Estimating the Extent of Intimate Contact between the Races: The Role of Metropolitan Area Factors and Union Type in Mate Selection

We use the 1990 U.S. Census Public Use Microdata Sample to assess the effect of union type and conceptions of marriage markets on patterns of intimate interracial contact. In so doing we challenge the assumptions that marriage is the only type of union that needs to be examined, and that people select partners in a national marriage market. Our results suggest that neither of these assumptions is correct. Cohabitations, which are a growing share of all unions, are much more likely than marriages to involve people from different racial groups. Similarly, consideration of local marriage markets reduces the gap between observed and expected levels of exogamy for whites, and narrows corresponding gaps in endogamy for blacks, Asians, and Latinos. We then conduct a reanalysis of black-white patterns of interracial partnering that incorporates what we have learned about union type differences and the geographic scope of marriage markets. A principal conclusion of the paper is that while race continues to be an important factor in union formation, it is much less important than previous work has suggested. It is widely accepted that the United States is undergoing a boom in interracial marriage. In 1990, nearly 5 percent of all marriages were between people of different races. This represents a dramatic increase in interracial marriage over the past 30 years (Harrison and Bennett 1995). Furthermore, recent evidence suggests that the rate of interracial marriage has continued to climb since 1990, with as many as 1 in 10 black grooms now walking down the aisle with a white bride (Besharov and Sullivan 1996; Suro 1999). The recent boom in interracial marriages has attracted both scholarly and popular attention. One central reason for this interest is that interracial marriage is often regarded as an indicator of social distance. The frequently attacked, yet pervasive, theory of assimilation posits that interracial marriage is an important step in the process of erasing group boundaries (Alba and Nee 1997; Gordon 1964; Park 1914). The very fact that people from different groups are able to form close relationships with one another and have these relationships sanctioned by the state suggests a growing similarity of cultures, as well as a decline in the saliency of group boundaries. The scholarly interest in interracial marriage is evident in a spate of recent studies. Kalmijn (1993) shows that there was a marked increase in black/white intermarriage between 1968 and 1986. Looking at national data for a broader array of racial groups, Qian (1997) finds that not only did the incidence of interracial marriage increase over the 1980s, but that this increase was concentrated among highly-educated couples (see also Pullum, Peri, and Bratter [1998] and Blackwell and Lichter [forthcoming]). Most recently, Farley (1999) extended Qian’s work by showing that in addition to education, one’s likelihood of marrying someone of a different race is also affected by region of residence and military experience. While these articles shed much light on the prevalence and patterns of interracial marriage, as a group they possess two key limitations that we seek to address in our work. First, while most analyses compare observed intermarriage rates to those that would be expected if people married at random, these estimates are based on unreasonable assumptions about marriage markets. Implicit in this work is an assumption of national marriage markets (i.e., one’s probability of marrying someone who lives down the