The Miracle of Microfinance? Evidence from a Randomized Evaluation Esther Duflo, Abhijit Banerjee, Rachel Glennerster, and Cynthia G. Kinnan, NBER Working Paper No. 18950, 2013 Presentation by Savanah Landerholm

About the Authors Esther Duflo Abhijit Banerjee Rachel Glennerster First three authors are from MIT and Cynthia is from Northwestern University This article supersedes the 2010 version, which reported results using one wave of endline surveys. Duflo looks into what works and what doesn’t (cutting edge) -- French Rachel Glennerster Cynthia G. Kinnan

Introduction Microfinance or microcredit is: “a type of banking service that is provided to unemployed or low- income individuals or groups who would otherwise have no other means of gaining financial services.” Microfinance institution (MFIs) have expanded rapidly over last 10-15 years 7.6 million families in 1997 137.5 million families in 2010 Originally viewed as a win-win opportunity, but recent backlash Primarily anecdotes are used for backing microfinance or speaking against it.

Literature Review Little to no hard evidence that it works or does not work There were no studies that had measured the longer-term effects of microcredit. While this study was underway, three other studies were too. Results were compared at the end. Primarily anecdotes are used for backing microfinance or speaking against it. Grammine started with women in Bangladesh (women give profits to educate their children)

Methodology Overview First randomized evaluation of the effect of the canonical group-lending microcredit model, which targets women Studies households over 3.5 years after the introduction of the system—the longest period of any study In 2005, 52 of 104 poor neighborhoods in Hyderabad, India, were randomly selected for opening of an MFI branch First endline survey was 15-18 months after introduction of microfinance in each neighborhood Second survey was two years later - Longest period of any study Hyderabad is 5th largest city in India and in the state where microcredit has expanded the fastest

Methodology Specifics Examined: Effect consumption New business creation Business income As well as human development outcomes: Education Health Women’s empowerment Sample size 6,864 households 90% maintained contact throughout the 3.5 year study

NO DIFFERENCE in monthly per capita consumption and monthly non-durable consumption Significant POSITIVE IMPACTS of the purchase of durables—households reduced spending on what they described as “temptation goods” INCREASE in the number of new businesses created, particularly by women However, these businesses are smaller and less profitable than the average business in the area (the vast majority which are already small and unprofitable) No significant growth Findings At First Endline Survey

HELPS households make different intertemporal choices in consumption With access to microcredit, families may sacrifice short- or even medium- term consumption in order to get the durable good, or to invest in a business NO SIGNIFICANT DIFFERENCE in total household expenditures Almost 70% of eligible households do not have an MFI loan—preferring to borrow from other sources, if they borrow (and most do) INSIGNIFICANT increase in business profits Findings At Second Endline Survey

NO EFFECTS on human development or women’s empowerment after 18 or 36 months NO EFFECT on health outcomes NO CHANGE in the probability that children/teenagers are enrolled in school NO EFFECT on the index of social outcomes (out of 16 social outcomes) NO PRIMA FACIA EVIDENCE that it leads to important changes in household decision-making or in social outcomes Findings On Human Development Outcomes

Result Comparison Consistencies Differences Insignificant impact on monthly consumption and non-durable consumption Decline in temptation goods, spending on parties, or both Expansion in self-employment activities In all but one, this activity did not translate into significantly higher profits or increase in profits of median business Impact is very small and insignificant at the median, however, it can help increase profits in 90th—95th percentile This study finds an increase in the purchase of durable goods used at home in last year On Women’s Empowerment 2 show no impact 2 do not report it 1 finds increase in women’s decision-making On Education 1 shows small positive impact 1 finds a negative impact On Health 1 shows less expenditures 1 shows more expenditures Are Social Impacts more influenced by the context? All 5 studied use of standard model of MFI programs.

Conclusion “Microcredit is not for every household, or even most households, and it does not lead to the miraculous social transformation some proponents have claimed.” “Its principal impact seems to allow some households to sacrifice some instantaneous utility to finance lumpy purchases, either for their home or to establish or expand a business.” “The only mistake that the microcredit enthusiasts may have made is to overestimate the potential of businesses for the poor, both as a source of revenue and as a means of empowerment for their female owners.”

Critique Evaluation Suggestions for future research: Are there results yet to be seen or experienced? Should we rethink the specifics to make the method more successful? Suggestions for future research: Find consistent ways to measure social outcomes Measure change in multiple places

Questions? Thank you for your time!