Optimal Performance Measures in a Task Assignment Setting

This study examines optimal task assignment and performance measurement under moral hazard in a two state setting where two agents are risk neutral and have limited liability. Both productive and incentive efficiency are affected by task assignment. The principal may assign to each agent either two heterogeneous, interrelated tasks (broad task assignment) or two homogeneous unrelated tasks (specialized task assignment). We provide conditions under which it is optimal to assign broad or specialized tasks and under which aggregate or disaggregate performance measurement is optimal. Under a broad task assignment approach, disaggregate performance measurement is sometimes preferred by the principal to aggregate performance measurement. Under a specialized task assignment approach, aggregate performance measurement is always optimal. Further, under conditions where disaggregate performance measurement is optimal under broad task assignment, the principal's welfare is greater if she uses specialized task assignment. Thus, if the principal can choose between broad or specialized task assignments, aggregate performance measures are preferred