Incentives and performance under two-dimensional moral hazard

This paper explores how an agent's incentives map into her performance. We let the agent choose how much effort to invest and which project to implement. We show that the relationship between an agent's incentives to perform and her expected performance could be negative and we characterize alternative conditions for a negative relationship. This also holds true when the principal instead of the agent makes the project choice. We thereby show that the you-get-what-you-pay-for principle is not robust and offer a non-behavioral explanation for the failure of incentives. The results have implications for regulation and ownership structures.