Poor Economics
Page 10
Since parents are able to respond to changes in the need for an educated labor force, the best education policy, for the demand wallahs, is no education policy. Make it attractive to invest in business requiring educated labor and there will be a need for an educated labor force, and therefore a pressure to supply it. And then, the argument continues, since parents will start to really care about education, they will also put pressure on teachers to deliver what they need. If public schools cannot provide quality education, a private-school market will emerge. Competition in this market, they argue, will ensure that parents get the quality of schooling that they need for their children.
At the core of the demand wallahs’ view is the idea that education is just another form of investment: People invest in education, as they invest in anything else, to make more money—in the form of increased earnings in the future. The obvious problem with thinking of education as an investment is that parents do the investing and children get the benefits, sometimes much later. And though many children do, in effect, “repay” parents for the investment by taking care of them in old age, many others do so only reluctantly, or they simply “default,” abandoning their parents along the way. Even when the children turn out to be dutiful, it is not always clear that the extra bit of money that they earn because they spent that extra year in school translates into that much more for the parents—we have certainly come across parents who rue the day when their children became rich enough to move out to their own house, leaving them to their lonely elderly lives. T. Paul Schultz, a Yale economist, talks about his father, the famous economist and Nobel Laureate Theodore Schultz, whose parents were against educating him, because they wanted him to stay back on the farm.
It is true that many parents do take pride and pleasure in the fact that their children are doing well (and in sharing the good news with their neighbors). In this sense they may feel more than adequately repaid even when they don’t get a penny from their children. So from the point of view of the parent, education is partly investment but also partly a “gift” that they offer their children. But there is also the flip side: Most parents are in a position of power relative to their children—they decide who goes to school, who stays home or goes out to work, and how their earnings are spent. Parents who are cynical about how much they would get out of a son’s earnings once he is old enough to push back, and who do not value education for its own sake, may prefer to take him out of school and send him to work when he is ten. In other words, although the economic return to education (as measured by the extra earnings of an educated child) clearly matters, lots of other things probably matter as well, things like our hopes about the future, our expectations about our children, even how generous we are feeling toward them.
“Exactly,” says the supply wallah. “This is why some parents need a push. A civilized society cannot allow a child’s right to a normal childhood and a decent education to be held hostage to a parent’s whims or greed.” Building schools and hiring teachers is a necessary first step to lower the cost of sending a child to school, but it may not be enough. This rationale explains why most rich countries simply give parents no choice: Children have to be sent to school until a certain age, unless parents can prove they are educating them at home. But this clearly does not work where state capacity is more limited and compulsory education cannot be enforced. In such cases, the government must make it financially worthwhile for parents to send their children to school. This is the idea behind the new tool of choice in education policy: the conditional cash transfer.
The Curious History of Conditional Cash Transfers
Santiago Levy, a former professor of economics at Boston University, was deputy minister in the Mexican Ministry of Finance from 1994 to 2000, tasked with reforming the intricate welfare system, which was made of several distinct programs. He believed that by linking the receipt of welfare payments to investment in human capital (health and education), he could ensure that the money spent today could contribute to eradicating poverty, not only in the short term but in the long term as well, by fostering a healthy and well-educated generation. This inspired the design of PROGRESA, a transfer program “with strings attached.” PROGRESA was the first conditional cash transfer (CCT) program: It offered money to poor families, but only if their children regularly attended school and the family sought preventive health care. They got more money if the children were in secondary school than in primary school and if it was a girl who went to school rather than a boy. To make it politically acceptable, the payments were presented as “compensation” to the family for the wages lost when their child went to school instead of working. But in reality, the goal was to nudge the family, by making it costly for the family to fail to send their children to school, regardless of what the family thought of education.
Santiago Levy had another goal—to make sure that the program survived the change of government every few years, since each new president usually canceled all his predecessors’ programs before launching his own. Levy calculated that if the program was demonstrably a great success, the new government would not find it easy to get rid of it. So he set up a pilot project, which was offered only in a randomly chosen group of villages, making it possible to rigorously compare outcomes in chosen and non-chosen villages. The pilot demonstrated beyond reasonable doubt that such a program does substantially increase school enrollment, particularly at the secondary level. Secondary school enrollment increased from 67 percent to about 75 percent for girls, and from 73 percent to about 77 percent for boys.11
This was also one of the first demonstrations of the persuasive power of a successful randomized experiment. When the government duly changed, the program survived, albeit renamed OPORTUNIDADES. But Levy probably did not anticipate that he had given birth to two new traditions. First, CCTs spread like wildfire all over the rest of Latin America, and subsequently to the rest of the world. Mayor Michael Bloomberg even gave them a try in New York City. And second, when other countries launch their own CCTs, they now usually also carry out a set of randomized trials to evaluate them. In some of these experiments, features of the program are varied, to try to understand how to design it better.
Paradoxically, it was one of these replications, in Malawi, that led us to rethink the success of PROGRESA. The conditionality in PROGRESA is based on the principle that increased income is not enough and that parents need to be given an incentive. Researchers and practitioners started to ask whether an unconditional program could have the same effect as a conditional transfer. A World Bank study found, provocatively, that conditionality does not seem to matter at all: The researchers offered the families of school-age girls a transfer ranging between $5 and $20 USD PPP per month. In one group, the transfer was conditional on enrollment. In another, it wasn’t. A third group (the control group) did not receive a transfer. The effects were large (after a year, dropout was 11 percent in the control group, and only 6 percent among those who benefited from the transfer), but they were the same for those who received the conditional transfer and for those who got the unconditional one, suggesting that parents did not need to be forced to send their children to school, they needed to be helped financially.12 Subsequently, another study that compared conditional and unconditional transfers in Morocco found similar results.13
Several factors probably explain why the financial transfer made a difference in Malawi: Perhaps parents could not pay for school fees, or could not give up the money their children earned. Of course, borrowing to finance the schooling of their ten-year-old based on what she will make at twenty is entirely a pipe dream. The income transfer, by moving parents out of extreme poverty, may also have given the mental space to take a longer view of life: Schooling is something where the costs are paid now (you have to nag—or drag—your children into school now) and it only pays off when they are older.
For all these reasons, income per se matters for education decisions: Jamal will get less education than John because his parents are poorer, e
ven if the income gains from education are the same for both. Indeed, in our eighteen-country data set we find that the share of spending on education increases as we move up from those who live on under 99 cents a day to those in the $6–$10 category. Given that the number of children born to each family goes down sharply with income, this means that education spending per child grows much faster than total consumption. This is the opposite of what we would expect in a world where education is an investment like any other, unless we are willing to believe that the poor are just incapable of getting educated.
This is important, because if parental income plays such a vital role in determining educational investment, rich children will get more education even if they are not particularly talented, and talented poor children may be deprived of an education. So leaving it purely to the market will not allow every child, wherever she comes from, to be educated according to her ability. Unless we can fully erase differences in income, public supply-side intervention that makes education cheaper would be necessary to get close to the socially efficient outcome: making sure that every child gets a chance.
Does Top-Down Education Policy Work?
The question, however, is whether this kind of public intervention, even if it is desirable in principle, is actually feasible. If parents do not care about education, isn’t there a risk that such a top-down education drive would just lead to a waste of resources? In The Elusive Quest for Growth, Easterly argues, for example, that the investment in education in African countries has not helped these countries to grow.
Once again, the best way to answer this question is to study what happened when specific countries tried it. The good news is that despite the poor quality of education, schools are still useful. In Indonesia, after the first oil boom in 1973, the country’s then dictator, General Suharto, decided to go on a school-building spree.14 It was the classic top-down supply-driven program: Schools were built based on a prespecified rule that gave strict precedence to areas where the number of unschooled children was the highest. If the lack of schools in this area reflected lack of interest in education, this program should have been a miserable failure.
In fact, the INPRES (Instruksi Presiden, or Presidential Instruction) program was a great success: To evaluate it, Esther compared the wages of adults who, as children, were young enough to have benefited from the newly constructed schools to what the immediately older generation (people who were just old enough to have missed their chance to go to these schools) was earning. She found that relative to the older generation, the wages of the younger one were significantly higher in areas where more schools were constructed. Putting together the effect on education and on wages, she concluded that every extra year of primary school due to the new school raised wages by about 8 percent. This estimate of the returns to education is very similar to what is commonly found in the United States.15
Another classic top-down program is compulsory schooling. In 1968, Taiwan instituted a law that made it mandatory for all children to complete nine years of schooling (the previous law only required six years of school attendance). This law had a significant positive effect on the schooling of both boys and girls, as well as on their employment prospects, especially for girls.16 The benefits of education are not only monetary: The Taiwan program had a large effect on child mortality.17 In Malawi, girls who did not drop out because of the cash transfer were also less likely to become pregnant. The same results were found in Kenya.18 There is now a significant body of rigorous evidence testifying to the far-reaching effects of education.
Moreover, this research also concludes that every little bit of education helps. People who are comfortable with reading are more likely to read newspapers and bulletin boards and to find out when there is a government program available for them. People who go on to secondary education are more likely to get a formal-sector job, but even those who don’t are able to run their businesses better.
It seems, then, that once again the polarized debate between philosophically opposed strategies largely misses the point. Supply and demand strategies have no reason to be mutually exclusive. Supply by itself does some good, but demand is important, too. There are indeed people who somehow find ways to get educated without any top-down help when the right jobs come to town, but for many others, the impetus from schools being built in their area can be critical.
None of this means that top-down strategies deliver as much as they could, or should. After all, as we saw, the quality of education delivered in public schools can be dismal. The fact that students are getting something out of them does not mean they could not work significantly better. Could it be that demand-based approaches would work better? Private schooling is the canonical demand-driven strategy—the parents must spend their own hard-earned money to put their children into one, even though free public schools are available. Have private schools cracked the problem of the quality of education?
Private Schools
There is a surprising amount of agreement that private schools should play an important role in the process of filling the gaps in the education system. India’s Right to Education Act, which was recently passed with strong support across the political spectrum (including the left, which, the world over, has traditionally opposed the role of the market), is a version of what is called voucher privatization—the government gives citizens “vouchers” to pay private-school fees.
Even before the education experts gave it the heads-up, many ambitious low-income parents around the world had decided that they had to get their children into private schools, even if they would have to scrimp for it. This has caused the surprising phenomenon of cut-price private schools all over South Asia and Latin America. The monthly fees in these schools can be as low as $1.50. The schools tend to be quite modest, often just a couple of rooms in someone’s house, and the teachers are often local people who couldn’t find another job and decided to start a school. One study19 found that an excellent predictor of the supply of private schools in a Pakistani village is whether a secondary girl’s school had been set up in the area a generation earlier. Educated girls, looking for an opportunity to make some money without having to leave the village, were increasingly entering the education business as teachers.
Despite their sometimes dubious credentials, private schools often work better than public schools. The World Absenteeism Survey found that in India, private schools were more likely to be found in villages where the public schools were particularly bad. Furthermore, on average, the private-school teachers were 8 percentage points more likely to be in school on a given day than public-school teachers in the same village. Children who go to private school also perform better. In India in 2008, according to ASER, 47 percent of government-school students in fifth grade could not read at the second-grade level, compared to 32 percent of private-school students. In the Learning and Educational Achievement in Pakistan Schools (LEAPS) survey, by third grade, children in private schools were 1.5 years ahead in English and 2.5 years in math relative to children in public schools. It is true that families who decide to send their children to private schools may be different. But this could not be entirely explained by the private schools’ attracting kids from richer families: The gap in performance between private- and public-school students was close to ten times the average gap between the children from the highest and lowest socioeconomic categories. And though it is not quite so large, there is still a sizable gap between children enrolled in public and private school even within the same family20 (this may still be an overestimate of the true benefit if parents send their most talented child to private school or also help that child in other ways).21
So children in private school learn more than children in public schools. This does not mean, however, that private schools are as efficient as they could be. We see that they are not when we compare the effect of being in private school to the effect of simple interventions.
Pratham Versus Private Schools
Pratham, the remarkable
educational NGO that runs ASER, not only exposes the deficiencies of the educational system but also tries to fix them.We have been working with them for the last ten years, evaluating almost every new edition of their program for teaching children arithmetic and reading. Our association started in the year 2000 in western India, in the cities of Mumbai and Vadodara, where Pratham was running what they called the Balsakhi (meaning “children’s friend”) program. The program took the twenty children in each classroom who most needed help and sent them to work with the balsakhi, a young woman from the community, on their specific areas of weakness. Despite an earthquake and communal riots, the program generated very large gains in test scores for these children—in Vadodara, about twice the magnitude of the average gains from private schooling that have been found in India.22 Yet these balsakhis were much less educated than the average private- (or public-) school teacher—many of them had barely ten years of schooling, plus a week’s training by Pratham.23