Monitoring Agents with Other Agents

Abstract

I investigate the multiple agency problem when agents can monitor the performance of other agents. A particularly interesting incentive scheme of this sort has been used by the Grameen Bank of Bangladesh and I use this example to motivate some general questions involving group incentive schemes. For example, I show that a principal prefers a monitor who can reduce the cost of desirable actions rather than increase the cost of undesirable actions. I also consider when it is beneficial to the principal for agents to mutually insure each other. Finally, I examine a sequential incentive plan in which agents form a group and first serve as monitors and later are monitored by other agents.