No Teacher Left Behind

One of the great successes of global development over the past 60 years has been getting kids in school. In 1950, less than half the world’s primary-school-age children were enrolled. Today that figure is trending rapidly toward 100 percent. More schooling is associated with all sorts of good things — not least higher earnings as an adult, lower fertility among girls, and lower mortality among their kids. And the world’s governments are responsible for educating the considerable majority of those in school — perhaps one out of 10 students at the primary level is in private school. Kudos to the ministries, aid agencies, educators, and parents that made all of this possible.

The bad news is that many of the billion-plus kids in school today aren’t learning very much. In fact, in public schools in the developing world, many are learning close to nothing — many kids leave school unable to read or do simple sums. If we’re going to convert more kids in class into more knowledge in heads, we’re going to have to turn our focus from the students to the teachers — ensuring they have the incentives to perform.

In terms of access to schooling, countries across Africa, Asia, and Latin America are massively ahead of where developed countries were only a little more than a generation ago. According to data from Robert Barro of Harvard University and Jong-Wha Lee of Korea University, Ghana’s population (ages 15 and over) had been in school for an average of nearly eight years in 2010. Zambia’s averaged nearly seven years, Bangladesh six, and Haiti a little above five. Now consider that France, Germany, and Spain — as recently as 1970 — were all below five years.

At the same time, developing countries have done a lot less well in ensuring kids actually learn something while sitting in class. Despite close to universal enrollment in primary schools in Bangladesh, over 50 percent of 11-year-olds are unable to write basic letters or numerals. International tests suggest the average math ability of Brazilian 15-year-olds is equal to that of the bottom 2 percent of Danish students. In South Africa’s Western Cape province, only two out of 1,000 sixth graders in predominantly black schools passed a mathematics test at grade level in 2005.

The problem isn’t that kids are incapable of learning. It is true that some children do arrive at school tired from a long walk, malnourished, or weakened by illness — and that can have an impact on test scores. But put even the most disadvantaged children in the right environment, and they learn lots very fast. Take one widely cited example: Sugata Mitra put a computer in the wall of a slum in New Delhi, and within days kids were surfing the Internet and playing games on Disney’s website — all without any formal instruction. In other words, slum kids in India can learn enough computer literacy to waste time online as fast as their Western counterparts.

If it isn’t that the kids can’t learn, is the problem that the teachers can’t teach? It is true that a study in southern Africa found many primary school mathematics teachers who actually scored lower than their students on math tests. But, as a rule, most teachers still know enough to help their charges learn. In fact, according to Abhijit Banerjee and Esther Duflo at MIT’s Poverty Action Lab, the same teachers in an Indian experiment who proved atrocious at providing an education during the semester in public schools turned out to be very effective at teaching literacy in summer camps. Put the teachers in the right environment and kids learn stuff.

This points to another possibility: Teachers in government schools just have too little incentive to teach. In fact, the problem may start with the problem that they have too little incentive to bother turning up to class at all. On an average day, some 16 percent of teachers are absent in Bangladesh and 27 percent in Uganda, for example. In schools in India’s Andhra Pradesh state, the chance that a teacher was in class and actively engaged in teaching during the school day was only 28 percent.

Even if they do turn up and bother to teach, public school educators are often encouraged to deliver a curriculum that is pretty much destined to leave all but the most well-prepared students behind. And, of course, in places like Tanzania and Bangladesh, they face large classes and atrociously limited supplies. Perhaps worst of all, many parents of their students may be incapable of teaching the basics or helping with homework at home — because they did not attend school themselves. But there’s plentiful evidence that if you get the incentives right, teachers in public schools can provide a better quality of education.

Some of that evidence comes from nonstate schools. All over the developing world, there are private schools providing an education for as little as $1.50 a month, suggest Banerjee and Duflo. They often operate out of a teacher’s house and are frequently run by educated girls who don’t want to leave their home village and can find few other opportunities. For all their limited stock of books and supplies, and for all that the teachers are frequently unqualified secondary school graduates, such schools often do a better job than the public system. In Pakistan, kids in private school are two-and-a-half years ahead of their public school contemporaries in test scores as early as third grade. And it isn’t just that richer kids go to private schools — the impact of being in a private school on test scores was nearly 10 times the impact of being from a rich family compared with a poor one. Similarly, in India, according to nationwide surveys, 47 percent of government-school students in the fifth grade could not read a second-grade textbook — compared with only 32 percent of fifth-grade private -school students.

This suggests that efforts to ensure teachers turn up and teach could generate returns in the developing world. Additionally, curricula flexible enough to allow them to teach to the level of their students could make a big difference. Imagine a system that actually rewarded educators if their kids showed advances in learning basic skills over the year. Compare that with the present system in much of the world, which pays teachers more purely on the basis of seniority and encourages them to finish the national curriculum lessons — however inappropriate that is for the skill level their students start the year with.

An additional approach is to help kids learn outside the classroom — akin to the Indian hole-in-the-wall computer experiment, but on a much larger scale. One example is putting subtitles on TV programs in the same language as is being spoken on the screen. The approach has been tried in India — a country with a TV audience of about 600 million. In 2002, the producers of Rangoli — a very popular program that plays songs from Bollywood musicals — started subtitling the hit videos. Survey evidence suggests young TV viewers who watched Rangoli at home had half the illiteracy levels of TV viewers who did not watch the show after five years of schooling and watching.

The good news of the past 30 years is that we’ve made immense progress in ensuring that all kids — and girls especially — go to school. The better news for the next 30 years is that we have some understanding of the ways to ensure those kids actually learn something. All that is left is the willingness to confront the political challenges connected with rewarding teachers for learning outcomes — and ensuring they have the tools to help deliver inside and outside the classroom. That bit should be easy, right?