A nonbinary Census? American Community Survey tests questions on sexual orientation and gender identity

What percentage of Americans identify as LGBTQ+? Academic institutes and polling firms have produced a wide range of estimates, but the U.S. Census Bureau will soon try and get the most accurate count ever. It will also potentially track the size of the population over time.

Keep up with the latest in LGBTQ+ news and politics. Sign up for The Advocate's email newsletter.

The bureau last week announced it would test the inclusion of questions about sexual orientation and gender identity in the American Community Survey, a regularly updated survey tracking the nation’s changing population and workforce. It’s a move that could mean strides for representation and recognition of LGBTQ individuals by the federal government.

Stephanie Galvin, assistant division chief of Social Characteristics; Social, Economic, and Housing Statistics at the Census Bureau, said at a Census National Advisory Committee meeting that the agency has drafted questions on gender identity and sexual orientation. Data will be collected in surveys and personal interviews. The federal agency hopes to gain valuable data about the number of LGBTQIA+ Americans and about the communities in which they live.

As questions get incorporated into data gathering, the agency has pursued input from the LGBTQI community, much of what was addressed at the national meeting.

D’Lane Compton, University of New Orleans’ Sociology Chair and a prominent sexual orientation demography expert, said they were happy to see the questions surveyed at the meeting.

“I'm very excited to see the results when available,” they said. “I was most excited to see that the trans and non-binary categories are not a single combined category in the gender identity series."

Compton advised that many transgender people may indeed be reluctant to offer too much information in the survey about their own gender identity.

“Past administrations and the current political climate has evoked fear among sexual and gender minority communities by promoting and proposing punitive legislation,” they noted. “Respondents may fear that they will be targeted.”

But Galvin said that interviewers will also try and address privacy concerns in order to see maximum participation.

“The Census Bureau already provides a lot of information to survey respondents about how we keep data confidential,” she said. “We do this you know for the ACS and for our other surveys.”

The bureau shared proposed questions to appear on the survey. One such question asks, ”Which of the following best represents how this person thinks of themselves?” Options include: gay or lesbian; straight, that is not gay or lesbian; bisexual; or a blank field for a different term.

In another example, the survey in a two-step question asks first the gender assigned at birth, offering only male and female as options, but then asking the person’s current gender, and offering transgender and nonbinary as choices, as well as offering an option for another term as provided by the respondent.

The American Community Survey typically gathers information from more than 3.5 million Americans, many of whom also answer for members of their household. The potential for differing family dynamics raises its own concern in terms of gathering information and ensuring accuracy. But census officials said interviewers will be debriefed on issues surrounding accuracy. Ultimately, the census plans to evaluate any problems faced such as a reluctance to answer gender or orientation questions, or to provide accurate information about family members.

Compton praised efforts to make the American Community Survey more gender-neutral throughout. The survey, for example, now asks about children, as opposed to simply sons and daughters. However, the professor pressed the agency on areas where options remain limited and bound by binary identity classifications. For example, options to identify family relationships with sons- and daughters-in-law remain in starkly binary.

Galvin stressed an “other relative” field will allow different answers. But she said terms like “child-in-law” generate confusion among respondents, even though the term “parent-in-law” seems generally understood.

Much of the development of questions and answers seeks to address apparent errors in the last decennial census. For example, the 2020 Census found about 1.2 million same-sex couples in the country, but researchers later found about a third of couples actually identified as different genders, Compton said.

New survey questions hope to simplify the process for respondents, reducing the chance of errors while improving accuracy.