Moral Hazard and Renegotiation in Agency Contracts

Ve consider the problem of designing a contract between a risk-averse agent and a risk-neutral principal when the agent's action is subject to moral hazard and the principal is free to propose a new contract after the agent has chosen his effort level but before the corresponding outcome is revealed.In this setting any optimal contract is equivalent to one that is "renegotiation-proof." A renegotiation-proof contract that induces the agent to choose high effort levels by promising a higher payment following good outcomes must also induce the agent to choose lower effort levels with sufficiently high probability that the contract would not be renegotiated.We show that for a range of utility functions for the agent, including exponential and logarithmic forms, the cost-minimizing renegotiation-proof contract for a given distribution of efforts is the same as the cost-minimizing contract for that distribution under commitment.Thus, the force of the renegotiation-proof constraint is not to change the way that given distributions are implemented, but rather to change which distributions are feasible.However, if the agent has constant relative risk aversion lower than one, the principal may prefer to give the agent an ex-ante rent in order to relax the renegotiation-proofness constraint, so that the optimal contract may differ from, that under commitment not only in the choice of distribution but also in the way that distribution is implemented.Our theory may shed some light on why compensation of managers and contractors is frequently insensitive to the information obtained after the relationship is terminated, and why executives have considerable discretion to adjust the riskiness of their compensation.1.