More Than Good Intentions

by Dean Karlan

  The rest is history. Since receiving a banking license from the government of Bangladesh in 1983, the Grameen Bank has grown steadily. Today it serves over six million clients with a total loan portfolio approaching $650 million. Along the way, Yunus and the Grameen Bank jointly picked up the 2006 Nobel Peace Prize for their efforts and, more important, inspired millions around the globe to follow their lead. Today, over a thousand microcredit institutions operate on six continents, serving some 155 million borrowers.

  As the numbers and accolades attest, people are excited about microcredit. Everyone is singing its praises, from UN secretaries general to rock-star economists to bona fide rock stars. Some see it as the storied “golden bullet”—the singular big idea that will solve poverty once and for all. Jeffrey Sachs, the noted economist and special adviser to the UN on its ambitious Millennium Development Goals antipoverty initiative, whom we mentioned in the introduction, is one of its most influential advocates. He writes, “The key to ending extreme poverty is to enable the poorest of the poor to get their foot on the ladder of development. . . . They lack the amount of capital necessary to get a foothold, and therefore need a boost up to the first rung.”

  Celebrities from other quarters are on board too. For instance, the actress Natalie Portman serves as Ambassador of Hope for FINCA, the same charity that finances Mrs. Potosí’s sweater venture. And antipoverty crusader Bono, lead singer of the rock band U2 and an outspoken ally both of Sachs and of the poor all over the world, adapts a proverb we met earlier: “Give a man a fish, he’ll eat for a day. Give a woman microcredit, she, her husband, her children, and her extended family will eat for a lifetime.”

  With so much hype about microcredit, what we need to do is drop our preconceptions and take a clear-eyed, unbiased look at the evidence. In this chapter we’ll do just that. We’ll see that there are real success stories, but that, as with “teach a man to fish,” it’s not as simple or as universal as we’d all like. In the following two chapters we’ll look at evidence for ways that microcredit programs might be improved, and we’ll conclude our foray into the world of microfinance by arguing that we should likely be paying a whole lot more attention to microsavings instead.

  Erlyn Drops Out

  Sari sari translates literally from Tagalog, the most widely spoken indigenous language of the Philippines, as “this and that.” It’s a phrase you will learn quickly if you visit the country, because you’ll find it emblazoned on signboards in every city and every village. The signboards will be red and rectangular, with Coca-Cola logos on either side, and in their centers will be white lettering in all caps declaring: SARI SARI STORE.

  True to their name, these are stores that contain a grab-bag of goods. Depending on the neighborhood, you might walk down to your local sari sari store to buy a plate of hot stewed pork and rice, some new pencils, a cup of hot coffee, an individual serving of laundry detergent, packs of dried spaghetti, prepaid mobile phone credit, or fresh-picked coriander.

  There is a method to the miscellany, though. The principle behind a successful store is simple: What do people want? The answer, of course, is always changing; but on any given day a fair approximation is written in the products on the narrow shelves and in the display cases.

  Jake and I met one particular sari sari store owner, Erlyn, in the summer of 2009. Evidently, the people who shopped in Erlyn’s neighborhood wanted pork rinds. Assorted pork rinds, in all different sizes and flavors and levels of crispiness.

  Erlyn was happy to oblige. Above the counter in her sari sari store a great waterfall of pork rinds was frozen in midcascade, suspended in plastic bags and in bandoliers of foil pouches, hung from the lintel with binder clips and fishing line. The other popular item was Tang, and it was also well-represented, a kaleidoscope of colored sachets strewn among the shelves.

  Erlyn did more than cater to her patrons’ tastes. She also found ways to accommodate their budgets so they could buy the things they wanted without breaking the bank. She sold individual cigarettes and half-servings of Coke, which were really just little plastic bags, each filled with a few ounces of soda and tied shut. (I first experienced soda-in-a-bag in Central America. Shopkeepers like the concept because they get to keep the glass bottle for the bottle refund, but an unintended consequence is that the buyer has to drink it almost immediately, as it is hard to put down a bag of liquid! Each bag was about fifty cents. I remember offering a store owner in Honduras a dollar to have the bottle, too, but evidently there simply was no price for such a luxury.)

  Through this myriad of goods, sold bit by bit, Erlyn built a successful business.

  Just as she had assembled a motley assortment of products to meet her customers’ needs, she had cobbled together a financial solution to meet her own. Well, to a point.

  At first glance Erlyn might seem like an ideal microcredit client, and for a while she was one, borrowing from one of the largest nonprofit lenders in the Philippines. She had been very successful with her first few microloans, and so her coborrowers and loan officer encouraged her to borrow more. She did. But when she came home with twenty thousand pesos (about four hundred dollars), her largest loan yet, she found she couldn’t put it all into the business at once. There simply was not enough room in the store for that many pork rinds. They would have been spilling out into the street. So she invested what she could in inventory, and the rest commenced to burn a hole in her pocket. There were opportunities everywhere: “When it is twenty thousand, then I would spend some in the house, on clothes or a TV. Then I know it is too much. It is so easy to spend!”

  The store had reached capacity. Erlyn could have simply capped her borrowing at the amount she could spend on inventory, but that wouldn’t have served her needs, either. The bank only made loans over six months, and the store needed restocking every two. It would not serve to borrow three times the restocking amount, since money lying around had a habit of disappearing. She was stuck.

  But not completely stuck. Formal microcredit is not the only source of credit for the poor. In fact, even in places where microcredit is widespread, we see individuals using credit from neighbors, family members, store owners, and yes, the reviled (but reliable!) moneylender. In their recent book Portfolios of the Poor, Daryl Collins, Jonathan Morduch, Stuart Rutherford, and Orlanda Ruthven use detailed analyses of households in South Africa and Bangladesh to learn about the plethora of options and mechanisms the poor use to save and borrow. The story clearly is not as simple as “microcredit provides the poor with loans that they otherwise could not get.”

  In this vein, Erlyn had a specific solution, and the solution made house calls. The local moneylender offered forty-five-day and sixty-day loans, and he came by the shop each day to collect payments. His interest rate was higher than the nonprofit lenders, but he could lend Erlyn just the amount she needed, and for just as long as she needed it. To her, it was worth the additional cost. She left the bank and has been borrowing steadily, and quite happily, from a moneylender for the past two years.

  This is not the way it’s supposed to work. According to the promotional pamphlets, microcredit is supposed to help you wrest yourself away from the usurious clutches of the local moneylender, not convince you that, on balance, he offers a service better suited to your needs. What can explain this mystery?

  Stripping Down to Bare-bones Loans

  Actually, it’s not such a mystery; there is less of a clear dividing line between microcredit and moneylending than you might imagine. People are often surprised to learn that the terms of many microloans around the world would violate the usury laws of most U.S. states. Consider a few examples from Mexico: The local affiliate of FINCA, a nonprofit microlender, lends at an 82 percent APR yield when all fees are included; Pro Mujer, another major nonprofit, lends at 56 percent. The for-profits aren’t charging any more (but they get more of the heat—why is that?): Compartamos, for example, a publicly traded for-profit company, charges 73 percent. That’s far worse than any American c
redit card. Even the low end of the microcredit interest rate spectrum, with annual rates in the neighborhood of 20 percent, is high by our standards.

  This begs the question touched on earlier: What is microcredit, if it’s not just another way of saying “small loans”? Despite the buzz surrounding the concept, that’s not an easy question to answer. Some modern incarnations of microfinance bear little resemblance to the system pioneered by Yunus with the bamboo craftswomen. Perhaps the best one can do is invoke former U.S. Supreme Court Justice Potter Stewart’s famous description of pornography: “I know it when I see it.” Still, there are some persistent features of microcredit—an explicit social mission, an emphasis on entrepreneurship, a requirement to spend loans on microbusinesses, group lending, frequent group meetings to make loan payments, a focus on women’s empowerment through borrowing—that are commonly thought to distinguish it from plain-vanilla moneylending.

  We can ease our way into the big questions about whether, how, and why microcredit works by stripping back all of these distinguishing features till we’re left with the bare essentials: a dollar amount, a maturity date, and an interest rate. If even bare-bones loans like these can be beneficial to borrowers, then there is good reason to be optimistic about microloans in general.

  In 2004, Jonathan Zinman and I were wrapping up a marketing and interest-rate study (which we saw in the last chapter) with Credit Indemnity in South Africa. The people there were friendly, smart, and fun to work with, but Credit Indemnity was not a warm, fuzzy microcredit operation. It was a for-profit consumer credit business with no social agenda—a closer cousin to payday-loans outfits in the United States, or Erlyn’s friendly door-to-door moneylender, than to Muhammad Yunus’s Grameen Bank. It didn’t target women or entrepreneurs, didn’t care what borrowers did with the money (so long as they paid it back!), and lent only to working people. And it charged about 200 percent APR. In short, it was in no danger of winning a Nobel Peace Prize.

  What we need to know is: Do these loans actually make people better off?

  Jonathan and I spotted a chance to find out. Over the course of our interest-rate and marketing study, we had been struck that Credit Indemnity spent a surprising amount of time turning away potential clients. In fact it was rejecting fully half of its applicants on account that they were too risky to lend to. But our analysis of the data suggested that clients who barely met the lending requirements were extremely profitable customers for the lender. So we had to wonder: Could it be that the barely rejected applicants would have been profitable too?

  After a bit of prodding and a lot of brainstorming with their credit team, we came up with a simple idea for an RCT that would serve everybody. It would help Credit Indemnity to improve its operations (and potentially its bottom line), and would also allow us to answer our question about whether borrowers benefited from credit. Where researchers usually face a struggle between answering meaningful questions and limiting interruptions to partners’ operations, this project struck a perfect balance.

  It worked by piggybacking on the existing lending process. When a new customer came in to apply for a loan, a staff member fed some basic information, like age, income, and number of years in their job, into a computer program, which instantly returned a basic recommendation about creditworthiness—either a thumbs-up, a thumbs-down, or a “maybe.” We modified the software so that some “maybes” would randomly be assigned a thumbs-up and others thumbs-down. While credit officers were allowed to ignore the computer’s recommendation, the net effect was that some marginally creditworthy applicants were randomly granted loans. By tracking all the applicants on the cusp—both those who were randomly assigned to be accepted and those who were randomly assigned to be rejected—and comparing their experiences, we could see whether receiving a loan made people better off.

  A year later, a coherent picture had emerged. Applicants who received a random thumbs-up were significantly more likely to have kept their jobs, and had significantly higher incomes to show for it. Clients’ families—not just the borrowers themselves—enjoyed greater prosperity too. The randomly approved applicants’ households earned more money overall and were less likely to be below the poverty line. Survey responses showed that they were also less likely to go to bed hungry.

  Most important, the strength of the results on income and job retention allowed us to effectively rule out the possibility that these loans were, on the whole, pernicious.

  This was great news for advocates of microcredit. Actually, it was great news for advocates of payday loans too. Lenders around the world were being attacked left and right for pushing their evil debt on borrowers, but much of the ammunition was invective and innuendo—not facts. Given the paucity of reliable information on the impacts of credit, any evidence showing it to be good—even at high rates—was a welcome addition to the conversation.

  Providing some hard evidence in favor of lending was a start, but the study with Credit Indemnity did more than that. It also showed us something interesting about the specific ways that loans can lead to prosperity. In many cases, we learned, the loans were used to deal with unexpected shocks.

  Two common stories emerged. First, many borrowers used the loan to pay for transport-related costs. They repaired brokendown cars and motorbikes and bought bus fares, all of which allowed them to get to work on time and avoid getting into hot water with their employers. Second, borrowers sent money home to needy relatives in rural areas. Had they been unable to send assistance, many would have been obliged to pull up stakes and go in person to help their loved ones, a move that would have spelled disaster for their steady jobs at home. But with the aid of credit—even the high-priced consumer variety—both stories had comparatively happy endings. The paychecks kept rolling in.

  Golden Eggs and the Case for Microcredit

  So far, so good. We have learned that bare-bones small loans can work for Credit Indemnity’s eligible borrowers, people in formal employment. Now what about the usual target market of microcredit, small-scale entrepreneurs?

  The basic idea behind microcredit is that the poor actually have great economic opportunities, but that they lack the resources to take advantage of them. Here’s a typical example: Lucia, a seamstress, makes her living sewing and mending clothes by hand. The $5 profit she makes each day is just enough to feed her family and pay their rent. With a $100 electric sewing machine she could double her output (and her profits), but that’s money she doesn’t have—until she pays a visit to a microlender.

  Lucia takes a six-month loan for $100, buys the sewing machine, and starts earning $10 per day. Even if the lender charges 100 percent APR—which, again, would be an unthinkably (and probably unlawfully) high interest rate in the United States, but is entirely realistic for a microloan—Lucia still comes out ahead, and by a wide margin. She has to set aside just under a dollar each day to make her monthly payment of $21.85, leaving her with $9 a day for her family instead of the $5 she earned previously. Once the loan is paid off, she keeps that last dollar for herself and goes home each day with the full $10. And so, without much fanfare, Lucia nearly doubles her income thanks to a 100 percent–APR loan.

  Simple, right?

  If such lucrative business opportunities do exist for the poor, then high interest rates on microloans are not such a problem; borrowers and banks can both win. But that’s a big “if.” A one-time hundred-dollar investment that doubles long-term profits, like Lucia’s sewing machine in our example, is a goose that lays golden eggs. Are there really, as microcredit advocates claim, so many of those golden geese waddling around the stalls of Dakar’s or Dhaka’s outdoor markets, or making themselves at home in the flooded rice paddies of Thai smallholders’ farms?

  We have to answer this fundamental question to know whether and when microcredit can be beneficial at all. Because one thing is certain: It can only work if clients pay back their loans. No matter what the interest rate, if clients rely on an increase in profits from borrowing to make th
eir payments, then the viability of the whole system depends on how much that increase will be. In economic jargon, the question to ask here is “What is the marginal return to capital for the enterprise?” In other words, if a microentrepreneur puts an extra dollar into her business, how much more profit will she earn?

  In 2005, three economists, Suresh de Mel from the University of Peradeniya in Sri Lanka, David McKenzie from the World Bank and IPA, and Chris Woodruff from the University of California at San Diego, set out for southern Sri Lanka to answer that question and find out what kind of business prospects the poor really had. How powerful were the economic engines of microenterprise, after all? Their strategy was simple and direct: They would inject money into some businesses and see how much additional profit was generated.

  Going door-to-door, the researchers found 408 microentrepreneurs. They were tailors, lace tatters, bamboo craftspeople, owners of small grocery stores, bicycle repairmen—the prototypical microentrepeneurs that we hear about every time we hear about microcredit. Half of them were randomly selected to receive a grant of either one hundred or two hundred dollars (the amount was also randomly chosen). This was a sizable amount, roughly equivalent to three or six months’ profits from the typical business.

  The researchers tracked the profits of all 408 businesses with quarterly surveys over the following fifteen months, and compared those of grant recipients with those of nonrecipients. Recipients’ monthly profits increased by about 6 percent of the grant amount on average. That is, investing an additional hundred dollars in the business generated six dollars more in profits per month, or seventy-two dollars more in profits per year. And potentially even more if the additional profits were reinvested back into the business. As a reference, if you put all your money into an investment with 70 percent annual returns (and kept reinvesting the profits), your wealth would almost double every year. That’s a golden-egg-laying goose if ever there was one.