Funding a Sector in Transition: Recent RCTs could reshape donors’ approach to microcredit
After a group of six studies published in January found that its benefits are modest at best, microcredit looks stuck at a crossroads. But that’s no guarantee that change is coming. With debate ongoing amongst practitioners about how to react to the powerful new data, two things are clear: great expectations are a thing of the past – but also, funding is safe. Some sources have dried up, but others remain enthusiastic. That latter group includes major development sources, where the results contain some good news as well as bad, but also the general public: the recent microfinance fundraiser at Whole Foods supermarkets offered no evidence of the existential debate bubbling underneath.
Perhaps the biggest change might be an end to the hype – microcredit may no longer be expected to solve the problems of the poor on its own. Many microcredit lenders rely on grants or loans from bigger donors to fund their operations. That’s mandated an endless marketing campaign, and responses to that have included skepticism and scrutiny. “A lot of microcredit groups have toned down their rhetoric,’’ Dean Karlan, a Yale University economist who played a principal role in the recently released studies, told me. Karlan, along with economists from the Massachusetts Institute of Technology, Esther Duflo and Abhijit Banerjee, lead the team behind January’s data deluge. They believe that the best conclusions about microcredit come from randomized controlled trials, which they conducted in Bosnia and Herzegovina, Ethiopia, India, Mexico, Mongolia and Morocco. This is a process similar to how pharmaceuticals are tested: a group of randomly selected people get a loan (instead of an experimental drug) and another group does not. After a set period of time, outcomes for the two groups are compared.
Qualitative observations, like the surveys that have driven past research on financial inclusion, have a place in this process only at the start – to shape the goals of the trials and the questions researchers will ask their subjects. This methodology is considered more scientifically rigorous than other types of fieldwork, Jake Kendall, a financial-inclusion program officer for the Bill and Melinda Gates Foundation, told me: “They are the only tool that allows us to go beyond correlations, which can be misleading, to focus on causality.’’
The scholarship these economists published is unprecedented thanks to a combination of its global scale, the consistency of the findings across locations, and its presence in peer-reviewed journals. During a February conference at The World Bank’s Washington, D.C. headquarters, the dominant reaction was less of a challenge to the results than a search for an appropriate reaction and course for reform. Changes could include experimenting with more-flexible forms of microcredit, such as offering borrowers customized terms, or reducing costs by using mobile-phone apps to reach more people, which is less expensive than opening offices and sending representatives to rural villages.
The use of randomized controlled trials is polarizing, however. This specific methodology has led to criticism that only some of the important factors at play can be accounted for in the trial systems designed, and the team’s conclusion that development cash should be diverted to other ideas is not widely accepted. “The studies have not found that microfinance is reducing poverty, but they are finding all sorts of ways that products and services can be improved,’’ Jennefer Sebstad, a microfinance specialist with the U.S. Agency for International Development (USAID), told me. This type of research is also new to the development world, and has been used thus far mostly to measure microcredit, so using RCTs to compare microfinance to other options for financial inclusion isn’t yet possible.
The results likely mean fewer people claiming that microcredit alone can lift people out of poverty – it’s more likely to be seen as one of a group of things that can help – but that isn’t likely to stop development funders from backing the practice. Private funders like Gates will decide based on their own criteria whether to support microcredit, but public agencies typically have rules to follow with their cash, and these will work in microcredit’s favor. One important distinction that shapes them is whether they are allocated taxpayer money for their operations. For example, USAID’s budget comes from Congress, whereas the Overseas Private Investment Corp. (OPIC) is a government body that must fund itself through profitable investments that generate social benefits as well. Groups that are allocated government money on a regular basis have increasingly moved on from funding microcredit, whereas their self-funding counterparts, generally referred to as development finance institutions, embrace it.
The majority of funding for microcredit now comes from these development finance institutions, such as the International Finance Corp., Germany’s KfW, and the European Bank for Reconstruction and Development. “Maybe microcredit isn’t a panacea, but I’m glad the studies didn’t show that microcredit was doing any harm, because it’s certainly an important access-to-finance activity for us,’’ Martin Holtmann, the IFC’s Head of Microfinance, told me. And an investment that does no harm is no small thing in the development world – the enduring voices of development critics such as New York University professor William Easterly or Zambian economist and author Dambisa Moyo are a constant reminder of the difficulty in turning development dollars into positive outcomes.
Many of the agencies that have already moved on from microcredit are now backing the “market infrastructure’’ in which poor people use financial-inclusion tools. Examples of this infrastructure include mobile phone apps, digital-payment platforms and other post-cash options. Yet only in a small minority of cases have the people offered these products taken to using them regularly – only six of about 150 mobile money programs worldwide had attracted more than one million customers as of a definitive 2012 study, and the majority of their activity was in sending money between each other and buying network airtime for their phones. (Other evidence shows that even among popular platforms, mobile money users tend to be at middle- or upper-level incomes). No randomized controlled trials have evaluated mobile money apps, and like microcredit long ago, there are some frothy claims. In early 2014 M-Pesa’s owner, the Kenyan mobile network Safaricom, said that 43 percent of the country’s GDP passes through the mobile app. The number spread fast over media, but then fell apart after someone applied scrutiny. In March 2014, a professor at the U.K.’s University of Bath, Susan Johnson, traced the statistic to its origin in data from the Central Bank of Kenya, and found that it was at least twice as high as it should have been thanks to double-counting transactions through the system. Given other flaws, she concluded that the number wasn’t actually a useful measure of anything at all.
In a way this looks like history repeating itself – like microcredit used to be what mobile money is right now: a fresh and exciting idea prone to hype and perhaps in need of some skepticism. Yet it is also easy to see the potential. Microcredit would benefit as well if development dollars redirected to mobile money yielded positive results. If mobile platforms do indeed allow financial inclusion products to reach more people and at a cheaper cost, then some of the savings for microcredit could be passed along to customers in the form of lower rates. Microcredit could become part of a package of interventions in financial inclusion, rather than trying to live up to the promise of being the one and only answer.
Matt Mossman is a development economics analyst and consultant based in Washington, D.C., specializing in African and Middle Eastern economies.