More or Less?

This post (and the next couple) take their inspiration from a favourite programme on Radio 4, “More or Less“, in which statisticians explore the numerical data behind headlines in the news or oft-repeated pieces of popular wisdom. In my brief exploration of Twitterspace, I found that debate raged hotly about some important statistics.

One of the most controversial topics in the debate about self-identification is the rate of suicide among transgendered individuals. A recent article in GQ presents a statistic that has been reported very widely in the media: it states that 41% of transgendered people attempt suicide, and that this figure is ten times that for the population as a whole. I have found this figure to be widely deployed on both sides of the self-ID war. For activists it is the proof required that self-ID is necessary – the wait for recognition of gender is driving people to suicide; and for gender-critical feminists, the claim of a 41% attempted suicide rate is plainly and self-evidently preposterous, and evidence of the preparedness of activists to say whatever they like, true or not, in their attempt to coerce society to collude with them in their misogynistic masterplan.

It is widely alleged by sceptics that the figure originates in a Stonewall report that contained no properly validated statistical data. I also suspect that the figure is false. I’ve been out and about on the trans scene for some time and nobody that I know has committed suicide. On the other hand, my experience also tells me that rates of mental illness are rather high among trans people. I’d say that compared to my colleagues at work, a much larger fraction of trans people suffer from mental illness.

A recent TV drama, “Butterfly”, explored the challenges experienced by transgender children and their parents. The drama was produced in consultation with the transgender children’s charity Mermaids. Mermaids believes that if a child says they are trans, then this is not a matter of debate or exploration, but an expression of an identity that must be respcted. If a boy says he is a girl then he must be a girl.

This position is founded on both an assertion of the right of the individual to self-expression, and also on a rejection of the eminently reasonable hypothesis that children are continually developing, so that what seems unalterable truth at one stage may prove to be a passing phase with the benefit of hindsight. Canadian psychologist Ken Zucker lost his job for propounding the hypothesis that many children who identify as trans may go on to become gay adults who are happy in their birth sex, reflecting the worrying trend that some ideas – though capable of precise, objective, scientific formulation – have been determined to be unthinkable. (Zucker recently received a public apology for his dismissal and a settlement.) Actor Rupert Everett recently described his own development from gender dysphoric child to gay adult: “I really wanted to be a girl. Thank God the world of now wasn’t then, because I’d be on hormones and I’d be a woman. After I was 15 I never wanted to be a woman again.” Of course, many of us who wanted to be women before 15 still want to be women long after the halcyon days of our youth have disappeared over the horizon, so I’d not like to suggest that Rupert Everett has the last word on this. But Zucker’s hypothesis is at least interesting and worthy of consideration.

Mermaids and other transgender organisations counter Zucker with a hefty emotional stick: rejecting a child’s gender identity will lead to the collapse of their mental health, and may very well drive them to suicide. To the parent of a gender dysphoric child, the thermonuclear character of this particular ordnance is clear: to question a child’s gender identity is in fact to put their life at risk. What parent would choose to do that? To parents grappling with these very complex problems, the notion that a false move may drive their child to suicide may be overwhelming: it is emotional blackmail of the first order.

Butterfly created a stage upon which the dramatic force of this argument could be played out for millions of viewers: in episode 1 the unhappy gender dysphoric child at the centre of the drama, whose father refused to recognise him as a girl, cut his wrists. The message could not be made more clear.

So what’s the truth? Is Mermaids playing fast and loose with data or is there a real problem with suicide among transgendered individuals?

Transgender Trend is an anti-trans lobbying group. It articulates scepticism about trans suicide rates very clearly, and has provided ammunition for those who would like to reject the notion of a high suicide rate. It reports summaries of two studies conducted in the UK that do seem to suggest that suicide rates are higher among transgendered youths.

Transgender Trend claims that the first of these studies, led by academics, was not based upon randomly selected subjects, but was based on responses to a questionnaire completed by only 27 transgender subjects. These are reasonable concerns, but may be levelled at many studies in the field. For example, Ray Blanchard’s seminal work expounding his theory of autogynephlia was based on studies of comparatively small numbers of subjects who presented in his clinic, and Lisa Littman’s paper on rapid onset gender dysphoria – which Transgender Trend reports with approval – was based on questionnaire responses from parents who had joined web sites specifically set up to support parents who believed their children had rapid onset gender dysphoria (let me think…might there be a possibility of confirmation bias?). So there is quite a bit of inconsistency here. As a physical scientist, I struggle to come to terms with the low threshold for “evidence” in much psychological research when compared with, for example, the rather exacting standards laid down for clinical trials of drugs. But human beings are extraordinarily complex; perhaps the poor statistics in much psychological research and the absence of appropriate controls are a reflection of this intrinsic complexity. If we apply strict standards of rigor, we must certainly apply them across the board. I think there is an important methodological debate here. The response to the Littman furore from many professional psychologists was to defend the methodology used as being normal in the field. This may be the case, but what if the professional standards of psychological research are rotten? When the pronouncements of professional psychologists are used to legitimate the formulation of public policy we have a moral obligation to ensure that they are founded on properly validated studies with rigorous methodology and proper control experiments.

The second study was conducted by a Cambridge University academic on behalf of Stonewall, and Transgender Trend makes the allegation that although the lead academic was asked to provide their methodology they did not do so. I believe this is the foundation for the argument that the Stonewall study is methodologically unsound. If the study was done without proper methodology, and the figures contained within it have been widely cited by Stonewall, then this is a serious state of affairs. However, a failure to respond to a request for information from Transgender Trend does not in itself invalidate a study.

As an academic I was keen to find peer-reviewed literature on the subject. Toomey et al recently published a study of 120 617 adolescents (ages 11 – 19) for whom data were collected over a 36 month period.1 They found that male to female transgender youths reported a rate of attempted suicide approximately three times that of male adolescents (29.9% compared to 9.8%), while female to male transgender youths reported a rate of attempted suicide of 50.8% compared to 17.6% for female adolescents. These data certainly suggest an increased rate of attempted suicide – for both MTF and FTM transgendered youths, the rate is approximately three times that associated with adolescents of the same birth gender. Perhaps the most shocking statistic, however, is that 14% of all adolescents reported a previous suicide attempt. This seems to me to be a worryingly large number – indicative of deep and widespread unhappiness and poor mental health among a very vulnerable sub-set of the population.

A further study by Barboza et al focussed primarily on the incidence of physical attacks upon transgender people.2 They noted that 37% of transgendered people questioned had experienced physical attacks, and that on average these people reported approximately 4 attacks. Such people reported higher rates of attempted suicide, suggesting a connection between victimisation of trans people as a potential cause of poor mental health. Overall they reported a rate of attempted suicide of 25%, with 38% reporting suicidal ideation.

It is clear that there is a shortage of peer-reviewed literature in this field. However, it also appears that critics are incorrect in claiming “foul” when transgender activists claim that rates of suicide are higher among the trans population.

A more difficult question concerns the big stick wielded by Mermaids: is a questioning approach to childhood expressions of gender dysphoria a likely trigger for suicide? This was clearly the message of Butterfly. However, the peer-reviewed work that I have cited suggests a more complex state of affairs. In particular, Barboza et al reported a connection between victimisation of trans people and attempted suicide, while Toomey et al, discussing the different rates of suicide observed among the groups identified in their study, note the importance of sociological factors (eg. previously published work demonstrating higher rates of discrimination against FTM transsexuals than among their MTF counterparts). If this kind of analysis is correct, it suggests that elevated suicide rates among transgender individuals are correlated with difficulties encountered in their interactions with wider society.

At the very least, one might say that there is no evidence in these studies that there is a direct connection between suicide and the adoption of the kind of “watchful waiting” approach favoured by the UK’s Tavistock Centre. Indeed, the Tavistock Centre published a firm response to the suicide scene in Butterfly, commenting:

Tragically, there is a higher risk of self-harm and suicidal ideation or attempts found in LGBT communities compared to their cisgender or straight peers. Suicidality in young people attending the GIDS is similar to that of young people referred to child and adolescent mental health services. It is not helpful to suggest that suicidality is an inevitable part of this condition.

To conclude, therefore, it seems that there is good peer-reviewed evidence that suicide rates are higher among transgender individuals than among the population at large, reflecting my own experience that many trans people suffer mental health problems, but not 10 times as large (as is widely quoted) – perhaps 2-3 times as large. A risk of suicide is not intrinsic to individuals with gender dysphoria, but is perhaps best understood as reflecting the increased tendency of marginalised groups to suffer from mental health problems. I leave the last word to the Tavistock Centre:

“Our focus is on supporting young people to be resilient and positive and many of the young people we see are doing very well. It is important to note that it would be very unusual for younger children referred to the service to make suicidal attempts.”


  1. Russell B. Toomey, Amy K. Syvertsen, Maura Shramko, Pediatrics 2018, 142, e20174218.
  2. E. Barboza, S. Dominguez and . Chance, Prev. Med. Rep. 2016, 4, 385-90. DOI:

2 responses to “More or Less?”

  1. When I saw that Toomey’s study had looked at 617 adolescents, I started to worry that this was a quite small number. If the incidence of transgendered individuals in society is less than 1% (as commonly cited), this means that the sample of them in this study would be about six (or alternatively the sample was biased towards transgendered individuals). Six is nowhere near enough to reasonably calculate the rates of suicide attempt reported by Toomey and quoted by Karen.

    Thankfully, Toomey’s study is better than that. In the abstract to the paper, the actual number of adolescents is reported at 120,617. (I can see how Karen misread it.) That means the sample of transgendered individuals in it would be at about the 1200 level, which seems more reasonable.

    Liked by 1 person

    1. Thank you for reading so carefully Justine! You’re absolutely correct – this was a copy and paste error. By the standards of this sort of study, N was a very large number thus giving significant confidence in the statistics.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s