Abstract

Limits on a government’s capacity to enforce laws can result in multiple equilibria. If most agents comply, limited enforcement is sufficient to dissuade isolated agents from misbehaving. If most agents do not comply, overstretched enforcement capacity has a minimal impact on behavior. We study the extent to which divide-and-conquer enforcement strategies can help select a high compliance equilibrium in the presence of realistic compliance frictions. We study the role of information about the compliance of
others both in theory and in lab experiments. As the number of agents gets large, theory indicates that providing information or not is irrelevant in equilibrium. In contrast, providing individualized information has a first order impact in experimental play by increasing convergence to equilibrium. This illustrates the value of out-of-equilibrium information design.