Racial Identification, Phenotype, and Ancestry Among Latinos: Comparing the Separate and Combined Question Formats

• In Event: 30251 - Shifting Racial Boundaries and Politics

Abstract

Following a recent modification to the Office of Management and Budget’s Statistical Policy Directive 15, the U.S. Census and other government datasets will ask respondents to report their race/ethnicity using one, combined question. The change has garnered criticism, some of which invokes the assumption that the earlier, separate items for race and Hispanic ethnicity captured meaningful variation––in terms of phenotype or ancestry––among self-identified Hispanics who selected different racial categories (e.g., as White or Black or Some other race). Among Hispanics, to what extent does racial identification on the separate items capture phenotypic variation, and how does this compare to the phenotypic variation captured by identification on the combined item? We examine these questions alongside another one: to what extent does racial identification among Hispanics on the separate versus combined items capture Hispanic–non-Hispanic ancestry? Results indicate that racial identification on the separate item captures some phenotypic variation among Hispanics, but not much. Racial identification on the combined item captures a similar amount of variation, but for a smaller group of people who identify as both Hispanic and as some other category on the combined item. However, the combined item brings a considerable strength that has not yet been appreciated: it is a strong signal that respondents have one Hispanic and one non-Hispanic parent. As more children are born to Hispanic–non-Hispanic unions, the ability to identify this population will become increasingly important.

Authors

• Maria Abascal, New York University

• Edward E. Telles, University of California-Irvine

• Samuel Berne Titus, University of California-Irvine

• René D. Flores, University of Chicago