Expert analysis of the results of making 'sex' a self-id question in the census, by @ProfAliceS, @lnmackenzie1 & @janeclarejones. We already have one example of the seriously misleading results of a survey based on self-id of sex (thread).

We compared the Stonewall School Report 2012 (data gathered by sex) with the report of 2017 (data gathered by self-id). The School Report 2017 finding that 45% of trans young people have attempted suicide, has had a huge influence on policy making. https://www.transgendertrend.com/stonewall-school-report-what-does-suicide-rate-mean/

We calculated the number of actual females who responded to the 2017 survey, and we found an anomaly. All other general population studies show a significant increase in rates of self-harm in girls since 2012. LGBT young people, in every other study, are particularly at risk.

So we would have expected the Stonewall report to show an even larger increase in rates of suicidal ideation in girls since 2012 than general population studies. But the Stonewall report, uniquely, showed an actual drop in rates of self-harm and suicide attempts for girls.

We looked for an explanation of this atypical result, and it didn't take long to find it. The ‘male gender’ category shows a significant rate increase in all three areas of self-harm, suicidal ideation and suicide attempts since 2012.

Girls (who previously may have described themselves as lesbian or bisexual) are now swelling the ranks of the 'male gender' category. The result of the Stonewall School Report 2017 actually shows a significant increase in the rate of suicidal ideation in girls.

It's just that they are now invisible.

The way the @ONS gathers data influences other agencies. If every agency gathered data based on self-id of sex, we wouldn't know that there has been an exponential increase in the number of teenage girls being referred to the Tavistock; data would show an increase in boys.

The Tavistock referral figures 2019/20 show that for a total of 103 children aged between 11-17, their sex is recorded as 'not known'. If sex is not recorded, the Tavistock, like Stonewall, will produce seriously misleading data. The @ONS must not collude.
https://gids.nhs.uk/number-referrals