Despite a widespread perception that education positively affects the social assimilation of immigrants and their children, the driving forces behind this relationship are not well understood. This paper examines the mechanisms through which education affects the probability of interethnic marriage, (1) one specific measure of social integration. Particular attention is given to the role of assortative matching on education in explaining marriage patterns of second-generation immigrants.
In a series of papers, Borjas explains the slow assimilation of immigrants using the concept of ethnic externalities. Put simply, if immigrants associate predominantly with members of their ethnic groups, then the descendents of immigrants growing up in better ethnic environments, as measured by average years of schooling, for example, will be more successful later in life. In this way, ethnic externalities can slow the upward mobility of members of low-skilled immigrant groups and the downward mobility of members of high-skilled groups (Borjas 1992, 1993, 1995).
Assimilation can be further slowed if social interactions are primarily between people with similar education levels. If high education members of low-skilled ethnic groups do not associate with their ethnic networks, then they will not serve as role models and mentors for the next generation. Similarly, if low education members of high-skilled immigrant groups do not associate with their ethnic communities, then the downward mobility of advantageous ethnic groups can also be further slowed. If we interpret intermarriage as a measure of the broader social integration of immigrants, then knowledge of the importance of assortative matching on education in marriage markets can inform our understanding of the intergenerational immigrant assimilation process more generally. For example, if high education members of low education ethnic groups systematically marry outside of their ethnicity because of a lack of similarly educated co-ethnics, the resulting decrease in their association with co-ethnics may lead to fewer high-skilled co-ethnic role models for the younger generation.
The relationship between education and intermarriage may also help to explain why intergenerational assimilation rates of certain immigrant groups appear rather low. Duncan and Trejo (2009) suggest that if education is positively related to out-marriage and if children of intermarried parents are less likely to identify with any specific ethnicity, then standard estimates of intergenerational assimilation may be biased downward. Duncan and Trejo find empirical evidence of this for Mexican Americans. Although my results are consistent with those of Duncan and Trejo for Mexicans, the relationship between education and out-marriage is negative for certain other ethnic groups. Knowledge of the different mechanisms through which education affects intermarriage can help researchers appropriately interpret estimates of intergenerational progress of different immigrant groups.
Previous empirical studies of the relationship between education and intermarriage have produced mixed results. A number of authors have found a positive relationship (Chiswick and Houseworth, 2008; Cohen, 1977; Lichter and Qian, 2001; Meng and Gregory, 2005; Qian, 1997). However, Hwang, Saenz, and Aguirre (1995) find that Asian women with lower levels of education are more likely to out-marry racially. Kitano et al. (1984) find no relationship between occupational status and out-marriage for Chinese, Japanese, and Koreans in California. Based on another set of studies, Lieberson and Waters (1988) concluded that the influence of education on ethnic endogamy is small. To reconcile all of these seemingly contradictory findings, it is necessary to examine the various pathways through which education affects ethnic endogamy.
Using the insights provided by Wong (2003) and Gullickson (2006) in their papers on racial intermarriage, I argue that the mechanisms through which human capital affects ethnic endogamy fall into three main categories. …