Flexible valuations for consumer goods as measured by the Becker–DeGroot–Marschak mechanism.

Economists have long been interested in mechanisms that lead to truthful revelation of the relative values individuals place on different goods. In this paper we take one of the most popular of such mechanisms, and show that valuations obtained using the Becker-DeGroot-Marschak (BDM) procedure depend on the distribution of prices presented to subjects when the mechanism is implemented. We show that this effect of price distribution occurs quite frequently, significantly impacts reported valuations, and that it is unlikely to be caused by misconceptions about BDM. This effect is the largest when pricing distributions show a large peak just above or just below an individual's average valuation of the good being considered. We also show that a simple non-incentive compatible subject rating of the desirability of goods can be used to predict the likelihood that pricing distributions will influence BDM valuations. Valuations for goods subjects report that they most want to purchase are most likely to be influenced by distributional structure. Our results challenge some of the dominant theoretical models of how BDM-like valuation procedures relate to standard notions of utility.