Urban Wire Separating Race from Ethnicity in Surveys Risks an Inaccurate Picture of the Latinx Community
Jorge González-Hermoso, Robert Santos
Display Date

Media Name: shutterstock_1379490593_crop.jpg

Whenever we—Jorge (a native Mexican) and Rob (a third generation native Mexican American)—fill out a survey, we face the inevitable paired set of questions asking about our ethnicity and race. We readily fill in “Hispanic/Latino” for the Hispanic ethnicity question, but the subsequent race question always gives us pause. What should we pick? “White”? “Amerindian”? “Mixed”? Or should we choose “Other” and fill in “Mestizo”?

Our struggle is shared, despite having different generational profiles. We are asked to clearly distinguish ourselves through complex concepts such as race, ancestry, geographic origin, and language among options defined by researchers who view life, race, and ethnicity through their own ethnocentric lenses, which are not the same for everyone.

As our Latinx population increases, we must reconsider how we collect survey data not only from Latinxs but also from the many others in our richly diverse US society, including the growing population of people of mixed race and ethnicity. This is especially important for the US Census Bureau and for the Office of Management and Budget (OMB), the federal entity that mandates the racial and ethnic categories used in all federal survey and statistical reporting.

Getting these racial and ethnic categories right is critical because they guide policy and affect communities through federal funding allocations, congressional redistricting, state and local budgets, and data-driven business and research decisions.

Why is detangling race from ethnicity difficult for Latinxs?

The term Hispanic dates back to the 1850s and refers to Spanish cultural heritage, mostly predicated through language. The terms Latino and Latina, popularized in the early 20th century, refer to ancestry in Latin American countries. Unlike Hispanic, the term Latino eschews Spanish ancestry, which could exclude people of Brazilian or Haitian origin. In recent years, the term Latinx has been adopted to embrace a gender-neutral alternative to Latino (we chose to use Latinx in this blog post).

For some Latinxs, Hispanic and Latino implicitly communicate a mix of Caucasian and Amerindian biological ancestry, also known as mestizo. This racial self-identification applies to some, but not all, Latinxs. For instance, Black and indigenous Latinxs may never use the term mestizo. Still, for a big share of Latinxs, there is a linkage between the Hispanic ethnicity and the mestizo race, which may explain why nearly half of Latinx chose not to identify with any of the race categories included in the 2010 Census (by filling in “some other race” or “two or more races”).


figure 1

The consequences of inadequate questions about racial and ethnic identity

In 2016, the Census Bureau reported to the OMB that a test using a combined race and ethnicity question yielded a substantially increased use of OMB standard categories among Hispanic or Latino respondents, as well as higher rates of respondents identifying as Hispanic alone, instead of other categories, such as White. Based on this evidence, the OMB considered replacing the separate ethnic and race questions with a single combined race-ethnicity question (plus the addition of a “Middle Eastern or North African” category) for the 2020 Census. But in January 2018, the Trump administration announced that it would continue with the same dual question format as in the 2010 Census.

Experts on Latinx polling, like Matt Barretoclaim that ignoring the OMB’s research findings is a way for the administration to evade a more accurate (higher) count of Latinxs, since many would stop checking “White” if given the combined format question (as well as those who would check “Middle Eastern or North African” instead). This issue is another in a long list of factors that could contribute to the misrepresentation of Latinxs in the 2020 Census.

Identity is dynamic and evolving—our data collection instruments should be too

Racial and ethnic categories are social constructs, defined and designed by those who have historically held positions of influence. Our data collection instruments continue to echo this influence, yet the relevance of separate race and ethnic constructs is lagging while society evolves to create new labels and categories that better reflect people’s views and experiences—labels like Latinx, Mestizo/Mestiza, Negro/Negra, and Afro-Latinx.

The policy implications of using inadequate methods to collect data on identity are not trivial. The Census Bureau’s research has revealed that adopting the single race-ethnicity question would improve survey response quality and ensure Latinx people are accurately identified. Furthermore, it illustrates the importance of conducting ongoing research to keep identity categories up to date.

The US is becoming increasingly diverse, a fact illustrated through the self-identities of authors Jorge and Rob—from first and third generations of Mexican immigrants, respectively. Jorge is Mexican; Rob is Chicano. But both of us are Latinx. That’s why we need data collection methods that evolve and adapt to ensure everyone is fairly represented.

Research Areas Race and equity
Tags Racial and ethnic disparities Immigrant children, families, and communities Structural racism in research, data, and technology Latinx communities
Policy Centers Metropolitan Housing and Communities Policy Center Statistical Methods Group
Research Methods Research methods and data analytics
Related content
EJ advocates and researchers aren’t at odds but contribute diverse skill sets, tools, and expertise—and neither corners the market on any one resource or skill.
Overhead shot of people walking