| This dissertation examines the phenomenon that most point-source sources of water pollution overcomply substantially with the pollution standards. I show that overcompliance with water pollution regulations exists in all sectors of the U.S. economy. I then try to explain how and why this overcompliance occurs. My data are discharges of Biochemical Oxygen Demand (BOD) over an eight-year period for all major polluters in the U.S. that release directly into rivers and lakes. All plants face plant-specific limits set according to rules promulgated under the Clean Water Act. Measurement of overcompliance is complicated by the presence of a random component in BOD discharges. Therefore, I define overcompliance in terms of the probability of violation using EPA guidelines. For the first time this dissertation measures the effects of random variations in discharges on compliance decisions. Lumpy changes in the plants' system, such as new technology, account for some but not all the overcompliance. I find evidence of a large degree of overcompliance before such changes take place. Average discharges are responsive to changes in the regulatory limits even when plants overcomply. This result holds both for manufacturing industries and wastewater treatment plants. Thus, I conclude that plants do not face zero marginal abatement costs. In other words, overcompliance is costly to plants. Community pressure, measured by relevant zipcode-level data, leads plants to reduce discharges. Residents in the vicinity of the plants who value clean water may be able to influence the plant's level of discharges. Plants in communities with socio-demographic characteristics that indicate a higher community pressure have a higher degree of overcompliance. |
