Government Policy and the Housing Bubble

Murray State's Digital Commons Murray State's Digital Commons

Honors College Theses Student Works

5-2023

Government Policy and the Housing Bubble Government Policy and the Housing Bubble

Hannah Reynolds

Follow this and additional works at: https://digitalcommons.murraystate.edu/honorstheses

Part of the American Politics Commons, and the Economic Theory Commons

Recommended Citation Recommended Citation

Reynolds, Hannah and Reynolds, Hannah, "Government Policy and the Housing Bubble" (2023).

Honors

College Theses

. 174.

https://digitalcommons.murraystate.edu/honorstheses/174

This Thesis is brought to you for free and open access by the Student Works at Murray State's Digital Commons. It

has been accepted for inclusion in Honors College Theses by an authorized administrator of Murray State's Digital

Commons. For more information, please contact [email protected].

Murray State University Honors College

HONORS THESIS

Certificate of Approval

Government Policy and the Housing Bubble

Hannah Reynolds

May 2023

Approved the fulfill the _______________________________

requirements of HON 438 Dr. Eran Guse, Associate Professor

[Economics]

Approved the fulfill the Honors _______________________________

Thesis requirements of the Murray Dr. Warren Edminster, Executive Director

State Honors Diploma Honors College

Examination Approval Page

Author: Hannah Reynolds

Project Title: Government Policy and the Housing Bubble

Department: Economics/Political Science

Date of Defense: May 5, 2023

Approval by Examining Committee:

_________________________________ _________________

(Dr. Eran Guse, Advisor) (Date)

_________________________________ _________________

(Dr. Marc Polizzi, Committee Member) (Date)

_________________________________ _________________

(Dr. Brittany Wood, Committee Member) (Date)

Government Policy and the Housing Bubble

Submitted in partial fulfillment of the requirements for the

Murray State University Honors Diploma

Hannah Reynolds

May 2023

Abstract

The housing market bubble was created by market forces and exacerbated by government

regulations. The government policies implemented incentivized certain behaviors by investors

that exacerbated the housing bubble beyond what would have occurred “naturally.” The policy

under analysis for this research is the Community Reinvestment Act (CRA) of 1995. The CRA

made affordable-housing initiatives more enforceable and results-driven. These revisions

produced unintended consequences in the economy in the form of incentives to give out risky

loans. Statistical analysis through difference of means tests show that there is a low level of

confidence in statistical significance between the CRA and increased real mortgage debt per

capita; a statistically significant relationship between the CRA and increased debt-to-income

ratio; and no impact on risk premium. These findings suggest that incentives were keeping the

risk premium low when it should rise under normal circumstances.

Table of Contents

Introduction …………………………………………………………………………..……Page 1

Literature Review……... ……………………………………………………………..……Page 4

Market Factors and Fluctuations ………………………………………………………..…Page 5

Monetary Policy……………………………………………………………………………Page 8

Government Legislation and Fiscal Policy ………………………………………………..Page 12

Analysis of the Literature…………………………………………………………………..Page 16

Research Design…………………………………………………………………………....Page 18

Discussion of Findings…………….………………………….…………………………....Page 22

: Risk Premium…………….…………………………………………………………..Page 22

: Mortgage Debt Outstanding, in terms of Real Debt per Capita…………….………..Page 24

: Mortgage Debt-to-Income Ratio …………...………………………………………..Page 26

Conclusion …………………………………………………………..……………………..Page 26

References………………………………………………………………………………….Page 30

iii

List of Figures and Tables

Figure 1…………...………………………………………………………………………..Page 19

Figure 2………………………………………………………………………...…………..Page 19

Figure 3………………………………………………………………………...…………..Page 24

Figure 4…………………………………………………………………………………….Page 25

Figure 5………………………………………………………………………………...…..Page 25

Figure 6……………………………………………………………………………...……..Page 26

Table 1…………………………………………………………………………….………..Page 23

Introduction

In 2006, US President George W. Bush stated that "If houses get too expensive, people

will stop buying them, which will cause people to adjust their spending habits ... Let the market

function properly" (The White House Archives 2006). This statement describes a normally

functioning market. However, a year prior to this, in 2005, even "bubble-deniers" had to

recognize that housing prices were increasing faster than income (Calculated Risk 2006). The

puzzle at this point is: why did buyers not behave in the expected way when prices were

increasing too much? Clearly, there were confounding factors at play that influenced the bubble

to expand beyond what would have been "normal" or reasonably expected. As such, to what

extent did government policy-making unintentionally expand the housing bubble of the 2008

financial crisis?

All decisions have a stated or intentional purpose, but all decisions also create results that

were unintentional and, oftentimes, unaccounted for during the deliberation process. The concept

of unintended consequences is the embodiment of the common sayings “hindsight has 20/20

vision” and “there are no solutions, only trade offs.”

There is no perfect economic model to predict the unpredictable. There are, however,

basic models that are known to be reliable and have stood the test of time that need to be

understood for this research. The principles of economics teach that the market naturally tends

towards equilibrium by finding the spot where supply and demand are balanced. This theory also

holds true within the housing market; the housing prices should be balanced with the amount of

housing available to purchase. If there is too much supply, the prices will drop to encourage

buying; if there is a scarcity of housing available to purchase, prices should rise to reflect that, as

well.

Market bubbles are normal occurrences in economic cycles. They are characterized by an

increase in the value or price of an asset and, after an unknown amount of time, are followed by

a sudden contraction, otherwise referred to as a crashing or bursting of the bubble (Kenton

2022). In bubbles, people tend to make overconfident decisions until the bubble bursts and their

actions catch up with them. Bubbles can crash for a variety of reasons, but, in general, the crash

is caused by a decrease in the availability of the money that was supporting the bubble in the first

place. There is a general consensus that the housing market bubble leading up to the financial

crisis of 2008 burst because of a large, unforeseen amount of loans that defaulted (Baily, Litan,

and Johnson 2008). While loan defaults are expected to an extent, the economy could not handle

the amount of defaults from the large quantity of risky loans that were given out during this time.

Risky loans and subprime loans are more susceptible to defaulting than a prime, or high-quality,

loan. The default of too many loans was the ultimate cause or precursor to the burst, but what

made that happen?

Three schools of thought will be looked at as possible explanations for the growth of the

bubble: market fluctuations, monetary policy, and government regulation and fiscal policy. The

literature shows that government and fiscal policy provides the most complete explanation for

the expansion of the bubble. The housing market bubble was created by market forces and

exacerbated by government regulation and fiscal policy decisions. These decisions created

unintended consequences by encouraging unsound banking practices in the form of relaxed

capital requirements, encouraging excessive borrowing through penalty-free refinancing, and

encouraging “greedy” lending by requiring more subprime lending. The effects of the 1995

Community Reinvestment Act (CRA) on the increased risk associated with mortgage loans in the

market will be tested through three separate difference of means tests.

The original 1977 legislation of the CRA intended to reduce redlining and increase the

availability of loans to low- and moderate-income lenders in the community surrounding a

banking institution (Federal Reserve 2018). This legislation encouraged institutions to lend to

these individuals through ensuring that there was an appropriate process for evaluating

applications. The amendment in 1995 was put in place to assess how well banks were actually

performing the stated goals of the CRA, not simply the process that the banks were using to

comply (Federal Reserve 2018). Institutions were subjected to lending tests to evaluate the ratio

of lending that the institution does in the low- to moderate-income neighborhoods (Brevoort

2022). In order to increase the amount of lending, the policy allowed for banks to either create

original loans, which is difficult, or purchase loans that already existed in the market, which is

easier and can lead to repetitive re-selling of said loans (Community Reinvestment Act

Regulations, 2001; Community Reinvestment Act Regulations, 2004; Brevoort 2022).

This topic is interesting to scholars because studying the connection between how policy

and economics influence each other can give a better understanding of policy consequences, and

it can expand upon the current understanding of other market bubbles. There are multiple schools

of thought concerning market bubbles, and it is important for scholars to study and understand

these. This topic is of interest to policymakers because, to develop the best laws and courses of

action, they should have a robust understanding of the potential consequences of their law-

making actions. It is important for policymakers to have research available to them in order to

understand the tradeoffs that may come from their decisions. This topic is of interest to citizens

because many people may still feel the effects of the housing crisis, think of it when making

financial decisions, and/or fear something similar happening. When market bubbles are talked

about today, the housing bubble from the 2000s is commonly talked about, as well. There is still

a fear of the possibility that a housing bubble burst could lead to another recession because

people remember the 2000s housing crisis and recession and because of the 2023 bank failures. It

has been previously mentioned that some citizens and policymakers blame market fluctuations

for the housing bubble and bubble burst; this is a prime example of why this topic and type of

research are so important (Wallison 2011). Some people may think it was purely a market

problem when there is a scholarly consensus that it was absolutely not just the market, although

there is still debate as to what caused the most effects. When some groups think that it was just a

market issue, that gives policymakers–monetary and fiscal–an opportunity to put the blame

somewhere other than themselves; but this does nothing to make sure these problems are avoided

in the future.

Literature Review

The housing bubble and crash of 2008 had many confounding variables, but the results of

government policy had more power to influence these other variables . There are three main

schools of thought that debate what influenced the housing bubble and burst the most: market

fluctuations, monetary policy, and fiscal policy and government regulation. Scholarship that

argues market fluctuations were the driving force of the housing bubble points to the influence of

outside investment, greedy capitalism, and unsound investment decisions (Wallison 2011;

Levitin and Wachter 2012; Acharya et al. 2009). These factors are important to understanding

the emergence of the market bubble, but, by themselves, are not enough to understand why the

bubble expanded the way it did relative to previous bubbles. Scholarship that focuses on

monetary policy points to the federal funds rate that was kept low by the Federal Reserve as the

driving force of the housing bubble. This argument is supported by evidence that the Federal

Reserve dramatically changed its tactics (Taylor 2014), but it is rebutted by evidence that the

federal funds rate stopped directly influencing the housing prices up to several years prior

(Greenspan 2009). Scholarship that argues fiscal policy and government regulation was the

driving force of the housing bubble points to government intervention in the form of housing

policy regulations that had poor interactions with the stated market forces. The theme of

expansionary policy comes up frequently in the literature and is generally associated with

monetary policy enacted by the Federal Reserve, but it is important to note that it does not

always have to apply to monetary policy. Expansionary policy depicts policy that increases

demand for and directly or indirectly increases spending on a particular variable (Rowley and

Smith 2009). The evidence provided by fiscal policy literature is the best explanation for the

expansion of the bubble because it shows what new incentives there were in the housing sector

that sets this bubble apart from previous, milder bubbles.

Market Factors and Fluctuations

There cannot be a market bubble without market actions and fluctuations. It is, then,

without question that market fluctuations impacted the housing bubble creation and eventual

crash. In order to debate the question of “to what extent did policy impact the bubble”, the

impact of market-driven factors must also be understood and further debated in importance.

Market-driven forces need to be examined in a way that compares their influence with the

influence of other driving factors. The main factors to look at for explaining the bubble

expansion and burst in this school of thought include outside investment and influence, greedy

capitalism, and unsound investment decisions (Wallison 2011).

First, the reference to “outside investment” in this context means the foreign investment

into the United States market. In many countries, there had been increased savings, which meant

that there were increased funds that could be invested. Foreign countries, such as China, chose to

invest into what were thought of as the most secure assets; this includes assets held by

government agencies such as mortgage-backed securities from Government Sponsored Entities.

This foreign investment kept long-term interest rates low for longer than they naturally would

have been (Levitin and Wachter 2012). What foreign investors did not know was that there were

other incentives at play that would encourage these seemingly sound institutions to take riskier

actions (Wallison 2010).

Further, there is an argument that faults the low rate of global long-term interest rates for

increasing housing prices and the continuation of the bubble by influencing domestic long-term

interest rates to be low, including the mortgage rate (Greenspan 2009; McDonald and Stokes

2011). Greenspan (2009) argues that this influence from outside factors is what sets this period

of time apart from other bubbles.

Second, some argue that the bubble was expanded unnecessarily when the main method

of getting a housing loan changed from credible bank lenders to private lenders, who were

assumed to be motivated by greed (Levitin and Wachter 2012). These greedy private lenders

were, then, the cause of the risky repackaging of loans and the originators of understating the

risks involved. Institutions of the shadow banking sector, the private lenders, that become of

consequence at this time will be referred to as “nonbanks” to distinguish them from traditional

banks. The nonbanks fell into a trap of lending, selling cheap, rescue investing, lending, etc

(Acharya et al. 2009). The largest problem arose once the short-term investors–whom the

nonbanks relied on for inflow of money–felt their assets were not safe, and made a run on these

already weak nonbanks. This stance argues that the failure of the shadow banking system is the

cause of the financial crisis (Acharya et al. 2009).

Third, unsound investment decisions encompass relaxed reserve requirements and

mortgages of understated risks. The banks’ abilities to avoid reserve requirements allowed them

to give out more loans than they could safely distribute (Acharya and Richardson 2009). This

over-lending knowingly put the institutions at financial risk of not being able to survive a

hypothetical run on the bank or a large number of defaulted loans. While this may seem similar

to the avoidance of established standards discussed later in the fiscal policy school of thought,

the former factor was not created by the same policy that made the latter. Another aspect of

unsound investment decisions was that financial entities/institutions repackaged risky mortgages

into securities that appeared to be less risky than they actually were. This factor might appear to

originate solely in the market, but it becomes convoluted with aspects of fiscal policy, since

fiscal policy is what fueled the need to have subprime loans given out. Repackaging them made

them stable enough to keep the cycle going (Acharya and Richardson 2009).

Another way of dividing these factors is by categorizing them as demand-side or supply-

side issues. Demand-side issues are characterized by there being too much demand in the market

for loans. It is established in the literature that there was an increase in demand for housing

during the period of the bubble; this explains why housing prices had to rise to combat the

demand (Taylor 2014). However, this does not account for why mortgage rates stayed low and

why there were so many opportunities to get loans. One could argue that since demand for

housing rose, both housing prices and mortgage prices should have risen. What actually

happened is that the supply of mortgages grew faster than the demand for them, which

necessitated that the price be lowered (Levitin and Wachter 2012). This means that there must

have been other factors that just normal market fluctuations influencing the amount of mortgage

opportunities and prices. Supply-side issues can be described as the excessive quantity and poor

management of financing options for housing and mortgages (Acharya et al. 2009; Levitin and

Wachter 2012). An example of this is that the nonbanks provided an excessive quantity of

financing options for housing and mortgages; this makes this market factor a supply-side issue

because there were too many opportunities made available (Acharya et al. 2009).

Monetary Policy

Scholarship on the influence of monetary policy argues that the Federal Reserve (“The

Fed”) kept the federal funds rate artificially low, thereby influencing the expansion of the

housing bubble. An argument is made describing how if the Fed increased rates, the bubble

would not have inflated to the unprecedented amount that it did, and, therefore, would not have

crashed as hard. The most debated variable in this school of thought is the federal funds interest

rate, which is an overnight rate, a short-term interest rate, and is controlled by the Federal

Reserve. In order to fully understand the impact the Fed is able to have, it is important to note

that monetary policy can be expected to take one and one-half years to translate into the market

(Peters 2022). Making low interest rates or keeping them low is a form of expansionary policy.

The Fed provided ease of access to credit by implementing expansionary policy and keeping

rates too low for too long, which allowed investors to ignore risk and over-invest (McDonald and

Stokes 2011).

Many scholars argue that the Fed keeping the interest rate artificially low for the period

between 2002-2005 exacerbated the housing bubble (McDonald and Stokes 2011; Taylor 2014;

Krugman 2009). The strongest argument made for this view is that the Federal Reserve strayed

too far away from the Taylor Rule, which encourages market equilibrium. The Taylor Rule is an

economic tool that is used as a guideline describing how the target or goal for short-term interest

rates should be chosen. The Taylor Rule states that the federal funds rate should be equal to

inflation plus the equilibrium rate that would have full employment plus the weighted average of

an inflation and output gap (Mishkin 2016). Basically, monetary policy creators should raise

nominal interest rates more than the amount of inflation (Mishkin 2016). Even though the Taylor

Rule accurately forecasted monetary policy implications in the prior decades, and, despite

previously following this policy rule, the Federal Reserve held the federal funds rate significantly

below these guidelines during 2002-2005. The year 2003 saw a record low federal funds rate

(Zandi 2009).

In 2003, inflation was two percent, while the federal funds rate was only one percent.

However, in the decades prior, inflation was two percent, and the federal funds rate was five and

a half percent (Taylor 2014). With the federal funds rate being below inflation, the policy being

used in 2003 was clearly different from what the Fed used to use; and it clearly was not

following the Taylor Rule any more. The Fed argued this was in part because of their concerns

with deflation, a condition that is considered more economically damaging to borrowers than

inflation. Even so, it should be expected that such a dramatic change in policy would have some

kind of negative consequence, even if the specific result cannot be predicted (Taylor 2014).

One study wanted to conduct an empirical test to determine if monetary policy had an

effect on developing the housing bubble by influencing the changes of housing prices. McDonald

and Stokes (2011) use granger causality analysis, an economic model that attempts to use one

variable to predict another variable, on time-series data to test if there was a causal relationship

between the variables housing price index and the federal funds rate. These authors conclude that

there is a granger causality between the variables that suggests a statistically significant

relationship. The researchers do note, however, that the relationship between the housing price

index and the federal funds rate did fluctuate when it was expected to be steady, suggesting that

something else also influenced the housing bubble (McDonald and Stokes 2011). This makes

sense with what the literature shows in other schools concerning the bubble from market and

policy forces.

Another point that Taylor (2014) asserts is that the Fed further contributed to the

exaggeration of the bubble because it misdiagnosed the problem that was occurring. Taylor

argues that the Fed mistakenly thought the problem in the market was a lack of access to

liquidity, so the Fed acted accordingly by using expansionary policy, including but not limited to

a low fed funds rate. However, when traditional banking institutions did not respond to the

expansionary policy and they began to fail, the first bailouts started. At this point, the bubble had

progressed so that it was too late for raising the federal funds to have a distinct impact. This

failure of lending institutions was the beginning of the financial crisis (Taylor 2014). When the

Fed bailed out Bear Stearns, which ignored the bankruptcy laws already in place, people

assumed that Lehman Brothers would also be bailed out. It was not. Subsequently, whatever

confidence had been left in the securities and loans was gone, and investors saw the gravity of

the situation. (Wallison 2010; Taylor 2014).

There are other scholarly opinions that contest this argument. Granger causality can be a

good guide that identifies that a relationship exists, but this model is still one of the weakest

causality relationships in economic models. While Granger causality can establish causality

between two variables, it cannot necessarily describe a primary causal factor, nor can it describe

if another variable had a more substantial effect on the housing bubble. Therefore, the

relationship it describes as being granger causal is not equivalent to the true “cause and effect”

that is searched for in the social sciences (Padav 2021). There is also disagreement among

scholars about the application of the Taylor Rule. The best person to argue why the Taylor Rule

did not apply in this situation would be the Chair of the Federal Reserve at the time, Alan

Greenspan.

Greenspan (2009) argues that the federal funds rate is not used as a determining factor in

setting housing-market rates. Instead, Greenspan argues that the long-term fixed rate mortgages

were the more relevant interest rate. The federal funds rate is an overnight rate, so it can reflect

the state of the current economy, but it does not dictate things such as mortgage rates. The value

of long-term assets, like real estate, are formed from interest rates that reflect the size and

lifespan of the asset. Greenspan argues that home mortgage rates had stopped following the

trends of the federal funds rate several years prior to the bubble burst. For decades, the mortgage

rate was in sync with the movements of short term interest rates, such as the federal funds rate,

and there was a statistically significant correlation of .85 (Greenspan 2009). From 2002-2005,

this relationship was found to drop to insignificance, indicating that the federal funds rate no

longer had predictive power over the mortgage rate. The strongest evidence supporting this

argument is that the Fed implemented contractionary policy several years before the bubble

burst, yet the home mortgage rates did not reflect that change (Greenspan 2009). Taylor rebuts

the idea that the federal funds rate has no impact on housing demand by arguing that, yes,

housing demand is largely impacted by long-term fixed rate mortgages, but it is still impacted by

short-term rates. One reason these were still connected is because the short-term rate allows for

the existence of low teaser rates which draw in customers (Taylor 2014). An example of low

teaser rates would be how credit cards and car dealerships advertise zero percent interest for the

first six months, or something similar.

Some other scholars agree with Greenspan’s claim that the federal funds rate became

detached from the mortgage rate, noting that the housing bubble was speeding up in 1999 despite

that the Fed was increasing rates. Further, even though the low federal funds rate was

simultaneous with the large increase of housing prices in 2001-2003, the housing price increase

lasted three times as long as the low interest rate phase (McDonald and Stokes 2011; Shiller

2009), implying that there must have been an outside factor also encouraging the growth of

housing prices. Unfortunately, monetary policy can never be perfect because it relies on

imperfect information and predicting an inherently unpredictable market (Sánchez 2011). This

school is limited in explaining the bubble because the impact that monetary policy could have

had on the market is insufficient to explain why the market reaction lasted three times longer

than expected. Instead, the argument that fiscal policy had a greater effect provides explanations

for the other incentives at play.

Government Legislation and Fiscal Policy

This school of thought argues that government regulation and interference had poor

interactions with uncontrollable market forces, resulting in unintended consequences that sent the

housing market into a crisis. This school of thought looks at implications of government housing

policy regulations and government sponsored enterprises in relation to how they affected the

bubble and ensuing crisis.

Rowley and Smith (2009) argue that expansionary fiscal policies by the federal

government–which can be in the form of cutting taxes or increasing spending–contributed to the

creation and continuation of the housing bubble and the ensuing financial crisis. They

specifically point to tax cuts and greater spending as causes, which are exemplified throughout

the literature for this school (Rowley and Smith 2009).

Relaxed capital requirements, penalty-free refinancing, and increased subprime lending

are all factors that incentivized buying into the housing bubble and making it bigger, which

enlarged the fallout that would come (Wallison 2009). Policies that were implemented were

designed to undermine and influence capitalism and, according to proponents of this school,

stemmed from the outdated and impractical idea that policy is supposed to fix economic

problems. Convoluted policies unnecessarily promoted subprime mortgages and risky

investments (Friedman 2009).

The two specific fiscal policies that are discussed the most are the Housing and

Community Development Act of 1992 and the Community Reinvestment Act revisions in 1995

(Wallison 2009 and 2011). The Housing and Community Development Act of 1992 increased

spending on low-income housing and attempted to create a check on government-sponsored

enterprises (Wallison 2011). The Community Reinvestment Act (CRA) was amended to

influence to whom financial institutions gave loans (Wallison 2010). The financial institutions

most impacted were Government Sponsored Enterprises (GSEs). GSEs are government-created,

privately-held financial institutions that act only in a specific area of the market (Segal 2022).

The GSEs that are of concern to this topic are Fannie Mae and Freddie Mac, which both operate

in the housing sector. Fannie Mae impacts the availability of loans, and Freddie Mac promotes

homeownership amongst the middle and low classes (Segal 2022).

To be more specific, elected officials amended the CRA to encourage more affordable-

housing initiatives. An affordable housing loan is a loan given to a borrower who has equal to or

less than the median income for the area (Wallison 2011). The government used this legislation

to change the regulations and quotas surrounding how institutions gave out affordable housing

loans. The government increased the amount of credit accessible to low-income borrowers

(Wallison 2011). The original legislation stated that 30 percent of the loans given out by these

institutions had to be affordable loans. This requirement kept increasing until, by 2007, 55

percent of loans had to be affordable housing loans. This requirement also had a subgoal that 25

percent of loans be for low-income borrowers in the area. In this case, low-income means the

borrower has an income equal to or less than 60 percent of the median income of the community

(Wallison 2011).

This resulted in Fannie Mae and Freddie Mac being required to have an increased amount

of affordable housing loans, the Federal Housing Administration (FHA) was required to provide

loans/insure mortgages for borrowers whose income was equal to or less than the median

income, and insured banks also had restrictions placed on them, with a requirement of providing

loans to borrowers who had 80 percent of median income (Wallison 2011). Fannie and Freddie,

FHA, and insured banks then had to compete for loans to provide to low to median-income

borrowers since they had a quota to reach. The competitive nature that was created by

government policy and placed on Fannie Mae and Freddie Mac shut out any other smaller firms

that could give median income loans (Wallison 2010). These main institutions found themselves

settling for subprime and low-quality loans in order to reach a government-required quota. By

2008, about two-thirds of low-quality loans were backed by government entities. This further

exemplifies that the amount of subprime loans that were administered was not a primarily

market-driven factor. By these calculations, only one-third of subprime loans in circulation were

non-governmentally required or promoted. An unprecedented amount of subprime loans that

were on the market were weak and risky investments. For Fannie Mae and Freddie Mac alone,

the subprime loans they were required to meet put them at risk for 400 billion dollars (Wallison

2011).

The aforementioned incentive of penalty-free refinancing includes tax preferences for

home-equity borrowing (Wallison 2009 and 2011). The continuation of the bubble and access to

credit allowed people to keep refinancing; this continued until they could not take out more

loans, and an unprecedented amount of defaults began. Institutions were unable to prepare for the

amount of defaults that would occur. Defaults began occurring due to rising interest rates, the

high cost of loans, falling real estate prices, and no more options for refinancing because of home

equity falling (Wallison 2011).

These policies created incentives for subprime lending and incentivized buying into the

housing bubble and making it bigger, which enlarged the fallout that would come (Wallison

2009). Because the bubble was artificially encouraged, it lasted twice as long and acquired more

capital investment than any previous bubble seen in the sector. It saw a real increase in housing

prices of 80%, whereas former bubbles had price increases of 10% (Wallison 2011). Previous

bubbles deflated in just a few years because poor-quality loans would default quickly, and the

bubble’s growth would be stunted. However, this housing bubble, which lasted for a decade, was

getting money and resources poured into it by government requirements. Intuitively, the longer a

bubble lasts, the more dangerous it becomes. The longer a bubble lasts, the more confident

people are that it will continue to grow, and the level of risk individuals are willing to take

(Wallison 2011).

This school of thought is still debated. Some claim that there is not enough empirical data

to assert confidently that the policies of the government heavily impacted the housing bubble. A

study conducted by Avery and Brevoort (2015) looks at the CRA policy and the GSE institutions

to try to see if one was more associated with detrimental loans. The study then uses regression

discontinuity tests to see if low-income areas had worse loan outcomes than the standard loan

area. The authors of this experiment argue that neither program outcome is significantly different

but that they cannot say from this data that they did not impact the bubble at all. According to the

researchers, this study was not a good judge of the loans from GSE lenders. The researchers also

find that low-income areas did have a worse outcome with the loans due to the interaction with

these loan policies, but these policies were not the only factor influencing that outcome (Avery

and Brevoort 2015). To determine that factor or factors in this study would require future

research.

Analysis of the Literature

Each of the schools considered has strengths and weaknesses in describing why the

housing bubble expanded dramatically, which inevitably led to a dramatic burst. It has been

mentioned that some citizens and policymakers blame unpredictable market fluctuations for the

crisis. The most referenced point of blame is that the predatory actions of greedy lenders caused

the increase in subprime loans given out, which subsequently weakened the whole system; yet

the school offers no explanation of why these lenders were any different from the greedy lenders

that exist at all times and in all market bubbles. Scholars recognize the importance of market

factors as the beginning of the bubble, but these arguments cannot explain why the size and

duration of this bubble was different from previous market bubbles.

The scholarship on monetary policy offers insight as to what could have influenced the

market. There is a consensus among scholars that if the Fed's monetary policy is found to have

had an effect on the housing prices, it was the actions or inactions taken through the federal

funds rate. The fed funds rate is the tool that provides the greatest and most direct power that the

Federal Reserve has over market bubbles. This is the primary tool that had been used to

influence the market for decades (McDonald and Stokes 2011; Taylor 2014; Krugman 2009).

Even scholars that argue that it was not the fault of monetary policy recognize that the Taylor

Rule is, in general, a good guide to follow (Mishkin 2016). Not everything the Fed did was

wrong, but their decision to keep the interest rate low was an important one (Taylor 2014).

Despite the fact that housing prices grew larger than can be explained solely by the fed funds

rate, Shiller (2009) argues that rate cuts in 2004 would have impacted the housing prices since

that was the year of fastest and unpredictable price growth; the rate cuts would have exaggerated

the growth that was going to happen anyway in 2004. It has been shown that expansionary policy

through the fed funds rate has the ability to impact interest rates and inflation in the economy

(Taylor 2014). However, it is also shown that there is not a direct and immediate correlation

between the fed funds rate and the housing market rate (Greenspan 2009; Peters 2022), which

necessitates considering what else impacted the low housing market rates.

The arguments that “policy is the problem” have more empirical evidence than the idea

that the problem resided in the market equilibrium being negative (Taylor 2014). In other words,

it makes more sense that the issue was because of policy rather than being a problem with the

market’s tendency to balance itself. The government’s promotion of subprime loans negatively

interacted with the market aspect of greedy lenders understating the risk associated with a loan.

This is evidenced by the fact that the GSEs reported 12 million of its subprime loans as prime

loans, which severely understated the risk that investors and borrowers were signing on to

(Wallison 2010). Fiscal policy's expansionary actions encouraged not only a greater demand for

housing loans but also a greater spending in the housing sector. What made this so detrimental to

the market is that these new requirements promoted an unforeseen amount of loans that were

largely comprised of weak and risky investments (Wallison 2011). The subsequent default of

these understated risky loans shocked the lenders, destabilized the already weak system of asset-

backing the loans, and drove real estate prices down faster and at a greater amount than seen in

previous bubbles (Wallison 2010).

Research Design

To what extent did government policy or intervention impact the expansion of the bubble

leading up to the 2008 financial crisis? This is a non-obvious answer because there is still debate

over a decade later as to what the driving factor was. It is clear that at least one factor, likely

many factors, made the bubble grow unnaturally large and at a fast pace. What is unclear is

exactly how these factors influenced the market, both individually and congruently. This

research strives to bring more light to the topic by looking at one individual aspect of the

situation, government policy’s role in the bubble. The 1995 change to the Community

Reinvestment Act promoted incentives for providing a greater number of risky, default-inclined

loans. This will be evaluated by three indicators of risk associated with mortgages. I derived the

data for this research from the FRED, Federal Reserve Economic Data, database. This is a

widely used, reputable source for economic data that is drawn from records of various

government agencies and other trusted sources.

The independent variable (IV) will be the 1995 change to the Community Reinvestment

Act (CRA). The range of years under consideration for the dependent variables will be through

2006, twelve years of the implemented policy change. To keep it even, the range for before the

policy implementation will begin with 1983. The first dependent variable (DV

) will be the risk

premium of mortgage loans for the time period. The second dependent variable (DV

) will be the

mortgage debt outstanding, in terms of real debt per capita. The third dependent variable (DV

)

will be the debt-to-income ratio.

Limiting the range to 2006 ensures that the data being measured is valid by only

measuring the possible effects of the independent variable through the expansion of the bubble.

The bubble ended in the 2nd quarter of 2006. This saw the peak of housing prices (see Figure 1)

and newly built homes (see Figure 2). There is a general consensus that after the peak in 2006,

the bubble started to fizzle, followed by a hard crash (Kosakowski 2023). This research is

concerned with what caused the build-up of the bubble to be excessively large, which,

consequently, caused the crash to be large. Looking after 2006 does not concern this research

because there are other incentives at play that influenced people's decisions to purchase or not

purchase homes and loans.

Figure 1 Case-Shiller Housing Price Index

The Case-Shiller Housing Price Index is the standard measure used for average house prices. The

graph is taken from the FRED, which sourced the data from S&P Dow Jones Indices LLC.

Figure 2 New Housing Construction in US

New housing construction in 2006 saw the last peak, marked on the graph, before the first major

market dip signaling the ending of the bubble. The graph is taken from the FRED and originally

derived from the U.S. Census Bureau.

The risk premium (DV

) indicates the amount of risk that investors believe exists in the

market; investors have to believe that the market is safe enough to put their money into it and

have a degree of assurance that they will get a return. The risk premium is measured by finding

the difference between the expected return of the asset and the risk-free interest rate, such as a

government-backed loan. The variable referring to mortgages will be measured by the 30-year

mortgage rate minus the government-backed 10-year treasury rate. This is a valid and accurate

indicator of the level of risk in the market because it looks at the long term mortgage rate at

different points in time in order to accurately depict the interest of loans that were being taken

on. It also looks at the long term government bond rate, which is the most accurate rate to

consider, even though the short term rates tend to follow the same trend. This is a reliable

measure because the data comes from a reliable source and this definition is widely used in

economics. The risk premium describes the risk investors see at a given time by depicting the

difference between the two rates. If the premium is increasing, then investors believe there is

more risk. In normal times, one would expect to see the premium rise as demand for loans rises

in order to control the growth of the amount of loans given out. However, this logic is not

consistent within a bubble because bubbles are inherently expansionary, which means that

investors tend to disregard risk. I hypothesize that government intervention in 1995 will lead to

little growth of the risk premium.

Mortgage debt outstanding shows the total amount of mortgage debt held, both active and

delinquent accounts. Alternative predictors for increasing mortgage debt outstanding are changes

in inflation and in population size. I plan to control for changes in inflation and population over

time by adjusting mortgage debt outstanding to the Consumer Price Index (CPI) and population.

Including CPI in the equation adjusts it for inflation and puts the data in real terms, making it

comparable over time. Including population controls for the fact that as population rises, total

debt should, intuitively, rise, as well. The mortgage debt outstanding, in terms of real debt per

capita (DV

) is measured by the percent change of mortgage debt outstanding minus the percent

change of CPI minus the percent change of population. Looking at the percent change of this

variable puts the data in terms of the growth rate, which makes it more intuitive to understand.

The analysis of this variable allows one to evaluate if the primary problem was demand-side or

supply-side. Real mortgage debt outstanding, per capita, is an indicator of increased risk taken on

by the individual borrower because taking on more debt without an increase in income means

that the debt taken on is subprime, or low quality. I hypothesize that after 1995, the mortgage

debt outstanding, in terms of real debt per capita will rise.

The debt-to-income ratio for mortgage loans (DV

) describes the amount of debt in the

economy compared to the amount of income. This shows how much income people actually

have compared to how much the loans are worth that people are taking out. Both of these

components of the equation will be in terms of percent change and nominal values. This ensures

that inflation as an alternate predictor is accounted for because inflation will influence the

nominal values in a similar way and cancels out in the long equation. This is a reliable and valid

measure because this is the widely used concept and equation for measuring debt-to-value ratios.

An analysis that produces a high ratio suggests that people are overextending themselves with a

loan that they do not have the means to pay sustainably. This is related to the previous variable,

, but it measures the concept differently. This will be interesting to see the outcome of both

to determine if the results are consistent, or if future research should look at them differently and

more in depth. I hypothesize that after 1995, the debt-to-income ratio will rise.

To test this hypothesis, I use an independent samples t-test. The method of statistical

analysis that will be used is a difference of means test, also known as an independent samples t-

test. The purpose is to determine if two sets of results are statistically different. This testing

method compares the average of the outcome affected by the variable to the average of the

outcome of the control group. This test is most effective when considering a nominal

independent variable and a continuous dependent variable. The years 1983 through 1994 are

coded as a 0. The years 1995 through 2006 are coded as a 1. I measure all of these dependent

variables in terms of annual averages. The unit of analysis is the year that the data points are

measured.

Discussion of Findings

: Risk Premium

My first model tests the relationship between the CRA policy and the growth rate of the

risk premium. Table 1, DV

depicts the change in growth of the risk premium across the two

periods. Based on my independent samples t-test, the two-sided p-value for this analysis is .902,

which shows that the policy change in 1995 did not have a statistically significant impact on the

risk premium. This analysis shows that there is no significant difference in the periods, so there

is no growth or contraction in the premium during the expansion of the bubble.

Table 1: Independent Samples t-test for Pre-CRA Period and CRA Period

t Mean difference

95% confidence

interval

Growth of risk

premium (DV

)

0.125

0.016

(0.127)

[-0.249, 0.280] 22

Mortgage debt

outstanding

(DV

)

-1.993

-2.844*

(1.427)

[-5.805, 0.117] 22

Mortgage debt-

to-income ratio

(DV

)

-2.513

-0.687**

(0.274)

[-1.263, -0.116] 22

Note: Standard error of the difference in parentheses

* p<.10; **p<.05; ***p<.01

Figure 3 displays how the rates of the 30 year mortgage loan and 10 year government

bond largely move in tandem with each other, showing that there was not an increasing nor

decreasing premium rate between the two. In a normal circumstance, it would be expected to see

an increase in the risk premium in order to combat the increasing demand for loans, yet this

shows that it is being kept low. This suggests that something prompted investors to allow the

high demand for loans to continue throughout the bubble. This data suggests a support of the

hypothesis that government intervention in 1995 led to little growth of the risk premium during

the bubble, which was at a time of high demand. If there were no outside influence on the risk

premium, it should have risen to attempt to curb this high demand that exacerbated the bubble.

Figure 3 Prime Rate

: Mortgage Debt Outstanding, in terms of Real Debt per Capita

Table 1, DV

displays the results of the effect of the CRA on the percent change of real

mortgage debt outstanding per capita across the two periods. The two-sided p-value for this

variable is .059, therefore, it does not meet the conventional 95% confidence level of

significance. However, the p-value is below the .10 suggesting 90% confidence in statistical

significance.

The simple error bar chart shows that the confidence intervals overlap, so we cannot say

with great certainty that these values are different from one another than by random chance.

Figure 4 Percent Change Real Mortgage Debt Outstanding Per Capita

The line graph shows the percent change in the variable through time in a more visual way.

Figure 5 Percent Change in Debt-to-Income Ratio

Mortgage debt outstanding, in terms of real debt per capita, allows researchers to evaluate

if the growth was influenced by an increase in the demand-side or supply-side. This is done by

recalling the rising prices for homes and comparing it to the growth rate of DV

. Intuitively, it

can only be supply driven if housing prices are decreasing as homes are being built; the reference

graphs previously discussed show that this is not the case. Therefore, the data suggest that the

possible increase in DV

after 1995 would be demand driven; this agrees with the intuitive

hypothesis that the CRA policy change created demand. Again, further examination of these

variables would be necessary to make these claims in more definitive terms.

: Mortgage Debt-to-Income Ratio

Table 1, DV

depicts how much debt, on a national average, people are taking out

compared to their income across the two periods. The two-sided p-value for this variable is .022,

which is within the conventional level of confidence. This relationship is statistically significant

at the 98th percentile. There was an increase in the percent change of the ratio after 1995. This

shows support for the hypothesis that after the 1995 policy, the debt-to-income ratio would rise.

The increase in this ratio, separate from the impacts of inflation, suggest that some external

factor had to be at play to encourage an increased amount of demand for mortgages. The line

graph below shows the percent change in the variable through time in a more visual way.

Figure 6 Percent Change Debt-to-Income Ratio

Conclusion

The statistical significance of the debt-to-income ratio (DV

) after the 1995 change to the

CRA supports the hypothesis and the related claims in the literature that the CRA amendment

artificially influenced the market to have more demand for housing loans. This was done without

an appropriate increase of new housing construction, otherwise, a decrease in housing prices

would be seen as well. There is a scholarly consensus that this is an accurate description of the

state of housing prices and construction, and it has been shown through the simple graphs

discussed earlier. The real mortgage debt outstanding, per capita (DV

) suggests a weak

correlation with the findings from DV

. It indicates a possible trend of increased mortgage debt

after the 1995 policy change, which supports the debate presented in the literature that the CRA

encouraged greater spending on housing loans. The risk premium (DV

) suggests that since there

was no significant change in the perception of the risk associated with giving out more housing

loans, contrary to what one would expect to see in an expansionary period of rising housing

prices, an outside factor influenced how investors behaved. It is reasonable to conclude that the

CRA change is the factor that provided investors with enough incentive to keep investing in the

housing market, even when it was clear that the bubble was continuing without the normal

checks between supply and demand.

Many factors contributed to the unexpected growth of the housing bubble, including

natural market factors and monetary policy decisions. Uncontrollable market components such as

foreign investment, greedy investors, and unsound investment decisions are always at play and

will always have a role in the market; therefore, this is, by itself, an insufficient argument to

describe the expansion of the housing bubble. The regulators of the Federal Reserve allowed

financial institutions to diverge from the well-established path of safe lending practices by

changing the short-term interest rate that encourages borrowing (Taylor 2014). This is also not a

complete argument, however, because it has been established that this short-term rate is not

influential enough to create the long-lasting reactions that were seen in the market. All of the

factors that are discussed in their respective literature were clearly further influenced by an

additional increase in incentives to participate in risky practices. This additional influence came

in the form of government intervention through regulation and fiscal policy. Regulators of

Fannie Mae and Freddie Mac allowed financial institutions to further deviate from the

established safe practices regarding risk and the amount of default-prone loans given out.

Still in this school is the discussion of government affordable-housing initiatives, namely,

the 1995 amendment to the Community Reinvestment Act. This change encouraged unsafe

lending practices by requiring an increased amount of subprime, risky loans to be given out.

This, in turn, encouraged lenders to make risky decisions based on these new requirements.

Incentives such as penalty-free refinancing and the abundance of repackaged loans encouraged

borrowers to make unsound investment decisions, as well.

The first major limitation of this research is that I am only running bivariate analyses.

This does not allow me to control for the alternative variables in the competing schools of

thought. It is possible that after controlling for alternative predictors in market forces and

monetary policy, there might be no relationship between government policy and the housing

bubble. To expand this study, future research should look at multiple causes.

One limitation I found was in choosing the dependent variable for my analysis because I

could not find consistent data on two of my ideal indicators. There is not a consistent tracking of

the amount of subprime loans, which is the ideal indicator of how risky loans are, until 2008

when the banking institutions realized it was a problem that should be looked at more directly.

Additionally, during the time period being considered, there were 12 million subprime loans that

were repackaged and reported as prime loans, therefore, that would not be an accurate measure

unless controlled for by a variety of variables. For this case, I, instead, chose variables that

indirectly describe risk in the market. The second limitation I found is that the data for mortgage

delinquency rates is not consistently tracked and reported, at least not where I can find it at this

time. For example, it is reported from 1953-1963, which is not relevant to this research question.

It is also reported from 1991-2022, which is relevant but does not provide enough data points for

before the policy change and year of focus, which, again, is 1995. Instead of this, I had to look at

total mortgage debt outstanding, which includes active accounts, whereas I would have preferred

to only look at the change in the amount of defaulted debt in the economy before and after the

CRA change.

There are many avenues for future research on this topic. From my data, real mortgage

debt outstanding per capita might suggest that the CRA influenced a rise in the amount of

mortgage debt per person, but further research would need to be done in order to say that with

true statistical significance. The first aspect that should be considered is if analyzing the data

based on quarterly instead of annual rates has a significant impact on the results.

Another interesting avenue for future research would be to look at how the policy

implication affected the recovery of the economy compared to other bubbles. This would also

mean taking into account policy implemented after the initial crash, which widens the scope of

the research.

References

Acharya, Viral V., and Matthew Richardson. 2009. “Causes of the Financial Crisis.” A Journal

of Politics and Society 21(2-3): 195-210. https://doi.org/10.1080/08913810902952903

Acharya, Viral V., Thomas Philippon, Matthew Richardson, and Nouriel Roubin. 2009. “The

Financial Crisis of 2007-2009: Causes and Remedies.” Financial Markets, Institutions,

and Instruments 18(2): 89-194.

Avery, Robert B., and Kenneth P. Brevoort. 2015. “The Subprime Crisis: Is Government

Housing Policy to Blame?” Review of Economics and Statistics 97(2): 352-363.

doi:10.1162/REST_a_00491

Baily, Martin Neil, Robert E. Litan, and Matthew S. Johnson. 2008. “The Origins of the

Financial Crisis.” Initiative on Business and Public Policy at Brookings.

Brevoort, Kenneth P. 2022. “Does Giving CRA Credit for Loan Purchases Increase Mortgage

Credit in Low-to-Moderate Income Communities?” Board of Governors of the Federal

Reserve System. Finance and Economics Discussion Series 2022-047.

https://doi.org/10.17016/FEDS.2022.047.

Calculated Risk. 2006. “Wikipedia: U.S. Housing Bubble.” Calculated Risk: Economics and

Finance. https://www.calculatedriskblog.com/2006/09/wikipedia-us-housing-bubble.html

(April 27, 2023).

Community Reinvestment Act Regulations (2001) “Joint Advance Notice of Proposed

Rulemaking,” Federal Register, 66(139), pp. 37602–8

Community Reinvestment Act Regulations (2004) “Joint Notice of Proposed Rulemaking,”

Federal Register, 69(25), pp. 5729–47

Federal Reserve. 2018. “Community Reinvestment Act (CRA).” Board of Governors of the

Federal Reserve System.

https://www.federalreserve.gov/consumerscommunities/cra_history.htm (May 7, 2023).

Friedman, Jeffrey. 2009. “A Crisis of Politics, not Economics: Complexity, Ignorance, and

Policy Failure.” A Journal of Politics and Society 21(2-3): 127-183.

https://doi.org/10.1080/08913810903030980

Greenspan, Alan. 2009. “The Fed Didn’t Cause the Housing Bubble.” The Wall Street Journal.

Kenton, Will. 2022. "What Was the Subprime Meltdown? What Happened and Consequences."

Investopedia. https://www.investopedia.com/terms/s/subprime-

meltdown.asp#:~:text=The%20housing%20boom%20of%20the,to%20individuals%20wi

th%20poor%20credit.

Kosakowski, Paul. 2023. “The Fall of the Market in the Fall of 2008.” Investopedia.

https://www.investopedia.com/articles/economics/09/subprime-market-

2008.asp#:~:text=In%202008%2C%20the%20housing%20market,defaulting%20on%20t

he%20home%20loans. (April 27, 2023).

Krugman, Paul. 2009. The Return of Depression Economics. New York: Norton.

Levitin, Adam J., and Susan M. Wachter. 2012. "Explaining the Housing Bubble." Georgetown

Law Journal 100(4): 1177-1258.

McDonald, John F., and Houston H Stokes. 2011. “Monetary Policy and the Housing Bubble.”

The Journal of Real Estate Finance and Economics (46):437-451.

https://doi.org/10.1007/s11146-011-9329-9

Mishkin, Frederic S. 2016. The Economics of Money, Banking, and Financial Markets. 11th ed.

Boston: Pearson.

Padav, Pallavi. 2021. “Granger Causality in Time Series Explained with Chicken and Egg

Problem.” Analytics Vidhya. https://www.analyticsvidhya.com/blog/2021/08/granger-

causality-in-time-series-explained-using-chicken-and-egg-problem/ (March 5, 2023).

Peters, Katelyn. 2022. “Response Lag Definition.” Investopedia.

https://www.investopedia.com/terms/r/response_lag.asp#:~:text=The%20impact%20of%

20tax%20cuts,any%20effect%20on%20the%20economy. (April 5, 2023).

Rowley, Charles and Nathanael Smith. 2009. Economic Contractions in the United States: A

Failure of Government. Fairfax: Locke Institute.

Sánchez, Manuel. 2011. “Financial Crises: Prevention, Correction, and Monetary

Policy.” CATO Journal 31(3): 521–34.

Segal, Troy. 2022. “Government-Sponsored Enterprise (GSE) Definition and Examples.”

Investopedia. https://www.investopedia.com/terms/g/gse.asp

Shiller, R. 2009. The Subprime Solution. Princeton: Princeton University Press.

S&P Dow Jones Indices LLC, S&P/Case-Shiller U.S. National Home Price Index

[CSUSHPISA], retrieved from FRED, Federal Reserve Bank of St. Louis;

https://fred.stlouisfed.org/series/CSUSHPISA (April 27, 2023).

Taylor, John B. 2014. “The Role of Policy in the Great Recession and the Weak Recovery.” The

American Economic Review 104(5): 61-66.

The White House Archives. 2006. “President Highlights Importance of Small Business in

Economic Growth.” https://georgewbush-

whitehouse.archives.gov/news/releases/2006/01/20060119-2.html (April 27, 2023).

U.S. Census Bureau and U.S. Department of Housing and Urban Development, New Privately

Owned Housing Starts in the United States by Purpose of Construction, Built for Sale

Total One-Family Units [HOUSTPFST1FQ], retrieved from FRED, Federal Reserve

Bank of St. Louis; https://fred.stlouisfed.org/series/HOUSTPFST1FQ (April 27, 2023).

Wallison, Peter J. 2009. “Cause and Effect: Government Policies and the Financial Crisis.” A

Journal of Politics and Society 21(2-3): 365-376.

https://doi.org/10.1080/08913810902934158

Wallison, Peter J. 2010. “Government Housing Policy and the Financial Crisis.” CATO Journal

30(2): 397–406.

Wallison, Peter J. 2011. “Three Narratives About the Financial Crisis.” CATO Journal 31(3):

535–549.

Zandi, Mark. 2009. Financial Shock (updated edition). Upper Saddle River: FT Press.