A Model Of Selection and Policy Concerns

Abstract A principal designs a policy by weighing the goals she shares with an agent and those that divide them. The policy’s success depends on the principal’s unknown quality and the agent’s decision to help. After the policy succeeds or fails, a committee awards a prize to the actor with the higher expected quality. In equilibrium, the agent helps to avoid blame for failure or share credit for success. The principal benefits from exploiting the agent’s selection concerns to shift the policy in her favour while obtaining the agent’s help, even if doing so lowers the probability she is selected.