If you’ve filled out a survey at any point in the last 25 years, chances are you were asked two questions about your race and ethnicity: Whether you are of Hispanic or Latino descent, and then separately, if your race is White, Black, Asian, Pacific Islander, Native American or another race.
A new proposal aims to change that, merging the two questions into one and adding a new category for people of Middle Eastern and North African descent. That would alter how the government – and by extension, the research community studying Americans’ demographics, opinions, voting habits and behaviors – measures and reports on the race and ethnicity of the American public.
The proposal put forth by a working group of government statisticians and methodologists is at least partly an effort to reduce the share of Americans choosing a nebulous “some other race” category that is required to be included in the decennial census and the American Community Survey, two of the key government studies measuring American demographics.
While some researchers say the proposed changes would improve the accuracy and depth of the data available on race and ethnicity, others – particularly those who advocate for the Afro-Latino community – fear the plan would make it harder to understand racially driven inequalities in the US.
Decisions about what gets measured and how reach far beyond the numbers that appear on the Census Bureau’s website: Data gathered through these questions drives the way racial disparities in housing, health care and employment are understood and tracked, how congressional districts are drawn, and how the resources of some government programs are allocated and assessed. It can affect policymaking at the federal, state and local levels.
“The simple fact is that if your community is not visible in the statistics, you are functionally invisible when it comes to political representation,” said Thomas Wolf, the deputy director of the democracy program at the liberal-leaning Brennan Center for Justice at the NYU law school.
The public comment period on the changes closes on April 27 after being extended. Nearly 18,000 comments had already been submitted on the Federal Register notice page as of Sunday morning. Once the comment period ends, the standards will be in the hands of the nation’s chief statistician, Dr. Karin Orvis. Final decisions on the standards are expected by the summer of 2024.
Here’s what to know about the proposals.
The Office of Management and Budget sets standards for both the wording of questions and the types of data government agencies and surveys must collect when they are gathering information about Americans’ racial and ethnic identities.
The existing standards, which have been in place since 1997, call for one question asking whether respondents have Hispanic or Latino background followed by a second question on racial identity, with options for American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, and White.
Because of a congressional law passed in 2005, the decennial census and the Census Bureau’s American Community Survey are also required to include a “some other race” category in the second question.
Over time, the Census Bureau has seen a notable increase in the number of people choosing that option. In the 2020 census, “some other race” was the second-largest racial group with 49.9 million people opting for it. That trend has raised questions about whether the two separate questions accurately capture the racial makeup of the country.
“The ‘some other race’ category is intended to be a residual category for people who do not identify with any of the minimum OMB categories,” Merarys Rios-Vargas, the chief of the ethnicity and ancestry branch of the Census Bureau’s population division, said during a webinar on the proposed changes hosted by the NALEO Education Fund last month. “But when the residual category is the second-largest response group, changes need to be made, and we have identified a solution with the combined question.”
If implemented, the new standards would merge collection of race and ethnicity information into a single question, expand the categories used to measure race and ethnicity, and mandate the collection of more detailed information on race and ethnicity whenever possible.
The proposed combined question measuring a respondent’s race or ethnicity includes seven broad categories: White, Hispanic or Latino, Black or African American, Asian, American Indian or Alaska Native, Middle Eastern or North African, and Native Hawaiian or Pacific Islander. Respondents can choose multiple categories from that list. The congressionally mandated “some other race” category would also continue for the decennial census and ACS.
Under the existing standard, respondents of Middle Eastern or North African, or MENA, descent were typically considered racially White. Census Bureau research conducted in 2015 suggested that without a distinct MENA category, roughly 12% of people who otherwise had been identified as MENA chose “some other race,” but that dipped to just 3% with the addition of a separate MENA category.
The proposed changes would also require the collection of more detailed information on national or tribal origin within each of the major racial or ethnic categories. An example provided by the working group includes checkboxes for some common subgroups (such as Italian under White, Puerto Rican under Hispanic or Latino, Korean under Asian, etc.) as well as an open-ended box in which respondents could write in any additional detail they wanted to share.
The proposed standards result from a review launched by the Office of the Chief Statistician of the United States last year, building on work conducted in the previous decade by the Census Bureau, the OMB and others. A working group of federal experts put together the proposed changes, and the OMB released the working group’s proposals for public comment in late January.
Part of the challenge in formulating these questions is that race itself is more a social than a scientific matter. As the Census Bureau puts it, the categories “generally reflect a social definition of race recognized in this country and not an attempt to define race biologically, anthropologically, or genetically.”
Because the questions used in government work set the standard for much other research, they can affect the way Americans classify their own racial and ethnic identity.
“The way that we talk about race in this country has been very much shaped by the way we ask about it,” said Mark Hugo Lopez, the director of race and ethnicity research at the Pew Research Center.
A Pew survey in January 2020 asking respondents to describe their race or ethnicity without offering categories found that about 8 in 10 gave responses that fit within the OMB’s race or ethnicity categories. When the same participants were separately asked about their race and ethnicity using questions from the 2020 census, nearly all respondents were consistent across the two formats, but the mismatch was significantly larger for those of Hispanic or Latino heritage.
The government’s working group noted that a “large and increasing percentage of Hispanic or Latino respondents” to both the Census and the ACS are skipping the race question outright or choosing “some other race.”
Recently released data from the 2020 census made public by the Census Bureau shows that 43.6% of the Hispanic population either skipped the race question or reported being “some other race” alone during the decennial count. The Census Bureau contends that its research shows this is because “a large proportion of the Hispanic population does not identify with any of the current Office of Management and Budget race categories.”
Wolf, of the Brennan Center, noted the challenge that type of mismatch could present to the usefulness of the data.
“If someone’s self-identification doesn’t map onto the categories that federal law recognizes, the data does not really help people activate and protect their civil rights,” he said.
Researchers outside the government are largely dependent on the OMB standards to frame questions on race and ethnicity in a way that allows comparisons with the gold-standard government studies that track American demographics. Some of these researchers are concerned that respondents who do not see themselves represented in the data may be less inclined to participate in surveys. Insights Association, a professional organization for market researchers, conducted testing on how to ask about race and ethnicity in a way that respondents prefer and found that a single question with more detailed response categories received the most positive feedback.
Cindy Neumann, the director of research for the Insights Association, said, “Where [respondents] feel that they’re included, we feel that they’re going to be a little bit more willing to participate in research, and engage a bit more.”
A 2015 test by the Census Bureau found that a combined question on race and ethnicity decreased the share of respondents choosing “some other race” or skipping the question entirely. For Hispanic respondents, a significantly higher share identified as Hispanic alone under the combined format, suggesting they could be less likely to select one of the race categories also offered in a combined question than they would have using separate questions.
Some are concerned that the proposed standards aren’t measuring the right information.
Many of the public comments submitted in response to the proposals or shared during a series of town halls OMB hosted in March have focused on the language used in the Black or African American category. A movement has emerged to add a category to measure those who are descended from enslaved people in the United States separately from people of African or Caribbean descent. The comments submitted reflect disagreement about the specific language and structure that would best capture the community, but suggestions have included adding categories for American Descendants of Slavery, American Freedmen, or Foundational Black American, separating Black American from African American, and adding a separate question asking whether a person is a descendant of enslaved people. Each could measure a part of the population that some feel is unrecognized under the current standards.
Among advocates for the Afro-Latino community, researchers worry that asking about Hispanic or Latino ethnicity within the same question as race could minimize the detail available about the racial makeup of the Latino community.
“If I, for example, a Black Latina, want to mark my Latinoness but also say that I’m a Black woman, then I have to choose Latino as my race and Black as my race and then I’m counted as multi-racial,” said Danielle Clealand, an associate professor at the University of Texas who studies Afro-Latino identity. “What it does is turn many of us who identify as Black or White or Native American as multi-racial, and that is not how we self-identify.”
Critics of the proposal say multiple questions are necessary to measure race, ethnicity and national origin, since a single question could muddy the measurement of those identifiers, even if responses related to each of those concepts are available for respondents to choose.
“You don’t measure two concepts with one question, and so by putting Hispanic ethnicity and race into one question, you are risking a huge undercount not only of racially stigmatized groups but also of the overall Latino origin population,” said Nancy López, a sociology professor at the University of New Mexico who directs and co-founded the school’s Institute for the Study of “Race” and Social Justice.
“It’s not going to help us know how you are treated, and if there’s an injustice that needs to be rectified,” she said.
The components of race and ethnicity that can affect how a person experiences the world may not be evident in their answers, according to critics of the proposal. A person’s racial or ethnic self-identification may not match the way they are perceived and treated by others, or may not align with their national origin or ethnic heritage. If the questions ultimately used in the government standards aren’t clear about which aspects they measure, their utility could be diminished, the critics say.
The stakes are extremely high. In making any changes to the way race and ethnicity are measured, the working group and the chief statistician will need to strike a balance between reflecting the ways Americans choose to identify themselves with fulfilling the need for data that allows the government to enforce its own laws.
“Does this allow us to do the things that the census is intended to do – voting rights, civil rights, allocation of congressional districts,” said Lopez from Pew. “Race and ethnicity is central to the work of folks who are in those spaces.”