COWLES FOUNDATION FOR RESEARCH IN
ECONOMICS Box 208281
COWLES FOUNDATION DISCUSSION PAPER NO. 1902 Nonparametric Analysis of Random Utility Models: Testing Yuichi Kitamura and Jörg Stoye July 2013 This paper develops new tools for the analysis of Random Utility Models
(RUM). The leading application is stochastic revealed preference theory, that is, the
modeling of aggregate choice behavior in a population characterized by individual
rationality and unobserved heterogeneity. We test the null hypothesis that a repeated
cross-section of demand data was generated by such a population, without restricting
unobserved heterogeneity in any form whatsoever. Equivalently, we empirically test
McFadden and Richter's (1991) Axiom of Revealed Stochastic Preference (ARSP, to be defined
later), using only nonsatiation and the Strong Axiom of Revealed Preference (SARP) as
restrictions on individual level behavior. Doing this is computationally challenging. We
provide various algorithms that can be implemented with reasonable computational
resources. Also, new tools for statistical inference for inequality restrictions are
introduced in order to deal with the high-dimensionality and non-regularity of the problem
at hand. JEL Classification: C14 |
