Optimal Regulatory Enforcement

Criminal law applies to everyone, but regulations constrain only those who choose to participate in a regulated activity. Firms considering such participation face three choices: comply with imperfectly enforced rules, violate them, or stay out of the activity altogether. Importantly, regulators not only punish violators but also reward firms with a strong record of compliance. We show that the presence of compliance-contingent rewards in a regime with optional participation unsettles several tenets of the traditional optimal deterrence model: (1) the maximal sanction may not be optimal, (2) it may be optimal to over-deter violations, (3) the probability and magnitude of sanctions may be complements rather than substitutes, (4) more frequent audits may have lower enforcement costs, and (5) the pursuit of a common enforcement objective of raising the rate of compliance may lead to lower welfare even if the overall deterrence level is too low. We illustrate the practical relevance of our findings and conclude that optimal regulatory enforcement is significantly more challenging than the basic deterrence model suggests. (JEL K10, K20, K42)