Penalizing Lies and Optimal Monitoring

Abstract

It has been established that monitoring schemes that allow self-reports of behavior can achieve superior results when the scheme induces truthful reporting (Kaplow and Shavell 1994). Often in reality, however, monitoring schemes do not enjoy the flexibility necessary to achieve truthful reports. In this paper we show how and why allowing self-reports can nevertheless improve on the monitor’s problem by way of penalizing lies when detected. Results show that, for any monitoring probability and sanction function, we can achieve actions closer to the first- best solution as well as higher levels of social welfare.