Last week I gave a talk at the Second European Research Conference on Microfinance in the Dutch city of Groningen (pronounced Grrrrrrrrrrrongn).
The organizers had a smart idea: pairing me with Milford Bateman in a plenary session. The potential for conflict made it inherently interesting. I don't think the event was recorded for posterity. His talk covered similar ground to this one he gave at the MFC conference in Prague last month. My text is below.
Given my past exchanges with Milford it is not surprising that the moment was a bit awkward. Still, I quite enjoyed the session. Good questions from the audience only added to the energy and interest. I expect I'm not the only one for whom it was a highlight.
After, I rented one of those one-ton, one-speed, wonderfully practical Dutch cruisers and explored the city's amazing bike path network. (In the city center, cars are the exception while bikes course by in droves.) I struck north and eventually found myself in the countryside. After a while, I had enough of cows, sheep, and intensive agriculture. I turned around and came back.
Thank you. This morning, I’m going to talk about two books by well-intentioned men thinking critically about microfinance. One is by Milford Bateman and is called Why Microfinance Doesn’t Work. I daresay it can be described as passionate. The other is by me. It is coming out this fall. It has the working title, Why Microfinance Sort of Works and Sort of Doesn’t Work, Depending on What You Mean by “Works,” and How to Make It Work Better. Well actually, it’s called Due Diligence: An Impertinent Inquiry into Microfinance, but you get the idea. I daresay it can be described as boringly dispassionate.
This is a research conference. But to me at least, a lot of the arguments in Milford’s book do not seem well-founded in research—neither in an understanding of what research has actually shown so far, nor in the ethic of research, which has to do with appreciating the difference between hypothesis and evidence and the difficulties in generalizing beyond any one piece of evidence.
Of course I could say the same about a lot of what has been written by microfinance promoters. So I would say: welcome Milford as a provocateur. Look to him for interesting, contrarian conjectures. There has been a lot of positive hype around microfinance, and it is good that people like Milford are giving those positive claims a critical look, and in a way that grabs attention. But the most negative claims about microfinance that we are hearing now also deserve scrutiny.
I’ll give you a few examples. The book contends (and I quote) that:
The increasing dominance of the microfinance model in developing countries is causally associated with their progressive deindustrialization and infantilization…A Morgenthau type plan is being implemented for real.
And by that he means the American plan to deindustrialize Germany after World War II, a plan that was abandoned.
That is a strong statement. It arguably contains four major generalizations: Supporters of microfinance want to deindustrialize and infantilize the developing world. [Note: After I spoke, Milford described this inference as “ridiculous.”] Their campaign has succeeded in elevating microfinance. Their success has come at the cost of alternative approaches to providing finance to fairly poor people. And those alternatives would reduce poverty much more effectively, i.e., prevent the deindustrialization and infantilization of the developing world.
I’ll grant one of these, that donors have elevated microfinance. But the other three beg evidence. I didn’t see much in the book. For example, history strongly suggests that it is hard for outsiders to influence the sources of economic growth, partly because they are poorly understood. So I don’t share the book’s optimism that there are alternatives waiting on the shelves that can work much better than microfinance.
What are the alternatives Milford has in mind? Admirably, the book devotes a chapter to examples. They come from Japan, Vietnam, China, Venezuela, India, and other places. But the evidence that these models contributed to the success of their countries is hardly more rigorous than is the evidence that microcredit “doesn’t work.”
And the evidence that support for microfinance has actively prevented other countries from following such alternative paths is also scant. In recent months, we’ve seen in Bangladesh and India how governments are quite ready to stomp on microfinance, in India’s case precisely in order to favor a government-led model.
It seems that Milford really is the anti-Yunus. Where Yunus promotes microcredit as a cure for poverty, Milford promotes alternatives as the key to local economic development. This is antisymmetry—which is a kind of symmetry. I see myself as being in the middle, rather dubious about the power of any approach to make such a big difference, but intrigued nevertheless by microfinance’s potential to do modest good for a lot of people at modest cost, i.e., with minimal subsidy. I’m also intrigued by Milford’s examples, though I know less about them. But I want to emphasize that I have no ideological opposition to them. I don’t see a binary choice between microfinance and other approaches.
Another place where the argument seems detached from research: Milford emphasizes the displacement effect, which is that those accessing microcredit may do better by expanding their vegetable-selling businesses, but at the expense of competitors who don’t borrow. The net effect for the local economy could be much less than the gross effect for borrowers, perhaps even negative overall. I think it is good that Milford reminds the public of this potential downside. I also think that researchers will recognize this as one plausible story of impact; will recognize that other stories, positive or negative, can be told about how microfinance can affect people’s lives; and will recognize that which effects dominate is ultimately an empirical question whose answer will vary over time and space. I don’t recall seeing that kind of subtle empiricism in the book.
Finally, I disagree with his book’s brief (two-page) review of the research on the impacts of microfinance. I think that almost anyone in this room who set out to summarize the literature would give prominence to the recent randomized impact studies. She might give a little space, or perhaps just a footnote, to the leading non-randomized studies by Mark Pitt and Shahid Khandker, and to the challenges to those studies from Jonathan Morduch and myself. In contrast, Why Microfinance Doesn’t Work does not mention the best (randomized) studies of whether microfinance works. It makes no mention of the finding that microcredit in Hyderabad increased business profits—not to mention that this result was measured at the slum level, weighing in both borrowers and non-borrowers, thus netting out the displacement effect. And the book makes no mention of the randomized study showing that a commitment microsavings accounts helped female market vendors in Kenya.
In fact, the conclusion that these two pages settle on is a misinterpretation of my own work. In the abstract of my working paper with Jonathan Morduch, in which we attempt to replicate the non-randomized studies I just mentioned, we write, “As for [Pitt and Khandker]’s headline results, we obtain opposite signs. But we do not conclude that lending to women does harm. Rather [the regressions] appear to fail in expunging endogeneity.” (That conclusion stands today, more strongly than ever, despite—indeed, thanks to—Mark Pitt’s recent criticisms.) Milford’s book turns that into “Roodman and Morduch obtained opposite signs—that is, their results suggested negative impact.” That is not what we wrote.
So, like I say, my advice to you, as researchers, is not to trust the empirical assertions in Milford’s book. But make the most of what has to offer. I think he has extensive first-hand knowledge of the microcredit bubble in Bosnia. As a contrarian, he offers provocative ideas, some of which must contain truth. He exposes us to alternatives that deserve more attention. To a point, this is a healthy antidote to the historical hype around microfinance. If he has inspired you to think critically about microfinance, then he has done some good.
Of course, Milford is not the only one to criticize microfinance at this conference. That inspires me to explain why I don’t come to a completely negative conclusion in my book.
I think it is entirely appropriate to challenge the hype about microfinance and look for the more complex truth. In particular, it is entirely appropriate to point out that there is a lack of rigorous evidence that microcredit reduces poverty, that it has overshot in some countries, and that group credit appears to reduce freedom, through peer pressure and high-pressure collection practices, as often as it enhances freedom through capital.
But I also believe deeply that in the social sciences, there are almost no simple truths. I admit that belief reflects my personality, having to do with how I survived my parents’ nasty divorce, by not taking sides. But—take this for whatever it’s worth—I also think it is a kind of wisdom. Families and villages and slums and nations are extremely complex, and no one has a monopoly on the truth about them. Those promoting black-and-white ideas—microfinance is a savior, microfinance is a disaster—have to be substantially wrong. I’m not proud to say this, but I’m someone who doesn’t get aroused by babies dying from preventable malaria. I do get upset by those who hurl condemnation based on sloppy reasoning. Dambisa Moyo, the author of Dead Aid, is an example.
And microfinance is full of ambiguities. One of those is around the profit motive. Yesterday I listened to Lamia Karim describe how microcredit had overextended itself and how it had led to problems such as women being trapped by loan collection practices that exploit the power of shame. Actually, I use some of her descriptions in my book. The scenes she described sounded a lot like India, where the commercialization of microfinance, in particular the involvement of for-profit investors, has been blamed. Except she was talking about Bangladesh, where microfinance institutions are mostly non-profit, or in the case of the Grameen Bank, cooperatively owned. So I think it is simplistic to talk, as Yunus and Milford both do, about the profit motive being all bad for microfinance.
Critics are also right to point out, as I have, that systems of credit can grow dangerously fast. But here is the hard question: when should we give up on a system of credit because of its tendency to overshoot? I think it would paternalistic and hypocritical to tell the poor that because a system of credit aimed at them sometimes gets out of control, their access to it should be cut off, while not telling the same things to ourselves in rich countries. Despite the mortgage crises, we still want mortgages. No system for providing credit will be perfect, and microfinance has done well enough, and demonstrated enough potential, that it should not be wiped off the map.
Every country that is today rich has had a financial crisis. Should we have banished credit in all these countries? On the contrary, crises appear to be regrettable but inevitable learning experiences. Consider that in the first country with a microcredit crisis, Bolivia, microfinance has survived and thrived. The two institutions that are offshoots of the microfinance pioneer there, BancoSol and Prodem, now hold 900,000 savings accounts between them.
Now, as I say, I have just finished up my own book. I wrote it in public, on my blog, where you can find the draft chapters. In the book, I look at microfinance from many perspectives—from the point of view of clients, microfinance managers, researchers. I also dig into the history of financial services for the masses, which turns out to be quite rich. A starting point for me was some writing by Hans Dieter Seibel, who is here in the audience. Over and over in history, people have found particular ways to mass-produce financial services for regular people, in the process demonstrating an eternal demand.
A few years ago, as I was working on that replication of the non-randomized Pitt and Khandker study I mentioned before, I visited some microcredit programs in Egypt. I remember in particular visiting a branch of the Lead Foundation in Cairo one day. It was disbursement day, so the lobby was packed with women and their children, who had come to wait hours for their new loans. In fact the crowd flowed out into the hall, down the stairs and onto the street. I got to meet some of the borrowers. They all told me that they would use their loans for petty trade—selling things like makeup and bed sheets to friends and family. I don’t know if they were telling me the truth. They were speaking to me in the presence of their loan officer, and may well have been required to state a business use to obtain the loans. But whatever they were using the loans for, they clearly wanted them. I thought to myself, should I tell the women that back at my hotel on my laptop—this laptop—that I was running conditional mixed-process recursive Maximum Likelihood regressions on a cross-section of household survey data from early 1990s Bangladesh…and I wasn’t so sure those loans were a good idea?
Of course not. But that meant I faced a paradox. On the one hand, I really did, and do, believe in the importance of good evaluation, in judging carefully what evidence to believe and what not. As a taxpayer supporting this microlender through USAID, I had a right to ask whether my money was making a difference. On the other hand, it seemed to me that something good was going on here, with all these women grasping these loans as a way to make their lives a little better.
The way I resolved that paradox forms the heart of my book. I realized that in the grand global conversation about microfinance there are at least three distinct and legitimate conceptions of success. Each really is a conception of the word “development,” at least in English. They are not mutually contradictory. As yardsticks for evaluation, each tends to generate different questions and lead to different kinds of evidence. I think that a full appraisal of microfinance, one capable of providing appropriate practical guidance, requires examining microfinance from all three perspectives.
The first conception is the one I associate with the Pitt & Khandker work and the recent randomized studies: development as escape from poverty. If the verdict here were unambiguously positive, or unambiguously negative, I suppose that would end the discussion. But it isn’t.
The second is “development as freedom.” This phrase comes of course from Amartya Sen the Nobel Prize–winning economist. For Sen, “freedom,” is not the freedom of American libertarians—freedom from interference. Rather, it means enhanced agency in one’s life. Education, income, health are all sources of freedom. Sen argues that freedoms tend to support one another, so that we should expect in a general way that enhanced control over one’s finances should lead to greater agency in other domains, for example, by making it easier for families to save up for school fees.
I use this theory to frame the work made most famous by Portfolios of the Poor. If you read that book, or Stuart Rutherford’s The Poor and Their Money, you won’t find claims that microfinance is lifting people out of poverty. But you will find powerful stories (and careful generalizations) about how people use many kinds of financial services to get a little more control over their lives—a little more freedom. So I think financial services inherently increase freedom—that is what they are for.
But they do not automatically do so, the obvious counterexamples being the borrowers who end up worse off for borrowing, perhaps even being “trapped” in debt. In reviewing the evidence, most of it qualitative, the overall pattern that emerged, if vaguely, is that the most famous kind of microfinance, group credit for women, does the least for agency. Some women do find empowerment in peer support. Others find constriction in peer pressure.
The third conception of development comes from Schumpeter. For a general audience, I call it “development as industry building.” The idea here, of course, is that what has really increased prosperity and freedom over the centuries is a constant process of economic churning, what we call industrialization. (There has been political creative destruction too, but that seems less relevant for evaluating microfinance.)
In this light, microfinance has been singularly successful in the worlds of foreign aid and philanthropy, not for turning its clients into creative destruction, as Milford rightly points out, but for supporting microfinance institutions that themselves enrich the institutional fabric of nations. Businesses and businesslike institutions such as the Grameen Bank are competing, innovating, growing, employing thousands and serving millions. They are extending the reach of the financial system. They are instances of Schumpeterian transformation.
Of course, here too there is room for critique. When is creative destruction more creative than destructive? And when is it opposite? In the book, I offer two general principles, drawn from ecology (or human valuations thereof), for judging when a microfinance industry is enriching the economic fabric. There need to be forces of restraint that limit the overshoot-and-crash dynamic. And the more ways that a microfinance institution connects to others in the economic web—taking investment domestically as well as from foreigners, doing savings as well as credit—the more constructive will its presence tend to be.
You might notice that in contrast to Milford, I don’t spend much time ascribing motives (whether of ideology or commerce) with phrases like “neoliberalism” and “faith healing.” I see myself as an empiricist. Whether a given loan is motivated by neoliberalism or not doesn’t seem to me to determine its impact. And what people intend to achieve and what their actions actually cause are often very different.
Overall, microfinance’s strengths lie in building dynamic services that deliver freedom-enhancing services for to millions of people with minimal subsidy. As I argued before, I don’t think the evidence is there to justify abolishing microfinance—unless you’re ready to give up your mortgages. The poor deserve safe access to the financial infrastructure just as they deserve access to clean water and electricity. The best way forward is to help microfinance play to its strengths. Among other things, that means pushing savings to the extent possible, through both low and high-tech means. And it means reducing the amount of money going into microfinance at the global level (though not necessarily to every country) in order to prevent more overshoots and increase incentives to take savings as an alternative source of capital.
I think this analysis puts in useful perspective some of the recent intellectual challenges to microfinance.
First, it tempers the contention that microfinance has been a complete disaster and ought to be abandoned, as I’ve already argued. Financial industries will always be flawed, but most are worth keeping anyway, and fixing.
I think my work also puts RCTs in perspective. These are sometimes touted as the way to find out what works. As I say, I suppose if the randomized trials were showing strong benefits or strong harm, that would trump other consideration. Since they have not, one must rely on various kinds of broadly credible theories—I rely on Sen and Schumpeter—to extrapolate beyond the evidence we have. These theories make relevant other kinds of evidence, such as financial diaries, qualitative studies of borrowers, and industry analysis. Imagine if randomized trials had been done of mortgages in rich countries five years ago. Would they have told us everything we needed to know about the role of mortgages in the economic fates of nations? Of course not, because financial industries are dynamical systems. The dynamics of the microfinance industry are part of its impact on development and should be studied and evaluated directly.
The core idea of the microfinance movement, it seems to me, is not that credit can end poverty, but that the poor deserve financial services; and that that need can be met in part through businesslike mass-production. Going forward, the hope and the challenge lie in realizing that vision while moving beyond the products, such as group credit, that got the microfinance movement started. Thank you.