Late last year, the U.S. Census Bureau announced that it would introduce questions to capture gender identity in the agency’s most comprehensive survey of people living in the United States, the American Community Survey (ACS). The ACS will feature the new questions alongside longstanding questions that ask about family life, income, commuting times, education, employment, and even about the number of sinks and toilets in each household.
The Census Bureau began piloting the questions in February 2024, using what public health scholars call the two-step approach. The first question asks all respondents to answer the following question: “What sex was [NAME] assigned at birth” with male and female as the only two options. In the second step, individuals 15 and older will be asked to provide answers to a follow-up question: “What is [NAME’S] current gender?” People can choose from these options: Male, Female, Transgender, Nonbinary, or “This person uses a different term” with a write-in option.
These questions can raise concerns
Public health experts have celebrated the two-step approach, saying it’s reliably capturing the number of transgender people in a population, particularly in healthcare settings where it is necessary to know one’s sex assigned at birth to provide proper care. However, a study of how LGBTQ respondents regard the two-step question outside of healthcare settings found that many transgender respondents perceive questions about sex assigned at birth as invasive.
Many respondents, for instance, stated that they would hesitate to provide information about their sex assigned at birth out of concern for their privacy. The recent wave of legislative attacks on transgender people might make transgender respondents especially hesitant to provide this information.
In addition to these specific concerns, research from sociology suggests that there might be analytic slippages embedded in these questions, which might result in the ACS failing to accurately capture gender. This has to do with the difference between what people understand to constitute “sex” and “gender.” In brief, many people consider “sex” to refer to biological status, with terms like “male” and “female” to refer to sex.
How to ask about gender identity
“Gender,” in contrast, refers to how one identifies in a socially and individually meaningful way – man, woman, transgender, nonbinary, to name just a few options. If the stated goal is to capture gender in society, then it reasonably follows that the ACS should use questions that provide options that are relevant to gender, not sex.
A final point: Research in trans studies suggests that offering only “transgender” as a survey option lumps trans women and trans men together. That would risk overlooking important power disparities between the groups.
Political theorist Heath Fogg Davis proposes addressing this issue in his book, Beyond Trans: Does Gender Matter? Davis argues that when it is necessary to ask about gender, the question should provide a definition of sex and gender as well as a brief explanation for how the answers will be used. This approach lets people understand the exact intention of a specific question.
The Census Bureau will spend $10 million piloting these questions over the next year. That leaves plenty of time to refine the questions before the 2030 census. Scholarship in trans studies and political science can hopefully point the way towards the best questions possible.