Delegation and Cheating : Why Amnesty Programs May Help to Motivate a Monitor

We consider a principal — agent model in which a principal delegates monitoring of cheating individuals to an agent. The principal can observe only the number of detection, not how many cheaters are deterred by the actions of the monitor. We show that conviction numbers should not be used for settings incentives for a monitor. Furthermore, we show that revenue maximization by the monitoring agency may lead to under-provision of monitoring effort. Increasing monetary incentives may have adverse effects on the monitor's effort. Counter-intuitively, the implementation of an amnesty program can reduce cheating among monitored individuals by uncovering the agent's effort afterwards and aligning the principal's and the agent's objective