Should Principals Reveal Their Private Information?

Abstract

We analyze a principal agent model with hidden action, limited liability and truth-telling constraints under the assumption that the principal has private information. We focus on whether the principal should reveal his private information to the agent. On the one hand, revelation allows to adjust the agent's effort to the best information available. On the other hand, the agent's marginal information rent can be reduced if the principal conceals his private information, because the signal can be exploited as a statistic on the agent's behavior. We show that hiding the signal is superior whenever the good signal is relatively uninformative, or when the agent's effort cost function is sufficiently convex.