Should we pay students for good grades?

Ed Robinson | May 24, 2016

Sections Economics

Cash rewards can help motivate students to achieve good results—but the effect may not last long enough to get some kids to graduation, according to a randomized field study involving high schoolers and their parents.

University of Chicago’s Steven D. Levitt and John A. List and UC San Diego’s Sally Sadoff conducted an eight-month experiment—dubbed the “Chicago Heights Miracle”—at schools in Chicago Heights, Illinois, a suburb 30 miles south of Chicago, whose school district has low student achievement and high dropout rates. The researchers set up a program to give money to freshmen who met goals related to grades, standardized-test scores, attendance, and behavior.

A number of recent research studies have looked into whether students respond to cash incentives. Levitt, List, and Sadoff went beyond prior research by involving parents as well as students, and by varying the kinds of rewards, making some fixed and others awarded by lottery.

Over a school year, the researchers paid cash incentives to four randomly chosen groups of students and their parents. Some received monthly fixed payments of $50 for meeting goals, while others were entered into a lottery with a 10 percent chance of receiving $500. The researchers tracked students for up to five years to measure the program’s impact.

Both incentives had a modest impact on students’ performance. Only a quarter of the students in the control group met academic standards, but students who qualified for the incentives did about 5 percentage points better. The results were driven most by students who qualified for the incentives and were on the threshold of meeting the academic standards. They did about 10 percentage points better than those in the control group, and were about 15 percentage points more likely to be on track to graduate during that first year.

However, those effects lasted only two years. In the second year, students who had been on the verge of meeting standards were still performing 12 percent better than their peers who hadn’t received any incentives.

“If the short-term effects had sustained themselves, it would have been at a cost of $1,200 per graduate,” says List. “But since the effects faded out, we do not have a significant impact on graduation.”