Commodity definition and content validity in stated preference valuation: a meta-analysis of water quality welfare estimates

Abstract

This paper applies a meta-analysis to investigate variation in willingness to pay estimates that arise from the use of different commodity descriptions in stated preference valuation surveys. To maintain commodity consistency, the data set for this meta-analysis is composed of willingness to pay estimates from contingent valuation, conjoint analysis, and choice experiment studies valuing water quality change in surface water bodies in the United States. The analysis uses an ordinary least squares regression with a cluster command to correct for potential correlation between observations drawn from the same study. The primary contribution of this study is the identification of systematic variation across stated preference studies resulting from changes in how the environmental commodity is presented and defined. By identifying the directional effect of these differences, this analysis provides insight into interpreting stated preference estimates and guidance for producing well-designed stated preference studies capable of eliminating bias and context effects.