In a three-part series, Roosevelt Institute Fellow Georgia Levenson Keohane explores India's microcredit crisis and what it teaches us about combating poverty. In her final post, Keohane questions the efficacy of microcredit. Does it really transform lives? How do we know?
Beyond yesterday's question of non-profit versus for-profit, the microcredit crisis in India has emboldened naysayers who question whether either model has proved itself the hoped-for panacea for global poverty. Does microcredit even work, they ask? And how do we know?
This spring, Esther Duflo and Abhijit Banerjee, the highly regarded MIT economists who run the Abdul Latif Jameel Poverty Action Lab (J-Pal), published Poor Economics: a Radical Rethinking of the Way to Fight Global Poverty. In it, they draw on their field research: hundreds of randomized control trials designed to examine which policies and practices (and under what conditions) successfully reduce poverty, and which do not. Duflo and Banerjee's empirical approach is widely credited with transforming the field of international development and the economics discipline more broadly. Moreover, their work on microlending finds "clear evidence that microfinance was working." Because Duflo and Banerjee, like other empiricists, also conclude that micro-lending produced little "radical transformation" in the lives of the poor people they studied, many have been quick to pronounce microcredit's failure.
The value of bold, persistent policy experimentation
Duflo and Banerjee insist otherwise. "The main objective of microfinance seemed to have been achieved," they write. "It was not miraculous, but it was working... In our minds microcredit has earned its rightful place as one of the key instruments in the fight against poverty."
The lessons here about what Franklin D. Roosevelt called "bold, persistent experimentation" are crucial for policy makers the world over. First, the absence of panacea does not amount to program failure. Second, the value of the 'controlled experiment' paradigm lies in its parsing power. These kinds of studies -- akin to the randomized contrail trials (RCT) of medical research -- offer a tool to pinpoint which components make a policy effective, which do not, and which can be improved to enhance service delivery and social benefit. Duflo and Banerjee suggest, for example, that most existing microcredit lending structures (for-profit or not for profit) do not permit the poor to borrow and invest sums large or long-term enough for higher risk and return projects that might actually transform their lives. The experiment indicates that creating access to this kind of credit is the next -- and more complex -- frontier in improving capital markets for the poor.
A different empirical tact has shown that microcredit works when loans are combined with other products or services, like savings or insurance. In Portfolios of the Poor, researchers Daryl Collins, Jonathan Morduch, Stuart Rutherford, and Orlanda Ruthven examined the financial diaries of the hundreds poor people in India, Bangladesh, and South Africa and determined that credit to build small businesses, though effective, was not enough. Borrowers also benefited from credit for things like doctor's bills, school fees, weddings, and funerals. Increasingly, microfinance institutions (MFIs) are experimenting with product and service innovations along these lines.
Portfolios of the Poor also describes how Grameen made enormous strides in learning from its own experience. In a series of reforms known as Grameen II, the bank began to offer a broader range of savings and credit accounts, and more flexibility as to when and how its clients could access them. A number of other Grameen inspired organizations continue to learn from these experiments. The Grameen Foundation, for example, promotes poverty reduction through microenterprise and technology, with recent innovations like Mobile Financial Services and Mobile Technology for Community Health (MoTech). Grameen America is adapting Yunus's original microlending archetype to serve the poor and unbanked in New York City.
Though microenterprise in developing countries has been an important testing ground for empirical research, the broader lessons about evaluation and experimentation are applicable across fields and are vital for American policy makers. In recent years, we have witnessed greater adoption of this approach in the U.S. in both the non-profit and public sectors. New York City's Center for Economic Opportunity (CEO), for example, aims to function as a kind of anti-poverty laboratory. Seeded primarily with philanthropic funds, the CEO pilots and evaluates innovative and untested social programs to assess which might be successfully scaled. The CEO has been cited as one of the models for the recent federal efforts in this area, including the new Office of Social Innovation in the White House, and its various funds and activities. In 2009, Peter Orszag, then the Director of the White House Office of Management and Budget, famously called for more rigorous and "evidenced based" evaluation of federally funded programs, advocating a kind of clinical trial methodology. Others have pointed to Duflo and Banerjee's J-PAL at MIT as an action lab template for other areas of public policy, from global climate change to domestic social programs.
Not surprisingly, a strict RCT approach raises a host of implementation concerns related to cost, ethics, and scope, and is not without detractors. However, the spirit of this kind of inquiry, and the success of its numerous and modified applications, has helped to shift policy makers towards more risk-taking experimentation and exacting evaluation, both essential in the fight against entrenched and persistent poverty in the U.S. and around the world. This, too, will be a focus of my research in the coming months, and the subject of future posts. I welcome your comments.
Georgia Levenson Keohane is a Fellow at the Roosevelt Institute.