Abstract

This paper uses previously unexploited data to examine polluters' compliance with point-source water pollution regulations, a line of inquiry motivated by widespread apparent overcompliance. We use a nationwide panel of monthly plant-level effluent concentrations from 1991-1999. These monthly data allow us to observe month-to-month variability in discharges and therefore to test hypotheses not previously examined.We find that plants that have higher discharge variability have lower median discharges. This is strong evidence for the purported "safety margin" effect. This result then implies that average discharges, a common measure of polluter behavior, are not an accurate indicator of polluter behavior.In response, we construct a plant-specific implied probability of a violation that accounts for discharge variability. We show that plants in richer communities have lower violation probabilities. We further argue that plants are indeed overcomplying with the regulations even when accounting for the safety margin effect.