A Proxy Contract Based Approach to the First-Order Approach in Agency Models

Abstract

To justify the first-order approach in principal-agent problems, the previous literature has focused on making the agent's expected monetary utility obtained from this approach concave in the agent's effort. However, relying on such concavity is overly sufficient. We propose new sets of conditions based on a novel doublecrossing property between a 'proxy' contract and the optimal contract derived from the first-order approach, extending the applicability of the first-order approach significantly. Due to the flexibility in choosing a proxy contract, our approach can be applied to a wider range of principal-agent problems including those with unbounded likelihood ratios (e.g., the normal distributions), in which the previous literature does not justify the use of the first-order approach.

Keywords: First-Order Approach, Proxy Contract, Double-Crossing Property

JEL Classification: C61, D86