Mechanism Design with Moral Hazard

This paper studies dynamic mechanism design in the presence of moral hazard. Revelation principle extends to models with moral hazard for both full commitment and limited commitment, but I also identify environments in which the principal doesn’t benefit from eliciting agents’ private information or beliefs. One-shot deviation principle requires the knowledge of agents’ private strategies after deviations, and I characterize the necessary and sufficient condition for all IC constraints that requires only the knowledge of agents’ equilibrium strategies. I also provide two sufficient conditions for smaller set of IC constraints that require obedience after a single-period deviation to be sufficient for all IC constraints. I illustrate how to apply revelation principle and the smaller set of IC constraints with an application allowing for endogenous state.