Erratum: Fatter attraction: Anthropometric and socioeconomic matching on the marriage market

Abstract

Alfred Galichon pointed out to us an error in our paper “Fatter Attraction: Anthropometric and Socioeconomic Matching on the Marriage Market” (Chiappori, Oreffice, and Quintana-Domeque 2012). The properties derived in the theory section (sec. III) are not sufficient to validate the empirical strategy developed in the following section; the latter requires more specific assumptions. The issue can easily be described in the TU (transferable-utility) case (sec. III.B).We use the same notation as in the initial paper. In particular, women (men) are characterized by a vector ðX, εÞ RL RK (ðY , hÞ RK RL), where X (Y) is a vector of observable female (male) characteristics and ε (h) is a random vector reflecting female (male) unobservable attributes. Proposition 2 actually implies that, for any stable matching, the conditional distribution of the female index I(X), given the male characteristics Y, depends only on the male index J(Y ), and conversely. This property can be used to empirically estimate these indexes even in the most general framework, a possibility explored in forthcoming work.