About the publication
Who Are We Now

Blaise Agüera y Arcas

From leading AI researcher Blaise Agüera y Arcas comes an exploration of how biology, ecology, sexuality, history, and culture have intertwined to create a dynamic “us” that can neither be called natural nor artificial.

A beautiful physical edition is available from Hat & Beard Press.

Order Book


Audiobook (Audible)
eBook (Kindle)

Identity politics occupies the frontline in today’s culture wars, pitting generations against each other, and progressive cities against the rural traditions of our past. Rich in data and detail, Who Are We Now? goes beyond today’s headlines to connect our current reality to a larger more-than-human story.

At the heart of the book is a set of surveys conducted between 2016 and 2021, asking thousands of anonymous respondents all over the United States questions about their behavior and identity, and especially about gender and sexuality. The resulting window into people’s lives is a bit like that of the Kinsey Reports, which scandalized postwar America more than 70 years ago. Today, the landscape is—in every sense—even queerer. Twentieth century heterosexual “normalcy” is on the wane, particularly among young and urban people.

The landscape outside has changed too. After millennia of being fruitful and multiplying, we’ve strained, and exceeded, planetary limits. Domesticated animals far outweigh wildlife, and many species are in catastrophic decline. Yet curiously, our own population is poised to begin collapsing this century too, our fertility now curbed by choice rather than by premature death. Is this the end of humanity—or the beginning?


“A fascinating, provocative account of the contradictions and complications of identity and community in the technological age.” Sonia Katyal, Distinguished Law Professor and Co-director, Center for Law and Technology, UC Berkeley School of Law

“This is that rare book that not only explores a fascinating topic, but that teaches you how to think more deeply about any topic. Insightful and original.” Tim O’Reilly, Author, publisher, and founder of O’Reilly Media

About the author

Blaise Agüera y Arcas is a frequent speaker at TED and many other conferences, winner of MIT’s TR35 Prize and Fast Company’s Most Creative People award, and a Vice President and Fellow at Google Research. He leads a 500-person team working on Artificial Intelligence (AI), large language models, smart devices, technology ethics, and privacy. Publicly visible projects from his team include Federated Learning, Artists and Machine Intelligence, Coral, and many AI features in Pixel and Android. In 2016, he wrote a widely read essay on the relationship between art and technology, and in 2017 he co-authored another popular essay on physiognomy and bias in AI and a refutation of claims that facial structure reveals sexual orientation. Some of this material has been incorporated into the book’s third and fourteenth chapters. His early involvement in large language models and generative AI prompted op-eds in the Economist and essays in Noēma, as well as inspiring the novella Ubi Sunt, also published by Hat & Beard Press.

Connect with Blaise

For press inquiries, please contact Noah Bethke ([email protected])

13 The return of Count Sandor

Many believe the trans community is tiny. This misconception became obvious when I ran a small side survey asking respondents for their estimates of some of the statistics the main survey measures—for instance, the percentage of trans people. Trying to keep things simple, I foolishly used a number widget on the web page that didn’t allow for responses lower than 1%. Yet 1% was the most common answer, and a number of respondents complained:

I wanted to put a smaller number for the transgender population around my age. (0.01%). 1

You should have allowed decimals on percentage trans. It is closer to zero than one, but it is above zero. I would say 0.001 percent. 2

Like any other anonymous identity, the real figure is a moving target. The second respondent above, born in the 1950s, actually would have been close to the mark had he given this answer as a teenager, in 1968. Back then, the US population was around 200 million, and American psychiatrist Ira B. Pauly, known for his research on the topic, estimated that there were just 2,500 transsexual people (as they were then called) in the country. 3

More recent figures seem to be all over the map. A 2011 study gave a “rough estimate” of around 0.3% of US adults, 4 while several 2016 studies put the number at 0.4–0.6%, though a meta-analysis does note that “Future national surveys are likely to observe higher numbers of transgender people.” 5 Indeed they do.

As elsewhere, our intuitive estimates tend to systematically undershoot for two reasons: first, because we form our models of the world when we’re young, and are often slow to update them; second, because if we look around at people our own age, we don’t see the whole picture. Big national polls, like those run by the Census Bureau, may be thrown off by anonymity concerns. Many studies also estimate overall averages without considering variation as a function of age and time. Yet those variations can be dramatic, as in responses to the question “Do you identify as trans?”

13.0
% by age

This is another of those cases where the changes in people’s responses over time are a lot more radical than a shift to the right by two years from the 2018–19 data to the 2020–21 data. In fact no amount of sliding the 2018–19 curve to the right will make it look like the 2020–21 curve. The effect seems to be accelerating, too; isolating the 2021 data reveals an even sharper rise, especially in the youngest cohort. Given the speed of the change, social transmission seems to be a major factor here—not just among the young, but across all ages.


Was David Reimer trans? While he was born a boy, he was raised as a girl. The medical procedures David underwent in his teens and twenties to align his body with his gender—phalloplasty and double mastectomy, plus lifelong hormone treatments—were the same ones undertaken by many trans people who medically transition to masculine bodies. The same is true of less tangible, but also weighty social transitions: name, pronouns, relationships, legal documents. In assessing the relevance of nature versus nurture in David’s ordeal, his biographer, John Colapinto, wrote:

None of this is to suggest that nurture plays no role in gender identity. Virtually every page of As Nature Made Him contains an environmental cue or clue that helped to reinforce what Brenda’s prenatally virilized brain and nervous system were telling her. […] I attribute the case’s final and complete collapse, however, to the pressing insistence of Brenda’s biological maleness—her awakening sexual attraction to girls; her inchoate but adamant aversion to possessing breasts and a vagina. For how many children, at the exquisitely awkward age of fourteen, will insist, upon threat of suicide, that they undergo full sex change, in plain view of neighbors, family, and friends? This almost incomprehensible act of courage on Brenda’s part speaks more convincingly than any other piece of evidence to the emphatic demands of our biology, and to the necessity that we—all of us—be allowed to live as we feel we must. 6

Clearly the entreaty to “be allowed to live as we feel we must” applies to many others. Although the evidence suggests that some intersex people, and some fraction of the population in general, may be quite gender-flexible, this certainly isn’t universally the case; and the assignment of intersex babies to a binary gender has often been arbitrary—skewing heavily to “boy” from the 1950s until the 1990s, then skewing heavily to “girl” in more recent years.

Consider the implications: out of the 2% or so of known intersex births, nearly 85% would have been assigned male in the old days, and are being assigned female today. We’re talking about a lot of people—in the US alone, it works out to about 5.6 million. That’s larger than the combined population of Los Angeles and San Diego (or of the seven least populous states—Wyoming, Vermont, Alaska, North and South Dakota, Delaware, and Montana). It seems a safe bet that a significant fraction of these arbitrary assignments have been flat out wrong, just as David Reimer’s reassignment in infancy was wrong. So, many people arbitrarily assigned a gender will likely seek to transition later in life, just as he did, making them both intersex and trans.

13.1
Intersexuality compared with trans identification
% by age

Exploring the overlap between the trans and intersex communities using the survey data is possible, but it requires stretching the analysis to its statistical limit. Only about one in a thousand respondents are both intersex and trans; hence error bars will be large. 7 Moreover, the rate of intersexuality plunges toward zero among the youngest respondents, requiring coarse age bucketing there. Still, the patterns revealed are interesting, and appear to be statistically significant. (Recall that the shaded regions show 90% confidence intervals.)

13.1
Overlap of intersex and trans
% by age

Between 5% and 15% or so of intersex people are trans, with that percentage rising steadily by age. For reasons that will soon become clear, people in this overlap often struggle for acceptance in either community, as with this 63-year-old woman with XXXY chromosomes assigned male at birth: “even within intersex communities there is great opposition to including transgender people like me.” Not all people who are intersex know they are, so the real overlap is probably higher, but this upward trend reflects an increasing number of intersex people who, at some age, feel that their sex assigned at birth wasn’t right, and decide to do something about it. They may only learn that they’re intersex later, when they seek to transition.

Fifteen percent is not an insignificant effect, especially given the implications among older respondents, vanishingly few of whom (trans or not) are non-binary with respect to their sex. Consider: if, when an intersex baby is born, male or female sex were assigned by flipping a coin, and 30% of the 50% assigned the “wrong” sex feel strongly enough about this to transition—and are in a privileged enough position to do so—that would give us 15%. Keep in mind, too, the high stigma associated with being trans, especially among the older generation; this may explain why the number rises only slowly with age. As a 2018 article in the Guardian put it, “Meet the trans baby boomers”:

[T]here are more than five times as many adult as child gender identity patients in the UK. Some are now having gender reassignment surgery not just in late middle age, but well into retirement. […] It’s perhaps only now that many older people feel comfortable coming out, having grown up in a time when being trans was so steeped in shame and silence that many couldn’t even put a name to what they felt. 8

Remarkably, among the youngest cohort of trans respondents, in their late teens and early twenties, at least 20% and perhaps as many as 50% are (and know they are) intersex. Yet at age 26 and above, that figure is only about 5%. It’s likely that more than one factor is in play here. First, only a small fraction of intersex people know they are at these young ages. But which fraction? It’s likely to include those born with especially non-binary genitals, that is, people from whom their intersexuality (and thus the arbitrariness of their sex assignment at birth) really can’t be hidden. Second, both sex and gender non-binariness are more readily embraced by young people, a phenomenon explored more deeply later in this chapter.

In general, being trans means something different for a 20-year-old than for a 60-year-old. The great majority of trans 60-year-olds think of themselves as either men who were assigned female at birth, or women who were assigned male at birth. On the other hand, a large and increasing proportion of young trans people think of themselves as not identifying with their sex assigned at birth—a broader category, since it includes a growing number with non-binary or fluid gender identities, or who reject gender altogether.

13.2
Overlap of non-binary and trans
% by age

This is evident in the pattern of overlap between trans and non-binary identities. Beyond age 60, the overlap is nil; the survey turned up noolder trans non-binary people, or non-binary trans people. Among 19-year-olds, though, 40% of non-binary people are trans, which likely implies more of a rejection of their (binary) birth gender than an identification with the opposite gender. Even more dramatically, fully 80% of 18- to 20-year-old trans people surveyed identify as non-binary.

13.3
Overlap of “they” pronoun users and trans
% by age

The overlap between exclusive users of “they” and the trans population follows a similar pattern. Vanishingly few older trans people use “they” pronouns exclusively, while about 20% of 19-year-old trans people do; and while 40% of 19-year-olds exclusively using “they” are trans, this, too, appears to fall off sharply with age—though with so few older respondents in either of these populations, estimating the overlap becomes dicey.

13.4
Overlap of “they” pronoun users and non-binary
% by age

As one might expect, the overlap between “they” users and non-binary people is very high at the younger end, but curiously, by age 70 the overlap is negligible. This is especially noteworthy given that, on their own, non-binary identification and the use of “they” both follow that U-shaped curve encountered in Chapter 10, with a decline followed by an increase among older respondents. 9 It seems that when older people use “they” (usually non-exclusively) or identify with non-binariness (usually not to the exclusion of identifying as men or women), they’re less consciously part of a social movement in which these answers go together, and imply a more wholesale rejection of gender.


Before delving into the decline in binary sex and gender identification among the young, let’s compare the numbers of trans people assigned female at birth versus male at birth.

13.5
Trans by sex assigned at birth
% by age

These curves tell a story at once remarkable and familiar. Among older people who don’t identify with their birth sex assignment (whether they identify as trans or not), male assignment at birth overwhelmingly predominates, while among younger people female assignment at birth is much likelier. This pattern should ring a bell: it’s the same as for intersexuality.

13.6
Fraction of intersex and trans people by sex assigned at birth
% by age

Directly comparing the birth sex assignment ratios for intersex and trans people reveals that they look nearly identical above age 25: by age 70, assignment is around 80% male at birth, while at age 25, sex assignment at birth is close to 70% female—an astonishing reversal. At the youngest ages, though, the curves diverge. Virtually all intersex people (at least, those who know they are) are assigned female at birth nowadays, for the reasons discussed in the previous chapter. Young trans people also tend to be assigned female at birth, although “only” by a 3:2 majority. I’ll return in a moment to the question of why this divergence has appeared among the young.

First, though, consider the trans statistics among the older population. These may seem familiar from experience—at least for those of us old enough to remember how things were in the last millennium. When I was a kid, in the 1980s and ’90s, words like “cross-dresser,” “transvestite,” and “transsexual” (which are generally out of favor today, along with some more offensive related terms) were strongly associated with people assigned male at birth who to one degree or another embraced femininity in their clothes, presentation, and identity. One could call the old tendency to assume trans people were assigned male at birth the “Priscilla effect,” after the 1994 surprise hit The Adventures of Priscilla, Queen of the Desert. 10

As Krafft-Ebing’s case history of “Count Sandor V” in the 1880s illustrates (see Chapter 6), there have always been people assigned female at birth who undertook the opposite journey in gender space. There must also be untold numbers of “Count Sandors” throughout history who, escaping the circumstances of their birth, successfully “passed” as men, leaving no medico-legal documentary record. Even so, the survey data confirm the intuition that among the older generation alive today, the “Bernadettes” far outnumber the “Sandors.”

Movie still from The Adventures of Priscilla Queen of the Desert, showing the three main characters, one of which is trans and the other two who are dressed in colorful, flamboyant drag outfits.
From The Adventures of Priscilla, Queen of the Desert, 1994.

Those assigned female at birth may have had less freedom to make choices about gender identity later in life, just as social and economic disadvantage suppressed same-sex attraction among women. It’s hard, though, to see the intersex curve overlaid on the trans curve and not suspect a connection. I think the relationship is causal: the reversal from a majority of trans people assigned male at birth to a majority assigned female at birth is, at least in part, a downstream effect of changing intersex medical protocols (though something further is afoot among the youngest cohort, which we’ll return to shortly).

It would be surprising not to see such an effect, given that in the old days, such an overwhelming majority of intersex babies were assigned male at birth—remember, we’re talking about millions of people—and the overwhelming majority are now assigned female at birth. Since John Money’s theory that gender can be freely molded by social pressure in the first couple of years of life has been disproven by cases like David Reimer’s, one would expect this change in policy to result in major shifts in the trans population—which is just what we see. Hence, among young people today, suddenly “Count Sandors” are handily outnumbering “Bernadettes,” though perhaps, in due course, a rising tide of non-binary trans people will outnumber both.

A grainy old photograph of a feminine-presenting person with curly hair wearing an intricately patterned garment. The person stands with their hands on their hips, contemplatively looking up.
“Invertido Sexual” from Gómez, La Mala Vida en Buenos Aires, 1908.

To my knowledge, this relationship has not been quantitatively established elsewhere, though as Jules Gill-Peterson notes in a 2018 book, Histories of the Transgender Child,

The persistence of the entanglement of intersex and trans life in the bodies of children has been underappreciated; in fact, it endured well into the 1950s, if not later. It lasted nearly as long as we have had the discourse of transsexuality, and yet it has radically faded from contemporary conversations about the plasticity of sex and gender. 11

Why hasn’t this entanglement been studied more closely in recent decades? One possible reason is discomfort with the topic on the part of both the intersex and trans communities, albeit for different reasons.


Consider intersexuality’s medical status over the past century. It has widely been thought about as a health condition, like diabetes or gigantism. In this spirit, many intersex people (who know they are) still strongly oppose the excluded middle-embracing approach taken by the Lohmans; they see medical intervention at birth as having allowed them to “grow up normal,” and feel thankful in much the way a child born with a cleft lip or palate might feel thankful for early reconstructive surgery.

A modern clinical term for intersexuality is Disorders of Sex Development (DSD), which some people find preferable to the term “intersex” or the older (and even more contested) “hermaphrodite.” It feels different to have a disorder of sex development than to be intersex. Medicine exists to manage or cure disorders—to make them, to the extent possible, go away. Conversely, an identity is emphatically not a disorder. 12

For intersex people who live as men or women but need access to medical care specific to their intersexuality, framing it as a medical condition that isn’t anybody else’s business makes perfect sense. The word “need” matters here, in that while treatment can occasionally be medically necessary (for example, the so-called “salt-wasting” variety of CAH makes the survival of babies like Rosie Lohman depend on daily medication), in most cases the point of the medical care is to bring an intersex person’s presentation in line with the gender binary. So, a self-fulfilling logic is in play here: when a person’s “untreated” intersexuality would be socially obvious, managing it as a treatable disorder is precisely what makes it possible for it to be “just” a medical condition—hence not socially obvious.

For many trans people, however, the medical framing is offensive. And with good reason. It carries stigma, implying that something is wrong with them—and “wrong” in the head, not “wrong” in the body. Remember that in the 1950s the DSM classified “transvestism” as a perversion, alongside homosexuality, “pedophilia, fetishism, and sexual sadism (including rape, sexual assault, mutilation),” and that these practices were all regarded as criminal. Some of the old “blue laws” criminalizing consensual non-heteronormative acts are still on the books today. 13 While pathologizing lesbian, gay, and bi people is no longer nearly as widespread as it was half a century ago, it’s still common to pathologize trans and non-binary people, as reflected in many comments on the survey:

I believe confusion about your gender is a mental disease that should be treated instead of encouraged. 14

If you actually believe you are anything but male or female you are in fact MENTALLY ILL. 15

I am normal […] [I] have female genitalia and chromosomes, no mental illnesses making me think i am not female. 16

Such views are commonplace even among the young and educated, as per this 22-year-old woman from Hammonton, New Jersey:

I don’t believe in gender identities opposite of birth gender, outside of those who are intersexual. Not to be rude, but I’ve been through many Psychology classes and see it just as a trendy misrepresentation of a mental disorder that everyone thinks they relate to.

This attitude is unsurprising considering that, when the DSM-III-R finished the job of de-pathologizing homosexuality, it merely reclassified transsexuality under a new heading, “Gender Identity Disorders,” where it remained until 2013. This dovetails neatly with the fact that John Money and his adherents needed a medical diagnosis to justify their surgical and hormonal meddling with intersex babies and children too young to make medical decisions for themselves. “Fixing” newborns can’t be medically justified unless we can claim something is wrong with them.


For trans people, a version of the same conundrum can also surface later in life. Unlike being lesbian, gay, or bi, being trans may in itself involve having expensive medical procedures done—though note that this applies only to a fraction of trans people today. Medical costs impose an especially steep economic barrier in countries lacking free universal access to healthcare—like the US. For insurance to cover a procedure, it needs to be regarded as medically necessary, which is usually interpreted to mean that it must address a medical condition. 17 This has led to rhetorical finessing in which “gender identity disorders” have given way to “gender dysphoria,” which is still considered an illness. So, trans people are assigned a status not unlike that of lesbian, gay, and bi people in the wake of the American Psychiatric Association’s 1973 memo—the one that threw shade on homosexuality, male chauvinism, and vegetarianism in the same breath (see Chapter 6). Media theorist Sandy Stone described this Catch-22 in an influential 1987 essay:

When the first academic gender dysphoria clinics were started on an experimental basis in the 1960s, the medical staff would not perform surgery on demand, because of the professional risks involved in performing experimental surgery on “sociopaths.” At this time there were no official diagnostic criteria; “transsexuals” were, ipso facto, whoever signed up for assistance. Professionally this was a dicey situation. […] [A] test or a differential diagnosis was needed for transsexualism that did not depend on anything as simple and subjective as feeling that one was in the wrong body. […] But even after considerable research, no simple and unambiguous test for gender dysphoria syndrome could be developed. 18

As of the last few years, being trans is no longer a disorder, according to the APA—yet getting trans medical care continues to require a diagnosis. It can be a bit head-spinning to think about getting “diagnosed” with something that is not a disorder in order to be “treated” for it.

Nuanced changes in official diagnostic language clearly don’t instantly flip the definitions in everyone’s minds, either. There’s much disagreement even within the trans community; for this 39-year-old man from College Station, Texas, for example, being trans remains strictly a medical issue:

Transgender is a medical condition I have, not a freaking identity or a gender in itself. Unlike these little self-diagnosed “genderfluid” snowflakes—aka poseurs—I actually transitioned years ago, complete with medical care and hormones and a name change and not picking a new gender or pronoun every four seconds. I am quite goth/androgynous looking and always have been. I am most definitely male. You don’t “identify” as trans any more than you “identify” as having diabetes. You get diagnosed with it, ffs.

It seems likely that in the coming years we’ll see this binary—disorder versus identity—start to dissolve for many.

For starters, medical techniques in general tend to become more sophisticated and safer over time. As that happens, we can start to shift away from thinking of healthcare purely in terms of treating disease. We’ve already seen a reframing of health broadly in terms of prevention and wellness, but what about proactive agency over our bodies? In years to come, medical choice will surely extend far beyond today’s superficial “cosmetic” surgeries. Hormone treatments are of course already more than cosmetic, as they hack the body’s own mechanisms for modifying form, function, and even behavior. Still, these remain imperfect technologies. Cross-gender hormones typically limit fertility, and can cause health problems. 19 They also require lifelong medical maintenance.

A more profound change is taking place without any need for transformative medical advances, though. An increasing number of trans people don’t opt to have medical procedures done; their transition is social and behavioral. This move marks a significant decoupling of identities from bodies—a topic I’ll return to in Chapter 15.


Most people in the trans and intersex communities agree that neither being intersex nor trans are “lifestyles,” but manifestations of something deeper in one’s makeup. Ironically, in this respect, they also agree with a lot of more conservative people in the general population; it’s just that those more conservative people tend to think of unchangeable “reality” in terms of sex chromosomes or genitals, whereas many trans and intersex people have a lived experience of unchangeable “reality” being their deeply rooted sense of who they are. As David Reimer could attest, that can’t just be ignored or wished away. At the same time, one’s sense of gender is an internal, subjective experience, which makes it easy for others to question.

To complicate the picture even further, this internal experience isn’t necessarily constant throughout life, or itself binary. 20 A number of survey respondents expressed a nuanced range of sentiments:

I do not consider myself trans, but I have been dealing with mild gender dysphoria. 21

previously diagnosed with gender identity disorder, I identify as a tomboy. 22

I am fine with being a woman. Wish I had been a man. Hated child birth and periods. Hate waxing and shaving and makeup. 23

wish i had been male. 24

A wish that things were different doesn’t necessarily equate with being trans, nor is it necessarily a disorder, but—either might be the case? This is the slipperiness the 1973 APA memo so clumsily struggled with.

To many trans people, the perennial (and possibly unanswerable) “nature versus nurture” question seems beside the point. As Juliet Jacques put it in her 2015 memoir, Trans,

I had no answer to this question, any more than I could convey to people why I was left-handed, and I didn’t think it was fair that I was constantly being obliged to answer it. It wasn’t my main concern: I didn’t think too much about nature and nurture when I was worrying about the possibility of having my head kicked in if I answered back to any of the people who yelled at me in the street. The scientists and the sexologists can argue about that, I thought.

It’s indeed both unfair and nonsensical to ask trans people to justify their lived reality. Still, just as for handedness, the scientific questions of cause, effect, and mechanism seem worth exploring, both because a better understanding of ourselves is valuable in its own right, and because ignorance breeds just the kind of bigotry Jacques describes.


Recent years have seen a sharp rise in gender dysphoria, especially among young people assigned female at birth. This isn’t just an American phenomenon. In March 2020, the Guardian reported that “Sweden’s Board of Health and Welfare […] confirmed a 1,500% rise between 2008 and 2018 in gender dysphoria diagnoses among 13- to 17-year-olds born as girls.” 25 Similar accounts have been reported in countries ranging from New Zealand 26 to Pakistan. 27

13.7
Referrals to NHS Gender Identity Development Service
Number by year

In 2022, England’s National Health Service (NHS) reported that its Gender Identity Development Service (GIDS), the UK’s only gender identity clinic for people under 18, would need to be shut down and replaced with regional centers to deal with skyrocketing numbers of referrals (more than 5,000 in 2021, doubling the 2019–2020 numbers 28 ), which had overwhelmed staff and created unacceptably long waitlists.

13.8
Referrals to NHS Gender Identity Development Service, share of total by sex assigned at birth
% by year

Prior to 2011, the majority of referrals had been assigned male at birth, but today, three quarters of referrals were assigned female at birth. This is consistent with the rapid changes evident in the US survey data.

An increasing number of young people are on puberty-blocking hormones, and some have undergone surgeries, occasionally as young as 12. 29 Advocates both within and outside the trans community argue about the appropriate age of consent for these procedures. In some respects, it mirrors the controversy about intersex surgery on infants, but whereas in that case a focus on preserving children’s agency militates against early intervention, the situation here is far less clear-cut.

The data offer some insights that push back on ideological arguments made on both sides. First, as we’ve seen, the shift toward defaulting to female for nearly all intersex babies likely accounts for a significant part of the rise in young people assigned female at birth seeking to transition. The overlap between young trans and intersex people offers an important clue, and the numbers suggest that many young trans men are intersex but unaware of it. This isn’t a small effect. However, it’s unlikely that latent intersexuality (even defined very broadly) accounts for the nearly 6% of young people assigned female at birth now identifying as trans—nearly half of whom also identify as non-binary. It’s important to look at social factors too.

Advocates for early access to medical treatment sometimes argue that gender identity is intrinsic and immutable, because if it were otherwise, it would be harder to justify prescribing life-altering surgery or hormones to children. Such advocacy has led some to deny or minimize accounts of “desisting” or “detransitioning,” in which young people change their minds about being trans, or seek to transition back after treatment with hormones or surgery (recall that only a fraction of young trans people undergo medical procedures). But lots of kids going through puberty are uncomfortable in their changing bodies, and many nowadays are identifying as trans, at least for a time. Desisting after puberty appears to be common, 30 and detransitioning sometimes happens too, though unbiased statistics about how often these occur are hard to come by. 31

Given the powerful evidence of social transmission pretty much everywhere in the survey—including sexual orientation, attraction, and family models—it would be surprising not to find such effects. It’s thus unfortunate that changing one’s mind can be perceived as invalidating, either by trans rights advocates or by their critics. Advocates may find themselves arguing against any element of free will or choice in gender identity, and against any evidence of social transmission—which is hard to square with the data, or with the entire notion of social progress. Conservative critics may argue, on the other hand, that evidence of social transmission proves that “gender nonconformity” is a mere fad or “lifestyle choice,” echoing older claims about being gay.

Just as it did then, this conservative argument has real-life consequences. Sonia Katyal, a legal scholar who has updated Kenji Yoshino’s 2001 critique of gay “contagion” 32 to address challenges faced by trans parents, notes that custody courts regularly express concerns about “gender contagion—the idea that normalizing gender transition or variance can introduce a level of instability into the child’s own gender identity.” In one custody case involving a trans parent,

[…] the court was left to wonder, rhetorically, “Was his sex change simply an indulgence of some fantasy?” Whether such evidence would have in fact made a difference in the outcome is unclear, but the court certainly implied that the immutable—or at least unavoidable—character of a parent’s gender identity could be entitled to greater consideration than a purely indulgent, selfish “choice” perhaps connected to sexual desire. 33

It would be helpful to move beyond the either/or argument that animates much of this debate today, and acknowledge that there’s no contradiction between people sometimes having “unavoidable” inner drives, exercising free choice, and being susceptible to the social transmission of ideas, including both behavior and identity.

Neither is being socially influenced something that only happens to unformed or impressionable young people. Every graph comparing 2018–19 with 2020–21 responses offers evidence that substantial numbers of older people are influenced by social transmission too. 34 Indeed, if they weren’t, human progress could only happen “funeral by funeral,” as an economist put it in 1975: “the old are never converted by the new doctrines, they simply are replaced by a new generation.” 35 Evidently, things don’t work this way; humans break the generational speed limit on evolution all the time.

It also seems clear from the data that different people have different degrees of flexibility with respect to their gender identity. Some older survey respondents reflect on how their identities might have differed if they were growing up in today’s environment:

I sometimes think I would have been happier if I was male. I never thought about this much when I was younger, as transexualism wasn’t spoken of much or widely known at the time. I would never want to transition now but I think if I was growing up in today’s world I might feel differently. I have always gravitated toward more classically ‘masculine’ hobbies and tastes. Then again, today there is somewhat less gender stereotyping and being a ‘tomboy’ is more normal. Even though I’m in my late 40s, I remain a bit confused about gender. I think this is mainly due to social conditioning. I do identify as a woman and very strongly as a lesbian. 36

Some of these effects are reminiscent of both handedness and sexual orientation. Recall that exclusively same-sex attracted people make up a remarkably consistent minority regardless of age; yet the data also reveal a rapidly growing cohort of younger people who are flexible in their attraction, and increasingly identify with the formerly narrower terms “gay” and (especially) “lesbian.” Gender identity follows a similar pattern. A rapidly growing cohort of gender-expansive young people have embraced the formerly narrower term “trans,” whether they opt for medical intervention or not. There is clearly an element of individual choice involved here, and that choice occurs within an evolving social context.

At risk of oversimplifying (for this, too, is not a binary), imagine a “hard core” within the trans community consisting of women born in male-assigned bodies, or (more commonly nowadays) men born in female-assigned bodies—a situation far more prevalent than Ira Pauly’s 0.001% estimate in 1968, but likely well below the nearly 6% of young trans people assigned female at birth today. Unless powerful environmental factors are in play (like the estrogen-mimicking pollutants described in Count Down), this “hard core” has likely remained fairly constant over time, though evolving intersex protocols seem to have shifted the tide from a majority of trans women to a majority of trans men. This population isn’t limited to—but does include—people with rather traditional views about gender:

As an older transsexual who had transitioned before the fad, I feel a lot of this stuff these days is invented and find it offensive and unhelpful. 37

A studio portrait of a bald and bearded Buck Angel looking intently at the camera. He is wearing a white tank top that reveals arm and chest tattoos.
Buck Angel (1962–), trans porn star, producer, activist, and educator.

On the other hand, a sizable and growing number of young people could be described as “gender abolitionists” who reject not only their own binary gender assigned at birth, but the binaries of gender and sexuality altogether:

Enby [non-binary], but don’t self-id as trans. My entire friend group of 13 or so is all trans or enby. Pansexual is the closest descriptor to my sexuality that was provided, non-gender-differentiated demisexual would probably be more accurate. Also the gender binary is fake. 38

Although this is a broad movement, it, too, is especially embraced by young people assigned female at birth. Perhaps this should be unsurprising given the more privileged status of men, and the way campaigns like #MeToo have drawn attention to the stubborn persistence of gender inequality despite centuries of feminist activism.


  1. An 18-year-old man from Fort Lee, New Jersey.

  2. A 63-year-old man from Phoenix, Arizona.

  3. Pauly, “The Current Status of the Change of Sex Operation,” 1968.

  4. Gates, “How Many People Are Lesbian, Gay, Bisexual and Transgender?,” 2011.

  5. Meerwijk and Sevelius, “Transgender Population Size in the United States: A Meta-Regression of Population-Based Probability Samples,” 2017.

  6. Colapinto, As Nature Made Him: The Boy Who Was Raised as a Girl, 279–80, 2000.

  7. To understand why, consider a population of N=100 people, some number X of whom have a property A. X will then be the percentage of people with property A, as shown in many of this book’s graphs. The mathematical details of how the uncertainty around X is calculated are described in the Appendix, but let’s suppose for simplicity’s sake that the uncertainty is 1%, that is, plus or minus one respondent. Given a property that holds for half of the population, a ±1% uncertainty doesn’t matter—49% and 51% tell much the same story (“about one in two”). However, if X=1, this same ±1% absolute uncertainty is a much bigger deal, in relative terms; it means the incidence could easily be twice as high (“one in fifty”), or half as high (“one in two hundred”), or even rarer (though given that it did occur, it can’t be zero). Now, suppose there’s a second property B that also has a 1% incidence rate. Calculating the percentage of overlap between A and B will be hard. If only one individual out of the 100 has property A, and a different individual is the only one with property B, the real overlap might be zero, or 50%, or something else—nobody can say. The error bar calculations take this uncertainty into account. Making the age bucketing coarser helps by increasing N within each bucket, allowing overlaps between small populations to be sampled better, at the expense of being able to distinguish differences across ages.

  8. Hinsliff, “‘Age Has Nothing to Do With It’: How It Feels to Transition Later in Life,” 2018.

  9. As shown at the beginning of Chapter 10, around 1–2% of 20-year-olds identify exclusively with they/them pronouns, a number that only declines for older respondents.

  10. Watch it if you haven’t. It’s a delightful road movie following three larger-than-life drag queens on a journey through the Australian outback. The eldest, Bernadette, is transsexual.

  11. Gill-Peterson, Histories of the Transgender Child, 17, 2018.

  12. Jones et al., Intersex: Stories and Statistics from Australia, 95, 2016.

  13. Wikipedia, “Sodomy laws in the United States,” 2022.

  14. A 35-year-old man from Yuma, Arizona.

  15. A 33-year-old man from Las Vegas, Nevada.

  16. A 47-year-old woman from Nesquehoning, Pennsylvania.

  17. The same argument is taking place in countries that do have socialized medicine, where distinctions are still generally made between medically necessary and elective procedures. Elective procedures tend not to be prioritized or even covered by national health plans.

  18. Stone, “The Empire Strikes Back: A Posttranssexual Manifesto,” 2006.

  19. Coleman et al., “Standards of Care for the Health of Transgender and Gender Diverse People, Version 8,” 2022; Hembree et al., “Endocrine Treatment of Gender-Dysphoric/Gender-Incongruent Persons: An Endocrine Society Clinical Practice Guideline,” 2017; Nahata et al., “Understudied and Under-Reported: Fertility Issues in Transgender Youth—A Narrative Review,” 2019.

  20. Singh, Bradley, and Zucker, “A Follow-Up Study of Boys With Gender Identity Disorder,” 2021.

  21. A 28-year-old man from San Jose, California.

  22. A 32-year-old woman from Granbury, Texas.

  23. A 43-year-old woman from Orange, Texas.

  24. A 59-year-old woman from Pensacola, Florida.

  25. Orange, “Teenage Transgender Row Splits Sweden as Dysphoria Diagnoses Soar by 1,500%,” 2020.

  26. Delahunt et al., “Increasing Rates of People Identifying as Transgender Presenting to Endocrine Services in the Wellington Region,” 2018.

  27. Irshad, Hashmi, and Aamer, “Between a Rock and a Hard Place – Gender Dysphoria and Comorbid Depression in a Young, Low-Income, Pakistani Transgender Man,” 2020.

  28. Gentleman, “‘An Explosion’: What Is Behind the Rise in Girls Questioning Their Gender Identity?,” 2022; Smith, “Referrals to the Gender Identity Development Service (GIDS) Level Off in 2018-19,” 2019; “Gender Identity Development Service Referrals in 2019-20 Same as 2018-19,” 2020.

  29. Tang et al., “Gender-Affirming Mastectomy Trends and Surgical Outcomes in Adolescents,” 2022.

  30. Drummond et al., “A Follow-up Study of Girls with Gender Identity Disorder,” 2008; Steensma et al., “Gender Identity Development in Adolescence,” 2013; Wallien and Cohen-Kettenis, “Psychosexual Outcome of Gender-Dysphoric Children,” 2008; Steensma et al., “Desisting and Persisting Gender Dysphoria After Childhood: A Qualitative Follow-up Study,” 2011. Some researchers, however, have pointed out methodological challenges with desistance studies: Temple Newhook et al., “A Critical Commentary on Follow-up Studies and ‘Desistance’ Theories About Transgender and Gender-Nonconforming Children,” 2018.

  31. Expósito-Campos, “A Typology of Gender Detransition and Its Implications for Healthcare Providers,” 2021.

  32. See Chapter 7.

  33. Katyal and Turner, “Transparenthood,” 2019.

  34. WEIRD societies place a strong value on personal consistency, which is perhaps why we consider it so embarrassing for our identities and beliefs to be influenced by others as adults. We’d prefer, perhaps, to somehow always be the influencers, yet for the ideas we spread to have sprung into our own minds out of nowhere.

  35. O’Toole, “Science Makes Progress Funeral by Funeral,” 2017.

  36. A 47-year-old woman from Bridgeport, Connecticut.

  37. A 49-year-old man from Traverse City, Michigan.

  38. A 24-year-old from Portland, Oregon.