Poor Economics

by Abhijit Banerjee

  THE HEDGE

  What can the poor do to cope with these risks? A natural reaction when faced with a drop in wages or earnings is to try to work more. But this may sometimes be self-defeating. If all the poor laborers want to work more when times are bad (for example, because there is a drought or input prices have gone up), they compete with each other, which drives wages down. The situation is intensified if they cannot find a job outside the village. The result is that the same kind of drought has a more negative effect on wages in those villages in India that are more isolated, where it is harder for workers to go outside to look for work. In those places, working more is not necessarily an effective way of coping with getting paid less.9

  If coping by working more after the shock hits is not really a good option, the best bet is often to try to limit exposure to risk by building, like a hedge-fund manager, a diversified portfolio, and it is clear that the poor invest a lot of ingenuity in doing so. The only difference is that they diversify activities, not just financial instruments. One striking fact about the poor is the sheer number of occupations that a single family seems to be involved in: In a survey of twenty-seven villages in West Bengal, even households that claimed to farm a piece of land spent only 40 percent of their time farming.10 The median family in this survey had three working members and seven occupations. Generally, though most rural families have something to do with agriculture, it is rarely their sole occupation. This can be a way to reduce risk—if one activity falters, others can keep them going—though as we will see, there may be other reasons as well.

  Holding multiple plots in different parts of the village, rather than one single large plot, also provides some risk diversification. When a blight or infestation hits one section of the village, other areas may escape; when the rains fail, the crops on plots with better access to groundwater have a better chance of surviving; and most surprisingly, different parts of the same village may actually have different microclimates, determined by exposition, slope, elevation, and moisture.

  Temporary migration can also be interpreted in this light. It is relatively uncommon for an entire family to move together to the city. Usually, some members—mostly men and teenage boys in India or Mexico, but older daughters in China, the Philippines, and Thailand—migrate, while the rest stay behind. This ensures that the family’s fortune does not rest entirely on one person’s job in the city, while also allowing the family to maintain its village connections, which, as we will see, often turn out to be useful.

  Another way the poor limit risk is by being very conservative in the way they manage their farms or their businesses. For example, they may know that a new and more productive variety of their main crop is available but choose not to adopt it. One advantage of sticking to the traditional technology is that farmers don’t need to buy new seeds—they just save enough seed from last season’s crop to replant—whereas the new seeds often cost a significant amount of money. Even if the new seeds repay the investment many times over when things go well, there is always a small chance that the crop will fail (say, because the rains don’t arrive) and the farmer will lose the extra investment he has made in new seed.

  The family is also used in creative ways to spread risk. Farming households in India use marriage as a way to diversify the “risk portfolio” of their extended families. When a woman moves to her in-laws’ village after marriage, this creates a link between the household she came from and the household she married into, and the two families are able to call on each other when in trouble.11 Farming households tend to marry off their daughters in villages that are close enough to maintain a relationship, but far enough away to have a slightly different weather pattern. In this way, if the rain fails in one village but not the other, they can help each other out. Another way to buy safety may be to have many children. Remember, Pak Sudarno had nine children to ensure at least one of them would take care of him.

  All of these ways in which the poor cope with risk tend to be very costly. This has been well documented for agriculture: In India, poor farmers use farm inputs in a more conservative but less efficient way when they live in areas where rainfall is more erratic.12 Poor farmers’ profit rates go up by as much as 35 percent when they live in areas where the yearly rainfall pattern is very predictable. Furthermore, risk affects only the poor in this way: In the case of the richer farmers, there is no relationship between farm profit rates and variability in rainfall, presumably because they can afford a loss of harvest and therefore are willing to take risks.

  Another strategy that poor farmers often adopt is to become someone’s share tenant, meaning that the landlord pays a part of the cost of farming and claims a part of the output. This limits the farmer’s risk exposure at the cost of incentives: Knowing that the landlord will take half (for example) of whatever comes out of the ground, the farmer has less reason to work very hard. A study in India showed that farmers put in 20 percent less of their own effort on land that they sharecrop compared to land where they are entitled to the entire crop.13 As a result, these plots are farmed less intensively and less efficiently.

  Having multiple occupations, as many poor people do, is also inefficient. It is hard to become a specialist in anything without specializing in it. Women who run three different businesses and men who cannot commit to a fixed job in the city because they want to keep the option of returning to the village every few weeks give up the opportunity to acquire skills and experience in their main occupations. By passing up these opportunities, they also pass up the gains from specializing in what they are really good at.

  The risk borne by the poor is thus not only costly once a shock hits: The fear that something bad might happen has a debilitating effect on poor people’s ability to fully realize their potential.

  Helping Each Other Out

  Another, and potentially much better, way to deal with risk is for villagers to help each other out. Most poor people live in villages or neighborhoods and have access to an extensive network of people who know them well: extended families, communities based on religion or ethnicity. Whereas some shocks may strike everyone in the network (a bad monsoon, for example), others are more specific. If those who are doing well now help out those who are having a bad time, in return for similar help when the roles are reversed, everyone can be made better off: Helping each other out does not have to be charity.

  A study by Christopher Udry shows both the power and the limits of such informal insurance. Over an entire year that he spent in rural Nigeria, Udry got his fellow villagers to record every gift or informal loan that they gave to each other, as well as the terms under which they repaid those loans.14 He also asked them every month if something bad had happened to them. He found that at any point in time, the average family owed or was owed money by 2.5 other families. Furthermore, the terms of the loans were adjusted to reflect the situations of both the lender and the borrower. When the borrower suffered a shock, he would reimburse less (often less than the original loan amount), but when it was the lender who had hit a rough patch, the borrower would actually repay more than he owed. The dense network of mutual borrowing and lending did a lot to reduce the risk that any individual was facing. Nevertheless, there was some limit to what this informal solidarity could achieve. Families still suffered a drop in consumption when they experienced a shock, even when the total income of everyone in their network put together had not changed.

  A large body of research on informal insurance, which has investigated this phenomenon in places ranging from Côte d’Ivoire to Thailand, finds the same thing: While traditional solidarity networks do help in absorbing shocks, the insurance they provide is far from perfect. If risk were well insured, it should be possible for a family to always consume more or less the same amount, dictated by its average earning capacity: In good times, they would help others, and in bad times, others would help them in turn. This is not what we usually see.

  Health shocks, in particular, are very badly insured. In Indonesia,
consumption drops 20 percent when a household member falls severely ill.15 A study in the Philippines documents that intra-village solidarity functions particularly poorly in the case of nonfatal severe illnesses.16 When a family has a bad harvest, or when someone loses his or her job, other families in the village come to the rescue. The affected family receives gifts, interest-free loans, and various other forms of assistance. But when individuals suffer a health shock, this is apparently not the case. The family is left to deal with it.

  The lack of insurance for health shocks is very surprising, given that families do help each other in other ways. In a previous chapter, we talked about Ibu Emptat, a woman we met in a small village in Java, whose husband had had a problem with his eyes. Her child had had to drop out of school because she could not afford the medicine for his asthma. Ibu Emptat had borrowed 100,000 rupiah ($18.75 USD PPP) from the local moneylender to pay for a cure for her husband’s eyes and when we met her, with accumulated interest she owed 1 million rupiah. She was very worried because the moneylender was threatening to take away everything they had. However, in the course of the interview, we discovered that one of her daughters had just given her a television. The daughter had just bought herself a new one for about 800,000 rupiah ($150 USD PPP) and decided to give her old one (which was still very nice) to her parents. We were a little surprised by this exchange: Wouldn’t it have made sense for the daughter to have kept her old TV and given the parents the money to pay the moneylender? We asked her, “Can’t one of your children help you with the debt?” Ibu Emptat shook her head and replied that they had their own problems, their own families to take care of—she implied that it was not for her to question the form of the gift. She seemed to think it was normal that no one would offer to help out with her health expenses.

  What stops people from doing more to help one another? Why are some forms of risk not covered, or not covered well?

  There are good reasons we may be unwilling to offer unconditional help to our friends and neighbors. For one thing, we may worry that the guarantee of help might create a temptation to slack off—this is what insurers call moral hazard. Or that people may claim that they are in need even when they are not. Or simply that the promise of mutual help may not be carried through: I help you, but when your turn comes around, you are too busy.

  These are all explanations for why we may want to hold back our help a little, but it is not clear whether this could explain not offering help to those who just became very sick, because falling ill is presumably not a choice. The other possibility is that the way most economists think about informal insurance, as situations where we help others because we might need their help in the future, may not be the whole story. It could be that we would help our neighbors in extremis even when we had no expectation of being in a similar position, for example, just because it is immoral to let your neighbors starve. Betsey Hartman and Jim Boyce’s book about life in rural Bangladesh in the mid—1970s17 describes two neighboring families, one Hindu and one Muslim, that were not particularly close to each other. The Hindu family lost its main earner and was starving; in desperation, the woman of that family would creep across the fence into the other family’s yard and steal some edible leaves from time to time. Hartman discovered that the Muslim family knew what was going on but decided to turn a blind eye. “I know her character isn’t bad,” the man said. “If I were in her position, I would probably steal, too. When little things disappear, I try not to get angry. I think ‘The person who took this is hungrier than me.’”

  The fact that people may help each other out in hard times out of a sense of moral obligation, rather than because they necessarily expect to be helped in the future, can help explain why informal networks are not equipped to deal with health shocks. When even a very poor household that has enough to feed itself sees a neighbor who does not, it just shares what it has. But helping people to pay for hospitalization, for example, requires going beyond this basic act of sharing: Many households would need to chip in, given how expensive hospitalization can be. As a result, it makes sense to exclude expensive health events from the basic moral imperative to help neighbors in need, because it would require a much more elaborate social contract to carry it through.

  This view of insurance as mainly a moral duty to help someone in need explains why, in the Nigerian villages, villagers helped each other out on an individual basis, instead of all contributing to a common pot, despite the fact that sharing risk in this other way would be more efficient. It might also explain why Ibu Emptat’s daughter gave her mother a TV but did not cover her health costs. She did not want to be the one child who was responsible for her parents’ health care (and didn’t want to presume the generosity of her siblings). So she chose to do something nice for them without biting off more than she could chew.

  WHERE ARE THE INSURANCE COMPANIES FOR THE POOR?

  Given the high cost of risk and the limitation of the insurance anyone can get from informal solidarity networks, one must wonder why the poor do not have more access to formal insurance, that is, insurance supplied by an insurance company. Formal insurance of any kind is a rarity among the poor. Health insurance, insurance against bad weather, and insurance against the death of livestock, which are standard products in the lives of farmers in rich countries, are more or less absent in the developing world.

  Now that microcredit is something that everyone knows about, insurance for the poor seems like an obvious target of opportunity for the high-minded creative capitalist (a Forbes op-ed called it an “unpenetrated natural market”).18 The poor face an enormous amount of risk and should be willing to pay a reasonable insurance premium to insure their lives, their health, their cattle, or their harvest. With billions of poor people waiting to be insured, even a tiny profit per policy could make it a tremendous business proposition, and at the same time, it would also be a big help to the world’s poor. All that seems to be lacking is someone to organize this market: This has prompted international organizations (such as the World Bank) and large foundations (likes the Gates Foundation) to invest hundreds of millions of dollars to encourage the development of insurance options for the poor.

  There are of course a number of obvious difficulties with providing insurance. These problems are not specific to the poor. They are fundamental problems, but they are amplified in poor countries, where it is more difficult to regulate insurers and monitor the insured. We have already mentioned “moral hazard”: People may change their behavior (farm less carefully, spend more money on health care, and so forth) once they know that they will not bear the full consequences. Consider some of the problems of providing health insurance, for example. We have seen that even without health insurance, the poor visit some forms of health providers all the time. What will they do if visits become free? And won’t the doctors also have reason to prescribe unnecessary tests and drugs, especially if they also own a lab (which a lot of doctors do, both in the United States and in India) or get kickbacks from the drugstore? It seems that everything pushes in the same direction: Patients want to see action, so they tend to prefer doctors who are prescription happy, and the doctors often make more money if they prescribe more. Offering reimbursement-based health insurance for outpatient care in a country where health care is at best weakly regulated, and where anybody can set up shop as a “doctor,” seems like the first step toward bankruptcy.

  Another issue is “adverse selection.” If insurance is not mandatory, those who know that they are likely to have a problem in the future may be more likely to get insurance. This would be fine as long as the insurer also knows that, because it could be factored into the premium. But if the insurance company is unable to identify those who are joining because they want care now, all they can do is raise the premium on everyone. The higher price, however, makes things worse, as it drives away those who know they will probably not need the insurance, thus exacerbating the original problem. This is why, in the United States, getting health insurance at reasonabl
e prices is very difficult for those who cannot get it through their employers. And this is why affordable health insurance programs tend to be mandatory—if everyone is required to join, the insurer does not get stuck with just the high-risk types.

  A third problem is outright fraud: What is to prevent a hospital from presenting to the insurer a large number of bogus claims or charging the patient substantially more than the cost of care received? And if a farmer purchases insurance for his water buffalo, what is to stop him from claiming that the buffalo had died? Nachiket Mor and Bindu Ananth from ICICI Foundation are the two people in the Indian financial sector most committed to designing better financial services for the poor. They recounted to us, with self-deprecating humor, their first disastrous attempt, many years ago, to provide cattle insurance: After the first lot of policyholders universally claimed to have lost their cattle, they decided that in order to claim that an animal had died, the owner would need to show the ear of the dead cow. The result was a robust market in cows’ ears: Any cow that died, insured or not, would have its ear cut off and the ear would then be sold to those who had insured a cow. That way they could claim the insurance and keep their cow. In summer 2009, we went to a meeting where Nandan Nilekani, founder and ex-CEO of the Indian software giant Infosys, who had been charged by the government to provide every Indian with a “unique ID,” was explaining his plan for unique identification. He assured listeners that ten fingerprints and a picture of the irises are essentially enough to uniquely identify anyone. Mor was listening intently. When Nilekani paused, he spoke up: “It’s too bad that cattle don’t have fingers.”