Improving College Instruction Through Incentives

Abstract

In a field experiment, we test the impact of performance-based incentives for community college instructors. We estimate that instructor incentives improve student performance on objective course exams by 0.2 standard deviations, increase course grades by 0.1 standard deviations, and reduce course dropout rates by 17%. The largest effects are among part-time adjunct instructors. To test for potential complementarities, we also examine the impact of instructor incentives in conjunction with student incentives and find no evidence that the incentives are more effective in combination. Our instructor incentives are framed as losses and distributed in the form of upfront bonuses that instructors pay back at the end of the semester if they do not meet performance targets. We elicit instructors' contract preferences and find that, at baseline, instructors prefer to work under gain-framed contracts with rewards distributed at the end of the semester. However, after experiencing the loss-framed incentives, instructors significantly increase their preferences for them.