Values and marginal preferences in international business

In a recent paper in this journal, Maseland and van Hoorn argued that values surveys tend to conflate values and marginal preferences. This assertion has been challenged by Brewer and Venaik, who claim that the wording of most survey items does not suggest that these elicit marginal preferences. This commentary responds to their claim, and restates our original argument. We stress that the distinction between values and marginal preferences is a conceptual rather than a semantic issue. Conflation of the two is a problem of stated preferences methods in general, regardless of question formats. What is more, we argue that support for the marginal preferences interpretation of values survey data does not come from the specific wording of questions used in surveys, but from a growing number of paradoxical relations between values ratings and outcomes in values research. Implications are twofold. First, we argue that, in order to measure values, we need to look at alternatives to the stated preferences approach, such as revealed preferences and experienced preferences, of which the latter seems most promising. Second, we find that values surveys eliciting marginal preferences are useful for understanding and predicting differences in behavior in a given context.