May 25, 2008
Paying Students to Improve Performance
Paying the Poor to Improve their School Performance-Becker
In the mid-1990's Mexico started an anti-poverty program, called Progressa, that revolutionized the way low income countries try to reduce child labor and the school dropout rate. This new approach typically pays poor parents to keep their children in school and to take them for regular health check-ups. The reasoning motivating this approach is that while poor parents may love their children as much as rich parents do, the need for greater income induces poor parents to take their children out of school, so that they can go to work and add to the family's income. Offering monthly cash payments if the children remain in school and performs well instead of going to work helps compensate these families for the loss in their children's earnings.
The results of Progressa are publicly available so that they can be objectively analyzed, and compared with a control group of similar families who were not invited into the program. Studies by economists in the United States and elsewhere clearly show that Progressa has succeeded in inducing the mainly rural parents in the program to keep their children in school longer than they would have. The budgetary cost of that achievement has been sizable; although the cost would have been much less if Progressa had offered the subsidies mainly to parents with children at the ages-usually when children were in the 6-8 grades-when poor rural Mexican parents typically took their children out of school.
For many years I have enthusiastic about using incentives to encourage greater school attendance by children from poorer families. I first wrote about Progressa, and similar programs in Brazil and elsewhere, in a Business Week article entitled " 'Bribe' Third World Parents to Keep Their Kids in School", Nov. 22, 1999. Such programs seem to be the most effective way to induce poor families in developing countries to reduce child labor by keeping their children in school much longer. Prior to the introduction of these programs, poor parents simply ignored laws against child labor, and those requiring children to stay in school until they either reached a certain age or attained a minimum grade level.
Until recently, programs similar to Progressa had spread to many countries, but all of them were low to moderate income countries. However, within the past year, New York City and a few other American cities have started experimental programs to adapt the incentive concepts behind Progress to the situation of poor families in the United States. The New York experiment is fully funded by private foundations and individuals, including Mayor Bloomberg- I will concentrate my discussion on this city's program. Since the children involved are older than those in Progressa, they rather than their parents are paid for good attendance and for raising their test scores. Their parents are also paid to improve the choices they are more directly responsible for, such as working longer hours, and taking their children more frequently for health checkups.
It is obviously too early to evaluate the benefits and costs of the New York experiment, but I am confident that it will raise the performance of the students participating. The reason is simply that boys and girls as well as adults respond to incentives, as every parent realizes time after time. Rewarding these poor students for better performance is similar to the tuition scholarships and stipends that colleges award to students with good grades. To earn the "pay" offered, students involved will skip school less often. They will also pay closer attention to their teachers during classes and do more homework, so that they can do better on the standardized tests that are being used to judge their performance. Whether this particular experiment has the most effective link between rewards and increase in performance on these test will only be clear with further experimentation, but a pioneering program of this kind has to start somewhere.
The New York program is not without many critics, which perhaps explains why it has been funded privately rather than by public resources from tax revenue. Some critics believe it is wrong to pay children and parents to do what they should want to do anyway in their own self-interest since doing better in school will be valuable in getting good jobs when they are young adults and enter the labor force. Most high school students do in fact recognize the importance of doing finishing high school and doing reasonably well, but the New York program is directed precisely to those who are performing badly, perhaps because they heavily discount the future, or are in dysfunctional families. Other critics content that change has to start with these dysfunctional families that are responsible for their children skipping school and their poor school performance. The family is surely important to the achievements of children, but children from these families and their mothers can still do much better now if they are given strong financial incentives to do so.
Another set of criticisms does not deny the importance of incentives to poor families and their children in rich countries like the United States. However, it argues that the existence of incentive programs, such as in the New York experiment, will encourage some children who have been doing well to lower their school performance, so that they can qualify for the program. All incentive-based programs with income or other cutoff points induce some families to change their performance to better qualify for the programs. One has to be mindful of this effect in designing a program for poor parents and their children to make sure that that it is not so generous as to attract many more families to qualify by worsening their performance. I believe that this is a greater problem with the payment system to parents than that to children, but further experience will inform us about that.
Yet such possible risks are no reason to delay incentive-based programs until families become less dysfunctional, or their children become more aware of the future benefits of better school performance. Too many children, especially of African-American and Hispanic backgrounds, are doing so badly in school, and they are dropping out of school in such large numbers, that we should be willing to try an approach that has been successful in developing countries. I commend New York for being willing to take initial steps in the direction of offering financial incentives to badly performing students that encourage them to work harder to get more out of their education.
Paying Children to Go to School--Posner's Comment
The Mexican and New York City programs are well described in Becker's post and in a recent article in the Financial Times by Christopher Grimes, "Do the Right Thing," May 24, 2008, www.ft.com/cms/s/0/a2f1b24a-292a-11dd-96ce-000077b07658.html?nclick_check=1. I cannot comment on the Mexican program; nor do I oppose social experiments financed by private money, as in New York. But I am skeptical about the New York program, and if I were a New Yorker I would be reluctant to support public financing of it.
Before Milton Friedman proposed to replace welfare programs with a negative income tax--that is, a cash grant with few if any strings attached--welfare programs were in part devices by which the government endeavored to buy good behavior from the poor. Hence food stamps, but not food stamps that could be used to buy liquor. Or money earmarked for health or education.
Friedman's criticism of such programs was that people have a better sense of their needs than government bureaucrats, so that if the government simply gave poor people money they would allocate it more efficiently than the welfare bureaucracy would do. This philosophy was eventually adopted by the federal government in the form of the earned income tax credit. The danger in giving the poor money (or anything else for that matter) is that it will reduce their incentive to work; this problem was addressed by the replacement of welfare by workfare at the state and later the federal level.
Friedman's analysis requires qualification, however, when the issue is the welfare of children. The reason is that not all parents balance their own welfare with that of their children in an impartial manner. That is why we have laws forbidding child neglect and abuse. It is also why we have compulsory-schooling laws and forbid child labor. These are paternalistic laws in a quite literal sense, but are justified to the extent that there is legitimate concern that not all parents are faithful agents of their children. Nevertheless, as a general rule parents both know better than welfare officials what is good for their children and love their children more than the officials, however well meaning, do, so any proposal to expand the role of government in controlling children should be viewed with caution.
Public school is both free and compulsory, and schooling adds considerably to a child's lifetime income prospects, so we must ask why some parents do not compel their children to attend school regularly. One reason might be that some of them do not value their children's welfare. Another that they cannot control their children. And a third that they do not think their children benefit significantly from regular attendance. I would guess that the second and third reasons are more common than the first.
Paying children to go to school would probably have at least some effect in countering all three cases. However, the benefits would be limited to children who, but for the payment, would attend school less frequently. I do not know how those children could be identified in advance, which means that the program would confer windfalls on some, perhaps many, children. (It would be odd to disqualify children on the basis of their good attendance!) In addition, there would be substantial costs, both direct and indirect, to the program. The direct costs would consist of the costs of distributing the money to the kids, making sure that it is not appropriated by the parents, and monitoring the children's school attendance. (So: more bureaucracy.) The indirect costs would include perverse incentive effects--some parents would spend less on their children to offset the payments that the children would be receiving for staying in school. Also, giving children their own source of income would reduce parental control and by doing so weaken already weak families. And some children contribute more to family welfare by occasional truancy than by consistent school attendance--for example, they may be older children helping to take care of younger siblings in households in which the parents work full time, or in which there is only one parent. Also, how does one end such a program? If the payments are suddenly withdrawn, will the kids feel aggrieved and resume truancy with a vengeance?
The largest indirect cost, I would guess, would consist in relaxed pressure to improve the public schools or to allow them to be bypassed by means of voucher systems. High rates of truancy may be due in significant part to low quality of schools. Paying children to attend school will reduce truancy rates some but without improving school quality, and perhaps without improving the education of the children receiving the payments. (School quality may actually decrease, with more crowded classrooms--crowded by kids who don't really want to be there.) Suppose that a school is in session 200 days a year, a student is truant 10 of those days, and if paid to attend would be truant only 5 days. Then the effect of the payment would be to increase the number of days the child was in school by only 2.5 percent. If it's a bad school, there might be zero benefit from this modest increase in attendance.
Granted, there are many children in New York who are truant for much longer periods. An article by Harold O. Levy and Kimberly Henry, "Mistaking Attendance," New York Times, Sept. 2, 2007, www.nytimes.com/2007/09/02/opinion/02levy-1.html?_r=2&ex=1189396800&en=1d2692cb89c474d7&ei=5070&emc=eta1&oref=slogin&oref=slogin, estimates that 30 percent of New York public school students miss a month of school every year. But they may be children who for mental or psychological reasons, or extreme family circumstances, cannot benefit significantly from additional schooling. The beneficial effects of paying children to go to school are likely to be concentrated on the kids who are casual rather than extreme truants, and those benefits, as suggested by my numerical example, may be slight.
Another component of the program is paying children for performing well on standardized exams. Such measures reward work more directly than paying for attendance, and also avoid the bad signal that is emitted by bribing people to do what the law requires them to do (i.e., attend school until 16 or 18, depending on the state), but they may largely reward intelligence rather than study. Working hard in school is no guaranty of getting good grades. Scholarships for promising students and awards for high performance have good effects, but the paid students are unlikely to qualify in competition with students who do not have to be paid to attend school.
Paying children to attend school is a band-aid approach at best. Far better would be a voucher system that would create competition among the public schools to serve children better.