15 June 2011

Impacts of micro-credit: a study from the Philippines

Studies of the impact of micro-credit/ micro-loans have painted a mixed picture of the benefits that they deliver. Proponents argue that microcredit mitigates market failures, spurs micro-enterprise growth, and boosts borrowers’ well-being.

Yale and MIT researchers - working with First Macro Bank in Manila, Philippines - set up the financial equivalent of a double-blind trial. First Micro Bank (FMB) provides three-month micro-loans that including fees has an annual interest rate of 60%. FMB is considered a reasonable example of a "second-generation" microloan bank, and it has received support from the US Agency for International Development.

FMB's risk evaluation software was set up to stratify potential borrowers into three categories: low risk, marginal, and high risk (rejected) candidates. The low-risk group got loans, but, for the marginal borrowers, the program was set up to randomly approve a subset, ensuring that there would be experimental and control populations of roughly equal risk. Bank employees had no idea that the selection process had an element of randomness, and in fact, accidentally made loans to five people the software rejected.

The loans went to over a thousand individuals and groups that were reasonably representative of the Manila population as a whole. Then the authors engaged in a little economic stimulation of their own, paying researchers at a local university to conduct follow-up surveys of the loan recipients starting 11 months after the loan. Given a 70 percent response rate, they ended up with information on over 1,100 borrowers.

  1. There did seem to be a real unmet need for micro-credit ($250 average).
  2. Business growth did not see any statistically significant boost following the loans. Rather, compared to the control group, those receiving loans ended up owning slightly fewer businesses, and had 31 percent fewer employees (these were small businesses, so 31 percent works ought to be only 0.27 of an employee).
  3. Subjective well being measures for those who received a loan saw a small and statistically insignificant drop compared to the control group who did not receive loan.
  4. There were almost no differences in any measures between loans to men and women. The one significant difference is that males seem to have found the loan very stressful (the authors don't comment on whether this explained the drop in subjective well-being).
  5. Loan recipients did feel that their ability to cope with risk improved, community ties were strengthened community ties and they had better access to informal credit from friends and family afterwards.

Thus, microcredit may work, but through channels different from those often hypothesised by proponents.

[Hat tip John Timmer, Science via Ars Technica , 'Jeepney' image from Wikipedia]