Back to Blog

Meeting at the intersection of race and ethnicity in federal data analysis

According to the Pew Research Center, only about half of U.S. residents believe the 2020 census questions about race and ethnicity represent their identity “very well.”  

However, this is likely to change due to recent Office of Management and Budget efforts to formally review and update demographic data standards by the summer of 2024. 

Questions that more precisely reflect the country’s racial and ethnic diversity would give federal agencies more representative and higher quality race and ethnicity data, enabling them to better perform intersectional data analysis, which connects various aspects of identity (i.e., looking at race/ethnicity and gender, or race/ethnicity and age). These efforts would help our government understand important demographic trends—and address persistent intersecting inequities—when it evaluates 2030 census data.  

Advancing federal race and ethnicity data standards  

Federal agencies have collected data about race and ethnicity based on uniform standards since the development of Statistical Policy Directive No. 15 in 1977. Subsequent updates, most recently in 1997, helped enforce civil rights laws and ensure uniformity and comparability across federal data sets.  

In January 2023, the Office and Management and Budget and an interagency working group proposed further revisions to the 1977 directive, recognizing the country’s increasing racial and ethnic diversity, and changes in how people identify over the past 25 years.  

The current standards require a two-question format, first asking about Hispanic ethnicity and then about race (see photo on left below). Some experts want to combine racial and ethnic categories into one question by asking about race, Hispanic origin and other ethnic identities in a single prompt. 

These groups argue that the update will better reflect people’s identities and improve data quality for all communities of color—citing that approximately 44% of Hispanic and Latino respondents did not identify with any of the five race categories on the 2020 Census. Instead, these respondents either selected “Some Other Race” or did not answer the question, which resulted in them being categorized as white. A new combined question would reduce this inflation of the white population, enabling agencies to better analyze racial and ethnic disparities. 

 Another proposal is to add Middle Eastern or North African, or MENA, as a new minimum race category, rather than including it within the definition of “white,” and to require people to identify more specifically within a racial or ethnic category (e.g., “Filipino” under “Asian”), unless it creates unjustified burdens or privacy risks (see photo on right below). 

Employing an intersectional framework in research and analysis  

Having a single race and ethnicity question will make it easier to break down data by a wide range of demographic information. This shift will allow analysts to conduct intersectional analysis that uncovers the perspectives of groups that are often excluded or hidden from official census tallies, especially within Hispanic and MENA communities.  

Federal agencies have already expressed interest in this approach. The Department of Health and Human Services, for example, created a tool to highlight the importance of using an intersectional lens when designing demographic studies and involving people with lived experiences as collaborators.  

 The White House has also moved in this direction, advocating for an intersectional approach to address overlapping forms of discrimination and bias in its strategy on gender equity.  

 This growing consensus on the need to develop more intersectional data standards suggests the time is now to combine racial and ethnic categories into single census and federal recordkeeping questions. This small tweak would lay the groundwork for transformative change, enabling our government to better understand and address intersecting inequalities in society.  

---