Environmental policy as a multi-task principal-agent problem

We use a multi-task principal-agent model with moral hazard to study environmental regulation of a private agent by an EPA that can also allocate its budget to an alternative project with environmental benefits. In a first possible optimum, the EPA imposes a flat fine that exhausts the agent's participation constraint. In the second, the EPA provides the harshest possible punishment for a "poor" observed environmental performance and the highest possible reward for a "good" observed environmental performance. Increases in the available budget and in the maximally allowed penalty have then an ambiguous e_ect on total environmental quality.