On Young Turks and Yes Men: Optimal Contracting for Advice
42 Pages Posted: 29 Aug 2018 Last revised: 17 Jun 2021
Date Written: June 17, 2021
We study optimal contracting for advice by an agent about how much a principal should invest in a project. The agent has to be provided with incentives both to conduct research and to honestly report her findings. To motivate research, the agent's compensation must include a contingent component linking her recommendation to the ultimate outcome of the project. This endogenously creates incentives for the agent to misrepresent the magnitude -- though not the direction -- of her findings. For high-cost (low-cost) projects she wishes to overstate (understate) the magnitude of her research findings. To restore incentives for honest reporting, the agent must receive an additional payment that depends only on her report. For high-cost projects the principal optimally mitigates the concomitant agency rents by committing to ignore extreme (Young-Turk) recommendations, while for low-cost projects he ignores mild (Yes-Man) ones. These findings are shown to be robust to several natural extensions of the model.
Keywords: information, p-hacking, reporting bias, scoring rule, research
JEL Classification: C73, D81, D82, D86, L14
Suggested Citation: Suggested Citation