Contracting for Research: Moral Hazard and the Incentive to Overstate Significance

Abstract

A principal contracts with an agent, who is protected by limited liability, to acquire information concerning the desirability of investing in a project. To motivate the agent to perform the required research, it is necessary to offer him a schedule of contingent rewards that depend on his reported unverifiable findings and on the project's ultimate outcome. While the contingent rewards can be calibrated to solve the moral hazard problem ex ante, they endogenously create an adverse selection problem ex post. In particular, they generate an incentive for the agent to exaggerate the significance of his research findings, leading to another source of agency rents. The principal mitigates these rents by committing to ignore reports of extremely positive or extremely negative findings; i.e. extreme reports of either kind are bunched. Thus, the principal commits to under-utilize some of the agent's potential information.