Valuing recreation, time, and water quality improvements using non-market valuation: An application to San Diego beaches

In this dissertation, economic values of beach recreation and water quality improvements for San Diego beach users are estimated using revealed preference and stated preference methods. The results provide insights into preferences for water quality and behavior of San Diego beach users useful to recreation managers and policy-makers. This research contributes to improving the methodology of recreation demand analysis, particularly with respect to the proper modeling of money- and time-constrained recreational choices, the treatment of the opportunity cost of recreation time, and the analysis of referendum with follow-up contingent valuation questions.;To analyze recreation behavior and value beach trips, an empirical random utility maximization-based recreation demand model is developed that is fully consistent with underlying time- and money-constrained choices. Under the assumption that the next best alternative to recreation time is time spent in work, the shadow value of leisure time (SVLT) is the opportunity cost of recreation time. Examination of the two-constraint utility maximization problem reveals that the SVLT is in general an unobserved function of choice parameters, although for individuals who can marginally adjust their work schedules it equals the wage rate assuming that work time is utility-neutral. The recreation demand models explicitly use labor supply decisions to estimate the latent SVLT function, which is specified in a manner consistent with the two-constraint recreational choice model. Two approaches are used, one that estimates the labor supply model separately from the recreational site choice model and another that simultaneously estimates the two models.;The willingness to pay (WTP) for stormwater retrofit projects that would improve coastal water quality is estimated from a contingent valuation model that analyzes referendum with follow-up responses. This double-bounded format for eliciting values may introduce two basic types of response effects: incentive effects that cause observed preferences to be systematically different from the true underlying preferences, and anchoring effects from respondents using the proffered bids as value cues. Both response effects are explicitly modeled and estimated jointly. This double-bounded model provides a generalization of existing models and shows that controlling for these response effects can lead to accurate estimates of WTP.