 |
This paper studies a simple model of incentive-contracting where (i) a principal learns an agent's ability before the agent himself, and (ii) both the agent's productivity with the principal as well as his outside option depends on his ability. I characterize the optimal contracts for the principal, defined to be the most profitable equilibrium outcomes among those satisfying the D1 criterion: Pooling at an earlier date is strictly optimal if the agent's outside option is sufficiently sensitive to the principal's private information, whereas separation at a later date is (weakly) optimal otherwise. Further, the principal's profit is shown to be neither continuous nor monotone with respect to the agent's outside option. Implications for unraveling in entry-level labor markets are also discussed.
