Optimal Information Asymmetry and Team Performance Measurement

Abstract

This paper studies the effect of agents’ access to pre-decision information, depending on whether only group performance is rewarded or individual performance is rewarded. Pre-decision information only affects agents’ incentives, which in turn depend on the type of performance evaluation system. With public information and group performance measures, the principal prefers a perfect signal if the high productivity state is more informative, and otherwise prefers no information. With group measures, only team success is rewarded, but cooperation is more costly in the higher productive states when the high state is more informative, so the principal prefers to be able to distinguish the states. With individual performance measures, the principal can replicate the group performance measure or induce competition, and in both cases prefers no information. A private signal exacerbates the control problem, and generally, with either performance evaluation system, the principal prefers uninformed agents. However, if agents can become at least slightly informed, then with individual performance measures the principal is indifferent to the level of information conveyed in the signal. Individual performance measures better balance the incentive effects of private information. With group measures, by definition the principal is restricted to a cooperative performance based pay and always prefers the lowest level of information.

Keywords: Pre-decision information, teams, performance measurement, principal-agent

JEL Classification: M41, D82, M52, J33