Monitoring optimistic agents

Monitoring is typically included in economic models of crime thanks to a probability of detection, constant across individuals. We build on recent results in psychology to argue that comparative optimism deeply affects this standard relation. To this matter, we introduce an experiment involving proper incentives that allow a measurement of optimism bias. Our experiments support the relevance of so-called comparative optimism in decision under risk. In the context of illegal activities, our results provide a guide into costless devices to undermine fraud, through well-designed information campaigns.