Educational Assortative Mating in Two Generations: Trends and Patterns Across Two Gilded Ages

PWP-CCPR-2014-015

Abstract

Patterns of intermarriage between men and women who have varying levels of educational attainment are indicators of socioeconomic closure and affect the family backgrounds of the next generation of children. This paper builds upon my prior studies with Schwartz on educational assortative mating in the United States, which document a long run increase in the educational resemblance of husbands and wives; the effects of the timing of school leaving and marriage on assortative mating; the greater educational resemblance of parents than of married couples as a result of differential fertility; and the effects of economic inequality among education groups on educational homogamy. It reports two extensions of this research. First it documents trends in the educational assortative mating of the parents of adults observed in the 1972-2010 General Social Surveys and 1973 Occupational Changes in a Generation II Survey. These trends, which describe the educational homogamy of couples whose children were born between 1900 and 1975, are compared to trends for prevailing marriages reported by Schwartz and Mare. This work shows whether the trend toward increasing marital homogamy that one sees in a time series of prevailing marriages appears as the time series of parents' educational attainments as well and extends observations of the trend in educational assortative mating back to the end of the 19th Century. These analyses document a dramatic “U-turn” in educational homogamy over the 20th Century. Spousal resemblance on educational attainment is very high in the early 20th Century, trends down to an all time low for young couples in the early 1950s, and increases steadily since then. These trends broadly parallel the secular compression and expansion of socioeconomic inequality in the U.S. over the 20th Century and support the few that widening economic inequality has implications for not only the wellbeing of individuals, but also for patterns of social organization.

Second, the paper examines a hypothesis about educational assortative mating, that marital homogamy is transmitted across generations. That is, are individuals who are raised by educationally homogamous parents more likely to marry homogamously themselves? Such an association may arise because parents socialize their offspring to prefer mates similar to themselves. Additionally, individuals raised by homogamous parents may be exposed to socioeconomically more homogenous potential marriage partners. The analyses reported in this paper suggest that this is indeed the case. Couples in which a partner’s parents were educationally homogamous are 5 to 10 percent more likely to be homogamous than couples in which the partner’s parents were educationally heterogamous. If homogamy in the parent generation leads to homogamy in the offspring generation, this may reinforce the trend toward increased homogamy over time. Thus intergenerational transmission may be a cause of the well documented increase in educational resemblance of spouses. Intergenerational transmission of marital homogamy may be both an instance of socioeconomic reproduction at the family level and also a mechanism for explaining aggregate trends in educational assortative mating.