Philanthropy Action

News & Commentary


The blogosphere has been abuzz this summer with talk about the published results from the first two randomized controlled trials of the impact of micro-credit on poverty. The first was conducted by researchers at MIT’s Poverty Action Lab (Abhijit Banerjee, Esther Duflo, Rachel Glennerster and Cynthia Kinnan) in the slums of Hyderabad; the second was designed and run by Innovations for Poverty Action (IPA) fellows Dean Karlan and Jonathan Zinman in the Philippines. Neither study showed any alleviation of poverty as the result of having access to micro-credit.

For a full explanation of the study design and results, I sat down with IPA’s Nathanael Goldberg, Project Director for Microfinance. The transcript of our interview is here.

These results will be disappointing to many who viewed micro-credit as a huge source of promise for poverty alleviation. Indeed, the World Bank’s PSD blog immediately released its conclusion that “The Verdict Is In on Microfinance…And Its Not Pretty”. But to take these results as an indictment of all of microfinance would be a mistake. These studies really focus on micro-credit—small loans—not microfinance as a whole. In turn, David Roodman of the Center for Global Development points out that the Philippines study randomized within a population of borrowers firmly above the poverty line, and as such is not measuring whether microfinance alleviates poverty per se. And the Economist called out some of the more nuanced results around changes in consumption, which point to increased investments in durable goods, a possible precursor to income improvements.

Part of what makes these studies good is how narrow and focused they are. All previous evidence in support of the argument that micro-loans alleviate poverty comes from non-experimental or quasi-experimental research, studies which measured any change that occurred after a person had access to a loan, but could not attribute that change to the loan itself. The new studies are the first to rigorously apply randomization techniques, which basically allows researchers to be sure that any improvement (or decline) is caused by the loan itself. In truth, they are very important for the rigor they apply to the sector, and for the ways in which they begin to unpack what micro-credit can and cannot do for the poor.


Remember my personal information

Notify me of follow-up comments?

Comments may be edited for length. Inappropriate comments will not be published.