Performance comparison and incentive contracts

Workers use publicly observable performance information to evaluate their and their peers' status, thereby feeling superior or inferior. We analyze how such performance comparison affects optimal incentive contracts. In our model, a principal employs agents who may gain or lose utility by comparing their performance signals. We establish the following results. First, the optimal contract of each agent is based only on his own performance signal and muted compared to the setting with no performance comparison. Second, agents' performance comparison has a nontrivial effect on the principal's expected payment. Third, in the labor market where the principal cannot observe agents' sensitivity to performance comparison, she can attract the most desirable agents under some reasonable assumptions. We discuss the practical implications of these results.