September 27, 2012
An Overview of David Roodman's Work Beyond MicrofinanceBy Thea Garon
On October 3rd, FAI will host a conversation with Jonathan Morduch and David Roodman, a senior fellow at the Center for Global Development (CGD). The conversation will focus on Roodman’s new book, Due Diligence, which has been widely praised (but you should also check out some of the critiques) for its detailed, evidence-based look at the state of microfinance today.
Those familiar with Roodman from his work in microfinance may be unaware of his influential work in other areas of development. We thought we’d provide a quick overview to the other sides of David Roodman (though all of the sides feature an exceedingly careful attention to detail and data).
In addition to his work on microfinance for CGD, David also manages the Commitment to Development Index, an analysis that ranks the world's richest nations based on their commitment to policies that benefit poorer nations. The index is widely recognized as the most comprehensive measure of wealthy countries’ dedication to aiding the developing world.
Roodman has also published a number of influential papers about the impact of international aid on development. His non-technical paper, Guide for the Perplexed, offers an accessible discussion of the most influential studies that attempt to quantify how international aid affects country-level outcomes such as economic growth and school attendance rates. Prior to this review, Roodman coauthored a 2004 American Economic Review paper with Bill Easterly and Ross Levine that challenged the findings of World Bank economists Craig Burnside and David Dollar that aid leads to positive growth in the context of a good policy environment. When Burnside and Dollar’s paper was published in 2000, it had enormous policy implications, as it implied that foreign aid should be selectively distributed to countries with sound fiscal, monetary and trade policies. By updating the dataset used in the original analysis to include data up to 1997, Roodman and his coauthors found that the World Bank economists’ conclusions were not robust.
Roodman’s masterful understanding of data and his meticulous approach to policy analysis has also reshaped an important discussion around the fungibility of health aid. In April 2010, Roodman critiqued a paper published by a team of authors from the Institute for Health Metrics and Evaluation (IHME) which claimed that health aid is highly fungible – that is, governments who receive health aid cut their own spending in this area, effectively siphoning off money designated for health to other uses. When two Stanford scholars published a paper similarly suggesting the original work was significantly flawed, Roodman took on that paper with a fine-toothed comb as well. He discovered that the Stanford scholars’ work itself included a flawed analysis of the data (which had been missed by the journal's editors and peer reviewers). Based on Roodman’s observations, the two Stanford authors acknowledged their mistake and retracted the paper.
In addition to his body of work on policy questions, Roodman has contributed significantly to the field of econometrics. Among econometricians, Roodman is best known for his computer program xtabond2, which is one of the most popular add-ons for the commercial statistics program, Stata. He has also written a number of influential papers (for instance this one) about econometrics, establishing him as an expert on analyzing and interpreting complex data sets.
Even before Due Diligence met with such acclaim, Roodman had made a splash in the microfinance sector with a working paper coauthored with FAI’s Jonathan Morduch, The Impact of Microcredit on the Poor in Bangladesh: Revisiting the Evidence. In this paper, Roodman and Morduch reevaluate a prominent microcredit study and conclude that the analysis does not convincingly prove that microcredit improves households’ financial wellbeing because it fails to rule out reverse causation. The authors show that a positive correlation between microcredit and household spending may simply be due to the fact that richer households borrow more. The paper calls into question a fundamental piece of evidence that had been used to buttress the marketing of microcredit as a high-impact intervention. The paper touched off a great deal of controversy over the proper methods to analyze the data set and critique results. You can read a good summary of that somewhat esoteric debate here. It's a great illustration of Roodman's ability to not just understand complex data sets, but to explain the issues simply so that non-experts can follow along.