A field experiment on intertemporal enforcement spillovers

Enforcement can affect compliance directly and indirectly, through spillovers. We study intertemporal enforcement spillovers by conducting a randomized field experiment in a university exam. The initial phase of the experiment is the exam itself. We induce variation in enforcement by randomly assigning students to rooms with different monitoring levels. The second phase is post-exam. In this phase, all students are subject to the same low level of monitoring. Our outcome variable is the probability that students steal a pen in the post-exam phase. We find that enforcement in the exam phase has a strong intertemporal spillover on compliance in the post-exam phase: students subject to a high monitoring level in the initial phase are about 33% less likely to steal the pen than students subject to low initial monitoring.