Multiple choice discrete data joint estimation: John C. Whitehead
Introduction For many researchers in other fields (i.e. transportation and marketing), the use of “attribute-based methods” (Holmes and Adamowicz 2003) is synonymous (although inaccurate) with data combination and joint estimation in environmental economics. Choice experiments have become a popular approach to estimating the value of environmental amenities, especially in the context of recreation choices. There are other sources available for description on choice experiment survey design and analysis (Louviere et al. 2000). In this book, we recognize that the empirical methods developed in the transportation, marketing and (later) environmental economics literatures can also be used to jointly estimate data other than that elicited in choice experiments. In this chapter we provide two illustrations. First, revealed preference recreation site selection choices are often elicited during the recreation season with on-site surveys. Revealed preference data can be gathered on the total number of trips taken to that point. In order to estimate seasonal trips, stated preference trip data can be elicited for the rest of the season. In another application we illustrate how dichotomous choice contingent valuation for site access can be combined with recreation site selection data in order to increase the range of costs considered by respondents and generate gains from joint estimation.