Comparing Dichotomous Choice Models UsingTruncated Welfare Estimates

The purpose of this paper is to compare general and truncated welfare measures for single and double bounded dichotomous choice contingent valuation using data from an empirical example. The case study involves the benefits resulting from preservation of the landscape of a group of natural parks located in the higher areas of Gran Canaria (Canary Islands). Results show that double bounded leads to more conservative estimates than single bounded for general welfare measures. However, normalized truncated welfare measures for the mean and the median reduce the divergences between single and double bounded welfare estimates and tend to produce more efficient results, especially if the distribution assumption for willingness to pay is lognormal.