Testable Implications of Bargaining Theories

Abstract

We develop the testable implications of well-known theories of bargaining over money. Given a finite data set of bargaining outcomes, where utility functions are unknown, we ask if a given theory could have generated the observations. When the data come with a fixed disagreement point, we show that the Nash, utilitarian, and the egalitarian max-min bargaining solutions are all observationally equivalent. These theories are in turn characterized by a simple test of comonotonicity of bargaining outcomes. When the disagreement point is allowed to vary, we characterize the testable implications of the equal gain/loss egalitarian solution. The main application of our result is to testing the tax code for compliance with the principle of equal loss. For other theories, we introduce a general method based on the study of real solutions to systems of polynomial inequalities.

Files

Additional details