Controlling for Observed and Unobserved SiteCharacteristics in Rum Models of Recreation Demand

Random Utility Maximization (RUM) models of recreation demand are typically plaguedby limited information on environmental and other attributes characterizing the available sitesin the choice set. To the extent that these unobserved site attributes are correlated with theobserved characteristics and/or the key travel cost variable, the resulting parameter estimatesand subsequent welfare calculations are likely to be biased. In this paper we develop a Bayesianapproach to estimating a RUM model that incorporates a full set of alternative specic constants,insulating the key travel cost parameter from the inuence of the unobserved site attributes. Incontrast to estimation procedures recently outlined in Murdock [21], the posterior simulator wepropose (combining data augmentation and Gibbs sampling techniques) can be used in the moregeneral mixed logit framework in which some parameters of the conditional utility function arerandom. Following a series of generated data experiments to illustrate the performance of thesimulator, we apply the estimation procedures to data from the Iowa Lakes Project. In contrastto an earlier study using the same data (Egan et al. [7]), we nd that, with the addition of afull set of alternative specic constants, water quality attributes no longer appear to inuencethe choice of where to recreate.[...]