LAUSR

ABSTRACT This paper studies the effect of performance measurement error and bias on the principal's preference for a leader, who signals private information about a favorable common shock to a follower. Without a leader, both agents are privately informed and relative performance evaluation is optimal due to its ability to remove the common shock. An increase in the conservative bias can increase or decrease compensation, depending on the likelihood of the common shock. With leading by example, joint performance evaluation can be optimal for the leader, reducing the leader's incentives to free ride on the follower and an increase in the conservative bias reduces compensation. The principal prefers a leader if the likelihood of the common shock is low, or if agents' outputs are more likely to be independent. Further, the more accurate the performance measure, the principal's preference for a leader decreases, but the effect of conservatism is mixed. JEL Classifications: D23; D82; J33; M41.