From 2007 to 2011, a fascinating experiment played out in five typically “low-performing” school districts across the country: Students at these schools were, essentially, bribed into getting good grades. All told, the project, organized by the Harvard economist Roland Fryer, paid out a total of $9.4 million to 36,000 children — and it was the children, mind you, who got the money, not their parents.
Which leads us to the real question here: Did it work?
The answer to that is complicated; money-as-motivator is never as straightforward as it seems like it should be. First, it’s important to point out that each city — New York, Dallas, Chicago, Washington D.C. and Houston — had slightly different incentives. In the New York project, fourth graders (9- to 10-year-olds) could earn $25 for doing well on tests, while seventh graders (12- to 13-year-olds) could earn $50 for each “A” grade they achieved. This allowed them to earn a maximum of $500 during the school year, half of which was put into an account, untouchable until they left high school, but the other half of which they could spend immediately. The Chicago scheme was similar. But in Washington D.C., the students received money not for high grades but for good behaviour, regular attendance and timely homework completion.
Conventional wisdom would suggest that whatever the ethics, such large incentives would yield results. And there was some success, though not uniformly in all five cities. In fact, the outcomes were decidedly patchy.
In New York and Chicago, the payments had little impact on grades overall, although in Chicago school attendance improved. In Washington D.C., results on reading tests rose for some students, but scores for other subjects didn’t. In Houston, math scores improved in the particular areas they had been paid to study, but it only made a small difference to their math grades overall.
In Dallas, however, things really did change, due to a crucial difference in the way the trial was run. Here, children were offered money not for obtaining good grades, but for completing particular tasks — in this case, reading books. They were much younger, at only 7 and 8 years old, and they were given $2 for each book they read. They could choose the books from a selection and could read them at their own pace. They took a comprehension test to prove they had read it, and while there was no direct reward gained for higher reading comprehension scores, that’s exactly what the children achieved — which is not surprising when you think about it. After all, it’s by reading that you get better at reading.
Now, there are a couple of things to note here. First, bribing children to read books might rob them of what’s known as intrinsic motivation. It could take away their love of reading, so that once the payments have stopped, they never pick up a book again. The researchers did look to see what happened next: In the year after the payments stopped, students were still doing well on reading tests, but there was only half the improvement. Gradually, the effect of the payments fades away.
However, it’s the second aspect of the Dallas experiment that is most important to our understanding of how financial incentives of this sort work. In Dallas, the children were paid to complete a task which they could all do as long as they put in the effort. In the other cities, by contrast, payment was not for effort but attainment — not for reading textbooks, but for getting high grades. Of course, by concentrating hard in lessons, doing homework and revision, some children could increase their chances of doing well in the tests. But for others, however much effort they put in, they just wouldn’t get the “A” grades that would earn them the payments. After each of the tests in the New York project, students were asked what they could do in order to get better grades next time, and couldn’t really answer the question. The students in the other cities were certainly keen to earn the cash, but by contrast in Dallas the task was specific, clear and within the control of the students, and that distinction appears to matter.
Despite achieving some success, the Harvard experiment has not been widely applied in the U.S. — or, indeed, in other advanced schooling systems. It seems there is still a certain moral queasiness about introducing monetary incentives into schools — a reaction which, as it happens, dates back almost 200 years. As early as the 1820s, the Society for Progressive Education tried paying New York school children for high academic results, only for the scheme to be abandoned after a decade amid fears that this was the wrong sort of encouragement for a child and that they might end up developing a “mercenary spirit.”
One of the issues, perhaps, is that in richer countries, the incentives offered are just not life-changing enough to make a difference. Perhaps in poorer countries they would have more impact; in fact, such schemes have been tried widely in lower income countries including Colombia, Costa Rica, Jamaica, Kenya, Mexico and Pakistan. In Mexico, the Progresa scheme paid families a large sum, as much as half their salary, to ensure their older children went to school every day. In terms of attendance, the scheme worked, particularly among the problem 14- to 17-year-old age group, where attendance rose from 64 percent to 76 percent. There was even an add-on effect on younger siblings, despite the payments not applying to them. Yet higher attendance at school only resulted in a 4–10 percent increase in attainment. An improvement worth having, but, just as we saw in Chicago in the Harvard experiment, getting children to go to school doesn’t necessarily get them to learn.
But a scheme introduced in the Colombian capital, Bogotá, in 2005, was a spectacular success. This time, students who graduated from high school received a $300 bonus, enough to pay for college fees. Before the bonus was brought in, only 22 percent of young people graduated from high school. With the bonus that rose to an impressive 72 percent.
How can this be explained? It could be that lots of talented students were failing to graduate from school because of financial pressures, but once the $300 graduation bonus was introduced the calculation changed. It was economically sensible for them to stay on at school and utilize the ability they already had to pass their final exams.
A key element, though, was the size of the incentive relative to earnings in the country: $300 in Bogota in 2005 was a lot of money. Getting the bonus could make a real difference to your life, enabling you to study in a way that you couldn’t before.
After comparing schemes from around the world, Professor Robert Slavin from York University concludes that financial incentives are only successful in richer countries if they target the very poorest people. Meanwhile, in low-income countries, the schemes work best when conditions are well-defined and concurrent efforts are made to improve teaching in the schools. What is also clear is that incentives are better than penalties, and payments are more successful in concrete subjects like maths than in other subjects. Even then, they are an expensive way of enhancing results, so it is arguably better to invest the money in improving the education system overall. (My apologies to the straight-A students out there hoping to make some extra cash.)
Excerpted from MIND OVER MONEY by arrangement with Harper Perennial, an imprint of HarperCollins Publishers. Copyright © 2016, Claudia Hammond.
