LAUSR

Abstract Large, unpredictably variable gaps between Willingness-to-Accept and Willingness-to-Pay (WTA-WTP gaps) are a methodological concern with the Contingent Valuation Method. The present study investigates WTA-WTP gaps as one manifestation of gain-loss asymmetry: the phenomenon that selling prices exceed buying prices. We investigate some causes of this asymmetry. Using a novel paradigm, and scaled goods, we obtain valuations of three nonmarket goods: one private (Health), and two public (Environment). We systematically explore the effects on gain-loss asymmetry of Good-type (Public vs. Private), Method (Price vs. Choice), and Payment Units (absolute valuation vs. percentage of income). Across methods, gain-loss asymmetry varies by Good-type: it is larger for public goods. We also find that gain-loss asymmetry shrinks, but does not disappear, from Price to Choice. We identify the mechanism of this shrinkage: a bidirectional, asymmetric valuation shift in Gain and in Loss. Eliciting valuations via income percentages instead of absolute values, and reversing the Price/Choice sequence, preserves these effects. Valuations show less variability when elicited as income percentages than when elicited as absolute values. Insofar as smaller and less variable gain-loss asymmetries suggest methodological validity, Choice seems preferable to Price. Further, Income Percentage seems an economically justifiable basis for anchoring contingent valuations.