“Gender identity” as a category of practice

Before embarking on my substantive analysis, I want to clarify my use of “gender identity” and specify more precisely my object of analysis. I do not use “gender identity” as a tool of analysis; I take the category, rather, as my object of analysis. I therefore use the term “gender identity” to designate the category, not what the category itself ordinarily designates. In addressing “gender identity,” that is, I take as my object of analysis the category itself, not the elusive phenomenon—often characterized as an inner sense of self as male, female, both, or neither—designated by the category. Moreover, in addressing the category itself, I am concerned with “gender identity” as a category of practice, not as a category of analysis: I take as my object the practical workings of the category—the work it does in its various instantiations in organizational routines—not the theoretical debates about the category.²

The distinction between categories of analysis and categories of practice is of course anything but a sharp one. Not only is there continuous traffic between the two in both directions,³ but theoretical reflections on the world are routinely colored by and enmeshed with practical interventions in the world. And analysis is undertaken not only by researchers perched outside the particular social worlds they analyze; it is internal to many social worlds, all the more so in a “knowledge society.” Yet it remains important to distinguish between categories developed and deployed in the first instance for analytic purposes and those used in practical contexts, including the medical, legal and bureaucratic, data-gathering, and pedagogic contexts that are the focus of my analysis.

Taking the category “gender identity”—its history, workings, and contestation—as my object of analysis enables me to bracket the deeply contested question of the ontological status of gender identity as a phenomenon. It enables me to sidestep the question of what gender identity ultimately is in order to focus on what the category does in various institutional contexts.⁴ For my purposes, it is immaterial whether the category picks out a clear, definite, and universally identifiable phenomenon—whether everyone has something called a gender identity as a core part of their self. What matters, for my purposes, is that the category itself indisputably exists in public discourse, legislative provisions, administrative regulations, organizational policies, court decisions, medical textbooks and guidelines, pedagogic routines, and data-gathering practices.⁵ Regardless of whether the category points to a deep psychological reality, its discursive elaboration and institutional embedding across a variety of domains has made it a powerful and consequential social reality.⁶

Medicine

As is well known, feminist scholars began to use the category “gender” in the social sciences in the 1970s and in history and the humanities in the 1980s.⁷ What remains less well known, despite considerable historical work on the topic, is that feminists were not the first to use the language of “gender.” It was psychologists and psychiatrists who introduced the category “gender identity”—and the language of “gender” more generally—in the context of clinical assessments of disorders of sex development or intersex conditions on the one hand and transsexualism on the other (the word “transgender” came into widespread use only much later, in the 1990s).⁸

In the intersex context, “gender identity” was first introduced in the early 1960s to denote a socially learned rather than an innate or constitutional sense of self as male or female. (Given the cultural salience of later “born that way” accounts of gender identity, the irony of this initially social constructionist notion of gender identity is worth noting). According to prevailing mid-20^th century protocols, intersex infants (often referred to at the time as “hermaphrodites”) were to be made to fit binary categories and were routinely subjected to surgeries that would “normalize” the appearance of their genitals (Kessler 1998; Karkazis 2008). “Gender identity” was introduced in this context by the hugely influential psychologist John Money to make a very specific and limited argument: that neither chromosomal sex nor prenatal hormonal exposure was decisive in determining whether such individuals would come to think of themselves as male or female. What mattered more, Money argued, was the “sex of rearing” through which individuals would learn the sex category to which they belonged and everything that conventionally went along with it. Once established in the first few years of life, this “gender identity”—this sense of oneself as male or female—would be “imprinted” and difficult or impossible to change thereafter, regardless of its relation to chromosomal sex.⁹ Gender identity was thus understood to be socially learned yet, once established, deeply rooted and indeed nearly immutable. In the intersex context, the category “gender identity” did practical work: it helped clinicians formulate recommendations about surgical interventions and sex of rearing.

In the transsexual context, “gender identity” was used to designate what had earlier been called a “psychological sex” or “sex of the mind” that did not match the “sex of the body.”¹⁰ As in the intersex context, the category denoted a deeply rooted and, past a certain developmental point, apparently immutable sense of self as male or female. The clinical finding that the mismatched gender identity appeared to be fixed and refractory to therapy helped make surgical reassignment thinkable as a humane way of relieving otherwise intractable suffering. It thus helped legitimize the gender identity clinics established at Johns Hopkins and several other prominent universities in the late 1960s to perform and assess the outcomes of such surgeries, which had hitherto been performed only rarely, in private clinics, and without any systematic study of their outcomes. The effort to legitimize the new subfield of medicine also involved a stringent screening of candidate patients to maximize the chances of a successful outcome in the new gender, with success being defined in terms of gender conformity, ability to pass, and heterosexuality (Fritz and Mulkey 2021; Shuster 2021: Chapter 1).

For young children, however, the practical work done by “gender identity” in the 1960s was anything but benign by today’s standards. In the intersex context, the category licensed what many now see as genital mutilation, guided by a narrow conception of acceptable genitalia, an intolerance for ambiguity, and an insistence on early binary sex assignment, before the looming deadline beyond which gender identity would lose its initial plasticity (Kessler 1998; Dreger 1998). In the context of psychotherapy, it licensed efforts to monitor and “correct” gender-nonconforming behavior—especially that of markedly “feminine” boys—in an effort to prevent the consolidation and “fixing” of a non-normative gender identity that would later be refractory to change.¹¹ Precisely because gender identity was seen as at least partly independent of sex, and as initially plastic, it was understood to need careful monitoring and early correction should it appear to be developing along non-normative lines.

In this initial phase, gender identity was established in contexts that were expressly understood as pathological. This understanding was symbolized and codified in the diagnostic category “gender identity disorder,” introduced in 1980 into the third edition of the Diagnostic and Statistical Manual (DSM), the authoritative handbook of the American Psychiatric Association.¹² Beginning in the 1990s, the transgender movement campaigned against this pathologizing perspective; it argued that gender identities at variance with sex—or what was increasingly called “sex assigned at birth”—should be understood as a matter of difference, not disorder. The campaign modeled itself on the gay rights movement’s successful campaign of the late 1960s and early 1970s to depathologize homosexuality, which led to the removal of “homosexuality” as a diagnosis from DSM-II in 1974.

The effort to depathologize non-congruent gender identities took much longer and led to a more equivocal outcome. The movement scored a symbolic success in 2013, when “gender identity disorder” was replaced by “gender dysphoria” in the fifth edition of the DSM (Hanssmann 2023: 91–92). Under this new diagnostic regime, a non-congruent gender identity by itself was no longer considered a “disorder.” Yet depathologization was limited by continued medicalization (Hanssmann 2023: 50). When a mismatched gender identity was accompanied by “clinically significant distress,” in the language of DSM-5, it continued to be construed as a disorder, under the new name “gender dysphoria.”¹³ And the increasingly consolidated subfield of transgender medicine—a subfield with its own textbooks, conferences, and professional associations—held out the promise of alleviating that distress through hormonal or surgical interventions.¹⁴

The continued medicalization of non-congruent gender identities was not just imposed from the top down by medical professionals; it was demanded from the bottom up by those who actively sought medical interventions to bring their sexed bodies into alignment with their gender identities and by advocates who deemed such interventions “medically necessary.” The result was a paradoxical combination of limited, largely symbolic depathologization and continued medicalization. The transgender movement remained—and remains—deeply entangled with the medical jurisdiction against which it initially rebelled.

The terms of medicalization, however, changed dramatically. Transgender medicine was built around and legitimated by the category “gender identity.” But the category came to be used in new ways. It was initially a category deployed and controlled by physicians and psychiatrists to guide professional decisions about appropriate treatment; it became a category deployed and controlled by patients themselves, who were increasingly empowered to make their own decisions. It was initially a category deployed by gatekeepers, used to stringently screen and select patients for treatment (just 1% of requests for surgery at the Hopkins clinic were accepted [Fritz and Mulkey 2021]); it became a category deployed against gatekeepers and gatekeeping. In this way, gender identity became, for laypeople, an actionable category: a category they could invoke to make claims for medical services.

This new understanding of gender identity as a patient-controlled, patient-empowering category of practice was codified in the paradigm of “gender-affirming” care. What is affirmed in gender-affirming care is patients’ understandings of and declarations about their own gender identity. This paradigm has been established in textbooks, clinical practices and guidelines, and standards of care documents. According to the most widely cited Standards of Care, published by the World Professional Association for Transgender Health (WPATH) and now in its eighth edition, “the goal of gender-affirming care is to partner with TGD [transgender and gender diverse] people to holistically address their social, mental, and medical health needs and well-being while respectfully affirming their gender identity.”¹⁵ And as a transgender medicine textbook observed, “by the early 2000s, identity-affirming care had become the primary orientation for specialists in the field of transgender health” (Poteat et al. 2019: 15).

In the 2000s and 2010s, the model of gender-affirming medical care was extended to children and adolescents. A pioneering Dutch clinic was exceptionally influential in both opening the door to medical treatments for gender dysphoric minors and providing longitudinal evidence of the apparent success of such treatments.¹⁶ According to the “Dutch protocol,” persistently gender dysphoric adolescents could be offered puberty blockers in the early stages of puberty to prevent development of secondary sex characteristics that could be experienced as distressing and (for natal males, then predominant among those referred to the clinics) would make it more difficult to pass should they subsequently decide to transition; the adolescents could then decide later whether to proceed to cross-sex hormones or surgery.

At a moment when the new landscape and language of gender identity was becoming increasingly salient for adolescents, the new Dutch model was quickly adopted elsewhere. But it was also transformed. The Dutch protocol had imposed strict eligibility criteria: gender dysphoria since early childhood that worsened at the onset of puberty; psychological stability; and minimum age requirements for gender-affirming hormones and surgery. These criteria were relaxed or abandoned as the model was exported to the US and elsewhere. For children and adolescents, just as for adults, gatekeeping yielded to the affirmation of asserted gender identity.¹⁷ Youth gender clinics multiplied, referrals soared, and the patient profile changed dramatically (Cass 2024: 88–97). Medical interventions for minors remain relatively uncommon, but the numbers have been growing rapidly; a recent analysis of US insurance data suggested that at least a thousand gender-affirming mastectomies were performed each year on girls under 18 in 2021 and 2022 (Sapir 2024). Gender-affirming approaches have been extended to younger children as well. For pre-pubertal children with gender dysphoria, social affirmation—sometimes involving a public social transition to names, pronouns, presentation, and activities matching one’s gender identity—has become the institutionally favored option in some contexts. These changes set the stage for the recent explosion of contestation about youth gender medicine that I take up in a subsequent section.

Law and Administration

Gender identity came to be inscribed in laws, administrative regulations, and organizational policies in connection with the legal or formal organizational regulation of sex, specifically in connection with identification and classification by sex, discrimination by sex, and segregation by sex. In a broadening range of contexts throughout the Western world, gender identity came to authorize people to change their sex classification in official documents and organizational records, to redefine the meaning of “sex” in legislative provisions banning sex discrimination, and to allow entry to single-sex spaces and activities.

This is a highly heterogeneous domain that includes laws, court rulings, administrative regulations and directives, bureaucratic forms and procedures, and organizational policies and practices in both public and private spheres. The heterogeneity is compounded in the US, where different laws and regulations exist in different jurisdictions at different levels of government; where court rulings in different parts of a complex, multi-level court system are often at variance; and where successive federal administrations have (since 2017) reversed the policies of their predecessors (and the new Trump administration, of course, has gone far beyond reversing Biden administration policies, launching an all-out assault on gender identity and transgender people). Developments in this domain are therefore much more complex than in the medical domain, and I have to simplify drastically.

Classification by sex. From its inception, the modern state has been in the business of identifying and classifying its subjects (and, later, its citizens). That classification practice has always included classification by sex; it continues to do so even as sex has lost most of its legal significance, including, recently, its significance in the regulation of marriage. Sex classifications were seldom contested before the 1960s; the rare instances of contestation concerned intersex individuals, typically in cases involving marriage, divorce, or annulment. In the latter part of the 20^th century, the availability of hormonal and surgical interventions and the social recognition of transsexualism generated more frequent demands to change sex classification and posed the question of the criteria and procedures for making such changes. The procedures that were established in the 1970s and ‘80s, however, were onerous and expensive. They often required surgical transition, letters from doctors and/or psychiatrists, and evidence of having lived in the desired gender for a certain amount of time.¹⁸

Beginning in the 1990s, the movement for depathologization, pressure from advocacy groups, the incorporation of gender identity into human rights principles and court decisions, and (by the 2000s) the greater public visibility and legitimacy of transgender people inaugurated a move toward simplifying and facilitating sex reclassification by giving greater weight to gender identity and by reducing or eliminating requirements for sex reassignment surgery, medical supervision, or trial periods. This shift culminated in the model of gender “self-identification” or “self-determination,” which grants full authority to the individual, makes subjective gender identity decisive, and allows sex reclassification through simple administrative declaration.¹⁹ This model has been adopted in about 20 countries, mainly in Europe and Latin America.²⁰

The shift toward gender self-determination can be traced in the US as well, though the story is complicated by the fragmented and polarized nature of the American polity, the lack of a centralized population registry, and the fact that different federal agencies (the Passport Agency and the Social Security Administration, for example) and different state agencies within the same state (those in charge of birth certificates and those in charge of driver’s licenses, for example) may have different policies and procedures for sex reclassification. A dozen US states follow the self-determination model in handling requests for changing sex markers on birth certificates.²¹ Rules for changing sex markers on driver’s licenses are generally more liberal; twenty-one states follow the self-identification model.²² And from 2022 to early 2025, the self-identification model governed changes of sex markers, including the possibility of choosing an X marker, on US passports.

Discrimination by sex. The most consequential inscription of gender identity into law and policy happened indirectly, via an expansive administrative and jurisprudential interpretation of what counts as sex discrimination. In the American context, courts and administrative agencies came to interpret the ban on discrimination on the basis of sex—in employment (per Title VII of Civil Rights Act of 1964) and, less consistently, in education (per Title IX of the Education Amendments of 1972)—as also forbidding discrimination on the basis of gender identity.

The reinterpretation of civil rights-era provisions barring sex discrimination was a roundabout way of achieving what activists initially sought to achieve more directly. Sixteen US states—as well as many cities and counties—had adopted anti-discrimination measures covering gender identity by 2012 (Taylor et al. 2018: 237). But the failure of efforts to enact similar provisions at the federal level through the Employment Non-Discrimination Act (and, in the last decade, through the broader Equality Act) led activists to turn to courts and administrative agencies, seeking a broad, trans-inclusive reading of Title VII’s prohibition of sex discrimination (Lee 2012: 425–6).

Their initial successes in the courts, beginning in the early 2000s, were built on the idea of gender nonconformity. An influential 1989 Supreme Court case, Price Waterhouse v. Hopkins, established that sex discrimination includes discrimination on the basis of nonconformity with sex-based stereotypes. Extending this argument to the transgender domain, courts began to treat discrimination against transgender plaintiffs as discrimination based on their failure to conform to stereotypes associated with their natal sex. This success, however, was deeply equivocal: transgender persons were protected against discrimination, but their gender identity was not recognized; they were treated instead as gender-nonconforming members of their birth sex (Lee 2012: 25; Schoenbaum 2020: 27).²³

The Supreme Court’s surprising 2020 decision in Bostock v. Clayton County, Georgia took a different tack. Holding that Title VII’s prohibition of discrimination “because of …sex” extended to discrimination on the basis of gender identity and sexual orientation, it did not draw on the argument about nonconformity with sex-based stereotypes. Instead, it argued on “textualist” grounds that one cannot discriminate on the basis of gender identity or sexual orientation without discriminating on the basis of sex: “An employer who fires an individual for being homosexual or transgender fires that person for traits or actions it would not have questioned in members of a different sex. Sex plays a necessary and undisguisable role in the decision, exactly what Title VII forbids.”²⁴

Segregation by sex. Anti-discrimination law in the domain of education brought into focus a new set of questions that did not arise in the domain of employment. In banning discrimination by sex, Title IX allowed single-sex institutions as well as single-sex housing and athletic teams; regulations adopted a few years later expressly authorized single-sex bathrooms and locker rooms (Melnick 2018b: 231). Courts and administrative agencies have therefore had to determine who would have access to these sex-segregated activities and facilities. To do so, they have had to directly address the question of the meaning of “sex” and its relation to gender identity, a question courts were able to sidestep in Title VII jurisprudence. Courts and administrative agencies have thus been drawn directly into making formal pronouncements about the authoritative meaning of one of the most fundamental terms of human social life.

In May 2016, the Department of Education’s Office for Civil Rights, together with the Department of Justice’s Civil Rights Division, sent a “Dear Colleague” letter to schools and colleges nationwide “summariz[ing] a school’s Title IX obligations regarding transgender students.” The letter announced that in interpreting Title IX, the two federal agencies “treat a student’s gender identity as the student’s sex.” To comply with their Title IX obligations under this interpretation of the law—and to be eligible for federal funding—schools would have to allow students access to all sex-segregated facilities and activities (with the partial exception of single-sex sports teams) “consistent with their gender identity” (US Department of Justice 2016).

The Dear Colleague letter marked the end of the phase of quiet, behind-the-scenes institutional embedding of gender identity in the administrative sphere. Its nationwide reach, coupled with the threat of powerful sanctions, provoked widespread opposition. The first Trump administration revoked the letter; the Biden administration reasserted the principle of treating gender identity as sex in interpreting Title IX; and the new Trump administration once again repudiated that principle.²⁵ Meanwhile, court cases concerning transgender students’ access to bathrooms corresponding to their gender identity have been accumulating, and federal circuit courts have split on the issue. Since Bostock, courts have also split on the applicability of that decision to other questions (especially access to sex-segregated activities and facilities) and to domains beyond employment. The circuit split makes it likely that the Supreme Court will intervene to resolve the dispute (Degler 2022).

***

The embedding of the category “gender identity” in a dense, mutually cross-referencing network of laws, administrative rules and guidelines, court rulings, and organizational policies made it an actionable category, one that could be used to change one’s sex classification in government documents and organizational records, to file harassment claims and discrimination lawsuits, and to gain access to sex-specific spaces, activities, and facilities. Administrative rules and guidelines, to be sure, are vulnerable to changes in administration, as was shown at the federal level by sharp changes in 2017 and 2021 and by a much more radical break in 2025.²⁶ Still, gender identity remains powerfully encoded in federal court rulings (especially the Supreme Court ruling in Bostock), in legislative provisions and administrative rules in blue states, and in the policies and procedures of many schools, universities, businesses, and other organizations.

Statistics

In the domain of statistics, gender identity has been embedded in a growing array of governmental and nongovernmental data-gathering undertakings, ranging from nationwide representative surveys to internal administrative data-gathering by schools, universities, employers, and other organizations. Data on sex have long been gathered routinely, and in recent decades the term “sex” has often, though not always, been replaced with “gender.” But what I want to highlight here is not this terminological shift; it is a more fundamental conceptual and political shift. This has two aspects. The first is the shift from asking a single question about sex (or gender) to asking a pair of questions, one about natal sex, the other about gender identity; the second is the treatment of gender identity as more fundamental than sex in social practices of counting and accounting.

The shift to a two-part question format responded to growing demands for the statistical inclusion of gender and sexual minorities.²⁷ As in the case of other demands for statistical inclusion, to be counted was to count, publicly, as a legitimate, visible component of the population. Being counted was also a presupposition for documenting and addressing inequalities of treatment and condition experienced by gender and sexual minorities: social exclusion could be quantified only through statistical inclusion.²⁸

The statistical inclusion of transgender people, according to a growing social science consensus, required a two-question format. Initial attempts to use a single question—to ask, for example, if the respondent was male, female, or transgender, sometimes with an additional category to accommodate non-binary people who might not identify as transgender—were seen as inadequate, since they could not capture trans people who identified simply as male or female rather than as “transgender.”²⁹ For this reason, the two-question format was strongly recommended in a major 2022 report by the National Academies of Sciences, Engineering, and Medicine (Bates et al. 2022). It was adopted in the most recent censuses by Canada and New Zealand and by several major surveys in the US, including the National Crime Victimization Survey, the General Social Survey, the Behavioral Risk Factor Surveillance System, and the Census Bureau’s Household Pulse Survey (ibid: 111, 114ff; Centers for Disease Control n.d.). The Census Bureau even tested separate gender identity (and sexual orientation) questions for possible inclusion in its large-scale monthly American Community Survey (US Census Bureau 2024). The two-question format is also widely used by colleges and universities and by some other organizations in gathering data for their own administrative purposes. The Common App, for example, used by applicants to more than a thousand colleges and universities, has included separate questions for sex and gender identity since 2016 (Beemyn et al. 2023).

The routine inclusion of gender identity in surveys and administrative data-gathering is a powerful means of institutionalization. It universalizes gender identity by casting it as something that everybody has, rather than something relevant only to a few. It recognizes new forms of gender difference and maps out a space of legitimate gender variation by placing all gender identities side-by-side and on the same level. It confers official visibility—and therefore public existence as a recognized component of the population—on gender minorities. And it establishes gender identity as a basic demographic category and, in Bourdieusian terms, a fundamental principle of vision and division of the social world (Bourdieu 1991b: 232). In short, it bestows symbolic power on gender identity.

Gender identity has not only been included alongside sex; it has been displacing sex as the primary datum—and the primary principle of vision and division—in a range of contexts (Munzer et al. 2025). In the Anglophone world, a “gender by default” principle has been gaining ground. According to this principle, when data on both sex and gender identity are collected—as is the case when the two-question format is used—reports should use data on gender identity, not data on sex, unless there is a specific reason for using sex data. This principle has been articulated most clearly by statistical agencies in Canada and New Zealand, but it has been gaining ground in the US as well. It was endorsed by the National Academies report mentioned above.³⁰ The Common App similarly suggests that gender identity, rather than sex, should be “visible in campus information systems,” and that sex should be made available only to campus offices such as health centers that may have specific need for it (Beemyn et al. 2023). And the federal Equal Employment Opportunity Commission (EEOC), which had long required all large employers (those with more than a hundred employees) to report the demographic composition of their workforce by race and ethnicity and by sex for the purpose of enforcing anti-discrimination legislation, indicated in 2023 that “if the sex [i.e. gender identity] reported by an employee during voluntary self-identification differs from the sex recorded in the employee’s employment records, the employer should report the former (i.e., self-identification) and not the latter (i.e., employment records)” (EEOC 2023).

Pedagogy

In the domain of pedagogy, gender identity has been embedded both in formal educational routines—in school curricula and adult training programs—and in informal campaigns to change the way people think about sex and gender.

The decentralized nature of the American education system—which allows for variation not only between states but between districts, schools, and even individual classrooms—makes the embedding of gender identity in education difficult to study and impossible to quantify. But a few observations can be made about the substantial corpus of resources that are available for teaching gender identity in the classroom—stories, videos, graphics, activities, and lesson plans, tailored to different age groups from kindergarten on—and about the embedding of gender identity in sex education standards.

Comprehensive and organized sets of teaching resources have been developed and made easily accessible free of charge by at least two major advocacy organizations. The Welcoming Schools Program of the Human Rights Campaign Foundation, the research and public education counterpart of the lobbying-oriented Human Rights Campaign, the largest American LGBTQ advocacy organization, offers some two dozen “teacher-friendly lesson plans… designed to help educators create classrooms and schools that are free of gender stereotypes and gender norms that limit all children” and to “affirm all students’ identities across the gender spectrum.”³¹ The lesson plans, aligned with Common Core educational standards, are geared to various elementary school grade groupings, starting with K-2.

Detailed lesson plans on gender identity are also available from Advocates for Youth as part of that organization’s comprehensive K-12 sexuality education curriculum. Advocates for Youth was one of three organizations partnering to develop the second edition of the National Sex Education Standards. The second edition of the Standards, published in 2020, gives far greater prominence to gender identity than the 2012 first edition did. The term appears 71 times in the second edition, but only 12 times in the first edition. And while gender identity figured in specific grade-level standards starting in grades 6–8 in the first edition, it figures in standards starting with grades K-2 in the second edition. By the end of grade 2, students should be able to “define gender, gender identity and gender-role stereotypes,” “discuss the range of ways people express their gender and how gender-role stereotypes may limit behavior,” and “demonstrate ways to treat people of all genders, gender expressions, and gender identities with dignity and respect.” By the end of fifth grade, students should be able to “distinguish between sex assigned at birth and gender identity and explain how they may or may not differ,” “define and explain differences between cisgender, transgender, gender nonbinary, gender expansive, and gender identity,” and “explain that gender expression and gender identity exist along a spectrum.”³²

An Advocates for Youth lesson designed for first grade introduces the concept of gender identity by suggesting the following script for the teacher:

Identity starts with an I. That’s how you can remember it. “I” feel, “I” know. Gender identity is that feeling of knowing your gender. You might feel like you are a boy, you might feel like you are a girl. You might feel like you’re a boy even if you have body parts that some people might tell you are “girl” parts. You might feel like you’re a girl even if you have body parts that some people might tell you are “boy” parts. And you might not feel like you’re a boy or a girl, but you’re a little bit of both. No matter how you feel, you’re perfectly normal!³³

Another widely used resource is the “Gender Snowperson” and its closely related variants, the “Genderbread Person” and the “Gender Unicorn.” A Welcoming Schools lesson plan for grades 3–8 uses the Gender Snowperson to distinguish gender identity (“who you are and how you feel as a person”) from sex assigned at birth, sexual orientation (“who you love or are attracted to”), and gender expression (“one of the many ways that we can show who we are through clothing, hair and mannerisms”). These concepts are “not binaries but spectrums.” Gender identity, for example, is “your internal sense of being a girl, boy, both or neither.” The concepts are used to define cisgender, transgender, and non-binary and to underscore that one cannot know a person’s gender identity or their sexual orientation from their gender expression.³⁴

Many lesson plans are built around children’s books that illustrate these points. A popular elementary school book—so popular that it became notorious in conservative circles and was listed by the American Library Association as among the top 15 most banned or challenged books of the 2010s³⁵—is I am Jazz, written by transgender media figure Jazz Jennings and coauthor Jessica Herthel. Jennings came to public attention when she was interviewed by Barbara Walters at age 6; she has starred in the reality TV series I am Jazz since 2015. The book tells Jennings’s own story in the first person. She describes herself as having “a girl brain but a boy body. This is called transgender. I was born this way!” As she got older, she elaborates, she “hardly ever played with trucks or tools or superheroes. Only princesses and mermaid costumes. My brothers told me this was girl stuff…. My sister says I was always talking to her about my girl thoughts, and my girl dreams, and how one day I would be a beauuuutiful lady.” The story’s narrative arc moves from Jennings’s unhappiness at having to wear “boy clothes” and “pretend” she was a boy outside the house—which “felt like telling a lie”—through medical validation and parental acceptance of her transgender status, to the freedom to be herself and to be treated as a girl at school (Herthel and Jennings 2014).

As in the domain of statistics, but in a more personal and concrete manner, these and similar pedagogic materials serve to universalize gender identity and make it relevant to all children, not just to a few. They often do so by expressly stating, as the Gender Snowperson lesson plan does, that “we ALL have a gender identity” and by inviting students to reflect on their own gender identity. Again, as in the domain of statistics, these materials map out a space of legitimate variation and place all forms of gender identity on the same level, and they grant recognition and visibility to various forms of gender difference.

Outside the classroom, gender identity is embedded in workplace training routines in some states. In California, for example, large employers must include training on harassment based on gender identity, gender expression, and sexual orientation in their state-mandated sexual harassment training. This requirement brings employers into the business of educating their supervisory employees about gender identity as something distinct from both natal sex and gender expression. Even where such training is not expressly required, employers in states with broad definitions of sexual harassment or sex discrimination that include harassment or discrimination on the basis of gender identity may provide such training as a way of avoiding potential legal liability.

Beyond formal education and training, gender identity has figured in less formal educational efforts. Some of these are organization-driven and may combine formal and informal education. Advocates for Youth’s Amaze program, for example, has produced animated educational videos on gender identity as part of a broader portfolio of sex education videos.³⁶ These are aimed on the one hand at teachers, and they can be incorporated into classroom use, but they are also aimed directly at adolescents and their parents through a strategy of “harnessing the power of digital media” and providing “free, engaging resources that can be accessed anytime, anywhere.”³⁷ Informal education also includes books marketed directly to the public—Barnes & Noble lists 142 books, many of which address gender identity, under the heading “LGBTQIA+ Children’s Books”—and journalistic explanations of gender identity.

But online platforms undoubtedly provide by far the most important channel of informal education about gender identity. Young people are summoned by prevailing cultural scripts to reflect on their gender identity, in tandem with reflection on their sexual orientation. Online platforms provide both the channels through which that summons is conveyed and an inexhaustible resource to which young people can turn for information, guidance, exploration, support, and community as they seek to locate themselves in this churning and often confusing category space. The new digital ecology of communication includes highly interactive spaces like Tumblr, where people can work out their own gender identities in interaction with anonymous others by locating themselves in the context of a dynamically evolving, collectively produced vernacular taxonomy of gender and sexual identity categories (Cover 2018; Garrison 2022; Amin 2023). It also includes platforms that allow users to broadcast materials to networks of friends (on Facebook), followers (on YouTube or Instagram), or algorithmically assembled audiences (on TikTok) and thereby to become influencers (that is, educators), while enabling other users to search for relevant material or relevant influencers to follow. Platforms have not only served as conduits for such informal peer-to-peer education about gender identity, they have themselves adopted and promoted the category: in 2014, Facebook introduced 56 “custom” gender options to “help you better express your own identity on Facebook” (Facebook 2014). Dating sites and other social media platforms followed suit.

A Quiet Revolution

The developments sketched in the previous sections amounted to a remarkably rapid establishment in core institutional domains of an entirely new principle of vision and division of the social world, one that could be understood as competing with, redefining, or displacing sex. Since sex is, in historical and cross-cultural perspective, arguably the most fundamental and universal organizing category of social life, this is no small development. Just as remarkably, this basic change in conceptual infrastructure was not contested in a sustained way until relatively recently; earlier critiques by some radical feminists and cultural and religious conservatives, especially the Catholic Church, did not gain broad public traction.³⁸ The intense public controversy about gender identity in recent years makes it easy to forget how little controversy accompanied the embedding of gender identity in each of these domains. Gender identity was embedded in professional practices, organizational routines, and institutionalized procedures not only without public debate, but without much public attention of any kind. How did this happen?

My argument in this section is that the behind-the-scenes, under-the-radar institutional embedding of gender identity resembles the dynamics of what John Skrentny has called the “minority rights revolution,” of which it was a belated part. What was striking, for Skrentny, about the rapid legislative, administrative, and judicial shifts that established a wide range of nondiscrimination rights not only for African-Americans but for Latinos, Asian Americans, Native Americans, immigrants, women, and the disabled between 1965 and 1975 was that it was not driven, for the most part, by clamorous, disruptive, bottom-up protests applying pressure from the outside. The outside pressure exerted by the civil rights movement had of course been decisive in the passage of the landmark Civil Rights Act of 1964. But the subsequent burst of legislative and administrative activity was a remarkably bipartisan and, for the most part, non-contentious development in which social movement representatives worked from inside Congress, administrative agencies, and the White House, especially in the new “institutional homes” established as legacies of the Black civil rights movement, notably the Equal Employment Opportunity Commission (EEOC), the Department of Health, Education, and Welfare’s (later the Department of Education’s) Office for Civil Rights (OCR), and the Department of Labor’s Office of Federal Contract Compliance. These agencies were sites not only of rights enforcement but also of rights advocacy and extension, and they allowed social movements to “concentrate their lobbying efforts to a sometimes very receptive audience, usually out of public view.” The story of the minority rights revolution thus turns not on “angry minority protests, raised fists, picket lines, and placards” but on people “wearing suits, sitting at desks, firing off memos, and meeting in government buildings to discuss new policy directions.”³⁹

This captures well the gradual, low-profile institutional embedding of gender identity up until at least 2015 (and well beyond that date in contexts that remained sheltered from public view and insulated from public pressure). The pattern holds not only in the legal and bureaucratic domain, on which Skrentny’s account focuses, but also in medicine, data-gathering, and pedagogy. There was no mass mobilization: there were no masses to mobilize. There was little noisy public protest; LGBT organizations mobilized behind the scenes. Insider activism was more important than outsider protest.

In the legal and bureaucratic domain, mobilization took the form of lobbying sympathetic administrations (and key agencies) at the federal and state levels for actions that could be taken by executive order or by shifts in the enforcement posture and regulatory interpretations of administrative agencies. (The focus on administrative agencies at the federal level, as noted above, emerged as a second-best strategy after the failure of efforts to pass federal legislation barring discrimination on the basis of gender identity along with sexual orientation.) Lobbying was directed primarily at the Obama administration, since it was in office when demands for change were gathering force and it was seen as receptive (Taylor et al. 2018: 164, 167). The Human Rights Campaign (HRC), for example, the major American LGBT civil rights organization, provided the incoming Obama administration with a “blueprint for positive change” in 2009 and noted with satisfaction the many small steps the administration took to implement incremental changes (Taylor et al. 2018: 164), such as “add[ing] gender identity to the categories protected under the equal employment opportunity policy for Executive Branch positions” (Human Rights Campaign 2011). The Obama administration, as well as various state governments, took many such small steps (Taylor et al. 2018; Chapter 6). A series of inconspicuous actions deep in the bowels of the bureaucracy, for example, ended up making it much easier for transgender people to change their sex on government documents to match their gender identity (Eilperin 2015).

The 2011 policy shift was one small part of a large change that ended up redefining the civil rights-era legislative ban on discrimination on the basis of sex to include discrimination on the basis of gender identity. But that change happened in piecemeal fashion, largely hidden from public view, through a process of what political scientist Shep Melnick has called “institutional leapfrogging,” in which “judges and administrators take a long series of small steps, with each claiming to rely on the authority of the other” (Melnick 2018a).⁴⁰

The culmination of this leapfrogging process, in the domain of education, was the OCR’s “Dear Colleague” letter of May 2016, mentioned above, which specified that Title IX’s ban on sex discrimination required schools and colleges to allow students access to sex-segregated facilities “consistent with their gender identity.” This immediately ignited controversy; in connection with other controversies emerging around the same time, it signaled the beginning of the end of the era of quiet, uncontroversial institutional embedding. What I want to underscore here, however, is Melnick’s point that in the span of just a few years, OCR had reversed its longstanding interpretation of Title IX’s ban on sex discrimination, without any public awareness or debate, and that it had done so in a self-effacing manner that denied the novelty of the shift. The story is far too complex to relate in detail, but one particularly striking bit of leapfrogging or circular referencing is worth noting. The Dear Colleague letter claimed it was simply clarifying existing law, not introducing any new obligations. In support, it cited a federal circuit court case that had found that a school had to permit a transgender student to use the bathroom corresponding to his gender identity. But that court decision rested in crucial part on the principle that courts should defer to administrative agencies’ interpretations of their own rules, and it cited a statement of OCR’s position—that in determining access to sex-segregated facilities, schools must “treat transgender students consistent with their gender identity”—that had been conveyed in a private letter responding to an email inquiry from an attorney and transgender activist who had been following the developing school controversy.⁴¹ A rapidly evolving agency position—spelled out for the first time in private correspondence—was taken as the key basis for a court decision that was, in turn, cited as a precedent for the agency’s public articulation of its new position. This allowed OCR to wrap a fundamental change in the cloak of continuity.

Gender identity was quietly yet pervasively adopted in private sector antidiscrimination provisions as well. By 2023, 90% of Fortune 500 companies expressly included gender identity in their internal nondiscrimination policies, and 73% offered transition-related health coverage (Human Rights Campaign Foundation 2023). Corporate adoption of these and other LGBT-inclusive policies and practices has been measured in the Corporate Equality Index (CEI), published by the Human Rights Campaign since 2002.⁴² The CEI offers reputational benefits to firms earning high or perfect scores; in touting those benefits and noting that transgender-inclusive policies are “necessary for talent acquisition and retention,” the HRC pursues what has been called the “politics of alignment,” a non-contentious mode of social movement activism in which activists seek to align their goals with those of their targets and make a business case for adopting their favored policies.⁴³

In the medical domain, the establishment of gender-affirming care—care designed to affirm an individual’s gender identity—likewise proceeded entirely outside the glare of public attention. The paradigm of gender-affirming care established its authority gradually, as it came to be endorsed by a growing number of professional associations as well as advocacy groups. A key role in bringing professional associations on board was played by the World Professional Association for Transgender Health (WPATH), a hybrid organization that bridges the worlds of activism and professional medical practice. In 2024, WPATH became embroiled in a scandal when internal documents made public during court proceedings revealed blatant political interference in the formulation of its clinical guidelines (Ghorayshi 2024) as well as the organization’s interference with the publication of what were supposed to be independent “systematic reviews” of the evidence concerning different forms of transgender medicine (Singal 2024). Before these revelations, however, WPATH enjoyed great authority and deference in the world of transgender health. Its Standards of Care occupied a central position in the network of clinical guidelines published by a wide range of professional associations, all of which cite other guidelines to buttress their authority. In a manner partly analogous to the institutional leapfrogging in the legal and administrative domain, WPATH’s latest (2022) version of its Standards of Care repeatedly cited several other guidelines to support its recommendations, guidelines that were themselves heavily influenced by the previous (2012) version of the WPATH Standards of Care (Taylor et al. 2024: s69-s70). This circularity helped create the appearance of a medical consensus regarding “best practices” despite an extremely thin evidentiary base (Cass 2024: 130; Buttons 2024).

Even the extension of the paradigm to children and adolescents happened outside the limelight; public discussion and controversy lagged by more than a decade. The paradigm of gender-affirming care for youth was long presented—and is often still presented today, even after becoming the focus of public controversy—as the fully consensual view of the medical establishment, supported by abundant evidence and settled science, not as a novel form of treatment about which caution might be warranted and for which further evidence might be needed. The establishment of identity-affirming care for youth was a low-profile, quiet, uncontroversial affair; it became a high-profile focus of public controversy only when critics eventually challenged the quiet revolution that had already occurred.

In the domain of data-gathering, too, gender identity was incorporated into institutional routines with almost no public discussion. And here too advocacy took the form of behind-the-scenes lobbying, not public claims-making. Like other groups seeking statistical inclusion and visibility, LGBT organizations urged census officials, survey organizations, universities and colleges, and business firms to routinely collect data on gender identity as well as sexual orientation, and they elaborated “best practices” for doing so. But insider activism—in the form of initiatives from inside the Census Bureau, other data-gathering agencies, universities, and firms to collect such data—appears to have been at least as important as advocacy group pressures.

A similar story, finally, can be told about the introduction of gender identity in the pedagogic domain. LGBT advocacy organizations developed and promoted a wide range of resources for classroom use, participated in drawing up a new version of the National Sex Education Standards, and sought to “educat[e] journalists and Hollywood” to help them “respectfully and accurately write about transgender people” (GLAAD. n.d.). But all of this activism was of the quiet, behind-the-scenes variety. As in the medical domain, the incorporation of gender identity in the domain of pedagogy was a low-profile, uncontroversial affair that became contentious only after a considerable lag.

Flashpoints of Contention

During the phase of quiet institutional embedding, there were of course, as I have noted, some critics of gender identity, including notably radical feminists and cultural and religious conservatives. But the issue was largely invisible to a broader public, and it did not figure in partisan politics at the national level. As late as 2016, it played almost no role in Trump’s first presidential campaign.⁴⁴ Yet since the middle of the last decade, and especially since the beginning of the present decade, gender identity has emerged as a major focus of public controversy. What changed? Why did noncontroversial, behind-the-scenes institutional embedding yield to clamorous public contention?

Numerous scholars and commentators, with good reason, have described the controversy as having been fueled by a “moral panic” (Slothouber 2020; Pepin-Neff 2023; Amery and Mondon 2024). Goode and Ben-Yehuda (1994: 156–159) defined moral panics in terms of five criteria: a heightened concern about a putative threat, manifested inter alia through media attention and proposed legislation; increased hostility toward those seen as responsible for the threat; widespread agreement, at least in certain segments of society, that the threat is real and serious; disproportionate or exaggerated representations of the threat; and volatility, characterized by a relatively sudden eruption of concern. Each of these clearly applies to the explosion of controversy about gender identity. The most striking indicator is the burst of proposed and enacted legislation. In February 2025, the ACLU was tracking some 400 pieces of proposed state legislation that they characterized as targeting LGBTQ rights (ACLU 2025),⁴⁵ most of them related in some way to gender identity. These bills propose to restrict gender-affirming health care, especially for minors; limit access to facilities like bathrooms and locker rooms and activities like school sports; restrict what can be taught in classrooms; limit the ability to update identity documents; and define “sex” to mean biological sex. Only a fraction of these bills has passed or will pass, and some that have passed are tied up in court challenges. Nonetheless, the wave of proposed legislation, beginning in 2015 and building after 2020, has led to the passage, according to one tracking site, of a couple hundred anti-trans bills, with many more no doubt to follow.⁴⁶ And of course Trump’s initial blast of executive orders both fed off of and greatly amplified the moral panic.⁴⁷

The intensifying controversy over gender identity can be understood, in part, within the context of a broader transnational conservative network of campaigns and movements in the domain of gender and sexuality over the last two decades. Although highly heterogeneous and sometimes in tension with one another, these campaigns have been discursively and symbolically united by their opposition to what has been called—in Europe and Latin America, though not, until recently, in the United States—“gender ideology” (Case 2019; Paternotte 2023). Campaigns against “gender ideology” in Europe and Latin America have not focused narrowly or even primarily on gender identity or transgender issues; they have addressed a broad range of issues including same-sex marriage, abortion, reproductive technologies, divorce, declining fertility, family structure, and education about sexuality, many of which have been addressed by American conservatives under the rubric of “family values.” The American right’s use of the term in recent years, by contrast, has focused on transgender issues. Yet the trans-focused assault on “extremist gender ideology” by the new Trump administration is no doubt linked to the resurgent neo-masculinist ethos of the right and its critique of the feminization of American institutions (Zerofsky 2025) as well as to broader critiques of progressive gender norms (Bulman 2024; Brechenmacher 2025).

Characterizing the controversy over gender identity as a moral panic, however, and situating it in the context of a broader conservative politics of gender and sexuality cannot explain why the controversy emerged when it did after an extended phase of quiet institutional embedding. The controversies that flared up over bathrooms and locker rooms, competitive sports, pronoun policies, classroom pedagogy, social transition in schools, and youth gender medicine were not conjured up out of nothing; they responded to something new. Gender identity had previously been understood as something that concerned a small minority, and them alone; it had been understood as a self-regarding matter. In the second half of the last decade, however, it ceased to be understood as self-regarding; increasingly, it now seemed to concern others. And this shift in public understanding, in turn, responded to substantial changes in the work done by the category “gender identity.”

The “bathroom wars” that began in earnest in 2015 were responding to the legislative and administrative codification of the principle that access to bathrooms, locker rooms, and other intimate facilities should be based on gender identity. This codification was enacted in state laws (notably in California in 2013), local ordinances (such as the Charlotte, NC ordinance that provoked North Carolina’s notorious statewide “bathroom bill”), and administrative rulings, most crucially the 2016 Dear Colleague letter, discussed above, from the Department of Education’s Office for Civil Rights, which imposed this principle on all schools in the nation. The letter was in effect only briefly, but it served notice that questions of access to bathrooms were no longer to be worked out quietly on a case-to-case basis, according to local circumstances; they were now codified across the board as a matter of binding administrative regulation. This codification was new, and it implicated others, not just the small minority for whose benefit the principle was codified. The change could be seen as affecting others’ interests, notably their interest in privacy, as this had long been understood and taken for granted. It was this wave of codification that provoked what can fairly be characterized as a moral panic (or, as Schilt and Westbrook [2015] put it, a “penis panic”).

The controversy over competitive athletics has a longer history. The International Olympic Committee and other bodies have long had to formulate rules and (often stigmatizing) procedures for determining the eligibility of persons with sex development variations and, in recent decades, for transgender athletes. But widespread public concern flared up here too in response to newly codified policies adopted by some states (including California in 2013) that allowed students to play on sex-segregated teams consistent with their gender identity. It also responded to stories about the competitive advantage enjoyed by elite transgender athletes such as runner Terry Miller or swimmer Lia Thomas (Hobson 2021). Like the use of sex-segregated facilities, participation in competitive sports in accordance with gender identity implicated others, and it could readily be understood as threatening their interests in fair competition (Coleman 2017).

Controversy over pronouns emerged in response to new informal norms and expectations—and especially new formal rules and laws—that request or require the use of people’s preferred pronouns as an aspect of respecting their gender identity (or preventing discrimination or harassment on the basis of that identity).⁴⁸ Unlike a change in dress or grooming, a change in pronouns cannot be accomplished on one’s own; it requires the cooperation of others.⁴⁹ Freedom of expression is one thing; the request for recognition, especially when backed by the force of law or organizational policy, is another. Being referred to by one’s preferred pronouns is a matter of recognition from others, not of self-expression. While rules basing access to bathrooms and sex-segregated competitive sports on gender identity can be construed as affecting others’ interests, rules requesting or requiring people to use preferred pronouns impinge on others’ behavior; both kinds of rules bring gender identity out of the sphere of self-regarding action.

Heightened concerns about the teaching of gender identity in schools arose in part through parents’ personal knowledge of teaching materials that they found age-inappropriate or otherwise objectionable. In the new digital media environment that had become well-established by the middle of the last decade, however, far more parents encountered such materials only indirectly, through conservative social media sites such as Libs of TikTok.⁵⁰ Yet such fear- and outrage-mongering on social media can be powerfully effective: one can show the offensive materials, rather than simply telling others about them (Brubaker 2022: 139–140). Even if such fears were mobilized from above—and greatly exaggerated through selective attention to unrepresentative instances—policy entrepreneurs like Chris Rufo could point to new practices in schools that many parents found disturbing.⁵¹ They could highlight not only new materials being used in classrooms, but also the fact that some schools had adopted policies of affirming students’ gender identity by aiding social transitions at school and—when students felt their parents would not be supportive of their transition—concealing this from the parents or using a student’s preferred name and pronouns even against the parents’ explicit objections.⁵² So it is not surprising that schools relatively suddenly became a focal point of heightened and of course often exaggerated concern: a concern about curricular indoctrination on the one hand and about aiding and abetting transition without parents’ knowledge or against their will on the other.

The most powerful spur to public controversy was provided by an accumulating series of changes in the practice of youth gender medicine: the surge in referrals to youth gender identity clinics; the increasing prevalence, among adolescents referred, of natal girls with no history of childhood gender dysphoria; the high incidence of serious mental health problems among those referred, notably autism spectrum disorder, depression, and anxiety; and the greatly relaxed criteria for providing puberty blockers or gender-affirming hormones or surgery to young people. The shift toward a more “affirming” stance, increasingly skeptical of psychological assessment and exploratory therapy, meant that an adolescent’s assertion of a gender identity at variance with their natal sex was now much more likely than before to lead to medicalization. The novelty of the treatments and the lack of data about their long-term safety and efficacy raised questions about safety, life-long medicalization, and bodily integrity, while the growing numbers of young people identifying as trans and the salience of adolescent-onset gender dysphoria raised questions about social contagion. These questions, touching parents’ deep concerns and fears, were eminently conducive to a moral panic.

Still, the growing controversy about youth gender medicine cannot be reduced to a moral panic. The fact that the issue has been blatantly instrumentalized and cruelly exploited by the right during the last five years or so at the state level and now at the national level as well should not blind us to the emergence of critiques of youth gender medicine from the liberal center during the same period. What the “moral panic” characterization misses is, inter alia, the fact that several northern European countries—including all four Scandinavian countries—have adopted a much more cautious and restrictive stance on youth gender medicine in recent years, following systematic independent reviews showing the evidence for the safety and efficacy of medical interventions to be extremely weak.⁵³ Nor is there strong evidence that young people with gender dysphoria are uniquely prone to suicide or that treatment decreases suicidality.⁵⁴

Critics also challenge the idea that gender identity is a deep and enduring aspect of the self, reliably knowable by even young children and adolescents. Data on “desistance” indicate that most children diagnosed with gender dysphoria (or what was previously called gender identity disorder) will outgrow their dysphoria without social transition or medical intervention.⁵⁵ More generally, the assumption of a deep and immutable gender identity is in tension with a growing recognition of the dynamic and changing nature of the ways in which people understand their sexed and gendered selves. Critics therefore endorse “watchful waiting” or exploratory therapy rather than immediate affirmation.⁵⁶

Critics see the increased prevalence of natal girls presenting with adolescent-onset gender dysphoria as further reason for a cautious, therapy-first approach. It is also reason for suspecting that peer influence, social media, and available narratives may play a significant role in shaping how adolescents interpret distress related to their changing bodies, emergent sexualities, and gendered expectations. Lisa Littman’s (2018) hypothesis that such influences contributed to clusters of “rapid onset gender dysphoria” was rejected out of hand by youth gender medicine advocates.⁵⁷ Yet dismissing social influences on young people’s understandings of who they are seems willfully unsociological.

Critics of the model of gender-affirming care do not categorically oppose social or medical transition for youth; they acknowledge that transition may be appropriate in some cases and may help to relieve longstanding suffering. But they believe that declarations of gender identity, taken on their own, are too flimsy a basis for highly consequential decisions about social transition, puberty blockers, hormones, or surgery. They plead for caution and careful assessment, as well as for addressing any underlying mental health issues, before any decisions are made about potentially irreversible interventions whose efficacy and long-term side effects remain largely unknown. Many youth gender medicine advocates continue to argue against medical gatekeeping; some have expressly argued, for example, that no psychological assessment should be required (Ashley 2019). But critics defend medical gatekeeping for children and adolescents, and a fortiori for those suffering from serious mental health issues, who may not be able to fully appreciate the ramifications of decisions with potential life-long consequences, concerning for example fertility, sexual function, or the need for long-term medical treatment (Baron and Dierckxsens 2022).

* * *

The relatively sudden shift from quiet and uncontroversial institutional embedding to fierce public controversy, in sum, responded to a sharp underlying shift in public understandings of gender identity. To the extent public attention had been paid to it at all, gender identity—specifically, a gender identity at variance with one’s natal sex—had long been understood as a private, self-regarding matter affecting only a small number of people. Even the significantly increased visibility of transgender people after the turn of the century did not initially alter the basic understanding that how transgender people chose to live their lives—through whatever changes in dress, adornment, presentation, bodily contours, and formal identity documents they might adopt—was a matter that concerned only themselves.

Beginning around the middle of the last decade, this changed dramatically. As declared gender identity became an actionable category in a broadening range of contexts —a rights-conferring legal and administrative category; an entry ticket for sex-segregated athletic competitions, locker rooms, bathrooms, dormitories, and prison facilities; a claim for puberty blockers, hormones, or surgeries on demand, even for young people; a sufficient reason for a school to support a student’s social transition, without parents’ knowledge or against their wishes; a claim to be referred to using one’s chosen pronouns; and a pedagogic category used to encourage children to consider the possibility that their “assigned” sex might not match up with how they feel on the inside—it ceased to be understood as a self-regarding category. Claims and practices based on gender identity were now seen as potentially encroaching on others’ interests in privacy and fairness, compelling others to alter their language practices, exposing others to potential “contagion” or “indoctrination,” and threatening parents’ interests in knowing fundamental things about their children’s lives and protecting their children from potentially irreversible harm. Conservative groups and politicians greatly exaggerated these threats, presenting extreme examples as representative in an effort to cultivate and mobilize fears; this warrants the “moral panic” characterization. But that characterization should not blind us to the emergence of new and non-trivial grounds for concern, nor of nuanced critiques that cannot be simply attributed to a moral panic. The explosion of controversy, in other words, did not simply reflect the gradual intensification of a long-standing conservative critique or the increasing visibility of what was previously hidden from public view; it responded to a substantial change in what could be done with the category “gender identity.”

Conclusion

A quarter century ago, Fred Cooper and I argued that “identity” tended to mean too much (when understood in a strong sense), too little (when understood in a weak sense), or nothing at all (because of its sheer ambiguity) (Brubaker and Cooper 2000: 1, 10ff). Our article was concerned with “identity” as a category of analysis, not as a category of practice, and it mentioned gender only in passing. But our observations about the tension between strong and weak understandings of the term are centrally relevant to the career of gender identity as a category of practice.

When gender identity was first established as a category of medical and psychiatric practice in the middle of the twentieth century, in connection with intersex conditions and transsexualism, it was understood in a strong sense as a deep, abiding, foundational self-understanding that was held to be fixed and unchangeable, at least beyond a certain early age. For transsexuals, “gender identity” denoted a deep, strong, and persistent disidentification with one’s natal sex and identification with—or even as—the other sex. (This was long before talk of non-binary gender identities became widespread.) This strong understanding of gender identity—as a core, foundational part of the self; as immutable; and as refractory to psychotherapy—is what legitimized then-novel medical procedures, understood as a means of relieving otherwise intractable suffering. There was no suggestion that everyone had a gender identity in this strong sense; gender identity was understood to be relevant only in unusual cases, which were expressly understood, in a medicalized framework, as pathological.

The embedding of gender identity in legal and administrative, data-gathering, and pedagogic domains, however, presupposes the universal applicability of the category; it only makes sense if all people have a gender identity.⁵⁸ Yet there has been no agreement about what such a universally applicable concept of “gender identity” might mean. It is implausible to think that everyone has a gender identity in the strong meaning of the term: a strong, foundational, immutable identification as and with one’s natal sex category or another sex or gender category.⁵⁹ Many people identify cognitively as members of their natal sex category without identifying emotionally with that category (or with any other sex or gender category);⁶⁰ the degree and manner of their identification with (or disidentification from) that category may change over time. As a result, various “softer,” looser, weaker definitions of “gender identity” have been proposed in legal and administrative, data-gathering, and pedagogic contexts.

The proposed federal Equality Act, for example, which would have barred discrimination on the basis of gender identity as well as sexual orientation, defined “gender identity” in a circular and expansive manner as “the gender-related identity, appearance, mannerisms, or other gender-related characteristics of an individual, regardless of the individual’s designated sex at birth.”⁶¹ The Biden administration’s final Title IX regulation, which “understands gender identity to describe an individual’s sense of their gender, which may or may not be different from their sex assigned at birth” (Office for Civil Rights 2024: 33809) is likewise circular, while the University of California’s definition is transparently so: “Gender identity: The gender(s), if any, with which a person identifies. An individual’s gender is their gender identity, which can be the same or different from their sex assigned at birth” (University of California 2020). California law also uses a circular, though more cumbersome definition: “‘Gender identity’ means each person’s internal understanding of their gender, or the perception of a person’s gender identity, which may include male, female, a combination of male and female, neither male nor female, a gender different from the person’s sex assigned at birth, or transgender.”⁶² The Ontario Human Rights Commission (n.d.), for its part, defines gender identity as “each person’s internal and individual experience of gender. It is a person’s sense of being a woman, a man, both, neither, or anywhere along the gender spectrum.” The reference to everyone’s “internal and individual experience” might be taken to mean that everyone’s gender identity is unique; at the very least, the reference to a “sense of being a woman, a man, both, neither, or anywhere along the gender spectrum” suggests a plethora of possible gender identities.

What most weak definitions have in common, apart from their vagueness and circularity, is their reference to some kind of internal “sense,” “understanding,” or “experience.” Unlike strong definitions of gender identity, these weak definitions do not require this sense to be deep, abiding, or foundational; there is no implication that gender identity is fundamental or unchanging. A further commonality of weak definitions is that since gender identity is unmoored from the framework of binary sex, the number of gender identities is potentially limitless. And while strong definitions construe gender identity as a pre-discursive core of the self, existing prior to and independently of cultural idioms and social influence, weak definitions cannot plausibly do so. Once one abandons the idea of a foundational and immutable aspect of the self, one is left with plastic self-understandings constituted in and through language and culture. The range and distribution of such self-understandings will depend on the culturally and socially available ways of thinking and talking about the self—on the stock of stories, cultural idioms, and interpretive templates through which people make sense of themselves and their troubles. That such stories, idioms, and templates circulate through friendship networks and digital media platforms seems evident. The denial or downplaying of such social influence (Turban 2024: 84–90, 113–122) in response to alarmist claims about “social contagion” is understandable as a form of strategic essentialism, but it risks turning its back on a century’s worth of research on the social formation of selfhood—including a substantial body of work in transgender studies and gender and sexuality studies.⁶³

Weak definitions have an affinity with more general “soft” or “postmodern” understandings of identity as multiple, fluid, and subject to choice, change, negotiation, and experimentation. This way of thinking and talking about identity, influenced by or at least compatible with queer theory’s critique of strong, essentialist notions of identity, resonates with many young people. But it stands in sharp tension with the strong understandings of identity as inborn and immutable that continue to be called on to do the work of justification and legitimation, especially in legal and medical contexts.⁶⁴

To be sure, the ambiguity of “gender identity”—and the ambiguity of “identity” more generally—may help explain why the category was so widely adopted in a variety of contexts. Epstein (2022: 85, 89) has underscored the “productivity of ambiguity,” which may enable the “successful uptake [of concepts] in diverse social contexts.” As he notes, this is consistent with work in science and technology studies on the “strength of loose concepts” (Löwy 1992), which can stitch together different constituencies, each of which may see something different in the term, and on “boundary objects” (Star and Griesemer 1989), which are sufficiently ambiguous and flexible to enable cooperation among heterogeneous actors from different social worlds. The successful career of “gender identity” was arguably facilitated by such productive ambiguity. The ambiguity enabled the category to escape the control of the psychologists and psychiatrists who initially put it into circulation, to find niches in different organizational ecologies, to move between professional worlds and lay publics, to accommodate different professional and political agendas, to do different kinds of work in different contexts, and to resonate and recombine with various strands of contemporary cultural discourse.

Yet the ambiguity of “gender identity”—and especially the tension between strong and weak understandings—has also left the category vulnerable to attack. The crux of the problem is that the weak definitions are too weak to do the work they have been asked to do. Strong definitions—particularly the notion of a foundational, objectively grounded, and immutable substrate of the self—provide a strong warrant for anti-discrimination provisions as well as for gender-affirming medical treatment.⁶⁵ Legal scholars advocating for trans rights have emphasized the immutability of gender identity, sometimes explicitly invoking an analogy to the immutability of race.⁶⁶ And a key federal court ruling on discrimination on the basis of gender identity highlighted this immutability.⁶⁷ Weak definitions provide no such warrant: if gender identity is fluid, contingent, limitless, purely subjective, and potentially unique to each individual, the case for making it a legally protected category like race or sex collapses. So too does the case for endowing weak conceptions of gender identity with strong powers: the power to gain access to sex-segregated spaces, to claim treatment as a matter of “medical necessity,” or to change the pronouns by which one is referred to in schools or workplaces.⁶⁸ The tension between strong and weak understandings of gender identity—the former too strong to be widely applicable, the latter too weak to provide a rationale for the powers that have been invested in the category—suggests that the medical, legal, administrative, organizational, statistical, and pedagogical edifice I have analyzed in this paper has been built on a shaky foundation.

Like identity in general, gender identity is an overburdened concept, asked to do a great deal of work. It is at once a category of analysis and, as I have shown in this paper, a category of practice. It is a vernacular category, used by ordinary people to make sense of themselves and their everyday situations and predicaments and to make themselves legible in particular social and cultural contexts. It is a political category, used to make claims and formulate demands. And it is an actionable organizational category, embedded in codified rules and policies, organizational routines, and institutionalized procedures.

As an organizational category—the focus of my discussion—gender identity requires or at least invites explicit, formal definitions. These explicit definitions, even when they are based, as they generally are, on weak understandings of the concept, tend to fix and freeze gender identity, to reify it, to treat it as a “thing.” They function as sorting and classifying devices, suggesting that people come in different “kinds”: the cisgender kind, the transgender kind, the non-binary kind, and many more gender-diverse kinds.⁶⁹ An ever-finer categorical mesh offers the dignity of distinct kindhood to ever-more-specific categories. Rather than making gender categories and their constraints less binding and less relevant to social life, as second-generation feminism sought to do, the establishment of gender identity as a category of organizational practice binds people more tightly to categories and “kinds.” It channels gender diversity into special named categories and thereby renders natal categories, by implication, less diverse and reinforces stereotypes about them (Reilly-Cooper 2016; Brubaker 2023: 163ff). Its ideal is not freedom from categories but rather the freedom to find a category that fits.⁷⁰

The institutional embedding of gender identity formalized, legalized, and standardized informal self-understandings and self-identifications. Formalization made gender identity legible to and operationalizable by organizations. Legalization inscribed it in law and administrative regulations and gave it binding force. Standardization made gender identity diagnosable and treatable, countable and teachable. Together, formalization, legalization, and standardization made gender identity actionable in a broadening range of contexts: they made it a category that enabled people and organizations to do a widening range of things. But since the things that could now be done with the category inevitably affected others, the process of formalization, legalization, and standardization also made gender identity a visible target of a broadening and intensifying critique.

I have sketched the largely unnoticed and uncontroversial emergence of this regime of institutionalized gender identity in medical, legal, administrative, data-gathering, and pedagogic contexts; I have highlighted the work done by gender identity in these contexts as an actionable category of organizational practice; and I have analyzed the belated and relatively sudden eruption of fierce public contestation over this regime, beginning in the middle of the last decade and intensifying in the last five years. Gender identity has been drawn into the partisan vortex of the culture wars and blatantly exploited for electoral purposes, most strikingly in the final weeks of Trump’s 2024 election campaign and the barrage of executive orders issued in his first weeks in office.

Yet despite the administration’s cruel assault on transgender people, the controversy over the category “gender identity” cannot be plotted solely on a left-right axis. The recent years of intensifying attacks from the right have also seen the emergence of critiques of gender identity from the center and center-left (even, in some cases, from the left), precisely at a moment when the barrage of red-state initiatives has made many liberals reluctant to air criticisms of gender identity. These critiques have been developed by gender-critical feminists such as Stock (2021) and Reilly-Cooper (2015) who, unlike radical feminist critics like Raymond (1979), underscore their acceptance of trans people and their support for trans rights but object to the displacement of sex by gender identity; by gay and lesbian critics such as Sullivan (2019, 2024) who likewise insist on the importance of sex and worry that the hyper-available language of gender identity may lead some gender-atypical children who would have grown up to be gay adults—especially those growing up in a homophobic milieu—to interpret their gender-atypicality as a sign that they may be trans; by transsexual activists such as Wu (2024) and Hayton (2024), who argue that the definition of transgender solely in terms of declared gender identity, without any reference to dysphoria or transition, has stretched the category to the point of meaninglessness and eroded public support for those who transition; by transgender scholars such as Amin (2022) and writer-activists such as Chu (2024) who have challenged the practice of basing trans rights and trans politics on what Chu has called the “thin peg” of gender identity; and, as I noted toward the end of the previous section, by a gradually coalescing array of critics—parents of gender dysphoric children, de-transitioners, dissident staff at gender identity clinics (Reed 2023), therapists (Ayad et al. 2023), researchers (Gorin 2024), and journalists (Barnes 2023; Singal 2021b)—who have challenged the uncritical insistence on affirming declared gender identity in the youth gender medicine context. These critiques point in different directions, but they converge in demonstrating that criticism of gender identity is by no means confined to the right.

The emergence of these critiques from the center and center-left and from within the transgender movement is a reminder that the controversy over gender identity cannot be reduced to a struggle between progressive defenders of the “civil rights issue of our time,”⁷¹ as President Biden put it, and reactionaries who think transgender people have “no right to exist.” It is a more complex struggle over how to think about sex and gender as principles of vision and division of the world: a struggle over the relation between nature and culture, between the sexed body and gendered feelings, between givenness and chosenness, between responsibility and autonomy, between safety and self-determination, and between other-regarding and self-regarding actions. It is on these complex and difficult questions that the debate on gender identity as a category of practice is likely to turn.

Acknowledgements

For helpful conversations and comments on earlier drafts of this paper, I would like to thank Carol Bakhos, Zsuzsa Berend, Paula Diehl, Jason Ferguson, Julieta Goldenberg, Gail Kligman, Mara Loveman, Abigail Saguy, Leor Sapir, Jesse Singal, John Skrentny, Coral Utnehmer, Loïc Wacquant, Roger Waldinger, Andreas Wimmer, and three anonymous Theory and Social Inquiry reviewers.

Competing Interests

The author has no competing interests to declare.