A SIMPLE FORECASTING MECHANISM FOR MORAL HAZARD SETTINGS

This paper studies a multiagent moral hazard setting. We resolve the tacit collusion problem that arises in our setting while employing a solution concept that makes less demanding behavioral assumptions than Nash. A simple mechanism is constructed that approximately implements the second-best solution in two rounds of iteratively removing strictly dominated strategies. Under our mechanism, the agents' best-reply correspondences are well defined, a small message space is employed, and the messages can be interpreted as divisional forecasts. (C) 1995 Academic Press, Inc.