ABSTRACT
Recent trends suggest a decline in the rate of intermarriage between Mexicans and non-Hispanic whites. In this paper, we argue that interpretations of this trend as a decline in preferences for intermarriage are misleading because of the lack of adequate data that captures both spatial and temporal variation in the level of intergroup contact. Using data from the Decennial Census (1980–2000) and the American Community Survey (2008–2011), we employ a novel methodological approach to disentangle the impact of spatial diffusion, ethnic replenishment, and shifts in preferences for homophily on Mexican ethnic intermarriage patterns across 543 Consistent Public Use Microdata Areas (c-PUMA). Once changes in the demographic composition of c-PUMAs are accounted for, multilevel models for repeated cross-sectional data provide no evidence of a change in the marital preferences of Mexicans over time. Trends in intermarriage rates are predominantly explained by compositional and structural changes.
Acknowledgements
The authors would like to thank the two reviewers for their excellent comments geared to improving our manuscript. The first author would also like to thank Eliana Feigl as well as the Lilge family for general support and the German Academic Exchange Service (DAAD) for financially supporting the research visit to the University of North Carolina. We are grateful to the Carolina Population Center and its NIH Center grant (P2C HD050924) for general support.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. This article will focus on Mexicans since they represent by far the largest origin group of all Hispanic migrants to the United States. Moreover, a focus on Mexicans is common in the literature as they also constitute one of the largest ethnic minorities in the United States.
2. The vast majority of studies describing and analysing intermarriage trends focus on broad racial and ethnic categories where Mexicans are subsumed under the panethnic label ‘Hispanic’ despite studies noting important variations in intermarriage across ethnic groups within panethnic categories (Okamoto Citation2007; Qian, Glick, and Batson Citation2012). Hence, although people of Mexican descent account for more than 60% of all Hispanics, this literature review can only be indicative of broad trends in Mexican/white intermarriage due to the literature's strong focus on the marital behaviour across panethnic groups. It should also be noted that there is little evidence to suggest that increases in the rate of intermarriage with other ethnic or racial groups can be considered a viable alternative explanation for the observed trends. As with other racial groups (see Okamoto Citation2007), unions between Mexicans and other non-Mexican, non-white individuals are rare (∼3–5%) and their frequency has changed little over time.
3. To test that an increase in preferences for endogamy results in a decline in the magnitude of relative group size on intermarriage rates (i.e. the negative coefficient on group size moves closer towards 0), we conducted a Monte Carlo experiment with simulated data under a wide range of conditions. The results confirm this expectation. Full results and computer code are available upon request.
4. Using a broader age range does not affect the results. These results are available upon request.
5. In the absence of data on date of marriage, excluding all 1st generation Mexicans who immigrated after the age of 17 may overstate the level of intermarriage among 1st generation immigrants who were single when they entered the United States. Nonetheless, because we focus on trends in intermarriage over time, this restriction should not affect our results unless there is something fundamentally different about the trends for those who immigrated as unmarried adults.
6. We approached the question of spline definition in two ways. The first, as described above, took the whole sample to define the quartiles. This approach approximates population weighted splines. We also defined splines based only on the 543 c-PUMAs effectively assuming that each c-PUMAs has an equal weight. Both approaches reach very similar conclusions (see for results based on the second approach). Additional robustness checks regarding the parametrisation of the time trend (quadratic or dummy variable specification) produced no difference in the substantive conclusions. We therefore decided to use the more parsimonious linear trend model formulation.
7. This classification of settlement areas is presented in .