Distortion of Preferences and the Nash Theory of Bargaining

Nash’s (1950) theory and related theories of bargaining are frequently criticized for their lack of empirical content, since their predictions depend on the bargainers’ von :Jeumann-Morgenstern utility functions, which are not directly observable. The purpose of this letter is to suggest that an approach recently taken by Kurz (1977, 1978) in a somewhat different context may provide a remedy for this defect. Specifically we observe that agents who engage in a Nash bargaining situation may desire to misrepresent their utility functions. We can then show that a dominant strategy in the preference distortion game is for each agent to report a linear utility function. The resulting Nash bargaining solution is therefore equal division, which is undoubtedly the most often observed bargaining outcome in experimental bargaining situations [cf. Nydegger and Owen (1974)].