Black Sheep or Scapegoats? Implementable Monitoring Policies under Unobservable Levels of Misbehavior

Abstract

An authority delegates a monitoring task to an agent. Thereby, it can observe the number of detected offenders but not the monitoring intensity chosen by the agent or the resulting level of misbehavior. We provide a necessary and sufficient condition for the implementability of monitoring policies. When several monitoring intensities lead to an observationally identical outcome, only the minimum of these is implementable, which can lead to underenforcement. A comparative-statics analysis reveals that increasing the punishment can undermine deterrence, since the maximal implementable monitoring intensity decreases. When the agent is strongly intrinsically motivated to curb crime, our results are mirrored, and only high monitoring intensities are implementable. Then, higher monetary rewards for detections lead to a lower monitoring intensity and to a higher level of misbehavior.

Optimal Law Enforcement with Sophisticated and Naive Offenders

“…This could make the analysis of the optimal enforcement policy (and the decision concerning revelation) more intricate due to a potentially non-monotone relationship between the enforcement effort and the number of apprehensions (see e.g. Buechel and Muehlheusser, 2016 ). Second, a similar effect on the optimization problem arises when the fine is no longer purely re-distributive as in our framework, but for example receives the same weight ( π ) as the offenders' gains.…”

Optimal law enforcement with sophisticated and naïve offenders

“…As argued byBuechel and Muehlheusser (2016), the number of detections might not be too informative about the underlying enforcement effort.…”