Once the brightest idea in development economics, microcredit is under attack. But Rohini Pande, a co-founder of Evidence for Policy Design at Harvard Kennedy School, is looking for ways to make this sector fulfill its original promise to the world’s poorest women.
In a household in a Kolkata slum, the afternoon sunlight streams in through gaps in the corrugated aluminum walls. Ten women sit in a circle on the rug, chatting, laughing, sharing recipes, punctuating their conversation with gentle adjustments to their brightly colored saris. This is the beginning of their weekly microloan repayment meeting.
Field representatives from the microfinance institution roll up and lean their bicycles in the doorway. The room becomes quiet as, one by one, the women dole out their 200-rupee loan installments on the rug. Once the field representatives have collected the bills and coins and headed off to the next house on their circuit, the room again fills with gossip and laughter.
The astonishing success of microfinance — financial services, including very small loans, made available to the poor — over the past three decades has depended largely on this feature of the loan contract: group repayment. The predominantly female clientele help — or pressure — one another to make installments. It is the group, not the individual, that is responsible for the loan, and no one wants to be empty-handed when her turn arrives.
In lieu of collateral, shared responsibility has resulted in repayment rates nearing 100 percent, allowing microfinance to reach the very poorest people who previously had no access to banking. From small beginnings in South Asia in the 1970s, microfinance spread rapidly across the developing world, eventually reaching more than 150 million borrowers. In 2006 Muhammad Yunus and Grameen Bank, the original microfinance providers in Bangladesh, won the Nobel Peace Prize. The Nobel committee’s citation read, “Micro-credit has proved to be an important liberating force in societies where women in particular have to struggle against repressive social and economic conditions.”
What was good for the borrower seemed good for the lender: In 2010, the microfinance institutions Compartimentos in Mexico and SKS in India went public with lucrative first offerings. Microfinance was not only charitable but also profitable and perhaps sustainable. Or so the story went.
Recently, however, microfinance has come under attack. Its once-lauded repayment methods have been criticized as coercive and based on humiliating women in their communities — to the point of allegedly causing a rash of suicides in India in 2010. Studies in India, Mexico, and the Philippines showed that microloans rarely led to business formation and failed to deliver the promised social impacts, such as poverty reduction and improvements in health and education. The most recent numbers show a first-ever dip in the number of borrowers. And although Muhammad Yunus continues to win international prizes for his work, the government of Bangladesh has removed him from his position at Grameen Bank.
Critics have been quick to offer reasons for the failures: Perhaps credit wasn’t the crucial gap after all, and resources should be directed elsewhere — toward training, for example, or encouraging the poor to invest in insurance or savings. Perhaps there just isn’t demand for all these microbusinesses — a community can support only so many little shops. Or maybe ancient stereotypes about “business sense” are correct, and it was a mistake to target women.
So, is microfinance charitable or coercive? A win-win for lenders and the poor, or a losing prospect for those with very little to lose?
Rohini Pande, Mohammed Kamal Professor of Public Policy at the Kennedy School and an authority on the subject, refuses to take a black-and-white view of microfinance. She summarizes with characteristic concision: “What we’ve learned is, there is no free lunch.”
Through a suite of studies exploring the nuances of microfinance in India, Pande has found that the sector may be worth fixing — perhaps through government subsidies and under government supervision — in part because it offers poor women an unexpected nonmonetary return.
Pande’s office in the Kennedy School’s Rubenstein Building doesn’t fit the image of an ivory tower retreat. There are no rows of leather-bound volumes. The desk is pushed back, making room for a round table in the center of the room. This table sees a lot of use during Pande’s day, as she brings in colleagues and students for impromptu meetings.
Although the office shows a few attempts at decorating — a single houseplant, vintage posters for foreign films featuring pistol-brandishing femmes fatales — Pande is very much the mobile scholar. She visits her native India frequently to monitor the studies she has going there, and she presents at conferences around the world. Having attended both Oxford — as a Rhodes scholar — and the London School of Economics, Pande has deep roots in London and spends part of her year there.
This floor of the Rubenstein Building is the headquarters of Evidence for Policy Design, or EPoD, the research initiative within the Center for International Development that Pande founded with fellow Kennedy School professor Asim I. Khwaja in 2008. Comprising eight HKS faculty members, six additional researchers, along with administrators and research fellows, EPoD is part of a revolution in the social sciences that for the past few decades has brought the rigorous research methods of medicine and the hard sciences into economics and government. In particular, EPoD applies these methods to the design of policy, not only theoretically, in classes at Harvard, but in practice, through partnerships with governments, ngos, and policymakers in developing countries. EPoD is also the hub for an array of large-scale studies on policy design (see sidebar), including those on microfinance that Pande coauthored with Erica Field, of Duke University.
Pande explains that an early strain of her work dealt with public banks in rural India. “There was a large push by the government for ‘social banking,’ where they opened branches in unbanked areas,” she says. “We showed that it had pretty strong effects reducing poverty, but they were unsustainable. Default rates were 40 percent. It was against that setting of nationalized banking for the poor — which really hadn’t worked in most countries — that people have moved toward thinking about microfinance.”
Pande speaks with a mild, self-deprecating smile, as if she can’t quite believe you’re that interested in what she’s saying. Her words come in rapid bursts, and as you listen, you realize that this rhythm is less the result of shyness than an attempt to check the rapid expansion of her ideas. Pande has been in and out of the classroom for 14 years; she is well accustomed to having to slow herself down for her listeners.
In the mid-2000s Pande was teaching at Yale and Field was there as a visiting scholar. “We started talking about microfinance, and Erica had been thinking about repayment flexibility,” Pande says. “That’s how we started working on it.”
Field explains, “Even before we did impact evaluations to see if microfinance was working, and the extent to which it was working, we had some ideas about how the design of microfinance could be improved to have a bigger impact. That’s exactly the kind of question the Centre for Micro Finance in India was interested in investigating, so we teamed up with them and started a series of field experiments.”
One of those studies, written by Pande and Field with Natalia Rigol and John Papp, and forthcoming in the American Economic Review, is titled “Does the Classic Microfinance Model Discourage Entrepreneurship Among the Poor?” — or “the grace period study” for short. Working with a microfinance provider in West Bengal and using the same blind randomization methods as in clinical trials, the team separated borrowers into two groups. They gave the first a standard microfinance contract, which required repayment installments to start immediately, and the second a contract that featured a two-month grace period before the first installment. They thought the grace period might allow clients to invest in more-profitable activities that take more time — for example, buying a sewing machine and fabric to make saris, rather than buying readymade garments from a wholesaler. They surveyed clients at the time of the loans and revisited them three years later to check on their businesses.
Their findings were striking: Clients who received the grace period invested 6 percent more in their businesses and were twice as likely to start new ones. Three years later, their weekly business profits were 41 percent greater and monthly household income 19.5 percent greater, and they reported roughly 80 percent more business capital. There was a drawback, however: Grace period clients were three times as likely to default.
This suggested that if microfinance was to achieve its aim of fostering entrepreneurship, it would have to take into account the real needs of small, fragile businesses, and deal with high default rates. But what about the other complaints against it — for example, that its repayment methods increase, rather than relieve, the psychological stress of the poor?
In another study, titled “Repayment Flexibility Can Reduce Financial Stress,” published last year in the journal plos one, Pande and Field with coauthors Papp and Y. Jeanette Park looked at how the frequency of payments affected borrowers’ self-reported stress levels. In Kolkata, one group of microfinance clients made the standard weekly payments, while another made larger, monthly installments. This study used technology to capture subjects’ thoughts in real time. Starting a few months after loan disbursal, team members interviewed clients regarding their state of mind every 48 hours via cell phones distributed specially for the study.
The surveys showed that clients who repaid monthly were 51 percent less likely to report feeling “worried, tense, or anxious” about repaying, were 54 percent more likely to report feeling confident about repaying, and reported spending less time thinking about their loans than did clients who repaid weekly. Additionally, monthly clients showed higher business investment and income, suggesting that the flexibility encouraged them to use their loans more profitably, which also ultimately reduced financial stress.
Taken together, these studies suggest that microcredit’s shortcomings might have to do with its standard contract — which can be adjusted — and also with expectations for its returns and impact. The changes it will take to make microfinance actually foster small businesses will probably raise interest rates. In fact, Pande’s calculations based on the grace period study showed that covering increased default would raise annual interest rates from 22 percent to 33 percent — which would scare off many potential clients. “Or bring in a different kind of client,” Pande adds. “One that is much riskier.”
The other option is for governments or charitable organizations to subsidize microloans. But is microfinance worth having the rest of society pay for? The answer might be yes.
Pande and Field collaborated with Benjamin Feigenberg — then a research assistant at EPoD — to observe returns to microfinance that other studies fail to capture. Their paper, “The Economic Returns to Social Interaction,” is forthcoming from The Review of Economic Studies.
The team randomly assigned clients of the West Bengali finance provider to repayment groups that met either weekly or monthly for their first loan and monthly for their second loan. Two years after the end of the second loan period, they used a lottery-based game to elicit subjects’ willingness to share winnings with those from their loan groups. They found that clients initially assigned to weekly groups continued to interact more often and have stronger networks long after the loan period. They were also three times as likely not to default on their second loan.
“We don’t see evidence that the group meetings put pressure on people to repay,” Pande says. “It’s not pressure; they’re helping each other repay.” To make sure this was the case, Pande and her colleagues added a study arm in which clients met weekly but paid at only one meeting a month. They saw that frequent meetings had the same strong positive effects on social networks and repayment whether or not payments were made during them.
Not only did these findings constitute the first experimental evidence on the economic returns from social interaction, but they provided an alternative explanation to peer pressure for the success of the group lending model: the positive force of social networks.
Pande points out that these borrowers are women whose social supports are lost when they marry and move to their husbands’ neighborhoods. Their interactions may be limited to husband, children, and mother-in-law. “Meeting in a group setting allowed them to get to know other women in the neighborhood,” she says. “In the longer run, they were more willing to risk-share, shown by their willingness to help their peers participate in a lottery, but also as observed by the fact that these women were less likely to default on future loans.”
So the most profitable part of the weekly meeting might be not those quiet moments as the women count their bills onto the pile — or gains made by their microbusinesses — but the gossip and the laughter. It depends on how you define “profit.”
Pande has no doubt that microfinance, which she calls a “large, vibrant industry,” will continue. “We all have a life with a lot of financial stress and a need for credit,” she says. “This is no different in India or in the United States. The issue is how to figure out products for the poor that help them have a financial life and make good financial decisions.” With Asim Khwaja, Pande will be offering a short course on this topic through a Kennedy School Executive Education program, titled “Rethinking Financial Inclusion.” Their purpose is to bring the design insights from impact evaluations directly to the decision-makers who can implement them.
Pande’s microfinance evaluations may soon loop back into policy in another setting. The Indian parliament is considering a bill that will overhaul the microfinance sector. Its members will be looking at grace periods, meeting frequency, and other ways to modify the standard contract.
Pande continues to collaborate with Field on new projects. Currently, they are working with the Kennedy School alumna Bindu Ananth MPA/ID 2007 and her organization, Kshetriva Gramin Financial Services (KGFS), to evaluate the rollout of banks in rural areas across southern India. Unlike the government banks that Pande studied at the beginning of her career, these branches are small, decentralized, and private, and will offer a range of financial services including savings accounts and insurance and also loans — but only after consultations with the clients to determine their true financial needs.
“Success for microfinance institutions means acquiring more and more clients,” Ananth says. “That’s completely different from what we’re trying to do.” KGFS tells its branches not to acquire more than 800 or 900 households per representative. “Your job is to do the best you can giving them financial services.”
This evaluation conforms with EPoD’s mission to feed the results of rigorous research into policy and then evaluate the modified policy. “Microfinance is a conservative sector — bankers don’t like to lose money,” Pande says. “Maybe the KGFS model is the way to go, coming up with a wider, more viable product, rather than subsidies. My personal sense is that it’s going to be very hard to get bankers to agree to policy changes that mean they’re going to lose money. It might be better to tell a household, ‘I’ll give you a grace period loan, and an insurance product to go alongside.’”
There’s a gentle irony to be found here: In a sector created in large part by men for the benefit of women, it will be three women leading the large-scale experiment that may determine its future direction. It may be women who ultimately decide which types of profits are worth pursuing. Erica Field suspects that the best role for microfinance may be “improving the lives of the poor marginally, in terms of giving them more consumption-smoothing tools to make their lives more convenient and easier, and to get them through rough times. Whether it will help them make investments that will bring them out of poverty is an open question.”
EPoD and the Policy Life Cycle
Evidence for Policy Design, or EPoD, is a research initiative based at cid that Rohini Pande cofounded in 2008. Its mission is to make government work better for the poor through its rigorous analysis. EPoD dips into policy at every point — diagnosing problems, proposing solutions, evaluating those solutions, and channeling the refined policy back through the process.
Diagnosing the Problem
To examine how environmental law affects people’s health in India, the EPoD researcher and Kennedy School associate professor Rema Hanna and Michael Greenstone, of MIT, assembled the most comprehensive data set ever compiled on air and water pollution and infant mortality in a developing country. They found that regulations reduced air pollution but not water pollution, and had little success lowering infant mortality. Their detailed analysis suggested that to reduce emissions, bottom-up techniques such as market-based mechanisms might work better than traditional government edicts. They also saw that agencies, despite their best intentions, simply didn’t have enough good information to go on when attempting to curb pollution.
EPoD researchers, with colleagues at MIT’s Abdul Latif Jameel Poverty Action Lab and the Indian Ministry of Environment and Forests, are piloting the first emissions trading system, or ets, in a developing country. Such a system — whereby the government puts a cap on total emissions and makes industries bid for how much they will pollute — was successful in bringing acid rain under control in the United States. To get the system under way in India, the team has spent a year working with technicians on a low-cost monitoring system, whereby devices (known as tribo-electric probes) installed in smokestacks of small industrial units will be used to gauge particulate pollution from burning fossil fuels and give readings in real time to regulators — and possibly the public.
EPoD’s cofounder Asim Khwaja has spent years analyzing the importance of information in a different setting: education in Pakistan. His work has dispelled myths, proving that the system is not given over to radical religious seminaries and that report cards on school performance can help parents make better decisions and improve market competition among schools. Khwaja’s research showed that small private schools, often based in people’s homes and run by women, outperform government schools. Now his project has evolved to evaluate how financial support and teacher training can foster this promising sector.
Teaching the Methods
EPoD’s pedagogical engagements run the gamut from in-country training seminars to Skype videoconferences, and from Harvard undergraduate “bootcamps” to teaching in Kennedy School master’s and Executive Education programs. The data collection and analysis in Rohini Pande’s projects on microfinance involved young researchers who went on to join PhD programs.
Vestal McIntyre is a freelance writer based in London.