On the optimality of full disclosure

Abstract

When is transparency optimal in principal-agent relationships? We consider the following setting. The principal has private, payoff-relevant information about the potential of the project. She can share this information with the agent and can commit to any information structure. Positive news motivate the agent, while bad news depress effort. Under rather mild and natural restrictions on the utility functions of the parties, we obtain interpretable and easily verifiable sufficient conditions for the optimality of full disclosure. We also show that full disclosure is optimal under some modeling assumptions commonly used in applied principal-agent papers.