Powering Small Businesses Final Report March 2013
3
Treatment A: Pricing option 1 offered
Treatment B: Pricing option 2 offered
Treatment C: Pricing option 3 offered
Treatment D: Pricing option 2 offered, but converted to option 3 ex post.
Treatment E: Pricing option 1 offered, but converted to option 3 ex post.
This allows us to study 3 main questions:
(1) What is the impact of electricity?
To understand the impact of electricity on small retailers, we are administrating two baseline surveys,
three follow-up surveys and one endline survey and comparing relevant outcomes between retailers
offered the possibility to purchase a SOlite-3 in the treatment group with the highest take up –likely to
be treatment group C. These surveys collect data on business outcomes such as profits, revenue,
operating costs and operating hours. Through this randomized design, we will quantify the impacts of
providing light and energy to small retailers. We expect these benefits to include higher profits due to
stores being open longer and phone charging services being provided to customers.
(2) What is the impact of mobile repayments and mobile enforcement?
The Solite-3 automatically collects usage data in real time and Angaza’s internal systems collect
repayment data. Angaza Design has agreed to provide us with full access to this information. By
comparing adoption rates, repayment rates, usage and other outcomes between treatment groups B
and D, we can quantify the treatment effect of enforcement. Retailers in both treatment groups B and D
have been offered a low per hour price at the condition of being financially punished if they do not use
the Solite for more than Kshs 150 per week. Although retailers in treatment group D have decided to
purchase the product at these same conditions, the punishment is not enforced ex post. By comparing
both of these groups, we can determine the differences in behaviors when enforcement is cut
(Treatment group D). Similarly, the comparison between treatment groups C and D highlights the
selection effect of enforcement. Indeed, retailers in both groups do not get financially punished even if
the weekly Kshs 150 threshold is not reached. Average outcome differences might however exist as
retailers in both groups might have different characteristics having initially chosen to purchase the
product under different pricing schemes (each group’s self-selection was under different pricing terms).
(3) What is the impact of varying the price structure?
Groups are offered different per hour energy prices (Kshs 10 or Kshs 15) and comparing usage,
repayment rates and business outcomes between these groups will help us understand the importance
of pricing and price discrimination for small retailers. What impact on take up does an increase on per
hour price have? Does a lower per hour price encourage clients to consume more electricity and repay
more quickly or does it select unworthy clients who default? As part of this process, we will work closely