Mandated Empowerment

Handing Antipoverty Policy Back to the Poor?


  • Abhijit V. Banerjee,

    1. Department of Economics and Abdul Latif Jameel Poverty Action Lab, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
    Search for more papers by this author
  • Esther Duflo

    1. Department of Economics and Abdul Latif Jameel Poverty Action Lab, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
    Search for more papers by this author

Address for correspondence: Abhijit V. Banerjee, 50 Memorial Drive, E52-252d, Cambridge, MA 02142.


The current trend in antipoverty policy emphasizes mandated empowerment: the poor are being handed the responsibility for making things better for themselves, largely without being asked whether this is what they want. Beneficiary control is now being built into public service delivery, while microcredit and small business promotion are seen as better ways to help the poor. The clear presumption is that the poor are both able and happy to exercise these new powers. This essay uses two examples to raise questions about these strategies. The first example is about entrepreneurship among the poor. Using data from a number of countries, we argue that there is no evidence that the median poor entrepreneur is trying his best to expand his existing businesses, even if we take into account the many constraints he faces. While many poor people own businesses, this seems to be more a survival strategy than something they want to do. The second example comes from an evaluation of a program in India that aims to involve poor rural parents in improving local public schools. The data suggest that despite being informed that they now have both the right to intervene in the school and access to funds for that purpose, and despite being made aware of how little the children were learning, parents opt to not get involved. Both examples raise concerns about committing ourselves entirely to antipoverty strategies that rely on the poor doing a lot of the work.


As a purely abstract, normative proposition, it is hard to argue with the view that the poor ought to have greater control over their lives and that social policy must help; as the popular expression goes, the poor help themselves. In practice, it is not always clear what this means—part of the problem is that you cannot really have a lot of control when you are living on less than $2 a day, and that is not about to change. In the policy conversations in today's developing world, nobody is talking about offering the poor the right to quality health care when needed, or the ability to send their talented child to a really good school, let alone the capital to set up the world-class factory of their dreams, if it were that they would dare to dream of something so distant. No government in the developing world thinks that these are things that they can afford, and indeed, they may be right, given the size of the problem and the limits on their ability to mobilize resources.

What is on offer, typically, are things of very different order: some fairly rudimentary health insurance that might cover you for a simple surgical procedure or a heart attack, some microcredit or microsavings, attempts to improve the public education and healthcare systems, or conditional income support programs. The Indian government's universal health scheme, for example, announced in the 2003–2004 union budget, offered each family below the poverty line a policy that would cover them (in return for a subsidized premium of $11 per year) for a maximum of Rs. 30,000 ($600 in 2003) per year for the entire family and a cap of Rs. 15,000 ($300) per illness. Rs. 15,000 is a significant amount of money in Indian terms (the daily unskilled wage was around Rs. 60 a day in 2003), and medical costs are low by global standards, but the cost of open-heart surgery at Narayana Hrudayalaya, India's best-known example of social entrepreneurship in the health sector, at its published rate for financially constrained patients, is Rs. 65,000. The insurance can, at best, bring them a small part of the way there.

Then there is microcredit: After resisting the idea for many years, governments across the developing world (and beyond) have now decided to make it a central tool in their fight against poverty. The Indonesian government's Small Farmers Development Program, the National Program for Financing Micro Entrepreneurs launched by the Mexican Federal government of Felipe Calderon in 2006, and the Self Help Groups promoted by the government of India, are a few of the hundreds of governmental, semigovernmental, and private programs that are now offering tiny (ranging from less than $100 to $1000 or so) uncollateralized loans to indigent would-be entrepreneurs. This is by no means free money. You have to pay interest (which can be as high as 5% per month) and there is often an obligation to attend weekly meetings. And repayment is, of course, a must if you want to continue in the program. Nevertheless, the terms are typically much less harsh than what the market offers comparably poor people, who are often excluded entirely from all formal credit markets, and the idea is that this will stimulate entrepreneurship among the poor.

The Education for All (Sarva Shiksha Abhiyan) program launched in India in 2001 in many ways typifies the trend in public services these days. There is a significant inflow of resources—in total, around Rs. 100 billion a year for the entire country. This looks like a lot of money, but there are almost a million primary and upper primary schools in India, and many of them need teachers, but teachers are relatively expensive. In place of teachers, the government offers the schools to hire shiksha karmis, who are local teachers who are significantly less qualified than the civil-service teachers, but also much cheaper. Importantly, the responsibility for applying for the money to hire these teachers, as well as the decision to keep them on, was in the hands of the Village Education Committee (VEC), constituted mainly of parents from the school. More generally, the new rules transferred considerable monitoring and other powers into the hands of the parents, as well as some discretionary resources.

The recent education reform programs in Kenya went a step further. To accommodate the increase in enrollment resulting from Kenya's introduction of free and compulsory primary education in 2003, a program funded by the World Bank provided school committees with the funds to hire local teachers on a short-term contract basis. They were also entrusted with monitoring these teachers and deciding whether they should be renewed on the job. Similarly, when Madagascar recently eliminated school fees for primary education, a government decree created “primary school boards,” consisting mainly of parents, and devolved to them the task of managing the newly launched capitation grants (caisse ecole), that are supposed to replace the funding they used to raise from the parents.

Conditional cash transfer programs, having made their reputation in Mexico, are now spreading across the globe (Indonesia has just launched one). In these programs, the financial assistance that families get from the state depends on their compliance with certain norms: children have to be immunized and sent to school, for example. PROGRESA, a federal government program aimed at “breaking the intergenerational transmission of poverty” in Mexico, launched in 1997, made cash transfers to eligible household—stipends for upper-primary and secondary school students, subsidies for school supplies, and a cash transfer for nutrition, averaging 22% of average income among participating households—conditional on the family behaving in a certain way. Children had to be immunized and to attend school at least 85% of the time for the family to eligible; high fertility was discouraged by delaying grants until children reached the age of seven and capping the family entitlement; additional stipends were offered to those families where children stayed on in school beyond the primary years and these stipends increased with years of schooling; and families that sent their daughters to secondary school were rewarded with larger grants. The program reached 2.6 million families (40% of rural families) with an annual budget of approximately $777 million, 0.2% of GDP. India has taken a more traditional approach: the National Rural Employment Guarantee Act, 2005, offers every family a right to100 days of manual labor on public projects every year at a set wage. The idea is that those and only those who need the money badly enough will want take up the offer.

What all of these recent social programs have in common is the idea that the poor need to do their bit to help themselves. They need to buy the health insurance policy and pay the premium, which is only mildly subsidized,a instead of getting an unconditional guarantee from the state that their major health expenses will be taken care of (as they would be under the British National Health Service or Medicaid in the United States, or indeed under the national health system mooted in India's sixth plan document [1972]). They need to make sure that they make enough money to keep financing their microloans and show up in the weekly meetings to make the payments. They need to involve themselves in the running of their children's schools—hire and fire teachers and keep track of whatever is going wrong. Moreover, to be eligible to continue to receive social assistance, they need to make sure that their recalcitrant teenagers show up in school regularly.

The presumption clearly is that poor are able and indeed happy to do all this (and more), given the right environment and a little bit of help. John Hatch, founder of the Latin American microcredit institution FINCA, took this view to its logical conclusion when he said, “Give poor communities the opportunity, and then get out of the way!”

Yet, other than the negative evidence of the failure of many top-down programs that do not involve the beneficiaries enough (do we really even know that this is the reason why they failed?), what reason do we have to believe that this attempt to shift the burden on to the poor will work better? How do we know that the poor are really raring to go, as so many of the supporters of the current policy bias would have us believe? How can we be sure, to quote the World Bank this time, “If the stakes are high enough, communities tackle the problem,” unless that is what defines “high enough”?1

The point of this essay is to try to take a stab at answering this set of questions, using some data that come out of our recent work. What we have to say is nowhere near definitive, but the question seems important enough that it is worth trying. The two examples we have—promoting entrepreneurship among the poor and public school reform—are, conveniently, very different. The first has everything to do with private profit maximization; the second is almost entirely about promoting the public good. Both demand activism, but of very different kinds.

Promoting Entrepreneurship among the Poor

The description by Dr. Yunus, the remarkable founder of the famous Grameen Bank, of the poor as “natural entrepreneurs” must be one of the most powerful images in the world of policy today. In combination with business guru C. K. Prahalad's exhortation to businessmen to focus more on what he calls the bottom of the pyramid (in part because he believes that the very poor are “resilient and creative entrepreneurs”) and Peruvian social critic Hernando de Soto's claim that the poor stay that way because so much of their capital is “dead capital” (that is, capital that cannot be pledged against a loan), it is helping to secure a space within the overall antipoverty policy discourse where big business and high finance feel comfortable getting involved. The traditional strategies of public action are being supplemented by private actions, often by corporate leaders (Pierre Omidyar of ebay, for example), directed at helping the poor realize their potential as entrepreneurs. Microcredit is one important piece of this new policy package, as are business training and market creation.

In what sense are the poor natural entrepreneurs? To answer this question (and many other questions about the poor), we looked at data about the income-earning strategies of the poor from household surveys conducted in 13 countries from three continents: Cote d'Ivoire, Guatemala, India, Indonesia, Mexico, Nicaragua, Pakistan, Panama, Papua New Guinea, Peru, South Africa, Tanzania, and Timor Leste. These surveys mainly come from the Living Standard Measurement Surveys conducted by the World Bank and the “Family Life Surveys” conducted by the Rand Corporation as well as two surveys that we conducted in India with our collaborators. The first was carried out in 2002 and 2003 in 100 hamlets of Udaipur district in the state of Rajasthan.2 The second survey covered 2000 households in slums of Hyderabad, the capital of the state of Andhra Pradesh and one of the boom towns of post-liberalization India.

From each of these surveys we identified the extremely poor as those living in households where the consumption per capita is less than $1.08 per person per day, as well as the merely “poor,” defined as those who live on less than $2.16 a day in purchasing power parity (PPP) terms at 1993 prices. In keeping with what is now an established convention, we call these the $1 and $2 dollar poverty lines, respectively.

What is clear from these data is that all over the world a substantial fraction of the poor act as entrepreneurs in the sense of raising the capital, carrying out the investment, and being the full residual claimants for the earnings. In Peru, 69% of the households who live under $2 a day in urban areas operate a nonagricultural business. In Indonesia, Pakistan, and Nicaragua, the numbers are between 47% and 52%. A large fraction of the rural poor operate a farm: except in Mexico and South Africa, between 25% and 98% of the households who earn less than $1 a day report being self-employed in agriculture.b Moreover, many of the rural poor—from 7% in Udaipur up to 36% in Panama—also operate a nonagricultural business.

Should we therefore assume that the poor are entrepreneurial, in the sense that they have a special inclination for seeking out business opportunities? One way to look at this question is to compare them with the middle classes. In one study, we compared the poor with two other groups from the same 13 countries: Those with daily per capita consumption between $2 and $4 at PPP and those between $6 and $10. The basic conclusion from that study is that the rural middle class is typically less likely to be self-employed in agriculture than the rural poor, but more likely to have their own nonagricultural business, so that on balance there is no systematic pattern in their likelihood of being an entrepreneur. The big difference comes in what they do when they are not entrepreneurs: the rural poor are much more likely to be casual laborers, while the middle classes typically hold steady jobs.3

In urban areas in contrast, the broad occupation patterns are remarkably similar across the different levels of economic well-being. The share of entrepreneurs stays roughly the same, as does the share of employees. Once again, the big difference is in what the employees do: middle-class employees are more likely to hold a salaried job.

All the countries in our study range from poor to middle-income. It is therefore no surprise that by the time we come to the $6–10 category we are talking about people who are actually quite high up in the distribution of income—in all countries they are above the median and in most above the 80th percentile. Hence, a way to summarize the previous paragraphs is to say that there is no systematic difference between the poor and most other people, in terms of their observed propensity to be an entrepreneur.

This is impressive: After all, everything else seems to militate against the poor being entrepreneurs. They have less capital of their own (almost by definition), less access to bank and other sources of inexpensive finance, and almost no access to any formal insurance. As a result, they are less able to make the investments needed to run a proper business and more vulnerable to any additional risk that comes from the business itself. The very fact that they are still about as likely to go into business as their richer counterparts may be seen as a sign of their entrepreneurial spirit.

On the other hand, it can also be read as evidence that the poor have worse alternatives: Clearly, the middle classes have access to better jobs than the poor, who, when they do not run their own farm or businesses, often end up as casual laborers. It is therefore possible that the poor become entrepreneurs because they have nothing better to do.

We have no way to definitively reject one of these views, but there is some evidence that weighs strongly in favor of the latter position. First, all over the world, the businesses that the poor run tend to be very small: Most of them are run by one or at most two family members and have no employees. Often they do not even exhaust the labor supply available within the family. A very substantial fraction of families that own a business either have multiple businesses or also have one or more members working as laborers. In rural Udaipur district, for example, almost everybody owns some land and almost everybody does at least some agriculture. In that sense everyone is an entrepreneur. Yet only 19% of the households describe self-employment in agriculture as the main source of their income. The most common occupation for the poor in Udaipur is working as a daily laborer: 98% of households living under $1 per day in rural areas report doing this and 74% claim it is their main source of earnings. Similar patterns are observed all over the world both among rural, and to a lesser extent, among urban populations.4

Of course, there is nothing necessarily wrong with businesses being small. This could be the right scale for the particular business. However there are several reasons why we discount this possibility: First, when people estimate the returns to investment in these businesses they usually find very large numbers. For example when De Mel et al. gave randomly selected owners of firms in Sri Lanka that were very similar to the firms in our study an infusion of capital equal to 100% to 200% of the capital stock, they report returns to capital of over 5% a month, or 80% per year.5 This is consistent with the second fact, which is that when these businesses borrow they often pay interest rates of 4% per month or even more. In the Hyderabad survey, for example the poor report paying 3.99% per month on average.c Given any reasonable assumption about the social cost of capital, it would make sense for them to expand their businesses substantially.

There is, obviously, nothing surprising about the businesses of the poor being inefficiently small. They do not have any capital of their own and borrowing more is not easy for them. What is more striking is that there seems to be no evidence that the average poor person is trying to grow their business so that it reaches a more desirable scale. For one, the businesses of the middle classes are not much bigger than that of the poor.3 The number of employees is still tiny (the businesses of those with daily per capita expenditure (dpce) between $6 and $10 in the median country have at most one paid worker), and they still seem to operate with very little by way of assets, such as machinery or something to transport people and/or goods, that could be useful for their business. Unlike for example, radios and televisions, ownership of bicycles, which are potentially an income-earning asset, does not increase as we go from the poor to the less poor.

Second, given how little capital they work with, it is easy even for the poor to double or even quadruple the size of their businesses by saving a little bit more. In Banerjee and Duflo, we do a simple but (hopefully) revealing calculation.3 We take someone who lives in a household with a dpce of $2, which puts him at the border between those we call the poor and those we call the middle class. In Hyderabad, the middle classes spend about 10% of their total monthly expenditure on health care, while the poor spend about 6.3%. If, instead of spending the extra 3.7% on health care, our person at $2 dpce used it to build up his inventory, we argued that he could double his inventory in a year. Alternatively, the family could cut down completely on cigarettes and alcohol and save about 3% of their dpce: This would allow them to double their inventory in about 15 months.

The fact that the poor are not trying desperately to grow their businesses also comes out in a study of fertilizer use in Kenya among farmers by Duflo, Kremer, and Robinson.7 According to surveys conducted over several years, just 40.3% of farmers had ever used fertilizer, and just 25% used fertilizer in any given year. Estimates from experiments carried out on people's own farms suggest that the average return to using fertilizer exceed 100%, and the median return is above 75%. Duflo, Kremer, and Robinson conducted actual field trials of fertilizer on the farms of randomly selected farmers, in the hope of demonstrating to the farmers the best way to use fertilizer and the rewards from doing so. They found that the farmers who participated in the study are only 10% more likely on average to use fertilizer in the next season after the study, and the effects fade after the first season.

When asked why they did not use fertilizer, most farmers replied that they did not have enough money. However, fertilizer can be purchased (and used) in small quantities, so this investment opportunity seems accessible to farmers with even a small level of saving. The main issue, once again, appears to be that farmers find it difficult to save even small sums of money. The program in Kenya offered to sell farmers a voucher right after the harvest, which is when farmers have money in hand, which would entitle them to buy fertilizer later.

This program had a large effect: 39% of the farmers offered the voucher bought the fertilizer; the effects are as large as a 50% subsidy on the cost of fertilizer. The voucher seemed to work as a commitment device to encourage saving. But a puzzle remains: a huge majority of the farmers who bought the vouchers for future delivery of fertilizer requested immediate delivery, and then stored the fertilizer for later use. Moreover, almost all of them used the fertilizer they bought. They apparently had no self-control problems in keeping the fertilizer, even though they could easily exchange the fertilizer for something more immediately consumable. But then they could have done it all on their own. Given that doing it has very high returns, why did they need all the prodding to make them do it?

It is hard to resist the thought that the enterprises of the poor are more a way to stay afloat in difficult circumstances than a reflection of a particular entrepreneurial urge. Many of the businesses get run because someone in the family has some time in hand and every little bit helps. And it is probably not always clear that the person who ends up running the business (often a woman) actually wants to do it: It is a huge, if low intensity, time commitment—the median full-time entrepreneur in Hyderabad works 11 hours a day, 7 days a week. But she may not necessarily be in a position to say no.d

None of this, it is worth emphasizing, says that the poor are not interested in making money. While fertilizer, for example, offers high returns, the actual absolute gain in income that it promises someone who has little land and less capital, is not huge. It is entirely possible, indeed likely, that if they were offered an opportunity to make some real money (say a thousand dollars or more) without having to do something too unpleasant or wait too long, they may have responded very differently. But this is not what, for the most part, is on offer.

However, and this is important to emphasize, this does not at all mean helping the poor to get more credit and better business skills is a bad idea. After all, even if 1 in 10 (or even 1 in 20) among them is a born entrepreneur who would have otherwise been lost, it has to be worth the while. Moreover, microcredit potentially has important dimensions that have nothing to do with entrepreneurship, though the microcredit movement is still conflicted about them. For one, consider someone who wants to buy a durable (say a TV) but has trouble saving, perhaps because she has no way to protect her savings from the claims of her relatives. The microcredit organization offers her the option of borrowing the money, buying the durable, and in the process committing herself to save the money (since she has to repay the loan).

What is at issue then is to what extent the focus on entrepreneurship needs to be qualified by the recognition that many poor people would prefer not to have to be an entrepreneur to make a living. The danger is that governments may be all too willing to jump on the entrepreneurship band-wagon, both because the elites like it, and also because, other, more old-fashioned ways of helping the poor, like providing a quality education and helping to create jobs, are inherently more difficult and more expensive.

Using Community Control to Improve Education

There is mounting evidence that government services are failing the poor. When a group of scholars from Harvard University and the World Bank sent observers unannounced on three separate occasions to 3700 public and private schools in India, they found that 25% of teachers were absent on any given day.8 Moreover, only 45% of the teachers present were actually teaching when the observers arrived. The rest were either drinking tea or talking to other teachers or reading the newspaper. And, lest you think this is some South Asian aberration, the absence rate they found in Uganda was even higher (27%), while other countries, such as Bangladesh (15%), did substantially better. The study did not try to assess the quality of teaching, but it is hard to imagine that teachers who do not want to come to school, and do not teach when they do, put their heart and soul into teaching when they actually do teach.

It is therefore hardly a surprise that students in these school systems are not learning very much. A recent nationally representative survey in India, found that only 43% of fifth graders could do simple (one digit) subtractions and divisions and only 60% can read at a second grade level. Yet 93% of 6- to-14-year-olds say that they go to school (though attendance is only about 70%, reflecting, perhaps, their level of enthusiasm).9

The numbers for health services were, if anything, worse. The same study found absence rates for government health providers in Bangladesh and India of 35% and 43% respectively.8

The World Bank's World Development Report 2004, Making Services Work for Poor People,1 was very important for its honest and frank acknowledgement of this problem. It also discussed a range of strategies for doing something about it, including most importantly given its political ramifications, greater reliance on community control.

The basic theory here is that the families of the children who go to public schools, more than anyone else, see what is wrong with the school and the teaching, and have the interests of their children at heart. Therefore, if they only had the power to reward performance and punish negligence on the part of the teachers, things would work much better.

The trouble is that there are community-run schools in India, and the same survey that reported on teacher absenteeism in government and private schools tells us that teachers in community schools actually come less often even than in government schools.8 This becomes less of a puzzle when one actually talks to the community. In a recent study, a group of us (from Pratham, the World Bank, and J-PAL at MIT) carried out a survey of households in Jaunpur district of the North Indian state of Uttar Pradesh aimed at gauging the role played by the community in overseeing.10 One of the questions we asked the surveyed households was whether there was any committee in the village that was meant to deal with issues around education. By law every village in Uttar Pradesh has to have a VEC, and these villages were no exception. Yet a startling 92% of parents of children in the government school responded that they did not know of any such committee. Of those that claimed to know that such a committee exists, only 2% could name actual members of the VEC.

Things began to fall into place when we interviewed the VEC members. In Uttar Pradesh, the VEC consists of an average of five members: The school headmaster, the pradhan (head of the village government), and three parents. Among the parent members, about one in four does not know that he is on the VEC, and of those who do know that they are a part of the VEC, roughly two-thirds are unaware of the Sarva Shiksha Aviyan, the big new program that is supposed to bring new resources to village schools.

In part, this reflects the fact that people in rural Uttar Pradesh seem not be particularly engaged with any of the institutions of local governance. Only 14.2% of respondents know of a household member ever having been to a Gram Sabha (village meeting), which are mandatorily held in every village. Most (over 90%) said that they do not know when or where the Gram Sabha is held.

But it is also true that of the households who do go to the meeting, education does not seem to be a priority. Of those who have attended a Gram Sabha, only 5.8% mentioned education when asked about which issues were covered in the previous meeting.e Parents are not particularly more interested in education than anyone else. When asked what they considered the most pressing issues in the village, education ranks fifth on the list of village problems, with only 13.9% of parents even mentioning it. This is despite the fact that they say that parents need to take responsibility for making the school run better.

One possibility is that people do not know how badly their children are being taught. In a recently completed randomized social experiment, we teamed up with an Indian NGO, Pratham, to expose parents in a randomly chosen set of 130 villages in Jaunpur district to exhortations about the importance of education, suggestions about where they can go to complain, and a general call to action. In a randomly chosen half of these villages, parents were also given information about their children's learning levels. After a year, Pratham went back and tested the children in these 130 villages; there was absolutely no difference in the performance of these children and the children in the 85 (randomly selected) control villages.10

On the other hand, it was not true that there was no way to affect the performance of the children within the given time span. In another group of 65 randomly chosen villages, in addition to the information and encouragement treatments described above, Pratham trained village volunteers to teach children basic reading skills. There was a significant improvement in the performance of the average child in these villages, even though only about 8% of the children in the village got coached by a volunteer. The children who ended up being trained show considerable progress in the span of a year.

If it is not ignorance of their children's performance, why is community action not more effective? One possibility is that parents are not aware of the benefits of education. Jensen recently carried out a randomized experiment in which he gave secondary school students in the Dominican Republic information about how much more money an educated person makes compared to someone who did not get an education.11 He found that teenagers substantially underestimate the returns to education, and that providing them with this information had a substantial effect on the children's likelihood to stay in school. In the treatment group, children were 5 percentage points (an increase of 10% since only 50% return) more likely to come back to school the year after the intervention. These results are confirmed by recent work from Nguyen, who provided parents in Madagascar information about returns from primary and secondary education.12 There, too, many parents underestimated the returns from education, and when they were informed about the measured returns reacted by making sure that their children went to school more regularly. At the end of the year their grades were also better.

However it is not clear why, if there is no demand for education, the volunteer-based teaching intervention worked so well. After all, the volunteers had to be willing to put their time into teaching and the children (and their parents) had to be willing to put in the effort to learn. In principle, in villages where there was no teaching option, these volunteers could have also put the same effort into doing other things that would help education in the village. Lobbying the VEC to hire a shiksha karmi or trying to persuade the current teacher to put more effort into teaching could have been other good uses of their time.

The parents, too, must have had to put some effort into making sure that their children went to the classes that the volunteers were running (children that age rarely go to extra classes on their own, so this must have been some work for them and/or the volunteers). Could they not have spent the equivalent time lobbying the VEC for an extra teacher? Why don't we see any effect of the other two interventions?

One possibility that seems consistent with everything we observe is that people in the village think, rightly or otherwise, that the VEC is essentially powerless because to have any effect it would have to move someone higher up to do something (say sanction the money for a shiksha karmi) and this would probably not just be a matter of the VEC writing a letter to someone higher up. Based on their past experience, it would probably take something like the whole village organizing a demonstration or a meeting before the bureaucracy outside (or the political representatives high enough up to have an influence on the bureaucracy) takes them seriously, and the challenge of organizing this might be an order of magnitude harder than just starting some evening classes. It would not be a matter of one show of force: things would have to be followed up and pressure maintained. Coordination of a large group over time is always hard, even if they live close by. Things are probably harder if the group is divided politically or along caste or class lines, as is often the case in Uttar Pradesh.

This is not to say that collective action by the poor can never work. A recent randomized evaluation of an intervention into village level healthcare delivery by the public health centers in Uganda, which was very similar to the study we (with Pratham) carried out in Jaunpur, found very strong positive effects on infant mortality, among other outcomes.13

There could be many reasons why their results were so different. One possibility that was suggested to us is that in Uganda, unlike in India, the health workers are very much a part of the local community and are therefore care more about how the community views them.

There are other possibilities, however. One is that people take their health more seriously than the education of their children and were therefore more motivated for change. A second possibility is that, in the case of health, one outcome of the meetings was that people knew which treatment each of them was entitled to get when they visited a facility, and at what price. Once they had this information they could just go by themselves and demand what they needed. In the case of education, there was nothing that anyone could do on their own. It would have to be collective action.

Both these possibilities, however, seem to suggest reasons for people to use the health facilities more. It is still not clear why either would make the providers show up to work more often, since the providers remained under the control of the health bureaucracy outside the village.f However, one other important difference between the VECs in India and the equivalent HUMCs (Health Unit Management Committees) in Uganda was that the HUMC managed a dispensary that served over 100,000 people. The VECs managed a school for 100 households or fewer. For this reason alone, the HUMC probably had much more power and influence. One could imagine that someone who ran the HUMC well might be able to use that as a basis for a political career in the district, whereas the VEC was at best one of the many venues in the overall governance structure of single village.

Figuring out the source of the difference is obviously key: if the main difference is in the relationship between public officials and the local community, then community control is unlikely to succeed in India without a radical realignment of the local political economy. If on the other hand, it is a matter of convincing parents that education is worth fighting for, then an entirely different approach may be called for before the problem can be thrown back to the community. Finally, if, after all, it comes down to the level of intervention (village versus sub-district, say) or some other detail of implementation, one could imagine community control working well, once the appropriate model has been figured out.

In the meanwhile, however, the government of India has gone ahead and announced the National Rural Health Mission, the latest, and by far the most expensive, of many attempts to reform the public health system. The NRHM, as it is popularly called, will spend almost Rs. 10,000 crores ($2.5 billion) just in the present financial year (2007–2008). The plan involves spending more money on the very sub-centers and primary health centers where there is widespread absenteeism and low utilization. What will make this plan work, when so many other programs have failed? The answer, to the extent the mission document offers one, is community control!


The point of all of this is to deny us the comfort of assuming that the poor will take care of it all: There are too many things that can go wrong in part because we do not fully understand what motivates the poor and in part because these policies are made without thinking hard enough about the institutional constraints that make it difficult for them to be effective. It is not to disparage the activism of the poor. Given how difficult their lives often can be—given that even the barest survival for them often means having to hustle—doing an odd job here or an odd sale there—what many poor people manage to achieve is a remarkable testament to their resilience and creativity. Nevertheless it seems unreasonable, and possibly even callous, to assume that they are raring to play a much greater role in pulling themselves out of poverty. After all, they already have to cope with the many challenges of living a life on very little money: what to do when a loved one falls ill and there is no money to pay the doctor; how to explain to a crying child that you cannot afford the toy he wants; how to explain to yourself that you cannot have a break from work, though your back is breaking and it is hard to keep your eyes open. Now they are being asked to start a new business, make sure that their difficult teenager does not play truant, or orchestrate a village-wide monitoring campaign.

It is not that we know that they will not or cannot respond: Indeed the one lesson from the recent literature on the choices of the poor is that we do not understand very well why they do not respond to some things but then react so readily to others. We already mentioned the fact that farmers in Kenya often do not use any fertilizer despite observing the very high returns that it has. Contrast that with the fact, also mentioned above, that parents and children react very strongly to simply being told the returns on education. What explains the difference between investing in education and investing in fertilizer?

Equally puzzling are the enormous observed effects of small changes in the price of investments in health. Going from price of zero to a very small positive price seems to lead to a very large drop in the demand for health goods. Kremer and Miguel found that take up of deworming drugs among school students in Kenya dropped by 80% when partial cost sharing was introduced.14 Cohen and Dupas found the same kind of result for the demand for bed nets.15 Along the same lines, going from a zero price to a small subsidy also seems to have very large effects. In a project we recently completed, we evaluated a program were an NGO offered small incentives for immunization in Rajasthan (a kilogram of lentils, worth about 60 cents, or half a day's wages) led to an increase in full immunization rates by 31 percentage points.16 Similarly, Thorton found that offering tiny incentives (10 to 20 cents) to those who agree to participate in Voluntary Counseling and Testing (VCT) for HIV/AIDS in Malawi increased the participation rate from 39% to about 65%.17

The puzzle in all these cases comes from the fact that the potential benefit of taking up the program seems an order of magnitude larger than the change in the price. Bed nets and immunization can save your child's life. VCT will not actually save your life, but it could potentially prolong it and it could definitely save the lives of your loved ones. Deworming will help your child grow stronger and get more out of his/her schooling. If people are still not willing to take them up, it ought to be because there are some commensurately large costs: The psychological cost of finding out that you have HIV, the fear that the evil-eye will find your child when you take him out to be immunized, or the fear that the immunization drug will make your children unable to give birth (as it is occasionally rumored) are all (plausible) examples of things that are invoked to explain the lack of take up. But if this is why they are being resisted, why would 10 cents get them to change their mind? Or even 60 cents? Yet we seem to be comfortable making policies that turn entirely on the poor having the necessary motivation.

Conflicts of Interest

The authors declare no conflicts of interest.


  • a

    The family would end up paying about $11 a year while the government spends $2.

  • b

    The low level of agriculture among the extremely poor in South Africa is easily explained. Under the apartheid regime, the black population, which contains almost all of the extremely poor people, was not allowed to own land outside the “homelands,” and most of the land in the homelands was not worth cultivating.

  • c

    For a survey of interest rates paid by small businesses across the world, see Banerjee.6

  • d

    On the other hand it could be (and has been) argued that she also gets something out of having her own little operation that she could not get otherwise—some cash of her own, an opportunity to go out occasionally, a chance to meet other people, a challenge.

  • e

    This number increases to 25 % when asked specifically about whether education was discussed.

  • f

    Unless we believe that providers enjoy their work when there is more demand for it (which is perhaps reasonable).