Does identity affect intermarriage decision? Evidence from NLSY

The marriage gap between black and white Americans has widened since 1950s. Black women are less than half as likely to marry than white women, across all education and income levels. In 2012, nearly 50% of black women in the age group 25-50 have never been married whereas only 20% of white women in that same age group have never married. It has been hypothesized in the literature that sex-ratio imbalance of African American is the major culprit of the phenomenon due to high incarceration rate of black males, however, this explanation is only partial. We pay particular attention to the fact that black female is also the least likely group to form interracial marriage, which has not been paid much attention in the literature. In 2010, only 9% of black females married someone outside of own race, whereas the number is 36% and 26% for Asian and Hispanic women, respectively. Building on the framework of Akerlof and Kranton (2000)’s identity model, we empirically test the role of identity, a sense of self, in black females’ decision to intermarry. Two identification strategies are used to test the hypothesis of identity effects: i) fathers’ region of origin, which was categorized according to the timeline of each states decision to repeal anti-miscegenation laws, was used as a proxy for identity. ii) third-generation migrants were used as a proxy for identity under the assumption that a lag exists between assimilation to the host culture and one’s identity. The results of the empirical analysis using NLSY data indicate significant effect of identity on black female’s inter-marriage decision, which take a partial role in the current diverging marriage gap in U.S.