Impact of Incentives on Academic Achievement and Household Behavior: Evidence From a Randomized Trial in Nepal

Abstract

I present evidence on the impact of financial incentives on students’ outcomes from a randomized field experiment in Nepal. My optimistic estimate of the impact of incentives is 0.09 standard deviations gains in average aggregate scores, while the pessimistic estimate of treatment effects is 0.07, where the former finding is marginally statistically significant and the latter is insignificant. The result is meaningful given that several experimental and institutional factors in the study make it less likely to find a significant treatment effect. Looking beyond the aggregate grades, I find large significant gains in some subjects, no gains in other subjects, and significant losses in yet other subjects. This observation is consistent with the hypothesis that instead of increasing total effort in response to cash incentives, students reallocate effort towards the subjects in which the returns to effort are the highest. There are no noticeable differences in gains between males and females and students from different socioeconomic background, and cash incentives have a relatively higher impact on students from the lower quartiles. My study contributes to the growing literature on academic incentives by recording household responses to the incentive scheme. I find that incentives increase the probability that a student receives private tutoring after school, but decrease the probability that a household member helps students with schoolwork. Finally, financial rewards do not have an adverse impact on students’ intrinsic motivation to learn.