Zero / wiki

By: Timothy Ogden

There's no question that the idea of One Laptop per Child is appealing. It has whiz-bang technology, support from the glitterati of Silicon Valley and the World Economic Forum, and emotional resonance — giving poor children something we all know and value — on its side. And for quite a while, progress looked good. If the project hadn't hit the threshold it aspired to — a $100 laptop computer that would work in Third World countries without comprehensive electrical grids — it had produced a widely praised first model and received orders for several hundred thousand laptops from Uruguay, Colombia, Mexico and Peru.

In the last year, though, the news on OLPC has sharply soured. Technical difficulties slowed the development of the second generation of the laptop. The second year of the “Give One, Get One” promotion — which offered the people a chance to pay $399 for two laptops, one that they would keep, with the other sent to a developing nation — dramatically underperformed. Then in early January, the nonprofit OLPC announced that 50 percent of staff were being laid off and a major restructuring was under way. To even its most ardent supporters, the project seems nearly dead in the water.

And that may be great news for children in the developing world.

In 2005, Nicholas Negroponte, the founder of MIT's Media Lab, announced the One Laptop per Child program at the World Economic Forum. The concept was simple and enormously appealing: Innovate a $100 laptop and distribute it to children in the developing world. Negroponte and other techno-luminati formed a nonprofit and lobbied national governments and international agencies like The World Bank to commit tens of millions of dollars to buying laptops for children.

For technology optimists like Negroponte, the payoff was obvious. If only children in poor countries had access to computers, they could take control of their education and expand their horizons, benefiting everyone. With access to a virtually limitless store of information, connectivity to the world and educational software, OLPC advocates argued, the Third World would suddenly abound in educational opportunities. There's no question that improving education in the developing world is necessary. While school attendance has been trending dramatically upward over the last decade, with attendance rates approaching 90 percent, there are still roughly 100 million children under 12 who are not in school, according to the United Nations Children's Fund, or UNICEF.

But those numbers only measure enrollment, not regularity of attendance or learning. The data on those scores are grim: Studies by researchers at MIT's Jameel Poverty Action Lab, or J-PAL, a highly respected center for evaluation of poverty-related programs, have found average student absenteeism rates in several countries to exceed 20 percent, while teacher attendance is only marginally better. Even when children and teachers are in school, they don't seem to be learning much: A survey of 11-year-olds in Bangladesh found that 58 percent couldn't identify five of the six letters they were quizzed on; the average score of Peruvian students on an international science exam is equivalent to that of the fifth percentile in the United States. “One problem that seems to be common across a lot of the [developing world schools] is that the curricula are just not adapted to the students. Maybe it is not rocket science, but children will be able to learn and teachers will be able to teach if the goal is achievable,” says Esther Duflo, a co-founder of J-PAL.

Such dismal figures are a powerful argument for taking dramatic action to improve children's education — and what better way to do that than putting a free laptop in the hands of children? Actually, though, if the goal is improving education for children in the developing world, there are plenty of better, and cheaper, alternatives.

Despite the instinctive appeal of distributing laptops to schoolchildren, there is precious little evidence that making computers available to children improves educational outcomes. The circumstantial evidence that exists certainly doesn't buttress the one-laptop-per-child approach.

The OLPC concept has been pioneered in a number of school districts in the United States over the last decade — and many of the schools, public and private, that adopted such programs have abandoned them, for instance in Broward County, Fla., Syracuse, N.Y., Richmond, Va., and Costa Mesa, Calif., citing high costs and no evidence of benefit. Under the auspices of the No Child Left Behind Act, the U.S. Department of Education and Mathematica Policy Research recently completed a rigorous, two-year test of reading and math software (using programs that had at least some nonexperimental evidence that they “worked”) in dozens of school districts nationwide. With one minor exception, the studies found that children using the software scored no better than peers who did not have access to the software. “I wouldn't say that technology doesn't work at all, but in our studies at least, the technology didn't work any better than a normal classroom teacher,” says Audrey Pendleton, project officer for the study at the Department of Education.

Two other recent studies conducted in the developing world are even more telling. Economists Ofer Malamud and Cristian Pop-Eleches studied a program in Romania that distributed discount vouchers for the purchase of home computers to low-income families. When they compared the families that used the vouchers to acquire computers with families that were just above the income cut-off to receive the vouchers, they found that computers had a negative effect on students' grades and educational goals. Leigh Linden, an economist at Columbia University, and Felipe Barrera-Osorio of The World Bank studied a program in Colombia that increased the number of computers in schools and provided curriculum support and training for teachers — and found no impact on student outcomes. “In this case, despite the curriculum support, it was clear that the teachers simply weren't using the computers,” Linden says.

Linden also led one of the few experimental studies to show a positive impact from the use of computers — a project in India that provided computers and education software to schools and randomly assigned some schools to use the software during school hours and others to encourage computer use after hours. This study found that using computers during school hours —essentially substituting computers for teachers — actually hurt learning, while using them after hours as a supplement to traditional classroom teaching had dramatic positive effects on the weakest students. Even this outcome doesn't really support the OLPC mission, though; the software evaluated is very much in the “drill and practice” model that Negroponte has explicitly derided.

It must be said that none of this evidence speaks directly to the OLPC project. One independent academic study of an OLPC program, conducted in Ethiopia by David Hollow of the University of London, found that teachers limited access to the computers and had not been adequately trained — though Linden's and Barrera-Osorio's research suggests teacher training and curriculum support are not silver bullets. Negroponte seems to be basing his claims on surveys of students who say that the laptops help them learn — a claim that would have many parents rolling their eyes.

The Inter-American Development Bank is funding an evaluation of an OLPC project in Haiti, but that study just kicked off. Repeated calls and e-mails to OLPC and Negroponte seeking comment on OLPC did not receive a response. OLPC's apparent lack of impact studies is ironic, at least; the nonprofit is headquartered literally around the corner from J-PAL, a leader in studying development programs in the Third World.

There are a number of simple, cheap programs that have been proven successful at getting children in developing countries into school and helping them learn more while they are there.

The simplest and least costly of these programs is deworming. Nearly 2 billion people around the world are affected by parasitic worm infections, with children disproportionately affected. While each variety of parasitic worm affects a person differently, they all take a substantial toll on growth, energy and attention, with entirely predictable impacts on school attendance and learning. Harvard economist Michael Kremer has studied the impact of mass deworming in Kenya and India. Delivering deworming medication costs 50 cents per child per year in Kenya but yielded a 25 percent increase in school attendance; a similar program in India cost $4 per student per year and yielded a 20 percent attendance gain. “This is a simple, cost-effective and yet tragically not-done program. It's a scandal that [deworming] hasn't been addressed,” Kremer says. There are spillover effects as well. “The most surprising thing about the study in Kenya was the widespread impact,” Kremer says. The program drove down infection rates for several kilometers around the schools, he says, and there were significant improvements in attendance for untreated students, in the treatment schools as well as in nearby schools not in the program.

The impact of deworming would potentially be much greater than just better school attendance, says Kari Stoever, managing director of Neglected Tropical Disease Control at the Sabin Vaccine Institute. One of philanthropy's historic successes is the Rockefeller Foundation-sponsored elimination of hookworm in the southern United States in the first half of the 20th century. “If deworming follows the pattern we saw in the U.S.,” she says, “we could see a 30 percent increase in lifetime earnings.”

Beyond deworming, there are technology-based approaches to improving student learning in the developing world that show more promise than one laptop per child. Duflo, the J-PAL co-founder, tested a program run by a nonprofit in India to improve teacher attendance at remedial schools. In the program, teachers were required to take date-stamped digital pictures of themselves with students each day in order to receive their pay. Unsurprisingly, this requirement had a major impact on teacher attendance, which rose by nearly 50 percent. And the teachers apparently didn't game the system by showing up and not teaching. At the end of the year, students in the schools where teachers photographed themselves showed a significant and measurable improvement in test scores and were 43 percent more likely to be accepted into a regular school. Duflo estimated the cost of each additional day of teacher attendance at just $2.20.

Remedial education, in general, appears to be an effective method of improving student achievement. The simplest approach —separating students based on ability rather than age, or tracking — has significant benefits for both relatively high-achieving and low-achieving students. A study in Kenya evaluated splitting classes into two sections, one for high scorers and one for low scorers. (The program was compared to splitting classes into two sections randomly, so teacher-student ratios were not a factor.) Students in both the low-performer and high-performer sections did better than students who were randomly assigned to different sections.

There were two additional findings of the study. First, the benefit to students who were just below the cutoff for joining the high performers was as great as it was for students who made it into the high-performers section. Second, it didn't matter which class had the more experienced teacher. The implication is that creating classroom situations where it's easier for a teacher to customize the curriculum to a level appropriate for all students is a key factor in improvement.

Although this program was relatively expensive when compared to deworming or tracking teacher attendance — it did, after all, require hiring an additional teacher — the added teacher's salary was less than a quarter of the cost of supplying laptops to the children. And there is a cheaper way of achieving similar results: providing tutoring to students who are behind in literacy and numeracy. In another experiment in rural India, tutoring was delivered to the lowest performing 15 to 20 students in a class by women in the community, who had relatively low levels of education themselves and were paid salaries less than one-tenth of the average teacher's. Still, there was marked improvement among students receiving the tutoring.

James Tooley, a professor at Newcastle University (U.K.), has explored another way that parents in developing countries are attempting to improve their children's educations: They send their children to private school. As ludicrous as this might sound at first, Tooley has documented the booming phenomenon of private schools for the poor in slums around the world.

Tooley estimates that more than 50 percent of urban slum-dwelling children, and nearly 25 percent of children in rural India, where per capita income is less than $2 a day, attend private school, even though public schools are nominally free. Poor families can afford to send their children to private schools because they are quite inexpensive. In India, fees are about $2 per month. Worldwide, Tooley estimates, average fees are $3 per month. And, as Tooley notes, “Those fees pay for everything: books, uniforms, lunch. Other programs, like computers, require someone to pay for all the costs of teachers, buildings and materials on top of the cost of the computer.”

Although there are only a few studies comparing these private schools for the poor to public schools in terms of quality of education, as Tooley notes, “the fact that parents earning less than $2 a day are setting aside funds to pay the fees speaks for itself.” Cutting the cost of such schools in half via subsidies or scholarships, enabling even more parents to be able to afford to send their children (or to send additional children), would cost just $18 per child per year, on average.

To be sure, deworming programs, tutors and support for private schools are nowhere near as sexy as OLPC, its successful, tech-savvy supporters and its colorful, friendly looking XO laptop. But as Abhijit Banerjee, a co-founder of J-PAL, says, “There is this presumption that big changes need big levers. There is no evidence that big changes were ever wrought by big levers.”