Motivation and incentives in education: Evidence from a summer reading experiment

Policymakers and economists have expressed support for the use of incentives in educational settings. In this paper, rather than asking whether incentives work, we focus on a different question: For whom and under what conditions do incentives work? This question is particularly important because incentives' promise relies on the idea that they might take the place of some cognitive failing or set of preferences that otherwise would have led students to make choices with large long-term benefits. In this paper, we explore whether that is the case. In the context of a summer reading program called Project READS, we test whether responsiveness to incentives is positively or negatively related to the student's baseline level of motivation to read. As a part of the program, elementary school students are mailed books weekly during the summer. We implemented this book-mailing program as a randomized experiment with three treatment arms. Students in the first treatment arm were mailed books as a part of the standard Project READS program. Students in the second treatment arm were mailed books as a part of Project READS, and were also offered an incentive to read the books they were mailed. Students in the third experimental group served as a control and were given books after posttesting occurred in the fall. We find that, if anything, more motivated readers are more responsive to incentives to read, suggesting that to the extent that incentives are effective, they may not effectively target the students whose behavior they are intended to change.