Polarization through Assortative Mating: Is Social Closure among the College Educated Increasing?

Iñaki Permanyer, Center for Demographic Studies (Barcelona)
Albert Esteve, Center for Demographic Studies (Barcelona)
Joan Garcia Roman, University of Minnesota

In the last few years there has been a recent upsurge of interest among researchers and social observers regarding the implications that increased assortative mating may have in terms of mounting social distance between social strata. There is a growing concern that the tendency of individuals to look for partners with similar characteristics might contribute to generate increasingly unequal, impoverished and polarized societies between those who are multiply advantaged and those who are multiply deprived. In this paper we investigate whether or not the disequalizing force of assortative mating is mitigated or even counterbalanced by two global phenomena that are sweeping the world: (i) the rapid process of education expansion and (ii) the closing and reversal of the gender gap in educational attainment. For that purpose, we have developed simple – yet effective – decomposition models that neatly assess the contributions of assortative mating, education expansion and the gender gap in education to the share of college educated couples and the corresponding polarization levels. Drawing from census microdata of the IPUMS project with a complementary use of different household surveys, we have assembled a database with more than 400 samples of 120 countries all over the world since the 1970s to the present day. Based on that dataset, our empirical results suggest that assortative mating and the gender gap in education have played a very limited role in determining the education polarization levels that are observed worldwide – the later being rather mechanically driven by the share of the population with college education. As college education expands, it turns out that the share of mixed couples (i.e.: those couples with only one of its members having college education) increases at the same pace as the share of college educated couples. Therefore, the feared scenarios predicting the shrinkage and gradual disappearance of mixed couples in favor of compartmentalized partitions between college educated couples and below-college educated ones are not occurring.