I've asked Mumsnet to remove my earlier post, since I erred in the most important figure. (Embarassingly, I copied a typo over from the blog post rather than going to the source). So here it is again, but I've changed the paragraph about the Australian court case.
Just wanted to weigh in with an observation on the desistance rates (sorry, can't remember everyone's names so will just address in general)
I've now been reading lots about this, watching videos, reading research papers etc. The statistics are a mess, no matter which way you look at it.
The average desistance rate of children until now was 80 to 95%
Until now is important, because until now the appropriate course of action has been to explore why the child might be dysphoric, get them appropriate help and support (this can consist of play therapy, behaviour modification - some dysphoric behaviour isn't trans, it's anxiety related or social problems etc, all kinds of therapy in short)
The treatment attitude was wait and see how it pans out, let the child express themselves at home. If the medical practioner felt - after exploring the whys and wherefores and therapy etc - that social transitioning might help, the child was then transitioned socially.
I don't have the number of how many of those who presented were transitioned, but the number of those who desisted after transitioning socially fell to 30%. In the research this was explained as proof that the wait and see approach was right and also that there was now a social cost attached to detransitioning (affecting both children and adults alike). From a medical perspective, that's too high a failure rate already - a chance of one in three of getting it wrong.
Once medical transition comes into the picture, the statistics are much harder. The Dutch clinic who pioneered hormone treatment had a 0% desistance rate. But I've found in the literature that 20% of all who had bottom surgery regretted it (I think the studies also said 20% desisted after surgery, but couldn't swear that it was the same 20% that regretted it).
There simply is not enough data on detransitioners after medical transition. We do know the rate is not zero because of a recent cross-sectional study (slightly touched upon in the new Transgendertrend report which quotes 93% of detransitioners saying they didn't get enough help) which found over 250 trans-identified females who medically transitioned and then desisted.
The 4% comes from this blog post:
growinguptransgender.wordpress.com/2017/12/03/the-end-of-the-desistance-myth/
Like someone said before me it employs the No True Scotsman fallacy (thank you for teaching me that expression!) and claims that the studies include children who would not now be diagnosed as transgender.
One of the reasons why this is terribly misleading is that although they claim that the new guidelines for clinically diagnosing dysphoria are stricter, given that affirmation is now the only acceptable course of action for a child presenting as transgender, we're now seeing children being transitioned socially who've never even been diagnosed.
This should actually lead to a higher number of socially transitioned children later detransitioning.
The blog post also quotes this Australian court case where the new evidence comes from
growinguptransgender.files.wordpress.com/2017/12/602fd-rekelvin30november2017.pdf
Here evidence is presented from one clinic in Australia which states that 96% of those diagnosed persisted into late adolescence but that they have no longitudinal data yet
As most who desist do so only after a few years of dysphoria, I looked into how many referrals this particular clinic had within the last four years (working on the assumption that these children would not yet have desisted, but may do so in the future).
The clinic accepts children under 8 up to 17 (17yolds are referred to adult services. It had a total of 86 referrals in the first ten years of its existence, the other 614 in the last 3.5. No figure is given as to how many of those referred are diagnosed.
If most children desist only after a few years, the 4% desistance rate relies on only a small cohort of children (i.e. smaller than 86 if not all are diagnosed and fully 40 of that 86 were referred in 2013 and may or may not have reached the stage where they would typically desist).
Most importantly, the gender clinic emphasised that they had no long-term data and looking into the treatment regimen, which they updated in 2017, they now affirm at all ages and seem to socially transition early. At onset of puberty they start blockers, around 15 or 16 cross-sex hormones.
TL:DR The 4% compares apples with oranges. It is based on a clinic whose treatment regimen involves very early social and medical transition and we know that the desistance rate falls sharply once this has been done. It is also based on data from a tiny number of children without evidence of longterm outcomes.
As someone said earlier, however, it does seem to confirm that once you're on the rollercoaster, it's much harder to get off. Of course we could all just sit back and wait twenty years, then we'd definitely know what the new desistance rates are. They'll just involve a lot of kids who've been messed up for life...
P.S. Thank you to OldmanOfTheWeb for seeing me right about the actual name of the fallacy. I've corrected it above.