DOES INEQUALITY AFFECT ASSORTATIVE MATING? EVIDENCE FROM THE UNITED STATES

Abstract

We postulate that increasing income inequality intensifies the degree of positive assortative mating (PAM). Our key motivation stems from Fernández, Guner and Knowles (2005), who develop a model in which equilibrium levels of PAM increase with skill premium (i.e. inequality) because of the higher cost of “marrying down”. We utilise a panel dataset of US marriage markets from 1990 to 2010 to investigate the presence and strength of the association between various income inequality measures and PAM at different education levels using fixed effects and instrumental variable models. Our instruments follow the Bartik specification, which averages national wages across industries using state-level industry employment shares as weights to construct a measure of state-level wages that is plausibly unrelated to state-level shocks. We find a positive and statistically significant impact of inequality on PAM, but only for individuals with at least some college education. This relationship is stronger for individuals with four or more years of college education compared to those with less than four years of college education. A one standard deviation increase in inequality is sufficient to move a state with median PAM levels to levels beyond the 75th percentile.