Non-point-source pollution regulation as a multi-task principal-agent problem

This paper considers a multi-task, principal-agent problem where risk-averse farmers possessing private information have two tasks, pollution control and corn production, but only direct incentive for corn production. Using a highly tractable reformulation of the standard uncertain production model, a general method for solving the associated constrained Paretian problem is developed and analyzed. The optimal solution is shown to obey a generalized inverse-elasticity rule, and the optimal solution is characterized under a number of assumptions about the underlying technology and the role that pollution emission plays in reducing farmer risk.