How many interracial marriages would there be if all groups were of equal size in all places? A new look at national estimates of interracial marriage

Popular and scholarly efforts to understand the role of race in American society have long focused on interracial marriage. The most sophisticated of these inquiries adjust observed intermarriage rates for individuals' opportunities to meet a potential spouse from a given racial group. In this paper we argue that most of these national estimates of homogamy-heterogamy odds ratios fail to account for variation in racial composition across marriage markets. As a result, geographic constraints on interracial marriage are misinterpreted as evidence of social distance between groups. Using data from the 1990 census, we find that national estimates of race effects on partnering decline by between 19 and 53 percent once we control for the racial composition of local marriage markets, as opposed to the racial composition of the US. We discuss the implications of our findings for estimates of the social distance between racial groups, as well as for studies of other forms of intermarriage.