According to this set of ‘Key Stats’ from Stonewall, ‘almost half (48 per cent) of trans people in Britain have attempted suicide at least once.’

This is an alarming statistic; and our hearts must surely be moved with compassion towards any person who feels driven to attempt to take their own life.  Moreover, we should search our consciences to see whether our own attitudes and actions may have contributed to that person’s misery.

But is there an element of alarmism, both in the figure itself and in the way it is often used?

This statistic does not simply remind us of the sad fact that some trans people attempt suicide; it claims that very many do so.  It is the sheer size of the problem which grabs headlines and leads to knee-jerk reactions in which:

• policy changes are be made that have not been sufficiently thought through.
• emotional pressure is exerted which says, in effect, to parents and others: ‘Allow this child to transition, or you will have their suicide on your hands.’  In the same way, attempts are made to silence critics.
• the rights of others are threatened.  Teachers and others will be forced to ‘affirm’ what in all conscience they cannot affirm.  Girls will have established protections (with regard to privacy, for example) removed.  Unfair advantages (as in sport) will become more common.
• suggestibility is increased.  As someone has put it: ‘If you are told 50% of people like you will be suicidal, when you are having a dark day, instead of thinking tomorrow will be better, are you not more likely to worry that you’re going the way of all those suicidal folks and therefore spiral downwards?’ (Source).  According to the Samaritans, ‘young people are particularly vulnerable to imitational suicides.’  This is one reason, among others, why Media Guidelines for the responsible reporting of suicides are available.

These are serious implications.

So, it is not with a view of minimising the impact of attempted suicide on those it directly affects, but rather with a view to questioning the reliability of the headline statistic, that I want to proceed.

I shall do so by examining two frequently-quoted studies.

The RaRE study

One frequently-quoted study suggesting that a very high proportion of trans young people attempt suicide is this The RaRE Research Report.  According to this report, 48% of trans young people have attempted suicide at least once.  But this figure is based on a sample of just 27 self-selected trans young people, and this is insufficient for informing public opinion or guiding public policy.

Moreover, it is of concern that the Mermaids charity has seriously misrepresented this study by stating that the research surveyed ‘more than 2,000 trans people in UK’.  This was not a one-off error, since it was repeated in this interview.  The figure 2,078 represents people of all sexualities and genders.

The authors of the study itself recognise that with regard to sampling:

given the nature of PACE [the LBGT+ mental health charity] and also of the topic of the research, it is possible that there is a disproportion amongst research participants of people with experience of or who are sensitive towards mental health issues, as well as of people with an LGB&T community attachment. It is also possible that the survey attracted heterosexual people who are sympathetic towards the LGB&T community, even if the publicity materials for the survey were designed to minimise their sexual minority focus. This potential bias is suggested, for instance, in the rates of suicide experience and ideation for the heterosexual cohort, which are higher than expected when compared with those of the general population (The NHS Information Centre for Health and Social Care, 2009).

and with regard to the generalisability of the findings, the authors add their own caveat:

Ultimately our findings can only be considered valid for our samples and are also subject to the specificities of the social and cultural contexts where the research was conducted.

(See this discussion of the research)

It is regrettable that these acknowledged limitations with the study have been ignored by so many, and that they have been in some cases falsified by those with vested interests.

Stonewall School Report, 2017

I mentioned at the beginning of this piece the research referenced by Stonewall:

‘Trans Mental Health Study‘ (ed McNeil et al, 2002).  (N.B. Stonewall slightly mistitles this, and provides a ‘dead’ url link).

Actually, there is a more recent research study, commissioned by Stonewall itself:

Stonewall School Report: The experiences of lesbian, gay, bi and trans young people in Britain’s schools in 2017 (2017)

This report comes to very similar conclusions as have previously been cited about attempted suicide rates among trans young people, giving a figure of 45%.  In commenting on this study, I’m drawing on this analysis.

The study surveyed 3,713 LGBT students aged 11-19.  Of these, 594 self-identified as trans.  Clearly, this is a much bigger sample than that surveyed in the RaRE study.  Nevertheless, a number of issues remain.

This research has not been published in a peer-reviewed journal, and therefore lacks academic scrutiny.

The researchers have not published their detailed methodology.  No technical report is available.  We do know that an online survey method was used.  However, we do not know what questions were asked.  This is important, since very much depends on the phrasing of the questions about attempted suicide.  We do not know what counts, either for the researchers or for the respondents, what counts as an ‘attempted suicide’.  Data may well be inflated by ‘yes/no’ questions, or if the question is not explained clearly.  In population studies, follow-up interviews with respondents typically reduce numbers quoted for attempted suicide by about half.

There is no evidence that this study controlled for other risk factors, such as existing mental health problems, autistic spectrum disorder, family background, being in care, or previous trauma or sexual abuse.

Students may be influenced by the thought – actively promoted in the media – that if they describe themselves as a risk from suicide they will be more likely to get the support and treatment they want.

No distinction is made between those who were already receiving medical treatment towards transition and those who were not.  Therefore, no account can be taken of the known side-effects of hormonal treatments and puberty blockers.

The report does not distinguish between biological sex and gender identity, nor between those who would consider themselves to be ‘straight’ and those who consider themselves to be ‘gay’.  This is significant, because, given that (among other things) lesbian culture is now ‘under siege‘,

If we fail to record the biological sex of young people we are unable to try to understand why suddenly such a disproportionate number of young women want to ‘identify’ out of womanhood. Is ‘identifying as a boy’ just the latest coping mechanism for girls brought up in a culture where images of women being sexually abused in porn are casually passed around in playgrounds, and the internet ensures no escape from the pressure on girls to be ‘perfect’? And is the world so dangerous for lesbians that being seen as a man is a safer alternative?

The data in all these studies refer, of course, only to attempted suicide.  Stats on actual suicide are unknown (we know that 90% of actual suicides are associated with a psychiatric disorder).

Finally, and very significantly, this research used ‘non probability sampling methodology’.  This is a research method more suited to qualitative research, and therefore to reporting and analysing the lived experience of respondents, rather than in presenting reliable quantitative data.

Conclusion

In social research, there is frequently a trade-off between validity and reliability; between qualitative data (such as respondents’ lived experience) and qualitative data (such as how representative these experiences are of a wider population).

By their very nature, the research studies just discussed score highly on validity and much less so on reliability.  The qualitative data should alert us all to the very real feelings and perceptions that some trans people have about their lives, and about how they are viewed and treated by others.

But these studies do not offer reliable data on the incidence of suicide attempts among these young people.  In fact, I have given reasons why we might think that the figures given have been significantly inflated.  Considerable restraint, therefore should be exercised with regard to the figures quoted for the incidence of attempted suicide in trans young people.  The same kind of restraint should also be shown by those who influence, and those who formulate public policy in this area.

It’s time to turn the volume up on the stories.  It’s time to turn the volume down on the statistics.