Essays in Nonmarket Valuation with Applications to Environmental Economics

Abstract

Non-market valuation techniques are vital instruments of policy. Providing quantitative welfare estimates validates public policy decisions. This research independently addresses three aspects of nonmarket valuation. Hedonic valuation of industrial accidents, limitations in recreational demand discrete choice modeling, and opportunities to expand urban planning research techniques.

A hedonic estimation of the 2006 Apex, NC chemical fire finds that this unexpected event impacted public welfare, but only mildly and for a limited time period after the accident. Important factors seemed to be the lack of long-term contamination and the zero chance of future contamination.

A mixed logit latent class estimation of recreational fishing trips in Wisconsin demonstrates the welfare implications of the random sampling of alternatives. This technique strategically reduces estimation time, but is difficult to implement in models where the independence of irrelevant alternatives (IIA) hypothesis is not present. By estimating a multi-class model using the EM algorithm, IIA is reintroduced in the model. This research evaluates the impact of various sample sizes on welfare estimates.

The final research chapter applies the mixed logit model in the context of housing demand and urban planning. An exploratory estimation of a multi-class model using past research as a guide for choice set construction shows promise for this method's applicability to practitioners.