Optimal design of peer review and self-assessment schemes

A principal must decide whether or not to implement a project that originated with one of her employees. Several employees have information about the quality of the project. A successfully implemented project raises the inventor's chance of promotion, at his peer's expense, but a failed project ruins the inventor's career. An employee who has a relatively good reputation (and therefore is happy with the status quo) must be encouraged to promote new ideas. An employee who has a relatively bad reputation (and therefore wants to change the status quo) must be prevented from exaggerating the quality of new ideas. We study incentive-compatible and renegotiation-proof mechanisms, and we find that self-assessment (without any peer reports) is optimal.