures may also be tolerated on the grounds that, on balance, the

programme successes are a price worth paying. The European Union’s (EU) Common Agricultural

Policy is a good example. It does achieve key goals such as EU self-suﬃciency and raising the

standards of farmers’ incomes, but has generated considerable criticism and controversy because

of its drain on the EU budget and production of huge surpluses. Supporters perceive it to be an

EU ‘ﬂagship’ policy while opponents portray it more as the Titanic.

Dealing with such failures may also compromise political success. For example, few are likely

to disagree with a diagnosis of failure in relation to squalid high rise/tower blocks which are

home for some of the poorest and most vulnerable people in society. Yet is not diﬃcult to see

why such failures may be tolerated, especially under conditions of economic stress and ﬁscal strin-

gency. Costs of reform would need to come from other programmes, potentially jeopardising

their success. Dealing with programmatic problems may also compromise political success. Exem-

plars are austerity measures, debt reduction programmes and eﬃciency drives. Japanese Prime

Minister Hashimoto’seﬃciency drive in the banking sector in order to revive a struggling

economy underestimated the adverse political eﬀects and was a major factor in his defeat at the

1998 general election. In 2011 similar political pressures were brought to bear on governments

in Greece, Spain and Ireland. The success of debt reduction programmes generates high risk of

political failure, at least in terms of popular support for governments.

Even high-level failures may be tolerated, or at least there may be attempts to suppress them,

if policy-makers consider that addressing the issues is likely to jeopardise programmatic and/or

political success. One such example is arguably the Guantanamo Bay detention camp under the

Learning from success and failure

489

Obama Administration, where innumerable ethical, human rights, legal and logistical problems

continue, but closure of the camp may well jeopardise the success of the incarceration programme

itself (what to do with approximately 170 potential terrorists when there is a lack of a strong

evidence base to convict?) as well as jeopardising the political goals of the Obama Administra-

tion to ﬁght terrorism in a climate where approximately six out of every ten voters in America,

do not want the camp to be closed.

A more general point needs to be made. Policy problems vary in complexity, from simple

well-structured problems and no clear solutions, to ‘wicked issues’ with multiple and complex

causes, and absence of widely accepted and authoritative solutions (Head 2008; Hoppe 2010).

Also, signals of failure are often vague and ambiguous, subject to multiple (mis)interpretations,

leading to policy-makers processing information disproportionately – sometimes over- or

under-reacting to signs of failure (Jones and Baumgartner 2005).

In sum, failures are often tolerated because to do otherwise would be to consume valuable

agenda time, as well as engage in costly reforms with no guarantee that they will work. Political

success in terms of easing the business of governing and keeping broader policy and paradigm

values on track, is often at the expense of tolerating failures.

Masking failure

It is not only that failures may be tolerated, but they may also be masked, intentionally or

unintentionally, by the activities of policy-makers.

The use of political language can push failures, to the lower realms of political agendas.

Edelman’s (1977) famous dictum, ‘words that succeed and policies that fail’, was used to refer to

political promises and reassurances which hid chronic inequalities, injustices and policy failure,

especially in the area of social welfare. Deborah Stone (2002) provides a more detailed analysis of

the ways in which metaphors, synecdoche and other linguistic devices, can be used to suppress

policy problems. Phrases such as ‘storm in a teacup’ or ‘freak events’ can be used to dampen the

salience of failures, while those that seek to elevate the issues can be labelled as ‘trouble-makers’

or even ‘enemies of democracy’. Indeed, the argumentative turn in policy analysis (see e.g. Majone

1989; Fischer and Forester 1993: Hodgson and Irving 2007), illustrates the importance of argu-

mentation and framing, with diﬀerent interests battling to elevate and even dominate political

discourse with their views.

Policymakers may also produce placebo policies (where there is a high symbolic content in

terms of appearing to tackle the problem, with a low likelihood that they will do so, see

McConnell 2010a). For example, Clarke (1999) in his work on contingency planning for crises

and disasters, argues that plans are primarily of symbolic value (‘ we are prepared and in control’)

rather than operational value. Sharman (2011) in his detailed study of the regulation of money

laundering, explores the rapid diﬀusion and copying of regulations as being ‘politically successful

policy failures’. They demonstrate ‘political success’ because they create the appearance of good

governance, strong regulation and zero tolerance of money laundering, but outcomes are patchy

at best and ineﬀective at worst, particularly in developing countries. For example, the small

Paciﬁc island of Nauru (population 11,000) has developed state-of-the-art anti-money laun-

dering legislation but it has no ﬁnancial sector of any kind – not even a bank or an ATM! Even

institutional reforms and restructuring can have strong symbolic and ‘feel good’ elements, without

necessarily minimising long-term, chronic problems. The creation of the Department of Home-

land Security (DHS) in the United States symbolised an integrated response to the threat of

terrorist attacks, but in reality has arguably done little to minimise the behaviour of actors in

policy subsystems (May et al. 2009).

McConnell

490

Often, when policies fail and/or when crisis and disaster strikes, inquiries are used in order to

ﬁnd out what went wrong and learn lessons for the future. We might expect them to reveal, rather

than mask, failures and their causes. Yet as argued by recent literature on the politics of inquiries

(Stanley and Manthorpe 2004; Kitts 2006; Prasser 2006; Birkland 2007; Boin et al. 2008), investi-

gations may be steered away from examining potential causal factors, as a consequence of terms

of reference, membership and funding. In essence, when failures occur, potential underlying

causal factors may be insulated from investigation, in order protect the values and policies which

sustain broader success trajectories. In 2005–6 when the privatised Australian Wheat Board Ltd

was investigated by the Cole inquiry for paying ‘kickbacks’’to Saddam Hussein’s regime in

contravention of UN sanctions, the terms of reference did not encompass inter alia the minimalist

regulatory regime emanating from the original privatisation, that may have helped explain why

the company had freedom to engage in illicit ﬁnancial activities (McConnell et al. 2008).

A related issue is blame games, involving the deﬂection of attention towards particular indi-

viduals, interests, institutions, technical and climatic/geophysical factors (see Weaver 1986; Hood

2011). The purpose in deﬂecting attention towards a particular actor or set of conditions, is also

to deﬂect attention away from other factors. In other words, blame games may be an attempt to

shore up the returns of policy success.

As Schattschneider (1960) and Bacharach and Bartatz (1970) argued convincingly many years ago,

political systems develop a ‘mobilization of bias’ which is capable of ﬁltering out grievances and

failures. The masking of failure is the ﬂip aside of dominant biases, geared towards particular

aspirations of programmatic and political success.

Generating the risk of failure

Policymakers cannot predict the future with absolute certainty. They can certainly engage in

small policy trials, use ex ante evaluations and attempt to model outcomes, but there are always

elements of uncertainty and contingency (Shapiro and Bedi 2007). All policy-making involves

an element of judgement, sometimes instinctive and sometimes the product of careful scenario

planning, in relation to the risks (May 2005; Althaus 2008). All policy processes carry some

possibility of failure. To put this in the language of this chapter, attempts to achieve process

success carry risks of programme and political failure (see McConnell 2010a). For example:

 log-rolling and horse trading carry the risk of diﬀering interpretations at the implementation

stage, actors not behaving as agreed, and changing socio-economic and political contexts

rendering the agreements unworkable;

 evidence-based policy-making brings the risk of policy-making being informed by skewed

or incorrect evidence, or being merely the legitimating basis for policy and political trajectories

that bring strong contradictions;

 deliberation and stakeholder engagement can be little more than the legitimating framework for

policy and political pathways, bringing the risk of inﬂated expectations which cannot be met;

 use of strong executive power to push through policies and legislation carries the risk of

political backlash, as well as programmes lacking suﬃcient scrutiny to ensure eﬀective

implementation;

 policy transfers from other jurisdictions carry the risk of being uninformed, incomplete or

inappropriate for the new policy context.

The list could be longer but the basic point is clear. Every strategic process strategy brings

opportunities but also risks. What constitutes process success for policy-makers in producing

Learning from success and failure

491

their desired policies through constitutionally and quasi-constitutionally legitimate means, can carry

the longer-term risk of producing programmes which fail to fulﬁl intended goals, while backﬁring

politically on government agendas and compromising its broader political goals and pathways.

Failure is not inevitable but failure is often a prospect, to varying degrees and depending on the

context such as the degree of politicisation surrounding the issue and likely levels of support/

opposition. Public policy-making and accepting the risk of failure go hand in hand.

Cultivating/creating/exploiting failure

The logic of the policy cycle approach is that policy-making is driven by problems in search of

solutions. However, as ‘garbage can’ theory indicates, the opposite may be the case. Policy can

be driven by solutions in search of problems in order to justify an intended course of action.

The literature on crisis and disaster makes similar points, that crisis conditions can be capitalised

upon and even exploited in order to achieve longer-term programmes and political goals (Klein

2007; Boin et al. 2009). For example, often cited allegations include the existence of ‘weapons

of mass destruction’ in Iraq as a means of legitimating the coalition invasion of Iraq in 2002, and

asylum seekers throwing their children overboard a stricken vessel bound for Australian waters,

in order to legitimate immigration reform. Crises can be an opportunity to boost political careers

and demonstrate ﬁtness for oﬃce. For example, the 2002 ﬂoods in Germany helped to provide

the foundation for the invigoration of Chancellor Schröder, some six weeks before the federal

election - helping his party to victory (Bytzek 2008). In the wake of the 9/11 attacks, George

W. Bush’s approval ratings soared to the highest in US polling history. As Rahm Emanuel,

Barack Obama’s Chief of Staﬀ, stated: ‘you never want a good crisis to go to waste’ (Washington

Post: 6 March 2009). Overall, therefore, ‘failure’ may be sought out and even cultivated, in order

to pave the way for programme and political success.

Conclusion

Conventional wisdom and governmental rhetoric lean heavily towards the promise that political

and policy elites will build on successes and learn from failures. Leaving aside the issue that what

constitutes learning is a value assumption, change and learning can certainly happen. However,

it is far from an automatic or reﬂexive process. The nature of policy success is contested and

riddled with methodological diﬃculties, requiring a value judgement in terms of attaching the label

of ‘success’ to outcomes, depending on the salience given to achievements and the signiﬁcance

given to shortfalls and ambiguities. There is a degree of malleability in the way that policies can

be framed, as successful or otherwise.

When we also factor in the existence of diﬀerent forms of successes in the process, programme

and political realms of policy, it is evident that failures – and lack of learning from them – are

not only commonplace, but often necessary because they have been disregarded in the pursuit

of longer-term success goals. Successful processes (such as a government rushing a bill through a

legislature) in order to achieve its goal, may run the risk (conceived of as acceptable risk by

policy-makers) of programme failure through ill-thought-out legislation, and even of political

failure if the issue is high proﬁle and politically partisan. Or, successful programmes, especially if

they are ruthlessly e ﬃcient in achieving highly contested goals, may carry the risk of political

failure. Alternatively, ineﬀective programmes tackling diﬃcult and perhaps ‘wicked issues’, may

be tolerated and even promoted, if it suits political success aspirations in maintaining control of

agendas through the appearance of tackling problems. Colloquially, we might refer to this as

‘good politics but bad policy’.

McConnell

492

As indicated by the argument and analysis in this chapter, failures are not always abhorrent to

policy-makers. Failures can be tolerated, masked, risked, cultivated and even exploited in order

to pave the way for process, programme and political goals. Policymaking involves trade-oﬀs

and risk taking, juggling the hope of some successes, with the acceptance and even desirability

of failure.

Bibliography

Althaus, C. 2008. Calculating Political Risk. Sydney: University of New South Wales Press.

Bacharach, P. and M. Baratz. 1970. Power and Poverty: Theory and Practice. New York: Oxford University

Press.

Birkland, T.A. 2007. Lessons of Disaster: Policy Change after Catastrophic Events. Washington, DC: Georgetown

University Press.

Boin, A., A. McConnell and P. ’t Hart eds. 2008. Governing after Crisis: The Politics of Investigation,

Accountability and Learning. Cambridge, Cambridge University Press.

——,P.’t Hart and A. McConnell. 2009. ‘Towards a theory of crisis exploitation: Political and policy

impacts of framing contests and blame games’, Journal of European Public Policy 16(1): 81–106.

——,P.’t Hart, E. Stern and B. Sundelius. 2005. The Politics of Crisis Management: Public Leadership under

Pressure. Cambridge: Cambridge University Press.

Bovens, M.P. and P. ’t Hart. 1996. Understanding Policy Fiascoes. New Brunswick, NJ: Transaction.

——,P.’t Hart, and B.G. Peters (eds). 2001. Success and Failure in Public Governance: A Comparative Analysis.

Cheltenham: Edward Elgar.

Bytzek, E. 2008. ‘Flood response and political survival: Gerhard Schröder and the 2002 Elbe ﬂoods in

Germany’, in A. Boin, A. McConnell and P. ’t Hart (eds) Governing after Crisis: The Politics of Investigation,

Accountability and Learning. Cambridge: Cambridge University Press, 85 – 113.

Clarke, L.B. 1999. Mission Improbable: Using Fantasy Documents to Tame Disasters. Chicago: University of

Chicago Press.

Davidson, E.J. 2005. Evaluation Methodology Basics: The Nuts and Bolts of Sound Evaluation. Thousand Oaks,

CA: Sage.

Dunleavy, P. 1995. ‘Policy disasters: Explaining the UK’s record’, Public Policy and Administration 10(2): 52–70.

Edelman. 1977. Political Language: Words that Succeed and Policies that Fail. New York: Academic Press.

Fischer, F. 1995. Evaluating Public Policy. Chicago: Nelson-Hall Publishers.

—— 2003. Reframing Public Policy: Discursive Politics and Deliberative Practices. Oxford: Oxford University

Press.

——and J. Forester. 1993. The Argumentative Turn in Policy Analysis and Planning. Durham, NC: Duke

University Press.

Gupta, Dipak K. 2001. Analyzing Public Policy: Concepts, Tools, and Techniques. Washington, DC: Congressional

Quarterly

Inc.

Hall, P.G. 1982. Great

Planning Disasters. Berkeley, CA, University of California Press.

Hay, C. 2002. Political Analysis. Basingstoke, Palgrave Macmillan.

Head, B.W. 2008. ‘Wicked problems in public policy’, Public Policy 3(2): 101–18.

Hodgson, S.M. and Z. Irving (eds). 2007. Policy Reconsidered: Meanings, Politics and Practices. Bristol: Policy

Press.

Hood, C. 2011. The Blame Game: Spin, Bureaucracy and Self-preservation in Government. Princeton, NJ:

Princeton University Press.

——and H.Z. Margetts. 2007. The Tools of Government in the Digital Age. Basingstoke: Palgrave Macmillan.

Hogwood, B.W. and B.G. Peters. 1985. The Pathology of Public Policy . Oxford: Clarendon Press.

Hoppe, R. 2010. The Governance of Problems: Puzzling, Powering and Participation. Bristol: Policy Press.

Howlett, M. 2010. Designing Public Policies: Principles and Instruments. New York: Routledge.

Jones, B.D. and F.R. Baumgartner. 2005. The Politics of Attention: How Government Prioritizes Problems.

Chicago: University of Chicago Press.

Kerr, D.H. 1976. ‘The logic of “policy” and successful policies”, Policy Sciences 7(3): 351–63.

Kitts, K. 2006. Presidential Commissions and National Security: The Politics of Damage Control. Boulder, CO:

Lynne Rienner.

Klein, N. 2007. The Shock Doctrine: The Rise of Disaster Capitalism. New York: Metropolitan Books and

Henry Holt.

Learning from success and failure

493

McConnell, A. 2010a. Understanding Policy Success: Rethinking Public Policy. Basingstoke: Palgrave Macmillan.

—— 2010b. ‘Policy success, policy failure and grey areas in-between’, Journal of Public Policy, 30(30): 345–62.

—— A.Gauja and L. Botterill. 2008. ‘Policy ﬁascos, blame management and AWB limited: The Howard

government’s escape from the Iraq wheat scandal’, Australian Journal of Political Science 43(4): 599-616.

Majone, G. 1989. Evidence, Argument, and Persuasion in the Policy Process. New Haven, CT: Yale University

Press.

Marsh, D. and A. McConnell, A. 2010. ‘Towards a framework for establishing policy success’, Public

Administration 88(2): 586–7.

May, P.J. 2005. ‘Policy maps and political feasibility’, in I. Geva-May (ed.) Thinking Like a Policy Analyst:

Policy Analysis as a Clinical Profession. New York: Palgrave Macmillan, 127–51.

——, J. Sapotichne and S. Workman. 2009. ‘Widespread policy disruption: Terrorism, public risks, and

homeland security’, Policy Studies Journal 37(2): 171–94.

Pawson, R. 2006. Evidence-based Policy: A Realist Perspective. London: Sage.

Prasser, S. 2006. Royal Commissions and Public Inquiries in Australia. Chatswood, NSW: LexisNexis Butterworths.

Schattschneider, Elmer Eric. 1960. The Semisovereign People: A Realist’s View of Democracy in America. New

York: Holt, Rinehart and Winston.

Schneider, A.L. and H. Ingram. 1997. Policy Design for Democracy. Lawrence, KS: University Press of

Kansas.

Shapiro, I. and S. Bedi. (eds). 2007. Political Contingency: Studying the Unexpected, the Accidental and the

Unforeseen. New York: New York University Press.

Sharman, J.C. 2011. The Money Laundry: Regulating Global Finance in the Criminal Economy. Ithaca, NY and

London: Cornell University Press.

Stanley, N. and J. Manthorpe. 2004. The Age of the Inquiry: Learning and Blaming in Health and Social Care.

London: Routledge.

Stone, D. 2002. Policy Paradox: The Art of Political Decision Making, 2nd edn. New York: W.W. Norton.

Weaver, R.K. 1986. ‘The politics of blame avoidance’, Journal of Public Policy 6(4): 371 – 98.

McConnell

494

Author index

Aaron, P. and Musto, D. 145

Abbott, A. 233, 407

Abbott, D., Brown, S. and Wilson, G. 378

Abelson, Donald 6

Aberbach, J.D., Putnam, R.D. and Rockman.,

B.A. 337

Ackleson, Jason 184

Ackrill, R.W. and Kay, A. 469

Adam, Barbara 455

Adams, B. 440, 441

Adams, Gordon 338

Adelle, C. et al. 246–47, 249

Adelle, Camilla xi, 244–54

Adler, E. and Haas, P.M. 237, 240

Adshead, M. 156–57

Agranoff, R. and McGuire, M. 361, 363

Agranoff, R. and Yildiz, M. 369

Agranoff, Robert xi, 361–73

Ahn, C.S. 337

Ahn, T.K. 121

Ahuja, A., Kremer, M. and Zwane, A.P. 415

Ajzen, I. and Fishbein, M. 125

Albæk, E., Green-Pedersen, C. and Nielsen, L.B.

172

Albert, M. and Laberge, S. 233

Albrechts, L. 233

Albright, Elizabeth A. 129, 133, 186

Albrow, Martin 332, 334

Alchian, A.A. and Demsetz, H. 239

Alesina, Alberto 85

Alford, J. and O’Flynn, J. 436

Aligica, P. and Boettke, P. 91

Aligica, Paul Dragos xi, 87–97

Allan, J.P. and Scruggs, L. 32, 37

Allcott, H. and Mullainanthan, S. 423, 424

Allen, Linda 121

Allio, L. 244, 246

Allison, G. and Zelikow, P. 338

Allison, Graham 21, 338

Almond, G.A. and Genco, S.J. 59

Almond, Gabriel A. 62

Alter, C. and Hage, J. 364

Althaus, C. 491

Althaus, C., Bridgman, P. and Davis, G. 434

Altman, J.A. and Petkus, E. 302

Ameringer, Carl F. 133

Anderson, J. 300

Andersson, Krister P. 120, 121

Andersson, Magnus 129

Andrews, M., McConnell, J. and Wescott, A. 376,

381

Anglund, S.M. 304

Angrist, J. and Pischke, J.S. 416, 419

Angrist, J., Imbens, G.W. and Rubin, D.B. 430n9

Ansell, C. and Gash, A. 341

Appleby, Paul 333

Araral Jr., Eduardo xi, 115–24

Argyris, C. and Schön, D. 106

Aristotle 102, 201

Armstrong, E.M., Carpenter, D. and Hojnacki.,

M.E. 280

Aronson, P.H. and Ordeshook, P.C. 80

Arrow, Kenneth J. 76, 78, 83–84, 466

Arthur, W.B. 402, 466

Ashford, Douglas E. 362

Atkinson, M. and Coleman, W. 19

Baccaro, L. 36

Baccaro, L. and Simoni, M. 36

Bachrach, P. and Baratz, M.S. 167, 491

Baiocchi, G. 442

Baker, R. 340, 341

Baker, W.H., Addams, H.L. and Davis, B. 440,

441

Bal, Mieke 103

Balfour, D.L. and Mesaros, W. 103

Balla, S.J. and Wright, J.R. 280

Balla, Steven J. 280

Banerjee, A.V. 415–16, 419, 427

Banerjee, A.V. and Duﬂo, E. 416, 419, 420, 426

Banerjee, A.V. et al. 415

Barabási, Albert-László 282

495

Bardach, Eugene 7, 10, 34, 244, 369, 370

Bardhan, P. 416

Barnes, M. et al. 438

Barnett, M.N. and Finnemore, M. 339

Barnett, P. et al. 22

Barrow, C.J. 9, 246, 248

Bartels, L.M. 206

Barthes, Roland 106

Barzelay, Michael 335, 340, 405

Basu, K. 416

Basurto, Xavier 120

Bateson, Gregory 192

Baumgartner, F.R. and Jones, B.D. 18, 139, 140,

141, 142, 143, 144, 145, 146, 147, 150n2, 167,

172, 175, 177, 178, 179, 193, 201, 205, 281,

452, 453, 474, 477

Baumgartner, F.R. and Leech, B.L. 282

Baumgartner, F.R., De Boef, S.L. and Boydstun,

A.E. 193, 201

Baumgartner, F.R. et al. 143, 144, 145, 146, 147,

148–49, 281, 452

Baumgartner, F.R., Green-Pedersen, C. and

Jones, B.D. 167, 190

Baumgartner, F.R., Jones, B.D. and Leech, B.L.

205

Baumgartner, Frank R. 146

Beach, D. and Pedersen, R.B. 481

Beard, Adrian 191

Bearman, P., Faris, R. and Moody, J. 481

Beck, James M. 335

Beck, Ulrich 63

Becker, Gary S. 85, 282

Beetham, David 23, 334

Behn, R.D. 34

Béland, D. 184, 403

Bendor, J. and Hammond, T.H. 21

Bendor, J., Moe, T.M. and Shotts, K.W. 321, 325

Bendor, J., Taylor, S. and van Gaalen, R. 84

Bendor, Jonathan 277, 283

Benington, J. 436

Benjamin, M. 353

Bennett, C.J. 34, 404

Bennett, C.J. and Howlett, M. 24, 35, 404, 407

Bennett, W. Lance 170, 178

Bennett, W.L. and Feldman, M.S. 103

Bennett, W.L., Lawrence, R.C. and Livingston, S.

170

Benson, J.K. 158

Berger, P. and Luckmann, T. 100, 103

Berry, F.S. and Berry, W.D. 144, 404, 406

Bertrand, M. et al. 431n23

Best, R. 177, 184

Bevir, M. and Richards, D. 161, 338

Bikhchandani, S., Hirshleifer, D. and Welch, I.

282

Biller, R.P. 34

Bina, O. 245

Bingham, G. 357

Bingham, L.B., Nabatchi, T. and Leary, R.O.

438, 444

Birkland, T.A. and Lawrence, R.G. 175, 176, 181

Birkland, Thomas A. xi, 143, 175–88, 206, 484,

491; garbage can model (GCM), work with

326–27

Black, Duncan 76, 78, 79, 80

Black, Julia 67

Blackburn, J. and Holland, J. 381

Blair, Tony 65

Blais, A. and Dion, S. 84, 336

Blalock, A.B. 436

blame games 491

Blankart, C.B. and Koester, G.B. 85

Blankart, Charles B. 65

Blankenau, Joe 184

Blau, P.M. and Scott, W.R. 332

Bloom, H. et al. 415

Blyth, Mark 451

Boardman, A.E. et al. 5

Bobrow, D. and Dryzek, J. 105, 217, 255, 256,

258

Bochel, H.M. and Duncan, S. 108

Bogason, P. and Toonen, T.A.J. 363

Boin, A., McConnell, A. and ‘t Hart, P. 484, 491,

492

Boje, David M. 103

Bond, A. and Morrison-Saunders, A. 249

Boner, P.A. et al. 442

Bonoli, G. 32, 34, 36, 37

Boom, H. 317n3

Borgatti, S.P., Everett, M.G. and Freeman, L.C.

241n2

Boston, J., Church, S. and Bale, T. 6

Boston, Jonathan 8, 340

Boswell, C. 405

Bots, Pieter W.G. xi, 255–69

Bourdieu, Pierre 107, 239

Boushey, Graeme xii, 138–52, 282

Bovens, M.P. and ‘t Hart, P. 484, 485

Bovens, M.P., ‘t Hart, P. and Peters, B.G. 484

Bovens, M., ‘t

Hart, P. and

Kuipers, S. 434–35,

443

Bowornwathana, B. and Poocharoen, O. 338

Bowornwathana, Bidhya 338

Boxer, Sarah. B. 200

Boyce, Mary E. 103

Boydstun, Amber E. 282

Boykoff, M. and Boykoff, J.M. 208

Boykoff, Maxwell 207, 208, 209

Boyne, George A. 84

Bradford, N. 405, 408

Braithwaite, J. and Draos, P. 63

Braun, D. and Gilardi, F. 34, 404, 406

Braybrooke, D. and Lindblom, C.E. 21, 287,

288–89, 291, 292

Author index

496

Braybrooke, David 289

Breeman, G. et al. 168

Brehm, J. and Gates, S. 280

Bressers, H.T.A. and O’Toole, L.J. 22

Breton, A. and Wintrobe, R. 84

Breton, Albert 79

Breunig, C. and Jones, B. 147, 149, 150n10

Breunig, C. and Koski, C. 144, 147, 148

Brewer, G. and DeLeon, P. 255, 256

Brewer, G.D. 33–34

Brinkerhoff, Derick W. xii, 374–83

Brinkerhoff, D.W. and Crosby, B.L. 379–80

Brinkerhoff, D.W. and Ingle, M. 377

Brinkerhoff, Jennifer M. xii, 374–83

British Pension Commission 409

Brooks, David 220

Brooks, S.M. 34, 404, 410

Brown, Gordon (and government of) 65

Brown, Lawrence D. 292

Brown, Steven R. 278

Brown, T.L. and Potoski, M. 363

Browne, William P. 54, 156

Bruner, Jerome 102

Bruns, Bryan Rudolph 120

Bryson, J.M. and Crosby, B.C. 299, 300, 302, 303

Buchanan, James 76–77, 78, 82

Buchanan, J.M. and Tullock, G. 78, 79, 80, 83,

Buckman, G. and Diesendorf, M. 22

Bulmer, M. 407, 408, 409

Bulmer, Simon 473

Burnham, James 335

Burnier, DeLysa 178

Burns, John F. 196

Burstein, P. 155, 156

Busch, P.-O. and Jörgens, H. 34, 37

Busch, P.-O., Jörgens, H. and Tews, K. 34

Busenberg, George J. 184

Bush, George W. 223, 492

Bushouse, Brenda K. 121

Butcher, Mike 197

Butler, R. 299, 302

Bytzek, E. 492

Cahill, A,G. and Overman, E.S. 20

Callahan, K. 437

Callon, Michel 106–7

Camerer, C.F. and Thaler, R.F. 276

Camerer, Colin F. 276

Cameron, David 196, 197

Campbell, A.L. 403

Campbell, C. and Szablowski, G. 337

Campbell, J.L. 469

Cao, X. 405

Capano, G. and Howlett, M. 454

Capano, Giliberto xii, 451–61

Carley, Michael 20, 246

Carlson, C. 354

Carlson, Deven 5

Carpenter, D. and Sin, G. 179, 185

Carpenter, D., Esterling, K. and Lazer, D. 159

Carpenter, Daniel 280

Carson, L. 442

Carson, L. et al. 438

Carstensen, Martin B. 19

Cartwright, N. 430n10

Cashmore, M., Bond, A. and Cobb, D. 249

Cashore, B. and Howlett, M. 144, 453, 474

Castells, Manuel 363

Castles, F.G. 32, 37

Castles, F.G. and Mitchell, D. 32

Cecot, C., Hahn, R.W. and Renda., A. 249

Cerny, Philip 45

Chaker, A. et al. 245

Chapman, R.A. 408

Chattopadhyay, Paresh 335

Checkel, J. 238–39

Chen, Huey-Tsyh 11

Chenoweth, E. and Clarke, S.E. 184

Cheung, A.B.L. and Scott, I. 340

Choe, Jaesong 121

Chomsky, N. and Herman, E. 101

Chong, Dennis 282

Christensen, T. 332

Chung, K.H. and Cox, R.A.K. 282

Chwieroth, J. 239

Clark, C.D. and Russell, C.S. 22

Clark, J. 406, 410

Clark, William R. 59

Clarke, L.B. 490

Clasen, J. and Clegg, D. 32

Clasen, J. and Siegel, N.A. 32, 37

Clayton, R. and Pontusson, J. 34

Cobb, R.W. and Elder, C.D. 18, 167, 175, 176,

177, 190, 200, 205

Cobb, R.W. and Elder, C.W. 5

Cobb, R.W. and Ross, M.H. 175, 193

Cobb, R.W., Keith-Ross, J. and Ross, M.H. 18,

177, 190

Cobden, R., Mitrany, D. and Ohmae, K. 63

Coglianese, C. 357

Cohen, Bernard C. 205

Cohen, M., March, J. and Olsen, J. 51, 54, 278,

303, 320, 321

Cohen, W.M. and Levinthal, D.A. 24

Cole, R. and Caputo, D. 338

Coleman, Eric 121

Coleman, J.S. 238

Coleman, J.S. and Skogstad, G. 235

Coleman, W.D. and Perl, A. 47, 52

Coleman, W.D. and Skogstad, G. 19, 47

Coleman, W.D. and Tangermann, S. 479

Coleman, W.D., Skogstad, G.D. and Atkinson,

M.M. 405, 474

Author index

497

Coleman, William D. 24

Collier, D. and Collier, D. 402

Collier, D. and Messick, R.E. 34, 406

Collier, R.B. and Collier, D. 323

Collins, M.L. and Kapuçu, N. 184

Compston, H. 153, 158

Conlan, T.J. et al. 324

Considine, Mark 336

Conybeare, John A.C. 84

Cook, D.D. and Campbell, D.T. 9

Cook, F.L. et al. 205

Cook, P. and Mosdale, S. 66, 67

Corbett, J.B. and Durfee, J. 208

Corburn, J. 352

Cortell, A.P. and Peterson, S. 62

Coulter, Ann 200

Crecine, John 280

Crepaz, M.M.L. 31, 37

Cropper, S. 369, 370

Crosby, N. and Nethercut, D. 441

Crouch, C. 469

Crouch, C. and Farrell, H. 469, 470

Crozier, M. and Friedberg, E. 299

Curtain, R. 438

Czarniawska, B. 103

Dahl, R.A. and Lindblom, C.E. 21, 218, 287, 290

Dahl, Robert A. 296

Daniels, M.R. 34

Dar, S. and Cooke, B. 374

Das, J., Shantayana, D. and Hammer, S. 416,

428–29

Daugbjerg, C. and Swinbank, A. 474, 477, 479

Daugbjerg, Carsten xii, 466, 473–83

Davenport, T.H. and Prusak, L. 368

David, P.A. 402

Davidson, E.J. 485, 487

Davis, Charles 184

Dawson, R. and Robinson, J. 125

de Bruijn, J. and ten Heuvelhof, E. 260

de Bruijn, J., ten Heuvelhof, E. and in ‘t Veld, R.

255, 260, 266

De Castro, Fabio 120

De Francesco, F. 246, 247

de la Porte, C. and Nanz, P. 407

de Ridder, W. 245

de Ridder, W. et al. 246, 248, 249

de Sousa Briggs, X. 350

de Vries, Michiel S. 452

Dearing, J.W. and Rogers, E.M. 205

Deaton, A. 416, 421, 422, 423, 424–25, 426

deLeon, P. and Hernández-Quezada, J.M. 34

deLeon, P. and Varda, D.M. 161

deLeon, Peter 34, 105, 175, 255, 257, 258, 260

Dempster, M. and Wildavsky, A. 410

Derrida, Jacques 100

Derthick, Martha 5, 402, 403

Dessler, David 59

DeYoung, Sarah E. xii, 175–88

DiMaggio, P. and Powell, W. 106

Djelic, M.-L. and Quack, S. 403

Dodd, V. and Davies, C. 195

Dolan, J. and Rosenbloom, D.H. 337

Dolan, Julie 337

Dolowitz, D. and Marsh, D. 8, 35, 404, 406

Dolsak, Nives 121

Donnelly, Kevin P. xii, 189–203

Donnelly, Shanon 121

Donovan, Mark 143

Doolittle, F. and Traeger, L. 415

Douglas, Mary 100

Dowding, Keith 20, 64, 158, 159

Dowie, J. 245, 250

Downs, Anthony 76, 78, 80, 81, 176, 177, 205,

331, 332, 333, 335

Drake, W.J. and Nicolaïdis, K. 232

Drezner, D.W. 34, 37

Dror, Y. 255, 263

Druckman, James N. 192, 206

Dryzek, John S. 102, 103, 260, 293, 294, 437,

438, 444n1

Dryzek, J.S. et al. 37

Dudley, G. and Richardson, J.J. 234

Duﬂo, E. 416

Duﬂo, E. et al. 419, 424, 425

Duggan, Mark 196

Duit, A. and Galaz, V. 299

Dunleavy, P. 484

Dunlop, C.A. and James, O. 239, 404

Dunlop, C.A. and Radaelli, C.M. 409

Dunlop, Claire A. xiii, 19, 229–43, 407

Dunn, Delmer D. 334

Dunn, W.N. 248, 255, 256, 259, 261

Dunsire, Andrew 22

Dunt, Ian 196

Durning, D. 260

Durning, D. and Osuna, W. 255, 263

Dutil, P.A. et al. 443

Dwyer, Devin 199

Dye, T.R. 29

Dyson, Kenneth 46

Eales, R. et al. 248

Easterly, W. 374

Easton, David 125

Eckstein,

Otto 5

Edelenbos, J.

255, 261

Edelenbos, J. and Klijin, E.H. 341

Edelenbos, J., van Schie, N. and Gerrits, L. 101

Edelman, Murray J. 178, 190, 207, 300, 490

Egeberg, Morten 338

Eiji, Kawabata 337

Eilders, C. 205, 206

Einsiedel, Edna F. 184

Author index

498

Einsiedel, E.F., Jelsøe, E. and Breck, T 438

Eising, R. 24

Eisinger, Peter K. 363

Eisner, Marc Allen 5, 407

Elazar, Daniel J. 362

Eldredge, N. and Gould, S. 150n2

Elkins, Z. and Simmons, B. 34

Elkins, Z., Guzman, A.T. and Simmons, B.A. 34

Elliott, C. and Schlaepfer, R. 234

Elliott-Teague, Ginger 121

Elmore, Richard F. 365

Elster, Jon 20

Emanuel, Rahm 492

Emerson, K. et al. 358

Engeli, I., Green-Pedersen, C. and Larsen, L.T.

170

Entman, Robert M. 178, 193, 206, 207

Epstein, S. 235

Esman, M. 375

Esping-Andersen, G. 34

Esty, D.C. and Porter, M.E. 31, 33

Etheredge, L.S. and Short, J. 405

Etzioni-Halevy, Eva 50, 334

Evans, M. and Davies, J. 404, 405

Evans, P., Jacobson, H. and Putnam, R. 53

Eyestone, Robert 5, 406

Faiola, Anthony 196

Falleti, T. and Lynch, J. 457

Farazmand, Ali 341

Farnsworth, S.J. and Lichter, S.R. 207

Farnsworth, Stephen J. xiii, 204–14

Feldman, Martha 301, 307

Feldman, M.S. et al. 103

Fenger, M. and Klok, P.J. 130

Fenno, Richard F. 231–32, 324

Fernandes, R. and Simon, H.A. 278

Ferrara, M. 36

Finlayson, A. 232

Finnemore, M. 240

Fischer, F. and Forester, J. 98, 102, 103, 190, 255,

257, 259, 435, 490

Fischer, Frank 102, 108, 194, 207, 250, 257, 259,

260, 261, 267, 301, 352, 436, 438, 485, 487

Fisher, R. and Ury, W. 348, 359n1, 359n3

Fisher, R., Ury, W. and Patton, B. 105

Fisher, Ronald A. 415, 422, 424

Fishkin, J. 260, 441

Fishman, Donald A. 185, 186

Fiske, S. and Taylor, S. 275

Flickinger, Richard 205

Foley, Henry A. 337

Folke, C. et al. 341

Forester, John 11, 21, 205, 352, 355, 436, 438

Foucault, Michel 101

Fowler, A. 376

Frank, D.J., Hironaka, A. and Schofer, E. 34

Frantz, J.E. 34

Freedman, D.A. 431n21

Freeman, Gary P. 24

Freeman, J. Leiper 18, 125

Freeman, R. 404

Fuente, David xiii, 415–33

Fuller, Boyd xiii, 347–60

Fung, A. 437, 439, 442

Furlong, F.R. 339

Gabaix, X. 282

Galaskiewicz, J. 157

Galland, D. and McDaniels, T. 185

Galton, Francis 282; Galton’s problem 406– 7

Gamson, W.A. 206–7

Gamson, W.A. and Modigliani, A. 206, 207

Ganghof, S. 37

Garcia-Zamor, Jean-Claude 340

Garvey, Gerald 334

Gastil, J. and Levine, P. 439, 442

Gauss-Markov theorem 424

Gawthrop, Louis C. 20

Gerring, John 457

Gerrits, L. 299

Geva-May, I. 34

Gibson, C.C., Andersson, K.P. and Shivakumar,

S. 121

Gibson, C.C., Williams, J.T. and Ostrom, E. 119

Giddens, Anthony 107

Gieryn, T.F. 239

Giest, Sarah xiii, 17–28

Gigerenzer, G. and Goldstein, D.G. 283

Gigerenzer, Gerd 283

Gilardi, F. 405

Gilardi, F., Fuglister, K. and Luyet, S. 406, 410

Gillespie, P., Girgis, M. and Mayer, P. 381

Giné, X. et al. 431n23

Givel, Michael 146, 150n2, 150n6

Glascock, Jack 178

Glennerster, R. and Kremer, M. 416

Goeller, B. et al. 259

Goetz, A.M. and Gaventa, J. 436, 444

Goffman, Erving 102, 107

Goldﬁnch, Shaun 8

Goodin, R.E. and Dryzek, J. 442

Goodnow, Frank J. 333

Goodsell, Charles T. 334

Gore, Al 209

Gormley, W.T. and Balla, S.J. 334

Gourevitch,

Peter 46

Graddy, E.A.

and Ye, K. 34

Granovetter, M.S. 282

Grant, W. 160

Gray, Barbara 366, 367

Gray, Virginia 34, 219

Green, M.T. and Thompson, F. 279

Green-Pedersen, C. and Krogstrup, J. 169

Author index

499

Green-Pedersen, C. and Mortensen, P.B. 168–69

Green-Pedersen, C. and Wilkerson, J. 169

Green-Pedersen, Christoffer xiii, 32, 34, 37,

167–74, 185

Greenaway, J., Salter, B. and Hart, S. 159

Greener, I. 403, 464, 468, 470, 471

Grifﬁth, Ernest 125

Grifﬁth, Samuel W. 121

Grindle, M.S. and Thomas, J.W. 379

Grossback, L.J., Nicholson-Crotty, S. and

Peterson, D.A.M. 405, 407

Guba, E.G. and Lincoln, Y.S. 261, 394, 435–36,

438

Gulick, Luther 375

Gulrajani, N. 377, 378

Gunningham, N. and Young, M.D. 22

Gunningham, N., Grabosky, P. and Sinclair D. 22

Gunter, Valerie J. 178

Gupta, Dipak K. 336, 485

Gusﬁeld, Joseph R. 179, 192

Ha, E. 37

Haas, Peter M. 19, 48, 229–30, 231–32, 233–34,

235, 236–37, 238–39, 240, 241n1, 405

Haas, P.M. and Adler, E. 235

Habermas, Jürgen 23, 104–5

Hacker, J.S. 34, 402, 403, 464

Hahn, R.W. and Tetlock, P.C. 245

Hajer, M.A. and Wagenaar, H. 103, 436

Hajer, Maarten A. 102, 442

Halachimi, A. and Holzer, M. 437

Hall, J.A. and Ikenberry, G.J. 61

Hall, P.A. and Taylor, R.C.R. 36, 463

Hall, Peter A. 35, 37, 49, 50, 60, 248, 322, 403,

404, 405, 453, 464, 467

Hall, P.G. 486

Hall, R.H. and Quinn, R.E. 332

Hall, Richard H. 332

Hall, T.E. and O’Toole, L.J. 365

Halliday, Josh 197

Halpin, D. 282

Hamilton, Alexander 90

Hammond, T. and Miller, G.J. 283

Hampton, Greg 103

Handa, S. and Maluccio, J. 430n5

Hanf, K., Hjern, B. and Porter, D.O. 366

Harrington, W., Morgenstern, R.D. and Nelson,

P. 248

Harrison, G.W. 430n4

Harrison, K.A. 233

Harsanyi, John C. 83

Hartz-Karp, J. 442, 443

Hashimoto, Ryutaro 489

Häusermann, S. 32, 36

Hawkesworth, Mary 59, 255, 258

Hawkins, K. and Thomas, J.M. 22

Hay, C. and Wincott, D. 59, 470

Hay, Colin 467, 468, 471, 477–78

Haydu, Jeffrey 456, 475, 476–77, 479, 481

Hayek, Friedrich 292

Hayes, Michael xiii, 139, 287–98

Hayes, Tanya 120

Haynes, Philip 299

Head, B.W. 437, 439, 490

Heady, Ferrel 338

Healey, Patsy 105

Heberlein, T.A. 440

Heckman, J.J. 416, 422, 426, 428

Heckman, J.J. and Smith, J.A. 419, 422, 423, 424,

426

Heckman, J.J. and Vytlacil, E. 426

Heclo, Hugh 125, 301, 323, 404, 405, 408, 451,

478

Hedberg, B. 404

Hedström, P. and Swedberg, R. 238

Heffernan, Richard 8

Heichel, S., Pape, J. and Sommerer, T. 34

Heikilla, T., Schlager, E. and Davis, M.W. 121

Heikkila, T. and Isett, K.R. 437

Henderson, Jeffrey 337

Hendriks, Carolyn M. xiii, 434–48

Hendriks, C.M. and Carson, L. 444

Henry, A., Lubell, M. and McCoy, M. 129

Henry, Adam 129

Heraclitus 451

Herman, E.S. and Chomsky, N. 207

Hertin, J. et al. 246, 247, 248, 249

Hess, C. and Ostrom, E. 121

Higgins, R.S. et al. 84

Hilgartner, J. and Bosk, C. 175, 181

Hinich,

M. and Munger,

M. 278

Hinrichs, K. and Kangas, O. 403

Hintze, Otto 60

Hira, A., Huxtable, D. and Leger, A. 66

Hirschman, Albert 191

Hix, Simon 338

Hjern, B. and Porter, D. 125, 365, 366

Ho, A.T. 437

Ho, Khai Leong 338

Hobbes, Thomas 78

Hodgson, S.M. and Irving, Z. 490

Hogwood, B. and Gunn, L. 255, 256

Hogwood, B.W. and Peters, G.B. 34, 402, 451,

484

Holliday, Ian 62

Holzer, M. and Kloby, K. 436, 437

Holzinger, K. and Knill, C. 34

Holzinger, K., Knill, C. and Sommerer, T. 34, 37

Hood, C. and Margetts, H.Z. 484

Hood, C. and Peters, B.G. 249

Hood, C. and Scott, C. 65

Hood, C. et al. 64

Hood, Christopher 22, 245, 363, 491

Hoogerwerf, A. 303

Author index

500

Hopf, T. 239

Hopkins, R.F. 339

Hoppe, R. 255, 257, 490

Horner, L., Lekhi, R. and Blaug, R. 436, 438,

440

Hotelling, Harold 80

House, P. and Shull, R. 255, 258, 259

House, Ruth Schuyler xiii–xiv, 115–24

Howard, C. 436, 439–40

Howarth, J. 235

Howlett, M. and Cashore, B. 37, 452

Howlett, M. and Lejano, R. 217–18

Howlett, M. and Ramesh, M. 20, 29, 31, 33, 107,

244, 336, 341

Howlett, M. and Rayner, J. 35, 105, 106, 406,

410, 473–74, 475, 476, 480

Howlett, Michael xiv, 4, 17–28, 34, 244, 300,

369, 456, 484–85

Huber, E. and Ray, L. 32

Huber, E. and Stephens, J.D. 32, 37

Hull, C.J. with Hjern, B. 366

Hunold, C. and Young, I.M. 438

Hurwitz, J. and Pefﬂey, M. 125

Husserl, Edmund 106

Huxham, C. and Vangen, V. 369, 370

Huxham, Chris 370

Ikenberry, G.J. 235, 238

Imbeau, L.M., Petry, F. and Lamari, M. 168

Imbens, G.W. 416, 425

Imbens, G.W. and Angrist, J.D. 431n17

Imbens, G.W. and Wooldridge, J.M. 430n12

Immergut, E.M. 407

Immergut, E.M. and Anderson, K.M. 36

Immergut, E.M., Anderson, K.M. and Schulze, I.

32, 37

Ingold, Karin 129, 133

Ingram, Helen 225, 365, 367

Ingram, H.M. and Mann, D.E. 23

Innes, J.E. and Booher, D.E. 105, 107, 370

in’t Veld, Roeland J. 22

Inwood, G.J. 405, 408

Iyengar, S. and Kinder, D.R. 207–8

Iyengar, Shanto 142, 193, 207

Jacob, K. et al. 246, 247, 248, 249, 250

Jacobs, S.H. 247, 248

Jacobson, H.K. 239

Jacobsson, K. 407

Jacques, Elliot 333

Jagger, Pamela 121

Jahn, D. 31

Jahn, D. and Müller-Rommel, F. 37

James, O. and Lodge, M. 406

Jameson, Frederic 103

Jänicke, M. 31

Jarvis, Darryl S.L. xiv, 59–75

Jasanoff, S. 239, 260

Jay, S. et al. 245

Jayasuriya, Kanishka 65

Jefferson, Thomas 101

Jenkins-Smith, H. and St. Clair, G. 129

Jenkins-Smith, H., St. Clair, G. and Woods, B.

128, 129

Jenkins-Smith, Hank 125, 130, 131, 134, 134n1,

255, 263

Jennings, B. 20

Jennings, W. and John, P. 168

Jennings, W. et al. 168

Jenson, J. 408, 409

Jervis, Robert 455, 476

Jessop, B. 310

Jeuland, M. and Whittington, D. 427–28

Jillson, C.C. and Wilson, R.K. 121

Jobert, B. 51

Jobert, B. and Muller, P. 49

John, P. 250

John, P. and Jennings, W. 168

John Henry effects 422, 431n15

Johnson, G.F. 442

Jones, B.D. and Bachelor, L.W. 278

Jones, B.D. and Baumgartner, F.R. 139, 141, 142,

143, 145, 146, 147, 148, 281, 282, 283, 490

Jones, B.D. and Williams, W. 278, 281

Jones, B.D. et al. 144, 147, 148, 149, 281, 283

Jones, Bryan D. xiv, 141–42, 146, 273–86

Jones, Charles O. 290

Jones, G. 310

Jones, M.D. and McBeth, M.K. 193, 195, 201

Jordan, A. and Liefferink, D. 37

Jordan, A. and Schout, A. 247

Jordan, A., Wurzel, R.K.W. and Zito, A.R. 37

Jordan,

Andrew xiv, 19,

34, 244–54

Jordan, G. and Schubert, K. 19, 48

Jordan, Meagan 144, 147, 148

Jørgensen, T.B. and Bozeman, B. 436

Jreisat, Jamil E. 342

Jupille, J. and Caporaso, J. 36

Kaelberer, M. 233

Kagan, Robert A. 22

Kahneman, D. and Tversky, A. 276

Kahneman, D., Slovic, P. and Tversky, A. 217,

219, 221

Kamenka, Eugene 333

Kamieniecki, S. 233

Kanbur, R. 416

Kaplan, Thomas J. 102, 104, 194

Karch, Andrew 144, 218, 219

Karlan, D. 415

Karlan, D. and Appel, J. 416

Karlan, D. and Zinman, J. 431n23

Karvonen, L. 405

Kassim, Hussein 52

Author index

501

Kathelene, L. and Martin, J. 440

Kato, J. 403

Katzenstein, P. 156

Kauneckis, Derek 121

Kavanagh, Steve 197

Kay, Adrian xiv, 402, 403, 456, 462–72

Keane, M.P. 416

Keck, M. and Sikkink, K. 234

Kelly, J.M. 439

Kemp, P. 464, 466, 468, 471

Kenis, P. and Schneider, V. 154

Keohane, R. and Nye, J. 46

Kerber, W. and Eckardt, M. 407

Kerr, D.H. 484

Kersbergen, K. van 37

Keynes, J.M. (and Keynesian ideas) 322, 323

Khagram, S. and Thomas, C.W. 108

Kickert, W.J.M., Klihn, E.-H. and Koppenjan,

J.F.M. 107, 299, 310, 438

Kilduff, M. and Tsai, W. 364

Kim, Doo-Rae 334

Kim, S. 370

King, C.S., Feltey, K.M. and Susel, B.O.N. 441

Kingdon, John W. 7, 18, 55, 105, 141, 143, 167,

198, 205, 206, 223, 278, 282, 402, 405, 452,

455; focusing events and policy windows

175–77, 178, 182, 184, 185; garbage can model

(GCM) and study of the policy-making process

320, 322, 323–24, 325, 326, 327; models for

research into decision-making processes 299,

300, 303, 304

Kirkpatrick, C. and Lee, N. 246

Kirkpatrick, C. and Parker, D. 246, 247

Kirkpatrick, C., Parker, D. and Zhang, Y.-F. 247

Kisby, B. 158

Kiser, L.L. and Ostrom, E. 117, 118

Kiser, L.L. and Percy, S. 91

Kitschelt, H. 32, 34, 37

Kitts, K. 491

Klein, N. 492

Klijn, E.-H. and Edelenbos, J. 367–68

Klingemann, H.-D., Hofferbert, R. and Budge, I.

173

Knill, C., Debus, M. and Heichel, S. 32

Knill, C., Schulze, K. and Tosun, J. 34

Knill, C., Tosun, J. and Heichel, S. 37

Knill, Christoph 22, 34

Knoke, D. et al. 155, 157

Knoke, David xiv, 19, 153–63

Knoke D. and Laumann, E.O. 155

Kohler-Koch, B. 239

Koliba, C. 335, 342

Kooiman, Jan 47, 361, 438

Koppenjan, J. 303

Koppenjan, J. and Klijn, E.H. 101, 363, 368, 438

Korpi, W. and Palme, J. 34, 37, 403

Kramer, M.S. and Shapiro, S.H. 420

Krane, D. and Wright, D.S. 362

Krasner, Stephen D. 60, 61, 474

Krause, G.A. and Meier, K.J. 341

Krauthammer, Charles 198

Krebs, R.R. and Jackson, P.T. 193

Kremer, M. 415

Kremer, M. et al. 415

Krislov, Samuel 337

Kristof, Nicholas 199

Krugman, Paul 198

Kübler, Daniel 134

Kuhn, Thomas 467

Kühner, S. 32

Kuklinski, J. and Quirk, P. 275

Kunda, Ziva 278

LaFollette, Robert 5

Laitin, David 276

LaLonde, R.J. 415, 416

Larsen, J.B., Vrangnaek, K. and Traulsen, J.M.

132

Larson, Stephanie 207

Lascoumes, P. and Le Galès, P. 245

Lasswell, Harold D. 17, 105, 207, 222, 245, 250

Latour, Bruno 100, 106, 107

Laudan, Larry 126

Laumann, Edward 155

Laumann, E.O. and Knoke, D. 48, 155, 157

Laumann, E.O. and Pappi, F.U. 156–57

Laumann, E.O., Galaskiewicz, J. and Marsden,

P.V. 157

Law, John 106

Lawlor, Andrea xv, 204–14

Lawlor, E. 256, 265

Lawrence, Regina G. 178, 181, 184

Lawrence, R.G. and Birkland, T.A. 175, 181

Laws, D. 355

Lax, D.A. and Sebenius, J.K. 359n1

Le Gales, P. and Thatcher, M. 48

Lee, N. 248

Lee, N. and Kirkpatrick, C. 248

Leech, B.L. et al. 282

Legaspi, Perla E. 65

Lehmbruch, G. 157

Lejano, R., Ingram, H. and Ingram, M. 103, 107

Lejano, Raul Perez xv, 98–112

Lemer, D. and Lasswell, H.D. 125

Lenschow, A., Liefferink, D. and Veenman, S. 34

Lerner, D. and Lasswell, H.D. 29, 257

Levi-Faur, David 64, 65, 67

Levitt, B. and March, J.G. 404, 405, 406, 407

Lewicki, R.J. and Gray, B. 352

Lewis, O. and Steinmo, S. 451

Lichter, S. Robert 208

Lichter, S.R. and Rothman, S. 208

Lidskog, R. and Sundqvist, G. 239

Limbaugh, Rush 200

Author index

502

Lindahl, Erik 79

Lindblom, C.E. and Woodhouse, E.J. 301

Lindblom, Charles E. 3, 21, 61, 104, 125, 139,

140, 250, 277, 287, 288, 290, 292–93, 294,

295–96, 297, 335, 405, 478

Linder, S.H. and Peters, B.G. 18, 217

Lipsky, Michael 7, 107, 339

Liu, X.S., Lindquist, E. and Vedlitz, A. 178

Lock, Karen 9

Lodge, M. and Taber, C. 278

Lohmann, Susanne 282

Loomis, Burdett A. 456

Lord, C., Ross, L. and Lepper, M. 126

Lord, W.B. 353

Lorenz-Meyer, Dagmar 233

Loughlin, John 362

Lovell, H. and MacKenzie, D. 235

Lowi, T.J. 402, 438

Lowry, William 185

Lubensky, R. 443

Lucas, Robert E. 85

Lukensmeyer, C.J., Goldman, J. and Brigham, S.

442

Lundquist, Lennart 59, 459

Lupia, A., McCubbins, M.D. and Popkin., S.L.

275, 276

Lynch, J. 32

Lynn, Lawrence E. Jr. 255, 341, 365

Lyotard, Jean-François 100, 103, 104

McCain, John 221–22

McChesney, Robert 207

McComas, K.A. 440

McComas, K.A. and Shanahan, J. 208

McComas, K.A., Besley, J.C. and Black, L.W. 441

McCombs, M. 206

McCombs, M. and Shaw, D.L. 205

McCombs, M., Shaw, D.L. and Weaver, D. 206

McConnell, Allan xv, 484–94

McCraw, Thomas K. 5

McCright, A.M. and Dunlap, R.E. 208

McDonald Commission in Canada 408

McGann, J. and Johnson, E. 6

McGann, J. and Sabatini, R. 157, 159–60

McGinnis, M.D. and Ostrom, E. 121

McGinnis, Michael D. xv, 87–97

McGoldrick, D.E. and Boonn, A.V. 22

McGuire, Michael xv, 361–73

McGuire, T., Coiner, M. and Spancake, L. 84

Mackay, Charles 282

McKelvey, Richard D. 79

Mackintosh, J.P. 310

McLendon, Michael K. 326

McLennan, Gregor 61

McQuail, D. 206

MacRae, D. and Whittington, D. 255, 259

MacRae Jr., Duncan 19

Madison, James 90

Mahoney, J. and Thelen, K. 452, 453, 475

Mahoney, James 452, 474, 475–76, 477

Mahwinney, Hanne B. 454

Majone, Giandomenico 24, 63–65, 65–66, 76,

127, 178, 190, 250, 435, 436, 440, 490

Mandell, Myrna P. 366

Mannheim, Karl 98

Mansbridge, J.J. 351, 355

Maoz, Z. 154

March, J. and Simon, H. 105

March, James A. 274, 275, 276

March, J.G. 404, 405

March, J.G. and Herbert, A.S. 332

March, J.G. and Olsen, J.P. 61, 106, 303

Marier, Patrik xv, 401–14, 439

Marin, B. and Mayntz, R. 19, 299

Markoff, J. and Montecinos, V. 405

Marmor, Theodore R. 337

Marquez, Lilian 121

Marsh, D. and McConnell, A. 484

Marsh, D. and Rhodes, R.A.W. 157–58, 160, 341

Marsh, D. and Sharman, J.C. 34, 404, 406, 409

Marsh, D. and Smith, M. 158, 364

Marsh, D. et al. 160

Marsh, D., Lewis, C. and Fawcett., P. 440

Marsh, David 156–57, 341

Martin, J. 32, 103

Martinez, J., Susskind, L.E. et al. 350, 356

Mashaw, J. and Marmor, T.R. 339

Mashaw, Jerry 339

Mason, R. and Mitroff, I. 261

Matthews, D.R. and Stimson, J.A. 282

Matti, S. and Sandström, A. 129

May, P., Sapotichne, J. and Workman, S. 185,

490

May, Peter 176, 177, 178, 404, 491

Mayer, Igor S. xv, 255–69

Mayer, Robert N. 205

Mayntz, Renate 23, 29

Mazmanian, D. and Sabatier, P. 125, 379

Mazur, A. and Lee, L. 208

Mazur, Allan 125, 208

Meier, K.J. 334

Meijerink, S. 234

Melis,

T.S. and Pulwarty,

R.S. 184

Melo, M.A. 233

Menard, C. and Shirley, M.M. 116

Merriam, Charles 274

Merton, Robert K. 282, 338

Meseguer, C. 35

Mettler, S. and Soss, J. 222

Mieszkowski, P. and Zodrow, G.R. 81

Migué, J-L., Bélanger, G. and Niskanen, W.A. 84

Miguel, E. and Kremer, M. 415

Mikkelson, M. 159

Miliband, Ed 196

Author index

503

Miliband, Ralph 295

Mill, John Stuart 101

Miller, G.J. 280

Miller, G.J. and Moe, T.M. 84

Miller, J.H. and Page, S.E. 282

Miller, S.M., Evers, A. and Winterberger, H. 208

Milleron, J.C. 79

Mills, C. Wright 191

Milward, H.B. and Provan, K.G. 367

Minicucci, Daniela 197

Minogue, Martin 67

Mintrom, M. and Vergari, S. 178

Mintrom, Michael xv, 3–16, 20, 178, 405, 407

Mintzberg, H. 299, 300

Mintzberg, H., Raisinghani, D. and Therot, A.

303

Miser, H.J. and Quade, E.S. 255, 259

Mitchell, N.J. et al. 232

Mitchell, William C. 79, 85, 93

Mnookin, R. 353

Molotch, H. and Lester, M. 177, 184

Mommsen, Wolfgang J. 335

Montjoy, R.S. and Watson, D.J. 334

Montpetit, Eric 127, 134

Mookherjee, D. 416

Mooney, Christopher Z. 13, 207

Moore, Mark Harrison 341, 369, 436

Morgan, G. 300

Morgan, S.L. and Winship, C. 430n6

Mortensen, P.B. et al. 168, 169

Mortensen, Peter B. xv, 167–74

Mosher, Frederick C. 334, 337

Mosley, P. and Dowler, E. 381

Mouritzen, Hans 339

Mowles, C. 377, 378, 381

Moyer, H.W. and Josling, T.E. 479

Mucciaroni, Gary xv–xvi, 320–28

Mueller, Dennis C. 77, 78–79, 80–81, 82, 83, 84

Muldoon, J.P., Aviel, J.F. and Reita, R. 339

Munger, Michael C. 10

Munro, G.D. et al. 126

Munro, John 129

Murray, D., Schwartz, J. and Lichter, S.R. 208

Murray, G. and Pacheco, D. 8

Mwangi, Esther 121

Myers, Samuel L. 9

Myint, Tun 121

Myles, J. and Pierson, P. 403

Nachmais, David 23

Nakamura, Robert T. 175

Nedergaard, P. 407

Nelson, Barbara 13

Ness, B. et al. 248

Nettl, J.P. 60

Neumann, J. van and Morgenstern, O. 98

Newell, A. and Simon, H.A. 277, 278, 283

Newell, Allen 274, 275, 277, 283

Niemann, A. 239

Niemeyer, S. 438

Nilsson, M. 249

Nilsson, M. et al. 246, 248

Nisbet, M.C. and Myers, T. 208

Nisbet, Robert 453

Niskanen, William A. 7, 8, 78, 84, 331, 340

Noakes, J.A. and Johnston, H. 192

Nohrstedt, D. and Weible, C.M. 126, 132

Nohrstedt, Daniel xvi, 125–37, 177

Nooteboom, B. 467

North, Douglass C. 67, 115, 116, 121, 462, 466,

469

Northrup, T.A. 353

Nussbaum, Martha 103

Nutley, S.M., Walter, I. and Davies, H.T.O. 250

Oakerson, R.J. 88, 90, 95n4

Oakerson, R.J. and Parks, R.B. 121

Oakley, M.R. 182

O’Faircheallaigh 434

Ogden, J., Walt, G. and Lush, L. 235

O’Leary, R. and Bingham, L.B. 341

Oliver, W. Hugh 7

Olsen, Johan P. 335, 408

Olsen, J.P. and Peters, B.G. 24, 337

Olsen, Mancur J. 76, 78, 295, 438

Orenstein, M.A. 404, 405

Orren, K. and Skowronek, S. 61, 62

Osborne, D. and Gaebler, T. 7, 340, 439

Ospina, S.M. and Dodge, J. 103

Ostrom, E. and Crawford, S. 119

Ostrom, E. and Ostrom, V. 87, 88, 91, 92,

94–95n1, 95n4

Ostrom, E., Gardner, R. and Walker, J. 95n4

Ostrom, E., Schroeder, L. and Wynne, S. 88, 121

Ostrom, E., Walker, J. and Gardner, R. 82

Ostrom, Elinor 60, 76, 82–83, 87, 88, 89, 91, 93,

95n4, 115–17, 117–18,

119–20, 121, 126,

299,

452, 463

Ostrom, V., Bish, R. and Ostrom, E. 88

Ostrom, V., Tiebout, C.M. and Warren, R. 89,

95n2

Ostrom, Vincent 89, 90, 94, 95n2. 95n3, 95n5,

95n7, 121

O’Toole, Laurence J. 361, 363, 366

Ouchi, William G. 335

Overbye, E. 34

Owens, S., Rayner, T. and Bina, O. 245, 249,

250

Oyerinde, Oyebade Kunle 121

Pacheco, Diego 120

Padgett, John F. 147, 283

Page, B.I. and Shapiro, R.Y. 172

Page, C. and Jenkins, B. 339

Author index

504

Page, S. 466, 471

Palin, Sarah 101

Palumbo, D.J. 250

Pappi, Franz 155

Pareto, Vilfredo 274–75

Park, Y.H. 409

Parkinson, J. 436, 437, 442

Parks, R.B. 91, 121

Parks, R.B. and Oakerson, R.J. 91

Parsons, Wayne 336

Patriotta, Gerardo 103

Patterson, M. and Monroe, K.R. 193

Patterson, Thomas E. 207

Patton, C. and Sawicki, D. 255, 259

Patton, Michael Quinn 9

Pawson, R. and Tilley, N. 431n24

Pawson, Ray 20, 108, 245, 250

Pefﬂey, M. and Hurwitz, J. 125, 126

Peltzman, Sam 85

Pemberton, H. 403, 464, 466, 468, 471

Pena, Jim 221

Pentland, Brian 107

Perl, A. and Dunn Jr., J.A. 45

Perl, A. and White, D. 50

Perl, Anthony xvi, 44–56

Perri 12

Perrow, Charles 179

Persha, Lauren 120

Persson, Torsten 85

Peters, B.G. and Van Nispen, F.K.M. 22

Peters, B.G., Pierre, J. and King, D.S. 403

Peters, B. Guy 333, 334, 335, 336, 337, 338, 339,

363

Peterson, J. 238

Peterson, M.J. 235

Picard, L.A. and Buss, T.F. 374

Pierce, Jonathan J. 129

Pierson, Paul 24, 32, 34, 284n3, 323, 402, 403,

452, 464, 466, 473, 474, 475, 476, 478, 480,

481n1

Plümper, T. and Neumayer, E. 34

Pollit, Christopher 456

Pollitt, C. and Bouckaert, G. 340

Poocharoen, Ora-Orn xvi, 331–46

Popper, Karl 104, 295

Poteete, A., Janssen, M. and Ostrom, E. 95n4

Power, Nina 196

Pralle, Sarah 144, 146

Prasser, S. 491

Premfors, R. 408

Pressman, Jeffrey L. 365

Pressman, J.L. and Wildavsky, A. 7, 106, 125,

127, 339, 365, 379

Price, V. and Tewksbury, D. 206

Prince, M.J. 182

Prince, R. 233

Pritchard, David 205

Pritchett, L. and Woolcock, M. 375

Pross, A. Paul 19

Provan, K.G. and Kenis, P. 161, 341

Przeworski, Adam 59, 60, 61

Putnam, R. and Leonardi, R. 121

Putnam, Robert D. 46, 53, 125, 126, 437

Quade, Edward S. 248, 259, 307

Quattrone, G.A. and Tversky, A. 126

Raadschelders, J.C.N. 462, 465

Rabin, J. and Bowman, J.S. 334

Rabinow, P. and Sullivan, W. 103

Radaelli, C.M. 35, 230, 238, 244, 246, 247, 249,

250, 404, 405, 406

Radaelli, C.M. and Connor, K. 232

Radaelli, C.M. and Meuwese, A.C.M. 249

Radin, B.A. et al. 363

Radin, Beryl A. 3, 255

Rae, Douglas W. 79

Ragin, Charles 455

Raiffa, H. 357, 359n1

Raiffa, H., Richardson, J. and Metcalfe, D. 357

Ramseyer, J.M. and Rosenbluth, F.M. 337

Ranney, Austin 125

Rapley, John 99

Rawls, John 83

Ray, James Lee 59

Rayner, S. 245

Razin, A. and Sadka, E. 62

Reddel, T. and Woolcock, G. 438, 440

Redford, Emmette S. 172

Rein, M. and Schön, D. 192, 193, 359n8

Renda, A. 248, 249

Renn, O., Webler, T. and Wiedemann, P. 437

Repetto, Robert 144

Rethel, L. 235

Rethemeyer, R.K. 160

Rhoads, Steven E. 7

Rhodes, R.A.W. 47, 62, 157, 158, 299, 310, 341,

362, 363

Rhodes, R.A.W. and Dunleavy, P. 310

Rhodes, R.A.W. and Marsh, D. 363, 474, 477

Richardson, Jeremy 157

Richardson, J.J. and Jordan, A.G. 477

Richardson, T. 248

Ricoeur, Paul 102

Riggs, Fred W. 338

Riker, William 76, 78, 80

Riper, Paul P. Van 334

Risse-Kappen, T. 52

Rittel, H. and Webber, M. 99, 303

Robbins, M.D., Simonsen, B. and Feldman., B.

442

Roberts, N. 439, 441

Robertson, D.B. 405, 410

Robinson, Scott E. 140, 144

Author index

505

Robinson, S.E. and Caver, F.R. 147

Robinson, S.E. and Eller, W.S. 327

Robinson, S.E. and Meier, K.J. 140

Robyn, Dorothy 324

Rochefort, D.A. and Cobb, R.W. 191

Rochefort, D.A., Rosenberg, M. and White, D.

201

Rochefort, David A. xvi, 189–203

Rodrik, D. 424, 426

Roe, Emery 102, 103

Roederer-Rynning, C. and Daugbjerg, C. 478

Romney, Mitt 200

Roosevelt, F.D.R. 279

Rose, R. and Davies, P.L. 468

Rose, Richard 24, 35, 36, 323, 402, 404, 405,

406, 473

Rosenau, Pauline V. 207, 310

Ross, Marc H. 194–95, 402, 403, 410

Rossi, P.H., Lipsey, M.W. and Freeman, H.E.

250

Roth, C. 233

Rothstein, B. 402

Rothstein, Robert L. 290–91

Rourke, Francis E. 337

Rowley, C.K. and Schneider, F. 77

Rudd, Kevin 488–89

Ruggie, J.G. 405

Russel, D. and Jordan, A. 245, 248

Rutgers, Mark R. 334

Ryan, Marie-Laure 103

Ryan, Paul 200

Saari, Donald G. 80

Sabatier, P.A. and Brasher, A.M. 129

Sabatier, P.A. and Jenkins-Smith, H.C. 48, 55,

107, 125, 126, 129, 130, 131, 132, 134n1, 195,

452

Sabatier, P.A. and Mazmanian, D. 31

Sabatier, P.A. and Weible, C.M. 126, 128, 129,

130, 132, 134n1, 359n7, 452

Sabatier, P.A., Hunter, S. and McLaughlin, S. 126

Sabatier, Paul A. 17, 19, 35, 120, 125, 126, 130,

134, 134n1, 175, 234, 235, 301, 307, 325, 453

Sabet, Daniel M. 121

Sacks, Jonathan 197

Saint-Martin, Denis 340

Saint-Onge, H. and Armstrong, C. 361

Salamon, Lester M. 22, 363

Saleth, R.M. and Dinar, A. 120

Salter, L. 438–39, 440

Sanderson, I, 20, 245, 250

Sandler, Todd 13

Sandström, A. and Carlsson, L. 160

Sato, H. 302

Satterthwaite, Mark A. 79

de Saussure, Ferdinand 100

Savas, E.S. 7

Sawyer, A. 121

Scharpf, Fritz W. 36, 48, 299, 303, 305, 363, 365

Scharpf, F.W., Reissert, B. and Schnabel, F. 304

Schattschneider, Elmer, E. 141, 143, 145, 167,

172, 175, 177, 178, 179, 190, 402, 491

Schelling, Thomas C. 98, 282

Scheufele, D.A. and Tewksbury, D. 206

Schmidt, Manfred G. 168

Schmidt, V.A. 230

Schmidt, Vivien A. 19

Schmitt, Sophie xvi, 29–43

Schneider, A. and Ingram, H. 102, 105, 143, 178,

179. 192, 217, 218, 219, 222, 223, 224, 225–26

Schneider, A.L. and Ingram, H. 437

Schneider, Anne L. xvi, 144, 217–28

Schneider, Saundra 179

Schoﬁeld, Norman 79

Schön, D.A. and Rein, M. 12, 49, 55, 102, 105,

193, 301, 352, 353, 359n8

Schön, Donald A. 11, 218

Schot, J. and Schipper, F. 235

Schout, A. and Jordan, A. 245

Schout, A., Jordan, A. and Twena, M. 101

Schrad, Mark Lawrence 278, 281

Schram, S. and Neisser, P.T. 103

Schröder, Gerhard 492

Schulman, Paul R. 19

Schwarz, Herman 466

Scotchmer, S. 81

Scott, Colin 63

Scott, James C. 99

Scott, W. Richard 332

Scruggs, L. 31, 32

Sebenius, James K. 234

Seeberg, Henrik 169

Seldon, Sally Coleman 337

Self, Peter 9

Selznick, Philip 99

Sen, Amartya K. 375

Sewell, William H. 476, 477

Shafritz, J.M. and Hyde, A. 332

Shafritz, J.M. and Ott, J.S. 332

Shanahan, E., Jones, M.D. and McBeth, M.K.

127, 134, 195

Shankland, A. and Hasenclever, L. 235

Shapiro, I. and Bedi, S. 491

Shapiro, Ian 13

Sharman, J.C. 490

Sheriff, P.E. 408

Shiffman, J., Beer, T. and Wu, Y. 144

Shiller, R.J. 282

Shipan, C.R. and Lowry, W.R. 37

Shipan, C.R. and Volden, C. 404

Shughart, W.F. and Razzolini, L. 77, 78

Shughart, William F. 77, 81

Shulock, N. 255

Siegel, J. 161

Author index

506

Silvia, C. and McGuire, M. 367

Silvia, Chris xvi, 361–73

Simek, J., Zamykalova, L. and Mesanyova, M. 233

Simmons, B.A. and Elkins, Z. 406

Simon, Herbert A. 20, 93, 118, 125, 126, 245,

250, 274–76, 277, 279, 283

Skelcher, Chris 363

Skinner, B.F. 279–80

Skocpol, T. and Weir, M. 322

Skocpol, Theda 61–62, 402, 466

Sköldberg, Kaj 103

Skowronek, Stephen 61, 362

Slovak, Paul 276

Smith, Adrian 48, 133

Smith, G. and May, D. 21

Smith, Gilbert 20, 439

Smith, James A. 6

Smith, Mark A. 145

Smith, Martin J. 19, 299, 310, 334

Smith, S.R. and Lipsky, M. 362

Snider, J.H. 440, 441

Sniderman, P.M., Brody, R.A. and Tetlock, P.E.

207

Somaiya, Ravi 196

Sommerer, T. 37

Sorenson, E. and Torﬁng, J. 363

Sornette, D. 282

Soroka, Stuart N. xvi, 172, 204–14

Sowell, Thomas 292, 293, 295

Spector, M. and Kitsuse, J.I. 207

Speers, K. 444

Spicer, Michael 292

Stanley, N. and Manthorpe, J. 491

Starke, P. 34

Starke, P., Obinger, H. and Castles, F.G. 34

Staronova, K. 246

Steed, B. 120

Stephens, J.C. et al. 235

Stephens, J.D., Huber, E. and Ray, L. 32

Stevens, Joe B. 9

Stevens, W.K. 208

Stewart, J. 435, 444

Stigler, George 85

Stillman, Peter G. 23

Stimson, James 220

Stinchcombe, Arthur L. 466

Stokey, E. and Zeckhauser, R. 10

Stone, D. and Denham, A. 6

Stone, D. and Garnett, M. 407

Stone, Deborah A. 11, 102, 103, 178, 179, 181,

191, 194, 196, 218, 223, 490

Stone, Diane L. 6, 35, 37

Strand, R. 299

Strange, Susan 62, 63, 64

Streeck, W. and Schmitter, P. 54

Streeck, W. and Thelen, K.A. 36, 452, 469, 475

Sulkin, Tracy 171

Sundquist, James L. 364–65

Sunstein, C.R. 353

Surowiecki, James 282

Susskind, L., Thomas-Larmer, J. and McKearnen,

S. 105

Susskind, L.E. 359n9

Susskind, L.E. and Cruikshank, J.L. 359n9

Susskind, L.E. and Field, P. 353

Susskind, L.E. and Thomas-Larmer, J. 354

Svara, James H. 333, 334, 337

Sveiby, K.-E. 368

Swank, D. 32, 36, 37

Swidler, Ann 51

Sykes, R., Palier, B. and Prior, P.M. 37

Tabb, William K. 13

Tabellini, Guido 85

Tamborra, M. 248

Tavits, M. 406

Taylor, M. 79

Taylor-Gooby, P. 37

Teisman, Geert R. xvii, 299–319

Teisman, G.R. and Klijn, E.H. 299

Teisman, G.R., van Buuren, A. and Gerrits, L.

299, 300

Termeer, C. 309

Tetlock, Philip 276

Tews, K., Buschand, P.-O. and Jörgens, H. 37

Thaler, Richard H. 276

Thatcher, M. 236

Thelen, K. and Steinmo, S. 474

Thelen, Kathleen 468, 474–75

Therborn, Göran 60

Thiel, A. 248

Thiel, A. and König, B. 249

Thomas, A. 374

Thomas, C.W. 233

Thomas III, Herschel F. xvii, 273–86

Thompson, Grahame F. 363

Thompson, J.A. 334

Thomson, A.M., Perry, J.L. and Miller, T.K. 369

Throgmorton, J.A. 105

Tiebout, C.M. 81, 82

Tilly, Charles 59, 457

Tocqueville, Alexis de 90

Todd, P.E. and Wolpin, K.I. 415

Toke, D. 233

Toonen, Theo A.J. 89

Topal, C. 441

Torgerson, Douglas 20, 194

Torriti, J. 248, 249

Tovey, Craig A. 80

Tribe, L.H., Schelling, C.S. and Voss, J. 353

True, J. and Minstrom, M. 9

True, J. and Utter, G. 144

True, J., Jones, B. and Baumgartner, F. 140, 142,

147

Author index

507

Trumbo, Craig 208

Tsebelis, George 36, 140, 407, 409

Tullock, Gordon 76, 81

Tuohy, Caroline 24

Tupper, A. and Doern, G.B. 22

Turnpenny, J. et al. 245, 247–48, 249

Turnpenny, John xvii, 244–54

Twaalfhoven, P. 263

Tyran, J.-R. and Sausgruber, R. 34

Urwick, Lyndall 375

Urzua, S. and Heckman, J.J. 431n17

van Bueren, A. et al. 300, 308

van Buuren, A. and Gerrits, L. 299, 301, 307, 308

van Buuren, Arwin xvii, 299–319

van Creveld, Martin 62

van Daalen, Els xvii, 255–69

van de Riet, O. 257

Van den Ven, Andrew H. 134

van Eeten, M. 257, 259

van Herk, S. et al. 301

van Laerhoven, Frank 121

Van Slyke, David 363

van Waarden, F. and Drahos, M. 232, 233

van Waarden, Frans 19, 47, 48

Vaughan, S.K. and Arsneault, S. 182

Vedung, Evert xvii, 22, 387–400

Verdun, A. 238

Verhulst, J. and Walgrave, S. 184

Vickers, Geoffrey 307

Vis, B. 32, 37

Vis, B. and Kersbergen, K. van 37

Voets, J., Van Dooren, W. and de Rynck, F. 160

Vogel, David 24

Volden, C., Ting, M.M. and Carpenter, D.P. 34

von Winter, T. 159

Wade, Robert 337

Wagenaar, H. 299

Walgrave, S. and Van Aelst, P. 205

Walgrave, S. and Varone, F. 141

Walgrave, S. and Verhulst, J. 178

Walgrave, S., Soroka, S. and Nuytenams, M. 206

Walker, Jack L. 125, 175, 217, 404, 406

Walker, K.J. 239

Walker, W. 257

Wälti, S. 31

Walton, Jeff 197

Wang, Jianxun 120

Wanna, J. 437

Warren, M. and Pearse, H. 442

Watts, Ronald L. 362

Weaver, K. 403, 409

Weaver, R.K. 402, 407, 491

Weber, Max 47, 60, 99, 154, 156, 331, 332, 335,

342

Weible, Christopher xvii, 125–37

Weible, C.M. and Sabatier, P.A. 129, 161

Weible, C.M. et al. 128, 131–32, 134n1, 236

Weible, C.M., Sabatier, P.A. and McQueen., K.

454

Weick, K.E. 103, 106

Weidner, H. and Jänicke, M. 37

Weimer, D.L. and Vining, A.R. 7, 10, 255,

256

Weir, M. and Skocpol, T. 402

Weiss, A. and Woodhouse, E. 293, 294, 295, 296,

297

Weiss, C. and Bucuvalas, M. 259

Weiss, Carol H. 9, 125, 235–36, 250

Weiss, Linda 63

Wenger, E. 234, 368

Wesselink, A. and Warner, J. 102

West, William F. 338

Westney, D.E. 406

Westoby, Adam 335

Weyland, K. 404, 406

Wheare, K.C. 407

Whitaker, G. 91

White, Leonard D. 255, 333

Whiteford, P. 32

Whitford, Andy xvii–xviii, 76–86

Whittington, D. et al. 427–28, 431n25

Whittington, Dale xviii, 415–33

Wildavsky, A., Swedlow, B. and White, J. 336

Wildavsky, Aaron 3, 6, 140, 255, 260, 277, 289,

293, 336, 455

Wilensky, H.L. 32

Wilkerson, John 146

Wilkins, L. and Patterson, P. 208

Wilkinson, D. et al. 248

Williams, Claire xviii, 3–16

Williams,

Rowan 197

Williams, Russell

A. 454

Williams, Walter W. 365

Williamson, A. and Fung, A. 441, 442, 443,

444

Williamson, John 7

Williamson, Oliver E. 82, 117, 120

Wilsford, David 459, 464, 471

Wilson, James Q. 332, 340

Wilson, Woodrow 333

Windhoff-Héritier, A. 29, 31

Witte, J.M., Reinicke, W.H. and Benner, T.

157

Wittgenstein, Ludwig, treatise on logic 100

Wofford, Harris 184–85

Wolf Jr, Charles 7

Wollmann, H. 250

Wood, Christopher 9

Wood, Robert S. 144, 145, 178

Woodside, K. 22

Woolcock, M. 376

Author index

508

Workman, S., Jones, B. and Jochim, A. 139, 140,

141, 280

Wright, K. 231, 232

Wu, Chen-Yu xviii, 153–63

Wu, X. et al. 22, 336

Yagade, A. and Dozier, D.M. 206

Yang, K. and Callahan, K. 444

Yanow, Dvora 102, 103, 104

Yergin, D. and Stanislaw, J. 7

Youde, J. 234

Young, Lori xviii, 204–14

Youssefmir et al. 282

Zafonte, M. and Sabatier, P.A. 128, 129, 130

Zahariadis, Nikolaos 167, 320, 323, 452;

garbage can model (GCM), work with

325–26

Zelikow P. 232

Zhong, L.J. and Mol, A.P.J. 440

Zito, A.R. 233

Zucker, H.G. 206

Author index

509

Subject index

action situations and arenas 117–18

actor promoted events (APEs) 184

adaptation, principle of 275

Administrative Behavior (Simon, H.A.) 274

administrative capacity, effectiveness and 245

administrative systems and policy outputs 335–36

Advocacy Coalition Framework (ACF) 125–34;

advocacy coalitions 127–30; aggregation of

actors 126; assumptions 126–27; belief

subsystems 126; belief systems, hierarchy and

stability of 129; bounded rationality 126;

bounded rationality, public policy

decision-making and 273; change, policy

dynamics and 452, 454, 455, 456, 457, 458;

coalitions, formulation of 128–29; coalitions,

level of conﬂict between 130–31; conceptual

categories and relations 127; development of

125; “devil shift” 126; disagreement and

alternative dispute resolution in policy process

352; focusing events and policy windows 175,

186; framework 126–27; hypotheses building

and testing 129, 131, 133; implicit theories and

assumptions 126–27; individuals, attributes of

131; interdependencies 133; international

dimensions and dynamics of policy-making 48;

issue networks 125; long-term time perspective

127; momentum of the ACF research program

134; organizations, distinction between

different types of 129; policy brokers 131;

policy change 125, 127, 129, 132–33, 133, 134;

policy network models 161; policy-oriented

learning 125, 130–32, 133; policy process

research 125; policy subsystems 134; policy

subsystems, composition of 128–29; policy

subsystems, policy processes and 126; political

behavior of advocacy coalitions 129–30;

scientiﬁc and technical information 127; scope

127; stimuli, attributes of 131; stimuli,

processing of 126; subsystem actors 126;

theoretical emphasis 133–34; translated beliefs

126–27

agenda-setting and political discourse 189–201;

agenda building, process of 190–91;

agenda-policy hypothesis 170–71; agenda

prioritization 193; agenda-setting perspective,

questions central to 189; argumentative turn in

policy studies 190; audience reaction, evocation

of 192–93; command issues 189; counting,

language of 191; description, dimensions of

190–91; discussion 200–201; focusing events as

aspects of agenda-setting 175–77; frame analysis

193; framing 190, 192–93; inﬂuential and

powerful interests 189; issue attentiveness,

agenda-setting and 205–6; metaphor 191;

narrative, explanatory function of 194–95;

narrative,

language and 190,

193, 194–95;

noncontradictory argumentation 193; Occupy

Wall Street, case of 195, 198–200; policy

analysis and agenda setting 5–6; policy design

192; preference ranking 192; problem

deﬁnition 190–92; problem description,

anatomy of 191; problem handling 191;

problem ownership 191–92; riots in UK, case

of 195–98; September 11th, narrative

explanation of 194–95; story lines within public

policy discourse 194

Agendas and Instability in American Politics

(Baumgartner, F. and Jones, B.D.) 149–50,

150n6, 167

aggregation: of actors in ACF 126; of harms 176;

rules, IAD and 119

alignment of streams in GCM 323–24, 325–26

ambiguities, dealing with 190, 199, 200, 326, 389,

486, 492

analytical styles 255–67; archetypal styles,

deﬁnition of 261–62; argumentative style 259;

arguments and values, clariﬁcation of 257, 258,

264; client advice style 260; combining

activities 262–63; democratization 257, 258,

264; design and recommendation 256–57, 258,

264; design of policy analysis 266–67; diversity

of, untangling and explaining 255–56;

510

evaluation criteria 263; evaluation of policy

analysis 267; interacting activities 256–58, 267;

interactive style 261; mediation 257, 258, 264;

participatory style 260; perspectives on ﬁeld of

policy analysis 265–67; perspectives on policy

analysis, framework for understanding and

design 255; policy analysis styles 258–62; policy

analyst, role of 263–65; process style 260–61;

rational style 258–59; reﬂection on policy

analysis 265–66; research and analysis 256, 258,

264; research approach 256; strategic advice,

provision of 257, 258, 264; values underlying

263–65

antecedent conditions in GCM 323

anti-rationalist criticisms of incrementalism

294–96

anticipated side-effects 391

applications of: Institutional Analysis and

Development (IAD) framework 120–21; policy

analysis 9–10

appraisal: processes as sites of political behavior

245; systems for, research on design of 249; see

also policy appraisal

archetypal analytical styles, deﬁnition of 261–62

argumentative style in analysis 259

argumentative tradition in evaluation 435

argumentative turn in policy studies 190

assessment of operation of policy appraisal 248,

249–50; see also policy appraisal

Association for Public Policy Analysis and

Management (APPAM) 10

asymmetrical power 154

attention: attention-driven choice in public policy

280–81; dynamics in executive speeches 168,

169; public opinion and policy change, linking

of 170–72, 173

audience reaction, evocation of 192–93

authority rules in IAD 119

availability in policy design and transfer 219–20

behavioral theory of choice 274, 276–77

behavioral theory of organization 278–79

belief subsystems in ACF 126

belief systems, hierarchy and stability of 129

benchmarks, multiplicity of 485–86

beneﬁciaries, empowerment of 402

beneﬁt-cost analysis 5

best alternative to a negotiated agreement

(BATNA) 349, 356, 357–58

Betuwe railway line: events in decision-making

on 314–15;

fourfold analysis of 310–16; phase

model in

analysis of 311; rounds model in

analysis of 312–13; stream model in analysis of

311–12; tracks model in analysis of 313

bias: biased results, risk of 34, 37; cognitive biases

and decision heuristics 219–22; contribution

bias 220; mobilization of 491; randomization

bias 421–22; selection bias 418, 419, 420,

422–23, 424, 431n18; substitution bias 421–22

boundary rules in IAD 119

bounded rationality: advocacy coalition

framework (ACF) 126; genesis and

development of 273–74; institutional analysis

and political economy 92; policy-making

process 20–21

bounded rationality and public policy

decision-making 273–84; adaptation, principle

of 275; Administrative Behavior (Simon, H.A.)

274; Advocacy Coalition Framework 273;

attention-driven choice in public policy

280–81; behavioral theory of choice 274,

276–77; behavioral theory of organization

278–79; bounded rationality, genesis and

development of 273–74; bounded rationality,

policymaking systems and 283–84; bounded

rationality, stochastic processes and 282–83;

bounded rationality, tenets of 274–76; heuristics

277–78; human choice, model of 273, 277;

human goals, pursuit of 273;

information-processing 279–80; Institutional

Analysis and Development Framework 273;

institutional friction, disproportionate

information-processing and 281; intended

rationality, principle of 274–75;

know-recommend strategy 278; lexicographic

strategy 276; Mind and Society (Pareto, V.)

274–75; motivated reasoning 278; organizations

278–79; positive feedback, disproportionate

information-processing and 281–82; problem

from solution spaces, separation of 277–78;

Punctuated Equilibrium Theory 273; trade-offs,

principle of 276; uncertainty, principle of

275–76; “wisdom of crowds” argument 282

Breaking through the Bureaucracy (Barzelay, M.) 335

building blocks of governance 367–68

bureaucracy and policy process 331–42;

administrative

systems and policy

outputs

335–36; Breaking through the Bureaucracy

(Barzelay, M.) 335; Our Wonderland of

Bureaucracy (Beck, J.M.) 335; bureaucracy of

small island states 340; bureaucrat bashing 84;

bureaucratic organization, creation of 402;

bureaucratic power and dilemma for democracy

334; bureaucratic structure, policy and 338–39;

co-production and co-creation 341; deﬁnitions

of bureaucracy 331–32; democratic control and

bureaucratic autonomy 334; future studies 342;

governance and bureaucracy 340–42;

government, bureaucracy as part of 333–35;

institutional structure of 333; international

policy-making 338–39; jurisdictional areas

332–33; knowledge-based power 336;

leadership role, bureaucracy in 337;

legal-rational foundation of bureaucracy

Subject index

511

332–33; lobbying 337; The Managerial

Revolution (Burnham, J.) 335; networks and

collaborations, notion of 341; New Public

Management (NPM) movement 338, 340–41;

organizational form, bureaucracy as 332–33;

policy implementers, bureaucrats as 339–40;

policy-makers, bureaucrats as 336–39; political

control of bureaucracy 249; political systems

and decision-making processes 334, 335–36,

337–38; political systems and economic

systems, interdependencies of 334–35;

politics-administration dichotomy debate

333–34; The Politics of Bureaucracy (Peters, B.G.)

335–36; public managers and public value 341;

Reinventing Government (Osborne, D. and

Gaebler, T.) 340; role of bureaucracy in policy

process 335–36; social institution of 333;

street-level bureaucrats 339; super

bureaucracies 337

business advantage in resources 296

calculation, learning as 239

capitalism: capital movements, liberalization of 46;

capitalist societies 296

causality: causal explanations, research methods

and 36–38; causal mechanisms 426–27, 452,

455, 456, 457–59; deterministic causal patterns

476; gold standard in causal identiﬁcation,

promotion of RCTs as 417–20; historical

causality, lack in GCM of 322–23; problem for

policy dynamics and change 454–55; social

causation 463

central agencies, strength of 7–8

change, policy dynamics and 451–60; Advocacy

Coalition Framework (ACF) 452, 454, 455,

456, 457, 458; causal mechanisms 452, 455,

456, 457–59; causality problem 454–55;

change, normal status of policy-making

451–52; dependent variable problem 452–53;

Institutional Analysis and Development (IAD)

framework 456, 457, 458; logic of direction

problem 453–54; mechanistic perspective,

positives of 458; Multiple Stream Approach

(MSA) 452, 456, 457; neoinstitutionalist

theories 452; Path Dependence Framework

(PDF) 452, 456, 457; policy actors 453, 458,

459–60; policy change, puzzle of 452; policy

change, understanding in order to grasp policy

dynamics 451–52; Policy Generations

Framework (PGF) 452, 456, 457; public

policy, dynamic phenomenon of 451;

Punctuated

Equilibrium Framework (PEF)

452,

454, 457; puzzle and agents 459–60; sequence

of policy dynamic development 456; temporal

problem 455–56; timing of policy dynamics

and change 455

circumstantial reactors 176

citizen-led performance management 436–37

citizen surveys 439–40

citizens’ juries 441–42

city design and, policy design and 219

Clean Air Amendments Act (1970) 290

client advice style of analysis 260

client-orientation: client focus in evaluation 436;

in evaluation, model of 392–94; in evaluation,

pros and cons of 393–94

climate change debate 207–8

club goods 81–82, 119

coalitions: advocacy coalitions 127–30;

co-production and co-creation 341;

formulation of 128–29; interest groups, experts,

advocacy coalitions, and bureaucrats with

vested interests 172; level of conﬂict between

130–31; see also Advocacy Coalition

Framework (ACF)

cognitive biases and decision heuristics 219–22

collaboration: collaborative web technologies

442–43; combining activities in analysis

262–63; convening, getting problem and parties

right 354–55; governance, networks and

intergovernmental systems 361, 367, 369, 370;

networks and collaborations, notion of 341

collective choice and action: collective action,

problematic circumstances of 94; collective

choice, public choice perspective and 78–79;

collective choice level rules 117; collective

goals 121

collegial models of evaluation 397–99

command issues 189

commissions: membership of 409; policy impact

of 408–9; role in defeating policy feedback and

policy learning barriers 407–9

common-pool resources (CPRs): institutional

analysis and political economy 92; public choice

perspective 82–83

communities of practice 368; attributes of

community 119–20; counter-epistemic

communities 234–35; see also epistemic

communities; policy communities

comparative approaches to study of public

policy-making 29–39; biased results, risk of 34,

37; causal explanations, research methods and

36–38;

comparative case studies, policy

feedback and 406;

comparative public policy

analysis 29–30, 38; comparative public policy

research, varieties of 30–36; demographic

change 32; generalizability 35, 36, 38–39;

impact of policies, measures of 30–32;

incomparability and incommensurability 31;

institutionalism 36; models for research on

decision-making processes, comparative analysis

of 308–9; outcomes of policies, measures of

31–32; outputs of policies, measures of 31, 33;

pattern-oriented studies 29–30, 33–35, 38;

Subject index

512

patterns of policy-making 29–30; pension

politics 32–33; policy agendas, comparative

studies of 168–70, 173; policy change literature

29, 37, 38; policy convergence 29, 34, 36, 37,

39; policy diffusion 34, 36, 37; policy

dismantling 29, 30, 34, 35, 37, 38; policy

learning 35, 36; policy termination 29, 30, 33,

34, 35, 36, 37, 38; policy transfer or emulation

35, 36; process-oriented approaches 29, 30,

35–36, 38; public expenditure, data on 32;

quantitative and qualitative studies, division

between 37; quantitative comparative

environmental policy research 33; replacement

rates, measurement of 32; research methods and

causal explanations 36–38; veto players,

concept of 37

competency trap 405–6

competitive public economies 87–88, 89

complexity: assumption of increase in 299–300;

constituencies, complex or unorganized,

representation of 350–51; of institutional

analysis 121; policy problems, variable

complexities of 490; of policy process 350

Compliance Cost Assessment (CCA) 246

conﬂicts: coalitions, level of conﬂict between

130–31; conﬂict assessment 354–55; dealing

with 486; policy communities, conﬂict over

policy and 51; of values 352–53; see also

disagreement and alternative dispute resolution

in policy process

Congressional Budget Ofﬁce (CBO) 6

congressional dominance 81

consensual knowledge 291–92

consensus building 347–48; mediation theory and

353–57

constituents, representation of 80–81

consumer producers 91–92

consumer protection 205

contested knowledge 351–52, 359

contribution bias 220

coproduction 91–92

Le Corbusier 98–99

counter-epistemic communities 234–35

counting, language of 191

crises as sources of nonincremental change

289–90

critiques: anti-rationalist criticisms of incrementalism

294–96; criticisms and limitations of GCM

321–25; of goal-orientation and clariﬁed

ends

293; incrementalism in

analysis, critique of

hostility to 293; policy learning, criticisms of the

literature on 406; of randomized control trials

420–29; rationalist criticisms of incrementalism

293–94; of reactive nature of incrementalism

294; on training for policy analysis 11

cross-national differences in issue politicization

169–70

decision-making: Betuwe railway line, events in

decision-making on 314–15; bounded rationality,

public policy decision-making and 273;

decentralized decision-making 5; decision

environments, typology of 291; decision

heuristics 219–22; democratic decision-making

76–77; disagreement and alternative dispute

resolution in policy process 356–57; expert

decision-making, theories of 217; horizontal

classiﬁcation of 306; imposed decisions, fairness

and authority of 437; models for research on

decision-making processes, comparative analysis

of 308–9; organizational settings,

decision-making and 21; over time, accounts of

465–67; phase model of 300, 302–3, 304, 306,

307, 308, 309, 311, 314, 315, 316;

policy-making process 20–21; political systems

and decision-making processes 334, 335–36,

337–38; representation of 300, 301–2; rounds

model of 300, 301, 302, 303, 304–6, 308, 309,

310, 312, 314, 315, 316; styles of 21; tracks

model of 302, 306–8, 309, 310, 313, 314, 315,

316; vertical classiﬁcation of 306; see also

bounded rationality and public policy

decision-making; models for research on

decision-making processes, comparative

analysis of

decision theory 220

deductive social theories 59–60

Deepwater Horizon oil spill 177, 178

deﬁnitions: bureaucracy 331–32; evaluation 387;

“institution,” multiple deﬁnitions of term

116–17; policy analysis 4; policy appraisal

245–46; randomized control trials 417–19

deliberation: deliberative forms of public

participation 442; ﬁnding and packageing

options 356

democracy: in arguments on evaluation 396; in

decision-making 76–77; democratic control

and bureaucratic autonomy 334; in social

planning 98

Democracy in America (de

Tocqueville, A.) 90

democratization:

analytical styles 257, 258, 264;

evaluation, public participation and 437

demographic change 32, 180, 423

dependent populations 226

dependent variables 236; problem of 452–53

description, dimensions of 190–91

design: designing process 218–19; of policy

analysis 266–67; principles in IAD,

identiﬁcation of 121; and recommendation of

analytical styles 256–57, 258, 264

deterministic causal patterns 476

development management and policy

implementation 374–82; constituencies,

building up 380; developing countries, context

of 381; development, management for and of

Subject index

513

374; development administration 374;

development management, perspectives on

375–76, 381–82; development management

debates 377–78; diagnostic and analytic tools

381; donor-driven development policy 381;

impact monitoring 380; institutional agendas,

means to 376; international development policy

implementation 379–80; legitimacy, creation of

379–80; modernization 375–76;

multidimensional deﬁnition of development

management 376–77; multiple modernities

376; Organisation for Economic Co-operation

and Development (OECD) Paris Principles

378; organizational structures, modiﬁcation of

380; planning and management toolkit,

provision of 376–77; policy implementation,

development management and 378–79; policy

implementation management 380–81; process

dimension of development management 377;

radicalism 377, 378; reformism 377, 378, 381;

resources, accumulation of 380; resources,

mobilization of 380; values associated with

development management 376

“devil shift” 126, 128

diagnostic and analytic tools 381

disagreement and alternative dispute resolution in

policy process 347–59; Advocacy Coalition

Framework (ACF) 352; best alternative to a

negotiated agreement (BATNA) 349, 356,

357–58; complex or unorganized constituencies,

representation of 350–51; complexity of policy

process 350; conﬂict assessment 354–55;

conﬂicts of values 352–53; consensus building

347–48; consensus building, mediation theory

and 353–57; contested knowledge 351–52, 359;

convening, getting problem and parties right

354–55; decision-making 356–57; deliberation,

ﬁnding and packaging options 356;

disagreement, natural feature of policy process

347; disagreement and negotiation 347, 358;

group polarization 353; implementation 357;

interests and positions, separation of 348–49;

mediation theory 347–48, 353–57; negotiation

theory 347; negotiation theory, consensus

building theory and 353–54; negotiation

theory, starting point 348–49; negotiations (and

disagreements)

in policy processes,

reasons for

difﬁculties with 349–53, 358–59; negotiations

and mediations in public sector, evaluation of

357–58; problem deﬁnition 348–49; process

beneﬁts 358; relationship beneﬁts 358; rules and

responsibilities, clariﬁcation of 355–56; scale of

policy process 350; Seven Elements Framework

(Harvard Negotiation Project) 349; stakeholder

identiﬁcation 355; substantive beneﬁts,

improved outcomes and 357–58; values, frames,

and belief systems 352–53

discussion: after focusing events, differences in

substance of 180–82; agenda-setting and

political discourse 200–201

donor-driven development policy 381

dual federalism in US 362

economic models of evaluation 388

educational products 91

e-governance 442–43

empirical analysis: empirical and experimental

research, IAD and 120–21; empirical

approaches in punctuated equilibrium theory

(PET) 146–49; of epistemic communities

230–36; focusing event theory and 180;

policy design and transfer, empirical perspective

on 223

Employee Retirement Income Security Act

(ERISA, 1974) 402–3

engagement 115, 189, 231, 239, 368, 381, 444;

economic and political engagement 63;

e-engagement 443; evaluation, public

participation and 444; innovative forms of

public engagement 441–43; public engagement

(or citizen engagement) 47, 50, 435, 437,

441–42, 443–44; stakeholder engagement 67,

438, 488, 491; transnational engagement 51

Environmental Impact Assessment (EIA) 246, 248

environmental media coverage in Canada,

expository analysis 208–11

epistemic communities 229–41; as analytical

frameworks, challenge of 232–33; calculation,

learning as 239; counter-epistemic communities

234–35; deﬁcit model of learning 236–37;

dependent variables 236; in dynamic policy

environments 233–34, 235; empirical analysis of

230–36; in European Union (EU) 233; expert

inﬂuence

over decision-maker learning

230;

governance, networks and intergovernmental

systems 368; identiﬁcation of epistemic community

belief systems, challenge of 230–31; independent

variables 238; instrumental, learning as 238;

knowledge consensus and temporally bounded

truth 235; knowledge elements within belief

systems 229–30; learning interactions between

epistemic communities and policy elites 238–40;

learning process 236–37; legitimacy, learning as

239; microfoundations of learning and 238;

persuasion and socialization, learning as 238–39;

policy cycles and 235–36; politics of ideas and

229–30; in real world of politics 234; relating

micro world of learning to macro world of

policy outputs 240; renaissance of knowledge

230; resonance across social scientiﬁc disciplines

231–32; theoretical challenges 236–40;

unreﬂective, learning as 239–40

European Union (EU): Common Agricultural

Policy (CAP) of 477, 479, 480, 489; epistemic

Subject index

514

communities in 233; European Commission

Impact Assessment system 246–47

evaluation, models of 387–400; anticipated

side-effects 391; choice of evaluation models

399; client-oriented evaluation, pros and cons

of 393–94; client-oriented model 392–94;

collegial models 397–99; deﬁnition of

evaluation 387; democratic arguments 396;

economic models 388; goal-attainment model,

strength and shortcomings of 389–90;

goal-attainment model (effectiveness model)

388–90; goal catalogs 390; group deliberations

396; haziness argument 389–90; indeterminate

goals 389; knowledge argument 396; merit

criteria 387–88; null effects 391; peer review,

self-evaluation and combinations thereof

397–99; perverse effects 391; relevance model

392; resource-demands of stakeholder model

396–97; responsive evaluation 394–97;

self-evaluation 397; side-effects model 390– 92;

stakeholder evaluation, upsides and downsides

of 396–97; stakeholder model 394–97; subject

matter to be appraised 388, 390; Swedish

National Agency for Higher Education

ten-steps evaluation policy 398; unintended

side-effects 390; utilization argument 396

evaluation, public participation and 434–44;

argumentative tradition 435; citizen-led

performance management 436–37; citizen

surveys 439–40; citizens’ juries 441–42; client

focus 436; collaborative web technologies

442–43; deliberative forms of public

participation 442; democratization 437;

e-governance 442–43; engagement 444; fourth

generation evaluation 436; imposed decisions,

fairness and authority of 437; innovative forms

of public engagement 441–43; knowledge,

accommodation of alternative forms of 444;

managerial efﬁciency 437–38; New Public

Management (NPM) 436; Organisation for

Economic Co-operation and Development

(OECD) 437; participation in policy evaluation,

role of 435–37; participation in policy evaluation,

who and how? 438–39; participatory mechanisms

439; participatory policy evaluation, challenges

of 443–44; performance management 436;

political nature of policy evaluation 435; public

meetings and hearings 440–41; public

participation in policy, drivers for 437–38;

public value 436; publics, communities and

438–39;

rationalistic or traditional models of

policy analysis 434–35;

stakeholder groups 438;

tensions, generation of 443–44; tokenism 443

evaluation criteria: analytical styles and 263;

evaluation and evaluation criteria, shifting from

government to governance approach 309– 10;

evaluation of policy analysis 267

event-driven policy change and learning, theory

of 186

event-triggered group mobilization 179

evidence-based policy, calls for 415

exogenous change and policy trajectories 469

exogenous problems in process sequencing 477

external variables in IAD 118–20

Exxon Valdez disaster 178, 181, 184, 185

failure: cultivation of 492; exploitation of 492;

generating risk of 491–92; masking of 490–91;

media interest in 487; methodological

difﬁculties of ascertaining “success” and

“failure” 484–87; toleration of 489–90; see also

learning, success or failure and

falsiﬁability 321

federalism and separation of powers, principles of 90

The Federalist 90

feedback: comparative case studies, policy

feedback and 406; feedback loop and

policy-making process 24; indicators,

interconnections with focusing events and

feedback 185; institutional choice, positive

feedback cycles and 142–44; institutions and

negative feedback cycles 140–41; negative

feedback cycles 138–39; policy feedback,

learning and precedent 478–80; policy

incrementalism, negative feedback cycles and

139–40; positive feedback, disproportionate

information-processing and 281–82; positive

feedback cycles 139; in process sequencing

478–79; see also policy feedback, learning

ﬁnancial feasibility 366–67

ﬁscal authority 63

focusing events and policy windows 175–86; actor

promoted events (APEs) 184; Advocacy

Coalition Framework (ACF) 175, 186;

agenda-setting, focusing events as aspects of

175–77; agenda-setting process, focusing events

in 175; aggregation of harms 176; circumstantial

reactors 176; Deepwater Horizon oil spill 177,

178; discussion after focusing events, differences

in substance of 180–82; empirical applications

of focusing event theory 180; event-driven

policy change and learning, theory of 186;

event-triggered group mobilization 179; Exxon

Valdez disaster 178, 181, 184, 185; focusing

event idea, broader applications of the 182–85;

focusing

events and policy

change 185–86;

focusing events in attracting media 206;

foundations of focusing event theory 177–78;

group activity 178; group coalescence and

mobilization, special role of 178–79; group

competition 175; Hurricane Katrina 177;

indicators, interconnections with focusing

events and feedback 185; internal mobilization

178; language, use of 178; negative attention

Subject index

515

177; non-actor promoted events (NAPEs) 184;

political mobilization, role of 176–77; potential

focusing events 177; September 11 terrorist

attacks 177, 180, 185; sudden events 176;

symbol propagation 176; Three Mile Island

nuclear accident 178, 179; Type One and Type

Two events 185, 186

frameworks: advocacy coalition framework (ACF)

126–27; Institutional Analysis and Development

(IAD) framework 116

framing: agenda-setting and political discourse

190, 192–93; frame analysis, agenda-setting and

193; framing effects in policy design and

transfer 222; framing theory 206–7

future research: bureaucracy and policy process 342;

international dimensions and dynamics of

policy-making 55; policy appraisal 251; policy

network research 160–61; process sequencing 481

game models: games and decisions, theory of 98;

institutional analysis and political economy 93

garbage can model (GCM), study of policy-making

process and 320–28; alignment of streams

323–24, 325–26; antecedent conditions 323;

Birkland’s work 326–27; criticisms and

limitations of GCM 321–25; falsiﬁability 321;

focus of GCM 320; hypotheses derived from

GCM, testing of 327; international dimensions

and dynamics of policy-making 51;

McLendon’s work 326; origins and description

of GCM 320–21; paradigmatic problems 325;

policy formulation and adoption 323; politics

and problems, links between 324; politics

stream 320–21; problems and solutions, links

between 324–25; problems stream 320–21;

randomness and separation of streams 323–25;

September 11 terrorist attacks 327; situational

versus structural levels of analysis 321–22;

solutions and politics, links between 324;

solutions stream 320–21; studies employing

GCM 325–27; Zahariadis’ work 325–26

gender analysis 9

generalizability: comparative approaches to study

of public policy-making 35, 36, 38–39; state

theory, regulatory state and 59

Germany, policy network analyses in 157

global warming debate 207–8

globalization: ﬁnancial crisis, policy analysis and 8;

global communication, velocity and volume of 51;

and global dynamics 44–45, 54–55; inﬂuences on

policy-making 46, 47; interest in, policy network

models

and 159–60; policy

analysis and 7

goal-attainment model (effectiveness model) 388–90;

strength and shortcomings of 389–90

goals: goal catalogs 390; goal-orientation and

clariﬁed ends 293; indeterminate goals 389;

policy design and transfer 223

governance, networks and intergovernmental

systems 361–71; building blocks 367–68;

characteristics of network functioning 364;

chartered networks 364; collaboration 361, 367,

369, 370; communities of practice 368; dual

federalism doctrine in US 362; epistemic

communities 368; ﬁnancial feasibility 366–67;

governance networks and added value 369–70;

governance networks and public policy

364–65, 371; intergovernmental relations,

transformation of 362–65. 370–71;

intergovernmental relations (IGR) 362, 363,

364–65; learning and network management

368–71; legal dimension, networked actions

and 366; market superiority 362–63; network

activation 366–67; network constructs 363–64;

network formation 365–67; network

management, learning and 368–71; network

operation 367–68; networks working 363–64;

New Public Management (NPM) 363;

organized actors outside government 362;

policy design and implementation 370; policy

implementation 365–66; politics of activities

between and among organizations 367; social

policies 362; welfare state 362, 364–65

Government Accountability Ofﬁce 6

granularity of perspective, issue of 468–69

groups: coalescence and mobilization, special role

of 178–79; deliberations within 396;

event-triggered group mobilization 179;

exchange of resources between lawmakers and

interest groups 158; group activity, focusing

events and policy windows 178; group

competition 175; interest groups, experts,

advocacy coalitions, and bureaucrats with

vested interests 172; interest groups,

participation of 144–46; polarization of 353;

stakeholder groups 438; treatment and control

groups, treatment effects and 417–19

Guantanamo Bay 489–90

Hawthorne effects 422, 431n15

haziness argument 389–90

health care 18, 49, 63, 91, 121, 138, 168–69,

173,

184–85, 193, 201,

362, 399, 405, 442, 464,

470–71

heterogeneity 424–25

heuristics: bounded rationality and public policy

decision-making 277–78; decision heuristics

219–22

historical institutionalism: historical institutionalist

theory, policy change in 469; process

sequencing and perspective of 474–75; rise of

the broad school of 462, 471

human choice, model of 273, 277

human goals, pursuit of 273

Hurricane Katrina 177, 220

Subject index

516

Hussein, Saddam 485, 491

hypotheses: derived from GCM, testing of 327;

hypotheses building and testing 129, 131, 133

IAP2 439

immigration, politicization of 169

impact monitoring: development management

and policy implementation 380; measures of

impact of policies 30–32

implementation: disagreement and alternative

dispute resolution in policy process 357;

structures for, policy design and transfer 224

implicit theories and assumptions in ACF 126–27

incrementalism 287–97; advantages of 288–89; as

alternative to rational ideal 287–89; American

context, special problems in 296–97; analysis,

critique of hostility to 293; anti-rationalist

criticisms 294–96; business advantage in resources

296; capitalist societies 296; Clean Air

Amendments Act (1970) 290; consensual

knowledge 291–92; crises as sources of

nonincremental change 289–90; decision

environments, typology of 291; federal role

breakthrough policies 292; goal-orientation and

clariﬁed ends, critique of 293; incremental

alternatives, limiting focus to 288; incremental

and non-incremental policy change 139;

incremental model of policy-making 21;

incremental rationalizing policies 292; life cycle

of policies 292; majority-building incrementalism

290; margin-dependent choice 288–89; mass

public opinion as catalyst for nonincremental

change 290; nonincremental change,

preconditions for 289–92; objectives to policies,

adjustment of 288; partisan mutual adjustment

288; policies resulting from, trustworthiness of

292–97; policy change, quadrants of 289, 291;

policy-making through social interaction, system

of 296–97; policy process, serial or repetitive

nature of 289; rational decision-making,

attainment of conditions for 290–92; rationalist

criticisms 293–94; rationality, breakdown of

287–88; rationalizing breakthroughs 292;

reactive nature of, critique of 294; remediality

288; sleeper effects, inability to deal with 294;

social fragmentation of analysis 288

independent variables: epistemic communities

238; policy-making process 20

individuals, attributes of 131

inductive social theories 59–60

informal institutions, importance of 121

informal processes, postpositivism and 107

information-processing 279–80

information rules 119

informational weight 405

inquiries: inquiry commissions, expanding learning

potential of government 407–8; politics of 491

Institutional Analysis and Development (IAD)

framework 115–21; action situations and arenas

117–18; aggregation rules 119; applications of

120–21; authority rules 119; boundary rules

119; bounded rationality and public policy

decision-making 273; change, policy dynamics

and 456, 457, 458; collective choice level rules

117; collective goals 121; community, attributes

of 119–20; constitutional choice level rules 117;

design principles, identiﬁcation

of 121;

empirical and experimental

research 120–21;

external variables 118–20; frameworks 116;

informal institutions, importance of 121;

information rules 119; “institution,” multiple

deﬁnitions of term 116–17; institutional

analysis, complexity of 121; institutional

analysis, political economy and 92–93, 95n4;

institutional change, studies of 120; institutions

116–17; linking arenas 120; material context,

attributes of 119; models 116; payoff rules 119;

position rules 119; public choice perspective

82–83; rules as “shared prescriptions” 116–17;

rules-in-use 119; scope rules 119; theories 116;

universal elements, identiﬁcation of 116;

Workshop in Political Theory and Policy

Analysis, Indiana University 115–16

institutional analysis and political economy 87–95;

bounded rationality 92; collective action,

problematic circumstances of 94; common-pool

resource (CPR) 92; competitive public

economies 87–88, 89; consumer producers

91–92; coproduction 91–92; Democracy in

America (de Tocqueville, A.) 90; educational

products 91; federalism and separation of

powers, principles of 90; game models 93;

Institutional Analysis and Development (IAD)

framework 92–93, 95n4; institutionalism 87;

monocentrism and polycentricity 89–91, 95n3;

New Public Management (NPM) 89;

organization of metropolitan governance in US

88; political economy and analysis of

non-market decision settings 87–89. 94–95n1,

95n2; political economy institutionalism

approach 88, 89, 92, 94; polycentricity,

monocentrism and 89–91, 95n3; public goods

and services 87; public policy analysis 93–94,

95n5, 95n7; rational choice theory (RCT) 93;

service paradox 91; service providers 88

institutional choice, positive feedback cycles and

142–44

institutional friction, disproportionate

information-processing and 281

institutional structure of bureaucracy 333

institutionalism: comparative approaches to study

of public policy-making 36; emergence of a

“new” across social sciences 462; institutional

analysis and political economy 87

Subject index

517

institutions: deconstruction of 101; Institutional

Analysis and Development (IAD) framework

116–17; negative feedback cycles and 140–41;

social construction of 100

instrumental learning 238, 249

integration of relationships, degree and patterns

of 47

intellectual laziness 220

intended rationality, principle of 274–75

interactivity: interacting activities, analytical styles

and 256–58, 267; interaction processes,

ingredients of 306–8; interactive style of analysis

261; with rounds model 305–6

interdependences: advocacy coalition framework

(ACF) 133; international dimensions and

dynamics of policy-making 46

interdisciplinarity, policy network models and 154

interessement 107

interest groups: experts, advocacy coalitions, and

bureaucrats with vested interests 172; interests

and positions, separation of 348–49;

participation of 144–46

intergovernmental relations (IGR): governance,

networks and intergovernmental systems 362,

363, 364–65; negotiations between governments

53; transformation of 362–65. 370–71

intergovernmental systems see governance,

networks and intergovernmental systems

internal mobilization, focusing events and 178

internal validity of randomized control trials 419,

431n15

international dimensions and dynamics of

policy-making 44–55; Advocacy Coalition

Framework (ACF) 48; American auto

manufacturers 45–46; capital movements,

liberalization of 46; cultural characteristics 46;

future research 55; “garbage can” model of

policy-making 51; global communication,

velocity and volume of 51; global inﬂuences on

policy-making 46, 47; globalization and global

dynamics 44–45, 54–55; integration of

relationships, degree and patterns of 47;

interdependence 46; intergovernmental

negotiations 53; international development

policy implementation 379–80; international

forces in policy studies 45; international

policy-making, bureaucracy and 338–39;

internationalized policy environments,

mediating inﬂuence of 51–54; loose couplings,

context of 54; policy communities 47–48;

policy communities, conﬂict over policy and

51; policy network typologies 48–49; policy

options, development of alternatives 50–51;

policy paradigms and governance networks

49–51,

55; policy subsystems 44–45, 46, 47,

48,

49, 50, 53–54, 55; political authority 48;

political boundaries, erosion of 46–47; political

boundaries, policy explanations and

permeability of 45–49; public power, sharing of

47; public sector activism 52–53; self-regulatory

and private regimes 54; sovereignty, pooling of

47; state and society, interactions between 47;

strategic maneuvering of policy actors 48;

supranational governing arrangements,

institutionalization of 52–53; transmission

mechanisms, policy paradigms and 50;

transnational policy communities 49, 53, 54,

55; transnational political forces 46

International Finance Corporation (IFC) 66

interorganizational relations 155

interpretive tendencies 485

Interstate Commerce Commission 5

interventionist and regulatory state, typology of

68–72

interventions of interest, randomization concerns

427

Iran-Contra affair to Iraq 99

issue attention cycle 205

issue framing: mass media and policy-making

206–7; punctuated equilibrium theory (PET) of

agenda-setting and policy change 145

issue networks 125

issue types 205–6

Johnson, Lyndon B. 99

Joint Committees in US 408

Journal of Policy Analysis and Management 10, 12

journalistic norms 207–8

jurisdictional areas 332–33

Kaiser Family Foundation 138, 150n1

knowledge: accommodation of alternative forms

of 444; consensual knowledge 291–92;

consensus and temporally bounded truth 235;

contested knowledge 351–52, 359; elements

within belief systems 229–30; know-recommend

strategy 278; knowledge argument in evaluation

396; knowledge-based power 336; renaissance

of 230; think tanks, knowledge base of 6

language: counting, language of 191; narrative,

language and 190, 193, 194–95; public language

of policy and political elites 484; use in focusing

events 178

leadership role, bureaucracy in 337

learning: conceptual learning 249; deﬁcit model of

236–37; educational products 91; epistemic

communities and learning process 236–37;

expert inﬂuence over decision-maker learning

230; as instrumental 238, 249; interactions

between epistemic communities and policy

elites and 238–40; as legitimacy 239;

microfoundations of learning and epistemic

communities 238; network management and

Subject index

518

368–71; networks, learning and management of

368–71; policy feedback and 401–10

learning, success or failure and 484–93;

ambiguities, dealing with 486; benchmarks,

multiplicity of 485–86; blame games 491;

conﬂicts, dealing with 486; discursive

tendencies 485; European Union (EU),

Common Agricultural Policy of 489; failure,

cultivation of 492; failure, exploitation of 492;

failure, generating risk of 491–92; failure,

masking of 490–91; failure, toleration of

489–90; failure and media interest 487;

Guantanamo Bay 489–90; inquiries, politics of

491; interpretive tendencies 485;

methodological difﬁculties of ascertaining

“success” and “failure” 484–87; nature of policy

success and policy failure 487–89; perceptions,

differences in 485; placebo policies 490; policy

effect, isolation of 487; policy-making as

juggling of success and failure 489–92; policy

problems, variable complexities of 490; policy

success, deﬁnition of 487; policy success,

phenomenon of 484, 492; policy success,

realms of 488–89; political success 488–89;

process success 488; process success and risks of

programme and political failure 491–92;

programme success 488; public language of

policy and political elites 484; rational-scientiﬁc

tradition 485; shortfalls, dealing with 486;

spectrum of outcomes, possibility for 487;

success for whom? 486; variance over time 486

legal dimension, networked actions and 366

legal-rational foundation of bureaucracy 332–33

legitimacy: creation of 379–80; learning as 239;

legitimate power, policy network models and 154

lexicographic strategy 276

life cycle of policies 292

linking arenas 120

lobbying 337

logic of direction problem 453–54

long-term time perspective 127

loose couplings, context of 54

macro and subsystem politics, conceptions of

171–72

macro-level systems, policy network models and

153–54

macro political attention, shifts in 138–39

macro-political institutions 170–71

majority-building incrementalism 290

majority rule 79

managerial efﬁciency

437–38

The Managerial Revolution (Burnham,

J.) 335

margin-dependent choice 288–89

marginalization of politics 407

markets: globalization and 62–63; market

superiority 362–63

mass media and policy-making 204–11;

agenda-setting and issue attentiveness 205–6;

climate change debate 207–8; consumer

protection 205; environmental coverage in

Canada, expository analysis 208–11; focusing

events in attracting media 206; framing theory

206–7; global warming debate 207–8; issue

attention cycle 205; issue framing 206–7; issue

types 205–6; journalistic norms 207–8; mass

public opinion as catalyst for nonincremental

change 290; media as policy actor, challenges of

207–8; media coverage of complicated policy

domains 207–8; media effects, timing of 204;

newsroom norms and long-term policy

domains 208; policy re-framing 207; political

communication 207; political issues in public

discussion 208; problem identiﬁcation 206–7;

thematic frames 207

mechanistic perspective, positives of 458

median voter theorem 80

mediation: analytical styles and 257, 258, 264;

mediation theory 347–48, 353–57

merit criteria 387–88

metaphor 20, 49, 53, 178, 184, 191, 193, 194,

231, 263, 279, 300. 303, 320, 321, 451, 463,

465–67, 474, 490

methodologies: difﬁculties of ascertaining

“success” and “failure” 484–87; methodological

individualism 77

microeconomics: parametric rationality and 467;

policy analysis and 10–11, 12

Mind and Society (Pareto, V.) 274–75

mobilization 18, 127, 130, 144–45, 176–79, 184,

190, 296, 320–21, 403, 438; of bias 491;

postpositivism and policy process 107

modalities

of governance 65–66

models

for research on decision-making processes

299–317; comparative analysis of models

308–9; complexity, assumption of increase in

299–300; decision-making, representation of

300, 301–2; evaluation and evaluation criteria,

shifting from government to governance

approach 309–10; horizontal classiﬁcation of

decision-making 306; Institutional Analysis and

Development (IAD) framework 115–21;

interaction processes, ingredients of 306–8;

interactivity with rounds model 305–6; phase

model, concepts within and principles of 300,

302–3, 304, 306, 307, 308, 309, 311, 314, 315,

316; policy adoption 309–10; policy

formulation 302; reconstruction of

decision-making, models for 299–302, 316–17;

research questions 300; rounds model 300, 301,

302, 303, 304–6, 308, 309, 310, 312, 314, 315,

316; search for and choice of solutions,

problems preceding 303; stream model,

concepts within and principles of 300, 301,

Subject index

519

303–4, 305, 306, 308, 309, 311, 314, 315, 318;

tracks model of decision-making 302, 306–8,

309, 310, 313, 314, 315, 316; vertical

classiﬁcation of decision-making 306; see also

Betuwe railway line

modernization and modernization theory 375–76

monocentrism and polycentricity 89–91, 95n3

motivated reasoning 278

multidimensional deﬁnition of development

management 376–77

multiple modernities 376

multiple parties, competition between 80

Multiple Stream Approach (MSA) 452, 456, 457

narrative: explanatory function of 194–95;

language and agenda-setting 190, 193, 194–95;

narrative approaches to policy process 103–4,

107–8; September 11th, narrative explanation

of 194–95; story lines within public policy

discourse 194

negative attention 177

negative feedback cycles 138–39

negotiation theory: consensus building theory and

353–54; disagreement and alternative dispute

resolution in policy process 347; starting point

for dispute resolution 348–49

negotiations: and mediations in public sector,

evaluation of 357–58; in policy processes,

reasons for difﬁculties with 349–53, 358–59

neo-institutionalism 60–61; neo-institutionalist

theories 452

networks: activation of 366–67; characteristics of

network functioning 364; chartered networks

364; collaborations and, notion of 341;

formation of 365–67; management of, learning

and 368–71; network constructs 363–64;

network theories, core assumptions of 154;

networks working 363–64; operation of

367–68; see also governance, networks and

intergovernmental systems; policy model networks

New Directions in the Study of Community Elites

(Laumann, E. and Pappi, F.) 157

New Public Management (NPM): evaluation,

public participation and 436; governance,

networks and intergovernmental systems 363;

institutional analysis and political economy 89;

movement for 338, 340–41; policy appraisal

250; policy network models 161; practices 7

New Zealand Business Round Table (NZBRT)

7–8

newsroom

norms, long-term policy

domains and

208

“Next Steps” programme 64

non-actor promoted events (NAPEs) 184

non-increasing returns mechanisms 466

noncontradictory argumentation 193

nonincremental change, preconditions for 289–92

normative perspective: path dependency 470–71;

policy design and transfer 222

null effects 391

Obama, Barack (and administration of) 198, 199,

486, 489–90

Occupy Wall Street, case of 195, 198–200

open subsystems 19–20

Organisation for Economic Co-operation and

Development (OECD) 437, 444; core

principles of regulatory reform 244–45, 246;

evaluation, public participation and 437; Paris

Principles 378

organization of metropolitan governance in US 88

organizational form, bureaucracy as 332–33

organizational learning 405

The Organizational State (Laumann, E. and Knoke,

D.) 157

organizational state model of policymaking

155–56

organizations: bounded rationality and public

policy decision-making 278–79; distinction

between different types of 129; organized actors

outside government 362; organized interests

145; settings within, decision-making and 21;

structures of, modiﬁcation of 380

Our Wonderland of Bureaucracy (Beck, J.M.) 335

outcomes of policies, measures of 31–32

outputs of policies, measures of 31, 33

paradigmatic problems 325

parliamentary systems 6

parsimony 420, 424–25

participation: deliberative forms of public

participation 442; interest groups, participation

of 144–46; participatory mechanisms 439;

participatory policy evaluation, challenges of

443–44; participatory style 260; in policy

evaluation, role of 435–37; in policy evaluation,

who and how? 438–39; see also evaluation,

public participation and

partisan mutual adjustment 288

path dependence: concept of 402, 403; distinct

approaches to temporal analysis 474–75,

480–81; increasing returns and 466; Path

Dependence Framework (PDF) 452, 456, 457;

path-dependent sequencing and process (or

problem solving) sequencing, distinction

between 476; policy trajectories and legacies

462–64, 464–65, 465–71

Patient Protection and Affordable Care Act

(PPACA) 138, 150n1

pattern-oriented studies of policy making 29–30,

33–35,

payoff rules 119

peer

review, self-evaluation and combinations

thereof 397–99

Subject index

520

pension politics 32–33

pension provision (UK) 464

perceptions, differences in 485

performance management 436

perspectives: on ﬁeld of policy analysis 265–67; on

policy analysis, framework for understanding

and design 255

persuasion and socialization, learning as 238–39

perverse effects 391

phase model, concepts within and principles of

300, 302–3, 304, 306, 307, 308, 309, 311, 314,

315, 316

phat dependence 466

phenomenological description 106

placebo policies 490

planning and management toolkit, provision of

376–77

policy actors 30, 33, 45, 47–48, 49–51, 52, 54,

158, 205, 305, 403; epistemic communities

233. 235, 238, 239, 240; policy dynamics and

change 453, 458, 459–60; policy evaluation and

public participation 435, 441, 443;

policy-making process 17, 18, 19, 23, 24;

policy problems and 477–80, 481; policy

trajectories and legacies 467, 469, 470–71;

postpositivism and the policy process 101, 103,

104, 105, 106, 107–8

policy adoption: models for research on

decision-making processes 309–10; policy

feedback, learning and 406

policy advice: inﬂuence of advisors 7, 8; policy

analysis 4

policy agenda-setting studies 167–73;

agenda-policy hypothesis 170–71;

agenda-setting approach, tradition of 167–68,

173; Agendas and Instability in American

Politics (Baumgartner and Jones) 167; attention,

public opinion and policy change, linking of

170–72, 173; attention dynamics in executive

speeches 168, 169; comparative studies of

policy agendas 168–70, 173; cross-national

differences in issue politicization 169–70;

government change, role of elections and 168;

health care 169; immigration, politicization of

169; interest groups, experts, advocacy

coalitions, and bureaucrats with vested interests

172; macro and subsystem politics, conceptions

of 171–72; macro-political institutions 170–71;

policy agendas project (US) 171–72; public

saliency, concept of 172; public spending as

measure of public policy 171; re-election,

orientation towards 171; societal problems,

provision of solutions to 168–69

policy analysis: agenda setting 5–6; American

constitution, checks and balances in 6; analytical

capabilities available to politicians 6;

applications of 9–10; Association for Public

Policy Analysis and Management (APPAM) 10;

beneﬁt-cost

analysis 5; central agencies, strength

of 7–8; Congressional

Budget Ofﬁce (CBO) 6;

critiques on training for 11; decentralized

decision-making 5; deﬁnition of 4; demand for,

evolution of 4–10; early representations 3–4;

gender analysis 9; global ﬁnancial crisis 8;

globalization 7; Government Accountability

Ofﬁce 6; government programmes,

effectiveness of 7; Interstate Commerce

Commission 5; Journal of Policy Analysis and

Management 10; microeconomics and 10–11,

12; New Public Management practices 7; New

Zealand Business Round Table (NZBRT) 7–8;

parliamentary systems 6; policy advising 4;

policy advisors, inﬂuence of 7, 8; policy analysis

movement, prospects for 13–14; policy analysts,

perceptions of 9; policy development 5–6;

policy problems, evolution of 5; policy sciences

4; policy solutions, appropriateness of

alternatives 5; policy studies 4; professional

socialization, teaching key components of

11–12; public policies 4; race analysis 9;

redistribution, government role in 5; rise of

3–4; role of appraisal in 244, 250; role of policy

analyst 263–65; sources of demand for 4; styles

of 258–62; supply of, evolution of 10–13;

supply of, trends in 12–13; technical skills,

development of 12; think tanks, knowledge

base of 6; tools of 5

Policy and Comparative Agendas projects 149,

150n5, 150n8

policy appraisal 244–51; administrative capacity

and effectiveness 245; appraisal processes as sites

of political behavior 245; appraisal systems,

research on design of 249; assessment of

operation of policy appraisal 248, 249–50;

bureaucracy, political control of 249;

Compliance Cost Assessment (CCA) 246;

conceptual learning 249; deﬁnition of 245–46;

Environmental Impact Assessment (EIA) 246,

248; European Commission Impact

Assessment system 246–47; future for 251;

instrumental learning 249; New Public

Management (NPM) 250; Organisation for

Economic Co-operation and Development’s

(OECD) core principles of regulatory reform

244–45, 246; policy analysis, role of appraisal in

250; policy analysis and 244; policy appraisal

systems, diffusion of 247; policy formulation

and 244; policy-making change and 244;

political behavior, sites of 245; positivist

approaches to research 247–48; post-positivist

approaches to research 248–49; practice of

246–47; Regulatory Impact Assessment (RIA)

246; research on 247–49;

research-practice

relationships 249–50; Strategic

Environmental

Subject index

521

Assessment (SEA) 246; Sustainability Impact

Assessment (SIA) 246

policy brokers 131

policy change: advocacy coalition framework

(ACF) 125, 127, 129, 132–33, 133, 134;

literature on 29, 37, 38; orders and rates of 464;

puzzle of 452; quadrants of 289, 291; stability

and 467–70; understanding in order to grasp

policy dynamics 451–52

policy communities: conﬂict over policy and 51;

international dimensions and dynamics of

policy-making 47–48; policy-making

process 19

policy convergence 29, 34, 36, 37, 39

policy cycle model: policy cycles and epistemic

communities 235–36; policy-making process 17

policy design and transfer 217–27; agenda-setting

and political discourse 192; availability 219–20;

beneﬁts to contenders 225–26, 227; city design

and 219; cognitive biases and decision heuristics

219–22; commonly copied policy designs

225–27; contribution bias 220; decision

heuristics 219–22; decision theory 220;

dependent populations 226; designing process

218–19; empirical perspective 223; expert

decision-making, theories of 217; framing

effects 222; fundamental elements of designs

223; goals 223; implementation, boundaries

between 106; implementation,

intergovernmental systems and 370;

implementation, structures for 224; intellectual

laziness 220; normative perspective 222; policy

design as noun (policy content) 222–25; policy

design as verb 219–22; policy designing as

process 218; policy diffusion 218–19; policy

invention, theories of 217; punishment and

discipline designs 226; rationales 224–25, 226;

rationalist approach 218; rules 224, 226; target

populations 223–24; target populations,

political powers of 225–26; tools 224, 225;

underlying assumptions 225; value allocation

224; wildﬁres in Arizona, McCain ’s response to

221–22

policy development: policy analysis and 5–6; as

policy style 24; process of 17; see also policy

formulation; policy-making process

policy diffusion: comparative approaches to study

of public policy-making 34, 36, 37; policy

design and transfer 218–19; policy feedback,

learning and 405

policy direction, notion of 468

policy dismantling 29, 30, 34, 35, 37, 38

policy domain: communication and resource

exchange networks 156; concept of 155

policy dynamics, change and 451–60

policy effect, isolation of 487

policy entrepreneurs 144–46

policy

evaluation styles 23–24

policy

feedback, learning and 401–10; American

Association of Retired Persons 403;

beneﬁciaries, empowerment of 402; British

Pension Commission 409; bureaucratic

organization, creation of 402; commissions,

membership of 409; commissions, role in

defeating policy feedback and policy learning

barriers 407–9; comparative case studies 406;

competency trap 405–6; Employee Retirement

Income Security Act (ERISA, 1974) 402–3;

Galton’s problem 406–7; informational weight

405; initial adoption of policies 402; inquiry

commissions, expanding learning potential of

government 407–8; Joint Committees in US

408; McDonald Commission in Canada 408;

marginalization of politics 407; organizational

learning 405; path dependence, concept of 402,

403; policy adoption 406; policy diffusion 405;

policy feedback 401–3, 409–10; policy

feedback, logic of 403; policy impact of

commissions 408–9; policy instruments, policy

diffusion and 405; policy learning 401, 403–7,

409–10; policy learning, criticisms of the

literature on 406; policy reversal 402; positive

and negative policy feedback, distinction

between 403; process sequencing and

precedent 478–80; qualitative research designs

406; Social Security Association 403; terms of

reference 409

policy formulation: adoption and 323; models for

research on decision-making processes 302;

policy appraisal and 244; policy-making process

18–20

Policy Generations Framework (PGF) 452, 456,

457

policy implementation: development management

and 378–79, 380–81; governance, networks

and intergovernmental systems 365–66;

incrementalism, negative feedback cycles and

139–40; instruments of, policy diffusion and

405; policy implementers, bureaucrats as

339–40; postmodern condition and institutions

104–5; styles of 22–23

policy invention, theories of 217

policy learning: comparative approaches to study

of public policy-making 35, 36; criticisms of

literature on 406; policy feedback, learning and

401, 403–7, 409–10

policy legacies 468

policy-making process 17–24; bounded rationality

20–21; bureaucrats as policy-makers 336–39;

change

and policy appraisal

244;

decision-making 20–21; decision-making, styles

of 21; feedback loop 24; incremental model of

policy-making 21; independent variables 20;

initiation, outside and inside 18; juggling of

Subject index

522

success and failure in 489–92; lesson-drawing

24; open subsystems 19–20; organizational

settings, decision-making and 21; policy

behavior in policy cycle, styles of 18–21; policy

communities 19; policy cycle model 17; policy

development as policy style 24; policy

development process 17; policy evaluation

styles 23–24; policy formulation 18–20; policy

implementation styles 22–23; policy-making

through social interaction, system of 296–97;

policy networks 19; policy subsystems 19;

power of ideas 19; procedural instruments

22–23; rational model of decision-making

20–21; social learning 24; substantive

instruments 22; systemic agenda 18

policy network models 153–61; Advocacy

Coalition Framework (ACF) 161; asymmetrical

power 154; exchange of resources between

lawmakers and interest groups 158; future

directions for policy network research 160–61;

Germany, policy network analyses in 157;

globalization, interest in 159–60;

interdisciplinarity 154; interorganizational

relations 155; key ideas in modern policy

network analysis 158; legitimate power 154;

macro-level systems 153–54; network theories,

core assumptions of 154; network typologies

48–49; New Public Management (NPM) 161;

organizational state model of policymaking

155–56; policy domain, communication and

resource exchange networks 156; policy

domain, concept of 155; policy network

analyses, history and development of 156–58;

policy network theory and empirical research

158–60; policy networks, as units of analysis

154; policy networks, deﬁnition of 153; power

imbalances, policy exchanges and 158;

rationality of actors, assumption of 158; research

on, future challenges 160–61; social network

theory 153–56; United Kingdom, policy

network analyses in 157–58; United States,

policy network analyses in 157

Policy Networks in British Government (Marsh, D.

and Rhodes, R.A.W.) 157–58

policy options, development of alternatives 50–51

policy-oriented learning 125, 130–32, 133

policy paradigms and governance networks 49–51,

policy preferences, society and 60

policy problems: evolution of 5; policy actors and

477–80; variable complexities of 490

policy process: deconstruction of 105–8; policy

process research 125; serial or repetitive nature

of 289

policy re-framing 207

policy-relevant social theories 59

policy reversal 402

policy sciences 4

policy solutions, appropriateness of alternatives 5

policy stability and rapid change, forces leading to

139

policy

studies 4

policy subsystems:

advocacy coalition framework

(ACF) 134; composition of 128–29;

international dimensions and dynamics of

policy-making 44–45, 46, 47, 48, 49, 50,

53–54, 55; policy-making process 19; policy

processes and 126

policy success: deﬁnition of 487; phenomenon of

484, 492; realms of 488–89

policy targeting 22, 23, 102, 143, 145

policy termination 29, 30, 33, 34, 35, 36, 37, 38

policy think tanks 98

policy thought, turning points in 100–102

policy trajectories and legacies 462–71;

decision-making over time, accounts of

465–67; dualism between stability and change

468; exogenous change and 469; granularity of

perspective, issue of 468–69; historical

institutionalism, rise of the broad school of 462,

471; historical institutionalist theory, policy

change in 469; institutionalism, emergence of a

“new” across social sciences 462;

microeconomics, parametric rationality and

467; non-increasing returns mechanisms 466;

normative aspects of path dependency 470–71;

path dependency 462–64, 464–65, 465–71;

path dependency, increasing returns and 466;

pension provision (UK) 464; phat dependence

466; policy change, orders and rates of 464;

policy change and stability 467–70; policy

direction, notion of 468; policy legacies 468;

policy trajectories and legacies, beneﬁts of path

dependency for understanding 464–65; policy

trajectories and legacies, problems of path

dependency for understanding 465–71; primary

care sector (UK) 467; social causation 463;

social efﬁciency 470; social housing (UK) 465;

social inefﬁciency 470–71; temporality, change

and history in social and political analysis,

interest in 462

policy transfer or emulation 35, 36

policymaking in an interpretive world 102–4

political behavior: of advocacy coalitions 129–30;

sites of 245

political boundaries: erosion of 46–47; policy

explanations and permeability of 45–49

political economics: analysis of non-market

decision settings 87–89. 94–95n1, 95n2;

political economy institutionalism approach 88,

89, 92, 94; schools of 84–85

politics: actions of politicians, public choice

perspective and 78, 80–81;

of activities between

and

among organizations 367; of ideas,

Subject index

523

epistemic communities and 229–30; issues in

public discussion 208; mobilization, role in

focusing events 176–77; nature of policy

evaluation 435; need for politicians, public

choice perspective and 78, 79–80; political

authority 48, 158, 408–9; political

communication 207; politics-administration

dichotomy debate 333–34; politics stream in

GCM 320–21; and problems, links between

324; re-election, orientation towards 171;

success in 488–89; systems of, decision-making

processes and 334, 335–36, 337–38; systems of,

interdependences of economic systems and

334–35

Politics (Aristotle) 102

The Politics of Bureaucracy (Peters, B.G.) 335 – 36

polycentricity, monocentrism and 89–91, 95n3

position rules in IAD 119

positive feedback: cycles of 139; disproportionate

information-processing and 281–82; positive

and negative policy feedback, distinction

between 403

positive political theory 80

postpositivism and policy process 98–109;

communicative model of rationality 104–5;

democratic social planning 98; enrollment 107;

games and decisions, theory of 98; informal

processes 107; institutions, deconstruction of

101; institutions, social construction of 100;

interessement 107; Iran-Contra affair to Iraq 99;

mobilization 107; narrative approaches 103–4,

107–8; phenomenological description 106;

policy appraisal and postpositivist approaches to

research 247–48, 248–49; policy design and

implementation, boundaries between 106;

policy institutions, postmodern condition and

104–5; policy process, deconstruction of 105–8;

policy think tanks 98; policy thought, turning

points in 100–102; policymaking in an

interpretive world 102–4; post-constructionist

understanding of policy process 108; practice of

policy, turning points in 98–100;

problematisation 106; rational policy-making

98–99; rationalization of society 99; social

construction, problem of 108–9; social

experimentation 99; social issues, wicked

problems of 99–100; social judgments 103–4;

spin doctoring 103–4; structural theory of

meaning 100; telos of policy 101; Tennessee

Valley

Authority 99; thinking

about policy,

intellectual revolutions on 100–101; Troubled

Asset Relief Program (TARP) 101, 102–3;

weapons of mass destruction (WMDs) in Iraq

108; Wittgenstein’s philosophical treatise on

logic 100

power: agenda-setting and inﬂuential and

powerful interests 189; asymmetrical power

154; beneﬁciaries, empowerment of 402;

bureaucratic power and dilemma for democracy

334; federalism and separation of powers,

principles of 90; of ideas 19; imbalances in,

policy exchanges and 158; knowledge-based

power 336; legitimate power, policy network

models and 154; location of 63; measurement

of 64; public power, sharing of 47; sources of

64; target populations, political powers of

225–26; transfer of 63

preference ranking 192

primary care sector (UK) 467

prisoner’s dilemma 78

problem deﬁnition: agenda-setting and political

discourse 190–92; disagreement and alternative

dispute resolution in policy process 348–49;

problem from solution spaces, separation of

277–78; problems stream in GCM 320–21;

solutions and problems, links between 324–25

problem description, anatomy of 191

problem handling 191

problem identiﬁcation 206–7

problem ownership 191–92

problematisation 106

procedural instruments 22–23

process-oriented approaches to study of public

policy-making 29, 30, 35–36, 38

process sequencing 473–81; deterministic causal

patterns 476; European Union (EU) Common

Agricultural Policy (CAP) 477, 479, 480;

exogenous problems 477; feedback 478–79;

future research 481; historical institutionalist

perspective 474–75; path dependency and,

distinct approaches to temporal analysis 474–75,

480–81; path-dependent sequencing and

process (or problem solving) sequencing,

distinction between 476; policy feedback,

learning and precedent 478–80; policy

problems and policy actors 477–80; policy

sequences, shaped and reshaped by creativity

and stubbornness 477; in public policy

476–77; reactive sequencing and path

dependency 475–76; strategic learning 477 – 78;

temporal connections between policy events

and policy changes 473; theoretical

underdevelopment of 481

process

style 260–61

process success:

learning, success or failure and

488; and risks of programme and political

failure 491–92

professional socialization, teaching key

components of 11–12

programme success 488

public choice perspective 76–85; bureaucrat

bashing 84; club goods 81–82; collective choice

78–79; common pool resources (CPRs) 82–83;

congressional dominance 81; constituents,

Subject index

524

representation of 80–81; democratic

decision-making 76–77; Institutional Analysis

and Development (IAD) framework 82–83;

majority rule 79; median voter theorem 80;

methodological individualism 77; multiple

parties, competition between 80; political

economics, schools of 84–85; politicians,

actions of 78, 80–81; politicians, need for 78,

79–80; positive political theory 80; prisoner’s

dilemma 78; public choice (and theory of)

76–77, 84–85; public goods 78, 81–82; public

goods research 82–83; public policy

implementation 78, 84; rational abstention 81;

Rawls’ theory of justice 83; social welfare

functions, constitutions as 83–84; societal

demands 78, 82–84

public expenditure, data on 32

public goods see public choice perspective

public meetings and hearings 440–41

public participation in policy, drivers for 437–38

public policies: dynamic phenomenon of 451;

implementation of 78, 84; policy analysis 4,

93–94, 95n5, 95n7; public spending as measure

of 171

public power, sharing of 47

public saliency, concept of 172

public sector activism 52–53

public spending as measure of public policy 171

publics, communities and 438–39

Punctuated Equilibrium Framework (PEF) 452,

454, 457

punctuated equilibrium theory (PET) of

agenda-setting and policy change 138–50;

Agendas and Instability in American Politics

(Baumgartner, F. and Jones, B.D.) 149–50,

150n6; bounded rationality and public policy

decision-making 273; disproportionate

information processing and positive feedback

cycles 141–42; empirical approaches in punctuated

equilibrium 146–49; general punctuation

hypothesis 143; incremental and non-incremental

policy change 139; institutional choice, positive

feedback cycles and 142–44; institutions and

negative feedback cycles 140–41; interest

groups, participation of 144–46; issue framing

145; macro political attention, shifts in 138–39;

negative feedback cycles 138–39; organized

interests 145; Patient Protection and Affordable

Care Act (PPACA) 138, 150n1; Policy and

Comparative Agendas projects 149, 150n5,

150n8; policy entrepreneurs 144–46; policy

incrementalism, negative feedback cycles and

139–40;

policy stability and rapid change, forces

leading to 139;

policy targeting 145; positive

feedback cycles 139; stochastic processes 146,

147, 148, 149; venue shopping 144, 145–46

punishment and discipline designs 226

Qaddaﬁ, Muammar 47

qualitative research designs 406

quantitative and qualitative studies, division

between 37

quantitative comparative environmental policy

research 33

race analysis 9

radicalism 377, 378

randomized control trials 415–31; arguments in

favor of 419–20; average impact vs.

distributional concerns 426; categories of policy

questions suitable for 428–29; causal

mechanisms 426–27; critiques of 420–29;

current dialog about 416; deﬁnition of 417–19;

development interventions, identiﬁcation of

415–16; development programs, evaluation of

416; evidence-based policy, calls for 415; gold

standard in causal identiﬁcation, promotion as

417–20; heterogeneity 424–25; internal validity

419, 431n15; interventions of interest,

randomization concerns 427; parsimony 420,

424–25; practical considerations, departures

from ideal conditions and 420–21;

randomization bias 421–22; scope of questions

suitable for 425–29; selection bias 418, 419,

420, 422–23, 424, 431n18; selection issues

422–24; substitution bias 421–22; summary of

main points 429–30; treatment and control

groups, treatment effects and 417–19; treatment

compliance 421–22; value in reﬁning a single

parameter 427–28

randomness and separation of streams in GCM

323–25

rational abstention 81

rational choice theory (RCT) 93

rational decision-making: attainment of conditions

for 290–92; policy-making process 20–21

rationales 224–25,

226

rationality: of actors,

assumption of 158;

breakdown of, incrementalism and 287–88;

communicative model of 104–5;

incrementalism, rationalist criticisms of 293–94;

policy design and transfer, rationalist approach

to 218; rational policy-making 98–99;

rational-scientiﬁc tradition 485; in style in

analysis 258–59; in traditional models of policy

analysis 434–35

reactive sequencing and path dependency

475–76

reconstruction of decision-making, models for

299–302, 316–17

redistribution, government role in 5, 63

reﬂection on policy analysis 265–66

reformism in development management 377, 378,

381

Regulatory Impact Assessment (RIA) 246

Subject index

525

regulatory state 63; conﬂation of debates on 62;

International Finance Corporation (IFC)

perspective on 66, 67; neo-liberal agenda and

64; techniques of 67; theories of 62–65

Reinventing Government (Osborne, D. and

Gaebler, T.) 340

relevance model of evaluation 392

remediality 288

replacement rates, measurement of 32

research: and analytical styles 256, 258, 264;

methods and causal explanations 36–38; on

policy appraisal 247–49; on policy network

models, future challenges 160–61; questions on

decision-making processes 300;

research-practice relationships 249–50

resources: accumulation of 380; business advantage

in resources 296; communication and resource

exchange networks 156; exchange of resources

between lawmakers and interest groups 158;

mobilization of 380; resource-demands of

stakeholder model 396–97; see also common

pool resources (CPRs)

responsive evaluation 394–97

riots in UK, case of 195–98

rounds model for research on decision-making

processes 300, 301, 302, 303, 304–6, 308, 309,

310, 312, 314, 315, 316

rules: policy design and transfer 224, 226; and

responsibilities, clariﬁcation of 355–56;

rule-making 63–65; rules-in-use 119; as “shared

prescriptions” 116–17

scope: in advocacy coalition framework (ACF)

127; of questions suitable for randomized

control trials 425–29; rules in Institutional

Analysis and Development (IAD) framework

119

selection: bias 418, 419, 420, 422–23, 424,

431n18; issues in RCTs 422–24

self-evaluation 397

self-regulatory and private regimes 54

The Semisovereign People (Schattschneider, E.E.)

190

September 11 terrorist attacks: focusing events and

policy windows 177, 180, 185; garbage can

model (GCM), study of policy-making process

and 327; narrative explanation of 194–95

service paradox 91

service providers 88

Seven Elements Framework (Harvard Negotiation

Project) 349

side-effects model 390–92

sleeper effects 294

social and societal demands 78, 82–84

social causation 463

social construction, problem of 108–9

social efﬁciency 470

social experimentation 99

social fragmentation of analysis 288

social

housing (UK) 465

social

inefﬁciency 470–71

social institution of bureaucracy 333

social issues, wicked problems of 99–100

social judgments 103–4

social learning 24

social liberties and freedoms 61

social network theory 153–56

social policies 362

social scientiﬁc theories 59

Social Security Association 403

social welfare functions, constitutions as 83–84

solutions: and politics, links between 324; social

problems, provision of solutions to 168–69;

solutions stream in GCM 320–21

sovereignty, pooling of 47

spin doctoring 103–4

stabilization function 63

stakeholders: engagement of 67, 438, 488, 491;

evaluation, stakeholder model of 394–97;

identiﬁcation of 355; resource-demands of

stakeholder model 396–97; stakeholder

evaluation, upsides and downsides of 396–97;

stakeholder groups 438

state theory, regulatory state and 59–72; deductive

social theories 59–60; ﬁscal authority 63;

generalizability 59; inductive social theories

59–60; interventionist and regulatory state,

typology of 68–72; markets, globalization and

62–63; modalities of governance 65–66;

neo-institutionalism 60–61; “Next Steps”

programme 64; policy preferences, society and

60; policy-relevant social theories 59; power,

location of 63; power, measurement of 64;

power, sources of 64; power, transfer of 63;

redistributive function 63; regulatory authority

63; regulatory function 63; regulatory state,

conﬂation of debates on 62; regulatory state,

International Finance Corporation (IFC)

perspective on 66, 67; regulatory state,

neo-liberal agenda and 64; regulatory state,

theories of 62–65; regulatory techniques 67;

rule-making 63–65; social liberties and

freedoms 61; social scientiﬁc theories 59;

stabilization function 63; state and society,

interactions between 47; state-centric

explanations 61–62; state theory, focus of

59–60; statutory bodies and regulatory agencies,

profusion of 64–65

stimuli:

attributes of 131;

processing of 126

stochastic processes 146, 147, 148, 149

strategic advice, provision of 257, 258, 264

Strategic Environmental Assessment (SEA) 246

strategic learning 477–78

strategic manoeuvring of policy actors 48

Subject index

526

stream model for research on decision-making

processes 300, 301, 303–4, 305, 306, 308, 309,

311, 314, 315, 318

street-level bureaucrats 339

structural theory of meaning 100

substantive beneﬁts, improved outcomes and

357–58

substantive instruments 22

substitution bias 421–22

subsystem actors 126; see also policy subsystems

sudden events 176

super bureaucracies 337

supranational governing arrangements,

institutionalization of 52–53

Sustainability Impact Assessment (SIA) 246

Swedish National Agency for Higher Education

ten-steps evaluation policy 398

symbol propagation 176

systemic agenda 18

target populations 223–24; political powers of

225–26

technical skills, development of 12

telos of policy 101

temporality: change and history in social and

political analysis, interest in 462; temporal

connections between policy events and policy

changes 473; temporal problem in policy

dynamics and change 455–56

Tennessee Valley Authority 99

terms of reference 409, 491

thematic frames 207

theories: behavioral theory of choice 274, 276–77;

behavioral theory of organization 278–79;

challenges for epistemic communities 236–40;

consensus building, mediation theory and

353–57; decision theory 220; deductive social

theories 59–60; expert decision-making,

theories of 217; games and decisions, theory of

98; historical institutionalist theory, policy

change in 469; Institutional Analysis and

Development (IAD) framework 116; median

voter theorem 80; mediation theory 347–48,

353–57; negotiation theory 347; negotiation

theory, consensus building theory and

353–54; negotiation theory, starting point of

consensus building 348–49; neoinstitutionalist

theories 452; network theories, core

assumptions of 154; policy invention, theories

of 217; policy network theory and empirical

research 158–60; policy-relevant social theories

59; positive political theory 80; public choice

(and theory of) 76–77, 84–85; rational choice

theory (RCT) 93; Rawls’ theory of justice 83;

regulatory state, theories of 62–65; social

network theory 153–56; social scientiﬁc

theories 59; structural theory of meaning 100;

theoretical emphasis of ACF 133–34;

underdevelopment of process sequencing 481;

see also disagreement and alternative dispute

resolution in policy process; punctuated

equilibrium theory (PET); state theory,

regulatory state and

think tanks: knowledge base of 6; policy think

tanks 98; thinking about policy, intellectual

revolutions on 100–101

Three Mile Island nuclear accident 178, 179

timing: decision-making over time, accounts of

465–67;

learning,

success

failure and variance

over time 486; longterm time perspective of

ACF 127; media effects, timing of 204; of

policy dynamics and change 455; see also

temporality

tokenism 443

tools: diagnostic and analytic tools 381; planning

and management toolkit, provision of 376–77;

for policy analysis 5; policy design and transfer

224, 225

tracks model of decision-making 302, 306–8, 309,

310, 313, 314, 315, 316

trade-offs, principle of 276

translated beliefs 126–27

transmission mechanisms, policy paradigms and

transnational policy communities 49, 53, 54, 55

transnational political forces 46

treatment and control groups, treatment effects

and 417–19

treatment compliance 421–22

Troubled Asset Relief Program (TARP) 101,

102–3

Type One and Type Two events 185, 186

uncertainty, principle of 275–76

unintended side-effects 390

United Kingdom: British Pension Commission

409; pension provision 464; policy network

analyses in 157–58; primary care sector 467;

riots in, case of 195–98; social housing in 465

United Nations (UN) 52, 53, 66, 160, 239

United States: American Association of Retired

Persons 403; auto manufacturers in 45–46;

Constitution of, checks and balances in 6; dual

federalism in 362; incrementalism, special

problems in American context 296–97;

Iran-Contra affair to Iraq 99; Joint

Committees in 408; Occupy Wall Street, case

of 195, 198–200; organization of

metropolitan governance in 88; policy agendas

project

171–72; policy network analyses in 157

universal elements in IAD, identiﬁcation of 116

unreﬂective learning 239–40

utilization argument in evaluation 396

Subject index

527

values: allocation of 224; arguments and,

clariﬁcation of 257, 258, 264; associated with

development management 376; frames, belief

systems and 352–53; single parameter in

randomized control trials and 427–28;

underlying analytical styles 263–65

venue shopping 144, 145–46

veto players, concept of 37

weapons of mass destruction (WMDs) in Iraq 108

welfare state 5, 31–32, 34, 37, 62, 170, 362,

364–65, 402–3

wildﬁres in Arizona, McCain’s response to 221–22

“wisdom of crowds” argument 282

Workshop in Political Theory and Policy Analysis,

Indiana University 115–16

World Bank 337

Subject index

528