Let’s say you want to “prove” that red M&Ms are twice as common as any other colour M&M.
Bear with me – I have a point to this, and it’s about the far right and science.
Now, what you’re wanting to prove here it isn’t true – red M&Ms aren’t twice as common as any other colour, they’re one of the less used colours in fact. That doesn’t matter, though – you just want a study saying that, and you want it to look like it’s real.
So you get a bowl, and you randomly select a bunch of M&Ms from a supply of M&Ms. This is now your “random” sample. Let’s get counting!
Sounds okay, right?
Except beforehand, you pre-filled the supply of M&Ms with twice as many red M&Ms as any other colour.
So then you study your “random sample” of M&Ms and discover – oh look, there really are twice as many red M&Ms as any other colour! So you publish, and all the people who really, really want to say there are twice as many red M&Ms as any other colour quote you, and use your work, as you intended. A few of them will also write “studies,” using your study as a source, hiding it in footnotes so it looks like they’re doing new and separate work when really, they’re just reference laundering.
(If they were referencing your work in good faith, and doing additional work to build on it, that would be different. But we’re operating in a world of hilariously bad faith here, and in general, no actual additional work is done. And besides; they know.)
This is how fundamentalist “science” has always worked. This is how Paul Cameron used to produce his “studies” showing that 54% of gay men “regularly consumed the feces” of their sexual partners, how he showed that “homosexual men” were overwhelmingly made up of “violent predators.” He did so by pulling his “random selections” from a specifically loaded pool.
In his case, he often chose his “random population” of “homosexual men” out of a collection of imprisoned sex offenders. Since in his mind all LGBT people were, definitionally, sex offenders for the crime of existing – he also published “The Death Penalty for Homosexuality” advocating, well, what it says – I’m sure he felt just fine with that decision.
Now let’s apply this to the Cass Report and see what happens.
They start with a blanket statement of fact that “the evidence base” for gender supportive care “had already been shown to be weak” before they even got started on their review. That’s on Page 20, if you’re curious, and it’s basically a political statement. Opponents had declared it to be “weak” – particularly the Tories running the NHS – and therefore, it was. It is a conclusion presented before the data collection or the analysis, and a base assumption of the report.
“It’s already weak, so let’s prove it.”
If you’re putting your conclusion before you get to data or analysis, you have already given away the game. You just have. A lot of people won’t get that, but those of us who have spent a lot of time studying fundamentalist / reactionary “science” will recognise it immediately, because “creation science” does the same thing; they don’t do research to “find out where life came from,” they do research to “show God’s Hand in Creation” and prove Creationism is real, and they say that right up front.
Same shit, different sentences.
But once you’ve stated your originating intention, how do you go about it?
Let’s say you want to be able to say that there are “no high quality studies” supporting gender affirming care for children. That’s very easy to do: you just declare all studies supporting gender affirming care for children to be “moderate” to “low quality.”
How do you do that in a way that looks fair?
You do it by by deciding what “high quality,” “medium quality,” and “low quality” mean. You need to be clever about it. You can’t just say “if it supports what we oppose, it’s low quality,” even if that’s what you want to do.
So they were clever about it, and found a way that sounds real good to the casual reader.
First, they decided that only double-blind research really qualified as “high quality.” That sounds reasonable, right? It’s a gold standard for a lot of studies and for very good reasons.
But here’s the trick:
You literally can’t do double-blind studies on children with significant medical issues. Particularly not trans kids for gender affirming care. Why not?
1. It’s well past ‘ethical nightmare’ and into ‘how nice to see you again, Dr. Mengele’ territory, and,
2. They will always know, and in this context, can’t not know.
It is actually literally impossible. Any attempt of any reasonable length or size will degrade out of double-blind before it can complete.
And even if it was possible – they’re children, for fuck’s sake. Lying to them and their parents about their medical care? Giving them fake medications that they will absolutely know are fake after not all that long?
What do you think’s going to happen?
Ask me how I know about this, too.
But that’s not discrediting enough; there really is a lot of work out there, and you can’t just chuck it all out.
So then they went through through a hundred-plus studies that failed to meet this functionally impossible bar, and built a set of metrics to judge them. The structure of this judging followed the form of the Newcastle-Ottawa protocol, but using a different set of scales from the AGREE II protocol. That’s fair; AGREE II is common and, used properly, I understand it to be a good metric.
But there are categories in it, and those categories can be exploited, if you want.
And so the next step was to declare older gold-standard sets of guidelines and best practices – the 2009 Endocrine Society and the 2012 WPATH guidelines in particular, the latter of which built on the first – as lacking “reliability” and to state that they can’t be applied to the NHS even if they had it.
They did the same thing for 2022 WPATH, but that one mattered less because it was too new to have much weight.
I know – we seem to be very, very deep in the weeds here, it’s foggy and confusing and nobody’s sure which was is north. I’m sorry, this is what it’s always like when getting into this particular area of politics. But we’re getting there.
AGREE II includes two categories which are in general highly relevant to clinical analysis: “applicability” and “editorial independence.”
If you declare, as they did, that 2009 ES and 2012 WPATH (and 2022 WPATH) are not applicable to the NHS, and…
If you also declare, as they did, that studies that relied on these protocols are not “building on previous work” but are instead not adequately “editorially independent…”
…then you can heavily downgrade every study which relied on either of those protocols in your AGREE II matrix.
And oh my, look at all those red numbers in columns five and six of their tables.
Isn’t that just the cleverest thing? So deeply buried! And yet, so important.
Out of over 100 studies, the Cass Report went through and evaluated 11 studies on social transition support, 50 on puberty blockers and the like, and 53 on hormonal intervention beyond puberty blockers. With their chosen guidelines and using their chosen selection criteria, modified by their choices about editorial independence and applicability, they managed to reduce every study but two down to either “moderate quality” or “low quality.”
And once you’ve done that, if you go looking for “high quality” studies showing gender affirming care works for trans children, then oh look – our research shows are almost no high quality studies supporting gender affirming care for trans children! And none at all for social support!
Yay! Objective achieved, we can publish, and the New York Times can eat it up and echo it forever while rightist politicians have a field day campaigning to hurt queer kids.
It’s the same fucking trick every goddamn time. Not one single note has changed, the same kinds of people keep using the same kinds of routines, and it it’s all
one
big
goddamn
lie.
And it’s all because a lot of people would rather have a dead son than a live daughter, and some of them are billionaires, and so they’re willing to move heaven and earth to make sure if that’s the choice, they’ll get the one they want, the lives of their own children be damned.
Just like a lot of fathers used to kill their gay sons, and just like fundamentalist families flooded the streets of my town and many others with their own children that they made homeless and destitute at their leaders’ urgings, because they’d all rather see their children die than not be what they want.
If this reminds you of anything, maybe it should.
In all honesty, in all fair disclosure, I can’t prove this report was this sharply agenda driven. I’m just saying that it sure looks to me like it is, because I’ve seen this pattern before, and I’m seeing it again, and in these issues it’s always been used for evil.
[ETA: Or maybe I can prove it: Turns out Cass met with American rightist anti-trans politician Ron DeSantis specifically on banning gender transition and was also recommending the “gender-critical” polemic Irreversible Damage which compares gender transition to the holocaust before starting the report and describing herself as having “no fixed views.” That would, clearly, be a lie.]
Plus there’s the part where they immediately used this report to do what they wanted to do anyway, and issued a blanket ban on gender affirming care for trans kids, and then doubled down on it on their way out of power using a method that meant it couldn’t even be appealed for months.
I’m not even saying these studies are somehow earth-shatteringly great; some of them aren’t. Some of them really aren’t. There’s a lot of problems with population size, there’s the near-impossibility of double-blind (tho’ turns out it’s not impossible, at least not with young adults in particular rare situations), there’s a lack of funding and a huge political investment in making sure this isn’t studied, or if it is, that it’s not studied fairly. None of this leads to optimal science! And all of these issues do lead to issues with process and rigour and those do matter.
But even with all that, the results are still pretty fucking stark.
If nothing else, ask anyone who’d rather have a live daughter than a dead son, or a live son than a dead daughter, ask them how it’s going with their trans kid.
And ask about how they’d feel about a report like this being used to shut it all down.
I’m pretty sure they’ll tell you.
Sometimes, even if they’re Republicans.
102 days remain.
ETA: holy shit, the daughter Elon Musk denies is going at his lies about her, over on Threads. I won’t link to threads much but this? This is an exception.
Posted via Solarbird{y|z|yz}, Collected.