A ÞrmÕs decision of obeying environmental regulatory standards depends crucially on its chances of being detected and on the costs it must bear in case of detection. We investigate the relationship between the amount of resources devoted to environmental monitoring and the extent of non-compliancy, using a game theoretical model to capture the strategic implications of the monitoring process. In our model a population of Þrms, each of whom decides whether or not to be compliant, and a monitoring agency, that can detect non-compliance only by monitoring signals, strategically interact (more precisely, each Þrm interacts both with the monitoring agency and all other Þrms). In particular, each Þrm produces a signal, the distribution of which is (not perfectly) correlated with its behavior, while the agency, that is resource constrained, chooses some (optimal) fraction of the signals to monitor; hence, the probability of being monitored for each Þrm depends crucially on the behavior of both the monitoring agency and all other Þrms. Simply put, if a large fraction of Þrms chooses not to obey regulatory standards, the probability of being monitored for non-compliant Þrms is small. The main consequence of the strategic interaction among Þrms is that a more aggressive monitoring policy may end up relaxing the resource constraint of the monitoring agency as long as enough Þrms, perceiving a higher chance of being detected, become compliant. In fact, while in a framework with no strategic interaction a more aggressive monitoring policy simply induces a larger fraction of Þrms to be compliant (we call this e!ect, recognized by Becker and Stigler in their seminal contributions, Òimpact effectÓ), in our model a more aggressive monitoring policy also implies a higher probability of being monitored for the remaining non-compliant Þrms, and, in turn, implies a further switch to compliancy. We show that this further switch, that we call ÒmagniÞcation effectÓ, can be very relevant; hence, when monitoring policies are to be designed, our advice is to take strategic interaction among Þrms in the right consideration.
