[mezzie2]

> For more anecdotal context, my other interests lean into a demographic/stereotype of an upper middle class pilates/health/cleangirl girliepop. Which from my experience they centrist/ambivalent maybe skew left but not inherently pro trans.

I'm not sure this is accurate, speaking as someone whose social circle has a lot of people in that group and who also sometimes ends up on that side of the algo. For various reasons, it's only socially acceptable to be outspokenly on the left in that group (more so for women over the age of 25/30), but there's a sizable minority who either disagree with certain topics or are more aligned with the right, but they aren't going to jeopardize their social circles or (particularly in the case of content creators) their audience.

It's the girl version of the men who put 'apolitical' on dating apps to hide being a conservative because they don't want to limit their dating prospects.

[AnEro]

I agree and disagree! To clarify, when I say centre/centre left I mean neoliberal and left being more aoc/burny leaning. It is very acceptable to neoliberal, but often not worth the political risk to go further than that and vice versa. Neoliberal isn't inherently protrans, and in general people see trans people in two ways, example for trans women a different type of male(biology) vs a different type of woman(social construct). Where depending on their ideology and their view as mentioned, it determines what trans issues they will come to bat and risk talking about.

[mezzie2]

Yeah, we're talking within the general Overton window. Views that are outside of that (and particularly outside a binary) aren't going to do well with the algorithm. (It's interesting to me that you instinctually divide the views on trans people into a binary!)

There are just a lot more secret conservatives in that group than one might expect from outward appearances. The center of the Overton Window is as far 'right' as they can go without losing their audience. I can usually pick up on it pretty clearly, but I also worked in political communications for a while and became very sensitive to signaling and find it a fascinating topic. I wouldn't be shocked if the algorithm picks up on the association and assumes some similarity between down-low conservatives and forthright conservatives and starts serving the latter to viewers of the former.