An empirical study of environmental policy and technology adoption : phasing out toxic antifouling paints on recreational boats

In marine areas throughout the world, copper-based hull coatings are used on recreational boats to kill algae and barnacles. Unfortunately, copper, a registered fungicide, harms other marine organisms and violates government concentration standards when there is a sizeable number of recreational boats moored in close proximity. An immediate nationwide ban of copper would cost over one billion dollars. Designing more efficient policies for phasing out copper requires an understanding of copper and non-toxic hull coatings as a dynamic capital replacement problem, an understanding of the behavior of utility-maximizing consumers, and how these two factors interact with possible pollution control instruments. Once the dynamic nature of the policy problem is established, heterogeneity in capital vintage and heterogeneity in consumer willingness-to-pay for environmental properties become key determinants of the overall cost of alternate policy instruments. This dissertation consists of three papers that analyze these types of heterogeneity in a dynamic context and explore their implications for the optimal design of pollution control policies. The first chapter introduces the policy problem and puts forth a conceptual framework for thinking about how to design and evaluate alternative policies to transition to non-toxic boat hulls. Many of the issues raised are broadly applicable to environmental problems where the solution involves a large -scale replacement of durable consumer goods. The second chapter analyzes the preferences of utility-maximizing boat owners in order to accurately evaluate policy options. I implemented a choice experiment with recreational boaters in San Diego Bay to build an econometric model of paint choice, and estimate willingness-to-pay for marginal changes in paint attributes and discount rates implicit in respondents' tradeoffs over time. I also consider how choice behavior would change with altered expectations of future policies. Using my results, I discuss a fifteen-year plan to phase out toxic paints in San Diego Bay at no cost. The third and final chapter presents evidence that consumer preferences toward environmental attributes can be highly heterogeneous, and can strongly influence the time-path of pollution abatement. Prior studies tend either to ignore the speed of compliance/adoption by conducting static analysis or ignore heterogeneity by taking a representative agent approach; even models that condition on individual-level covariates are generally plagued by the fact that unobservable characteristics can drive the behavior of interest. Using data from my choice experiment, I estimate the distribution of environmental preferences. This heterogeneity will cause initial abatement to occur more quickly than would be expected without it, but the ultimate target will be achieved more slowly than in a homogeneous population. The greater the heterogeneity, the more this effect is exacerbated. Failure to understand and account for heterogeneity will prevent policymakers from achieving targets in desired time frames. I discuss implications of this finding, coupled with results from my first two chapters, for the optimal design of pollution control policie