Summary: use of “sex assigned at birth” in data collection ignores the impact of how references to that assignment are used as rhetorical devices to enable violence against trans people, especially trans women. We should interrogate the efficacy of the question and consider the construction of both sex and transness more broadly.

[In this essay, I use “we” to refer to two separate in-groups relative to myself: trans people and people conducting demographic research on trans people. Whether any given “we” includes the reader is left deliberately indeterminate.]

A 2014 report by the GenIUSS Group at the Williams Institute advises asking two questions to identify transgender respondents among survey participants. The first question is “What sex were you assigned at birth, on your original birth certificate?” with response options “Male” and “Female.” The question is “How do you describe yourself (check one)” with response options “Male,” “Female,” “Transgender,” and “Do not identify as female, male, or transgender.” Respondents are transgender if their self-description, what we might call their gender identity, does not correspond to their sex assigned at birth. This recommendation has been adopted as the best practice by various programs in California under the guise of cultural competence and inclusion, which has motivated this writing.

We live in a different political climate from 2014, when the two-step method was found to have the highest reliability among a select sample of trans informants. Doctors are still assigning sex to newborns – that is largely continuing as before, with the exception that some hospitals are no longer performing surgery on intersex infants, such as Lurie Children’s Hospital in Chicago and those in the Indian state of Tamil Nadu. What’s changed is that the notion of sex, often explicitly defined as sex assigned at birth, has gained mass traction as an anti-trans rhetorical device. Conservative legislative activists have weaponized conflicting ideas about whether sex is mutable, and to what degree.

Between 2014 and now, conservative activists have used sex assigned at birth to exclude trans children from sports and prevent them from accessing medically necessary care. Families with trans children are fleeing Arkansas after its Republican-controlled legislature overrode a gubernatorial veto and banned gender-affirming care for trans youth. Gavin Grimm v. Gloucester County School Board, which began in 2015 after a trans student in Virginia was made to use the wrong bathroom, was just settled this term, not a week after the ACLU announced it would sue Tennessee for mandating trans-antagonistic signage in inclusive (i.e. normal) restrooms. A proposed rule for HUD, since revoked by the Biden administration, gave guidance on how to exclude homeless trans women from shelters. Another Trump-era policy sought to allow discrimination against trans people in medical care. And this list is not the result of specific curation, just memory.

The concern for the researcher then, is that asking about sex assigned at birth in data collection parallels the attempted and enacted violence in contexts where other options are adequate and available. In the rest of this essay, we will examine how anti-trans legislation weighs in on the unsettled nature of sex’s permanence in legal and colloquial understandings, discuss additional problems with the two-step “sex and gender identity” approach to transgender demographics, and offer solutions that sidestep those issues.

A history of sex changes

Why do the Williams Institute guidelines specify an “original” birth certificate when asking about sex assigned at birth? Sex generally and sex assigned at birth in particular can be changed, at least sometimes. 

It helps to start from an earlier point than we might expect. Medicine has long understood sex as mutable. In Histories of the Transgender Child, historian Jules Gill-Peterson documents the field’s shift in understanding of sex from a paradigm in which humans were understood as fundamentally dually-sexed, containing both female and male elements expressed in different proportions, to one wherein children are initially “plastic” and should develop into idealized binary endpoints of male or female embodiment. This shift occurred by the early 1900s, first on the basis of animal experimentation and later as the result of experimentation on intersex and other children pathologized by medicine.

Note that prior understandings of sex encompassed both of what mainstream queer parlance now calls sex and gender identity. For example, in 1966, pioneering endocrinologist Harry Benjamin wrote of his trans women patients, “Their anatomical sex, that is to say, the body, is male. Their psychological sex, that is to say, the mind, is female.” What Benjamin calls “psychological sex” is now gender identity. The point of highlighting this lack-of-distinction is to identify that even the premise within the two-part sex and gender identity question that they are discrete is not fixed as fact.

Americans in the mainstream have understood sex as at least partially mutable since at least 1952, when Christine Jorgensen returned from Denmark to much buzz about the removal of her penis and testicles. Copy on the poster for the 1953 movie Glen or Glenda, inspired by her life to that point, reads “I changed my sex!” “What am I… male or female! The strange case of a “man” who changed his sex.” Gill-Peterson writes that “[u]ntil well into the 1950s and 1960s, many publicly trans figures in the United States and Europe used the language of intersex or endocrine ‘abnormality’ to legitimize their transitions in the public eye as a question of medicine ‘repairing’ mistakes made by nature,” (Histories, p. 80) indicating that sex was medically malleable in popular discourse.

Jorgensen herself moved away from “transsexual,” which names sex as the phenomenon being crossed, using “trans-gender” [sic] by the end of her lifetime. However, common parlance has not caught up. Operations on genital configuration are still called sex changes despite the existence of euphemisms like “gender-affirming surgery” and are the focus of many patients who fall under the medical designation of transsexuality. Designations for transsexual patients like male-to-female (MtF) and female-to-male (FtM) name sexes, not gender identities, as endpoints, and include specific understanding of genital surgery as desirable, with terms created for operative status or desire (pre-op, post-op, no-op). Many of these terms are considered dated and would not be used by, say, the ACLU or Lambda Legal, but they retain currency in medicine, medical research, and among trans people, especially older ones.

While LGBTQ 101 trainings today would shudder at the conflation of sex and gender identity, popular understanding does not care – all of these things are still labeled with “sex,” which is something that can be changed.

Birth certificate business

The Williams Institute specifies an “original” birth certificate because the legal system acknowledges the mutability of sex. Many states (e.g. Arizona, Texas, Maine, Michigan) require sex-reassignment surgery to change the sex marker on legal documents like birth certificates or ID cards, but others do not. In many states, a person can amend their birth certificate to one of male, female, or nonbinary (marked with an X) without any medical intervention. What this means is that people who are physically identical along the typical variables for defining sex may nevertheless have different sexes listed on their birth certificates after updating. Thus, sex assigned at birth is a legal construction with imperfect predictive power over biological and psychological reality.

California is one such state that requires only self-designation as the criterion for updating sex markers. The form to change legal name and sex is called the “Petition for Recognition of Change of Gender and for Issuance of New Birth Certificate” and is numbered NC-300. The form uses “gender” and “gender identity” throughout, yet Californians can use the court order to change the field labeled “sex” on birth certificates, California driver’s licenses, and Social Security Administration forms. (Are we transgender to the county but transsexual to the state and to the federal government?) Whether intentional or historical accident from the legal system’s sloth relative to changing terminological norms, the result is a paradigm where sex assigned at birth is wholly and always mutable at the behest of gender identity.

The National Center for Trans Equality notes that in California, “If you receive a new birth certificate after amending your gender or your gender and your name together, the agency creates a new birth certificate and seals the old record.” A trans person who completes this paperwork and receives a reissued certificate from the California Department of Vital Records becomes legally assigned their current gender identity at birth. (Yes, this means that you can be legally assigned non-binary at birth.) By asking about the contents of the original, a researcher is seeking information that requires a court order to verify. That is a level of security that requires trust that most researchers do not earn or deserve.

Conservative activism on sex and gender

While the Biden administration is removing medical barriers for sex marker changes on passports, certain corners of government in the United States are nevertheless actively moving away from understanding sex as alterable. These are legal attempts in the direction of codifying biological essentialism, a notion that holds that true or authentic membership in gender categories is or ought to be restricted by sex assigned at birth, or alternatively that “gender is sex assigned at birth.”

In the 2020 Bostock v. Clayton County decision, the Supreme Court’s majority opinion says “an employer who fires an individual for being homosexual or transgender fires that person for actions it would not have questioned in members of a different sex.” The Court writes that Aimee Stephens being openly trans at work was specifically a case of “presentation as a different sex from the one assigned at birth” (emphasis mine). This narrowing of sex to birth circumstance is unforced; both parties in Bostock agreed that sex was “status as either male or female [as] determined by reproductive biology” or even less helpfully, “biological distinctions between male and female.” Neither of these definitions is incompatible with the above discussion of how the legal system sees sex assignment as impermanent.

Efforts are underway in statehouses across the country to render “sex” as equivalent in meaning to “sex assigned at birth” and wholly discard gender identity as the relevant factor in social organization of gendered activities. There is little bashfulness here; this is frequently the explicit intention: Montana HB 112, which has been signed into law, is titled “Require interscholastic athletes to participate under sex assigned at birth.” Florida S1028, also signed, requires designation of sports teams in public secondary and postsecondary schools to “be expressly designated… based on the biological sex at birth of team members.” Wisconsin SB 323 uses “sex” to mean “the sex determined by a physician at birth and reflected on the birth certificate.” Sometimes these efforts are clumsier. Other bills like Maine LD 926 and Arizona SB 1637 allow a physician to verify that a contested athlete is “female” using all of: internal and external reproductive anatomy, naturally occurring levels of testosterone, and analysis of a student’s chromosomes. However, they never specify what ranges or configurations are satisfactory; as previously discussed, the paradigm is a legal construction and not a physical one.

The long-term project of merging “sex” and “sex assigned at birth” ultimately enables trans-antagonistic rhetoric and action. Specifically, equivocation between “men” or “males” and “people who are assigned male at birth but are not men” is a long-running tactic to demonize trans women. Trans women and others have discussed this at length because we wade through the muck of these misgendering, biological-essentialist arguments online all the time. (If this is all confusing to you, Julia Serano has written a Medium article on the unworkability of biological sex arguments related to trans people.) Although “assigned male at birth” was coined by trans women to describe political differences between trans women and non-trans men, the phrase has been appropriated with its analogue “assigned female at birth” to push a narrative that “male-bodied” people oppress “female-bodied” people or “penis people” against “uterus people” – these examples are from a twitter thread by a trans woman who has had to lock her account multiple times following harassment. This genre of claims uses sex assignment and its purported permanence to assert that trans women maintain a position of power under patriarchy, misogyny, or any other sex-based system of your choosing, despite all socioeconomic evidence to the contrary.

Some recent examples from the right: “A man who identifies as a woman” features as a rhetorical bogeyman in the statements made by R.G. & G.R. Harris Funeral Homes, a defendant who lost in Bostock. Justice Alito quotes similar arguments in his dissent in Bostock, specifically referencing, and arguably telegraphing then-forthcoming action on, sports. Determined to make antagonism toward trans women into a pattern, Justice Alito called a hypothetical trans girl “a student who is biologically male but identifies as a woman” during oral arguments for Mahanoy Area School District v. B.L., better remembered as the “swearing cheerleader” case. Many of these comments are repeated without criticism or fact-checking from purportedly neutral-to-left-leaning outlets like NPR and CNN. As these same outlets have taken to pushing back against baseless claims of election fraud in 2020, it is apparent that the biology of gender and sex is not considered settled enough to warrant taking a position.

In this midst of this toxic maelstrom, asking for sex assigned at birth, however innocently intended, does not happen in a vacuum. Researchers of trans people cannot claim ignorance of or separation from the current political environment that wields sex assigned at birth with intent to erase, harm, and in the most extreme cases, cause the social death of, trans people.

Yet this is not the only problem with the two-step method.

Strain the strainer

In building a definition of transness that designates people as trans in any other way but asking “Are you trans?” we encounter problems. This is the issue that undergirds this whole discussion: Never mind whether we’re retraumatizing people as we try to figure out if they’re trans, are we even right in designating them as trans? How do we combine a set of externally observable metrics with the weight owed to individual identification?

This is a long-running problem. Can you be transgender if you say you aren’t? In the blurb for Imagining Transgender, David Valentine describes the how in the late 1990s “many of those labeled ‘transgender’ by activists did not know the term or resisted its use. Instead, they self-identified as ‘gay,’ a category of sexual rather than gendered identity and one rejected in turn by the activists who claimed these subjects as transgender.” As with the invention and enforcement of transsexuality and the sex-gender split by the medical establishment in endocrinology and psychology, even “transgender” itself has been pushed on people in its history.

Are you transgender if a researcher says you are? This is complicated by the boundary policing and general fuzziness surrounding categories like “nonbinary” that overlap imperfectly with transness. It is possible for a person to have a nonbinary gender identity that specifically involves not identifying with the gender associated with sex assigned at birth while not identifying as transgender. I know people in this situation – many are lesbian. The construction of transness posed by the Williams institute recommendations lumps in those who fit the operational definition of transness against their self-identification, which is a violation of community norms. (In another feat of historical weirdness, nonbinary was reportedly originally populated by people “assigned male at birth,” per community wisdom relayed via trans twitter, but has become associated in popular parlance with those assigned female at birth, especially to weaponize imagined white cisgender girlhood. More ironically, some early use of “nonbinary” on message boards is in the context of waiting for sex reassignment surgery among people who called themselves transsexuals.)

Self-knowledge and self-identification are pillars of contemporary queer understanding. To respect this under the two-step method, transness as a set of behaviors and transness as an identity would need to be split, perhaps under a different designation. There is much trans community-internal squabbling about whether any set of behavioral criteria is necessary for transness, but that’s beside the point here. We in fact already do this sometimes for other areas of LGBTQ research. For example, there is a distinction between “men who have sex with men” as a behavioral classification and “men whose social identity includes attraction to men,” who might call themselves gay, bisexual, or queer, regardless of whether they are having sex with men. So too can “people whose gender does not correspond to normative expectations based on sex assignment” and “people who describe their experience of gender as trans” be distinguished.

In conclusion

There are reasons for using a two-step method as proposed by the GenIUSS Group. Suen et al. 2020 note that there are populations who would be popularly understood as trans and may even understand themselves as trans in some way, but do not openly associate with transness as an identity:

African-American women of transgender experience noted how their communities would be less likely to openly adopt the “transgender” label to describe themselves and would only do so if relevant to medical contexts. This may be due to safety concerns given the high rates of discrimination and violence experienced by transwomen of color, wanting to increase internal gender affirmation, and/or to avoid stigma associated with “not passing” (Sevelius, 2013).

Asking about sex assigned at birth and current gender identity would capture these women of trans experience. Yet, this reluctance to openly identify with labels due to violence and discrimination is precisely what we have explored in this essay, but with “assigned male at birth” instead of “transgender” as a problem term!

Not everyone listens to the Williams Institute, so there are some alternatives already out there in California. The California Healthy Kids Survey, which has produced studies of trans youth with enviable sample sizes, shows one such method. The demographic questions ask, “What is your gender?” and “Are you transgender?” Its execution is by no means perfect – only the binary options are presented as responses to the gender question. Its explanation of what transness is, however, is technically sound. The full text of the trans question explains transness thusly: “Some people describe themselves as transgender when their sex at birth does not match the way they think or feel about their gender.” While this grates against the legal possibility of changing sex at birth, something which can be done by minors in California, it does not forcibly define a respondent into a trans category. [Since writing this, I have learned of problems with this methodology, but those problems have to do with comparisons among trans people, and not in determining who is trans.]

While trans people are, as a non-monolithic class, not unanimous on this or any issue, since the direction of institutional advocacy is in favor of asking about birth assignment, here are some select recent research findings and arguments against it. Puckett et al. 2020, a qualitative survey, “found some participants noted that being asked about their sex assigned at birth was troubling: ‘It should ask [if] we are trans, rather than our birth assignment, because having to identify in any way with our sex assigned at birth can be very painful.’” The subheading for an April 2021 BMJ article about the medical incompleteness of sex assigned at birth as a shorthand for normative binary sex information qualifies the practice as “unhelpful, potentially harmful, and should be abandoned.” A respondent in Suen et al. 2020 says, “Not to like, short circuit the whole thing, but I feel like the question that really ought to be asked is, ‘What do we as medical practitioners or researchers need to know about your sex and gender identity in order to provide appropriate care?’ Like, that’s the question that all of these are trying to get at.”

The quotes above are all discussions of sex questions in medical care, where birth assignment often serves as a way of bundling features like hormone dominance and genital configuration into normative categories. For researchers whose primary concern is not even medical, birth assignment is even less relevant for putting participants into buckets of “trans” and “not trans” when “are you trans?” is an option. The highlights from a more recent report by the Williams Institute advise that “Gender minority communities should always be engaged in the development of research and data collection about those communities.” CVS did this belatedly when it removed the question asking sex assigned at birth from the COVID-19 vaccination registration forms after trans people and allies objected. We do not need to and should not stop asking about transness altogether, but we can ask in a softer way.