We investigate the moral hazard problem in which a principal delegates multiple tasks to
multiple workers. The principal imperfectly monitors their action choices by observing the
public signals that are correlated with each other through a macro shock. He divides the workers
into two groups and makes them compete with each other. We show that when the number of
tasks is sufficiently large, relative performance evaluation between the groups accompanied
with absolute performance evaluation results in eliminating unwanted equilibria. In this case,
any approximate Nash equilibrium nearly induces the first-best allocation.
|