Monitoring versus incentives

This paper analyzes the trade-o! between monitoring and incentives in a principal-agent relationship with moral hazard. We derive general results on the optimalmonitoring - incentives mix for the case where both parties are risk-neutral and the agentfaces a limited liability constraint. We show that the principal uses less monitoring andstronger incentives if the agent's liability limit is relaxed or if monitoring costs increase.To induce more e!ort on the part of the agent, the principal resorts to more monitoringor to stronger incentives, or both. In particular, there are cases where the cheapest way toinduce more e!ort is to use lower-powered money incentives, but with much more precisemonitoring.