1. contact@e-mfp.eu www.e-mfp.eu Measuring the Impact of Microfinance Meritxell Martinez- CGAP European Microfinance Week 29 Nov – 1 Dec.

2. After this presentation you will: Refresh your knowledge of monitoring and evaluation Get an idea of the type of evaluations being done in microfinance Have an update on experimental research findings

3. Monitoring vs. Evaluation in MF How is my program doing vis-à-vis the set goals? MONITORING (Social Performance – PPI, PAT, scorecards, MIX SP indicators) Are my clients doing better because of the program? EVALUATION

4. Evaluation QUALITATIVEQUANTITATIVE Non Experimental (participants to non participants) Quasi- Experimental (comparison group) Experimental (control group) i.e. Focus groups, structured interviews, direct observation More selection biasLess selection bias

5. The problem Anecdotes Monitoring data (Social performance indicators) Non or quasi - experimental studies IMPACT ? Problems establishing attribution! ≠

6. Impact evaluation using RCTs Randomized Controlled Trials (RCTs) are showing us the importance of truly understanding: -who benefits and who doesn’t and by how much -counterintuitive results -the diversity of clients RCTs are not just about impact, but the methodology can be used for market research and product testing.

7. CONS Expensive time and money (USD 200/400k) Politically or ethically difficult Short timeframes & contagion risk Results applicable to the context where they are done? PROS Selection bias is zero on average and mean impact is revealed Less subject to methodological quibbles More likely to be convincing to program funders and/or policymakers RCTs applied to product feature testing and behavioral research RCTs

8. What are RCTs telling us about microfinance so far? Highlights from the Microfinance Impact and Innovations Conference, Oct 21-23

9. . Findings on credit Positive ImpactNo Impact Existing business that took loans generated higher profits (India, Philippines) No impact on health, education, empowerment (Morocco, India, Philippines) Loan-takers also spent more on home-based durable goods (India) Credit not helping to cope with shocks - consumption not decreasing in places with access to credit. (Morocco) Change in activity – diversification – through buying more types of animals (Morocco) Credit not helping borrowers to start new businesses (Morocco) Cut down on "temptation goods" such as cigarettes and tea and reduction in social festivals (India, Morocco) No significant effect on consumption (Morocco), or household expenditure (India) * 3 studies

10. . Findings on savings Positive ImpactNo Impact Women faced shocks better (bought medication, positive effect on food expenditure. (Kenya) Accounts had minimal impact on men. (Kenya) Business investment for women increased by 40%. (Kenya) Not possible to estimate other effects. (Kenya) * 1 study

11. . Findings on (rainfall) insurance Positive ImpactMixed Impact Increased farm investment: farmers bought more fertilizer, planted more acreage, hired more labor, had higher yields and more income, missed fewer meals and their children missed less school. (Ghana, India) Adoption rates are modest but growing (Ghana, India) Farmers moved from planting subsistence, low value crops to higher risk, higher value cash crops that increased their incomes. (Ghana, India) Design improvements are important- payouts too slow. (Ghana, India) * 2 studies

12. Findings on financial education Positive ImpactMixed Impact 5.4% of customers reduce their use of payday lenders after seeing actual cost information on the loan in comparison to credit card charges. (US) APR disclosures did not result in reduction of use of payday lenders. (US) Basic training using “rules of thumb” lead to improvements in business practices and outcomes. (Dominican Republic) Accounting training did not result in improved business practices. (Dominican Republic) * 1 study

13. . Findings on product testing Positive ImpactMixed Impact Savings: Voluntary commitment devices and phone reminders helped increase savings (Peru, Philippines, Bolivia) Grace Period: Offering (2-month) grace periods increased investment and business growth after two years, but at the cost of more default in the short term -12% more. (India) Biometrics: Fingerprinting led to increases in the repayment rates of about 40% for those with higher probability of defaulting. (Malawi) Borrowers’ sensitivity to interest rates: revenue gains from additional lending at lower interest rates did not fully offset the revenue losses from the lower rates (Mexico)

14. Debate Do we understand the financial needs of the unbanked? What else needs to be researched? Is this evidence new? How does social performance help us understanding the impact of microfinance? What are the “evidence needs” of 1- double bottom funders, 2- policymakers, 3- providers? Is the reputational risk of investing in microfinance increasing?