Optimality of Straight Talk: Information Feedback and Learning

Kimiyuki Morita (Osaka University of Economics)

Abstract: This study examines the interaction between information feedback and learning in a principal-agent model. The agent implements a project, and its success probability depends on the agent’s ability level which is uncertain. While the principal cannot offer any monetary incentive, she has superior information about the agent’s ability level and can provide feedback. The key feature of the model is that the agent may develop his ability level after receiving feedback, but before the project’s implementation. If the principal observes bad news and tells it truthfully, then (i) it hurts the agent’s incentive to implement the project, but (ii) it induces the agent to develop his ability. I derive the condition under which the principal tells the bad news truthfully.