Three Essays on Choice -Based Estimations of the Economic Benefits of Contaminated Site Cleanup

Abstract

We offer three different approaches to estimate the economic benefits of contaminated site cleanup. All approaches are choice-based, use information from both market and survey data, and are first-time application in the property value context. The first approach uses the standard conditional logit model. We integrate the market transaction data ex ante into the choice data which results in more realistic choice sets and avoids the problem of perfect collinearity in market data. The second approach focuses on the role of both observed and unobserved heterogeneity and uses the random parameters logit model. The recognition of parameter randomness and the inclusion of attitudes and perception data improve the result from the standard conditional logit model. The welfare measures are found to be sensitive to attitudes and perceptions. The third approach uses a latent class model to represent preference heterogeneity. We estimate the class membership probabilities prior to estimating the utility function within each class. We are able to address the income effect by incorporating the income category in the utility function rather than as a covariate in the class membership function. The results suggest an order of magnitude of benefits in the range of $200 million to $400 million in added residential homeowner property value for the City of Waukegan. The willingness-to-pay patterns in all methods are the same: the WTP for full cleanup is higher than the WTP for partial cleanup, and the WTP for additional pollution is always negative.