This paper reexamines the issue of relative versus joint incentive schemes in a multi-agent moral-hazard framework. The model allows a full analysis of the information and dependence structure. An important result is that the widespread notion that greater correlation in outcomes calls for more...