Does College Education Mitigate Assortative Mating by Social Origins?

• In Event: Social Mobility

Abstract

This study tests whether college education mitigates assortative mating by social origins, a core implication of the modernization thesis. Using data from the Panel Study of Income Dynamics (PSID) and How Couples Meet and Stay Together (HCMST), we compare spousal correlations in social origin, as measured by intrinsic association coefficient (IAC), between college graduates and non-graduates. Results show that college graduates exhibit significantly weaker spousal correlations in social origins measured by parental education (0.31 vs. 0.61 in PSID; 0.26 vs. 0.42 in HCMST). Using a novel extension of conventional causal estimands to adjust for potential confounders and selection, we still find a sizable gap, suggesting that college education itself causally weakens origin-based assortative mating. We further find that the estimated gains from completing college are larger for non-graduates had they attended than for those who actually did. Mediation analyses indicate that about one third of the difference in origin-based assortative mating between graduates and non-graduates is explained by differences in meeting channels and confounding factors. These findings contribute to debates on higher education as a social equalizer by demonstrating its role in weakening—though not erasing—ascriptive boundaries in the marriage market.

Authors

• Fumiya Uchikoshi, Harvard University

• Lai Wei, The University of Hong Kong