Pollution control policy and social welfare: a theoretical and empirical analysis

Abstract

The impact of pollution control policy upon social welfare is a central theme to the discipline of environmental economics. Theoretical analyses focus on minimizing social welfare losses when selecting a policy instrument. The efficiency of a price or quantity policy depends on the nature of the marginal benefit and marginal cost curves. Their relative slopes are key to the selection of the more efficient instrument. In addition, uncertainty of the curve estimates can affect policy instrument choice. Since policymakers and even polluters rarely have accurate estimates of the abatement benefits and costs, the nature of the marginal curves is not well known. Such complications and how to account for them can affect the choice of the more efficient policy instrument. Typically theoretical analysis assumes an additive and symmetric distribution about the marginal curves. While these assumptions provide ease of use, they can lead to the use of the improper instrument. In the first two chapters of this dissertation, the influence of these error assumptions is examined. Chapter 1 compares a simple additive error structure with a more complex and realistic multiplicative error. When the marginal benefit and marginal cost curved are relatively flat, the difference between selection criteria can be substantially different. The ability of the price instrument to efficiently minimize welfare losses is compromised with the multiplicative error assumption. Chapter 2 focuses upon the role of informational bias upon policy instrument choice. With biased estimates, the assumption of a symmetric distribution about the marginal curves becomes invalid. Chapter 3 is an empirical analysis of air pollution in forty metropolitan areas of the United States. The criteria pollutant ozone is used as the dependent variable. Two hypotheses are tested: that 1) automobile use and 2) economic sector composition affect ambient ozone levels. Econometric results strongly support the first hypothesis and weakly support the second.