> This theory is related to the curious fact that, on average, the more feminist your society, the fewer women there are in math and science — which makes total sense if you assume that on average women are good at math but uninterested in it.
This isn't a complete explanation: we can see this by looking at other STEM fields. The early years of computer programming were dominated by women, yet nowadays, women are proportionally uninterested. You don't get such a dramatic demographic shift because of innate tendencies, but this was contemporaneous with a shift from programming being considered low-status to high-status work. Is this perhaps social, rather than directly economic?
To take an example from elsewhere in the thread (https://news.ycombinator.com/item?id=41718072): I can see the “you must use this method” prescription hitting girls harder than boys, since girls tend to drift towards copying / collaborative play, and boys tend to drift towards competitive play. This prescription might make mathematics seem less like play, to girls – which would be ironic, since real mathematics is an incredibly collaborative endeavour.
(Which raises the question: do girls inherently prefer copying play, and boys inherently prefer competition play? Who knows? I suspect not, but I think it'll be a long time before we find out.)
> The early years of computer programming were dominated by women, yet nowadays, women are proportionally uninterested.
We eliminated the job women were dominating (programmer) by combining it with the one men were dominating (analyst).
At the same time law and medicine were seeing huge increases in the proportion of women practitioners, so the status thing doesn’t make a ton of sense as an explanation. Besides, it was not high status in the 80s when this was going on (or the 90s… arguably it’s still not, just high pay)
> We eliminated the job women were dominating (programmer) by combining it with the one men were dominating (analyst).
Gender disparity is usually shown as a percentage, but years ago I ran across one for programmers that used absolute numbers and the pattern showed a different story than usual - which this reclassification could probably explain.
I don't remember what year exactly the flip was, but before the flip the number of men and women were increasing at around the same rate. After it, the number of men skyrocketed while the number of women kept increasing at the same rate as before. As a percentage this looks like women lost interest or got pushed out, but the absolute numbers look more like men flocked to it without pushing anyone out. Or, perhaps, got grouped into it.
I communicated that badly. It doesn't matter so much how things are seen outside the workplace, but within it. Programming is definitely considered high-status in a software firm: ever heard of the concept of a "rockstar programmer"?
Low-status tasks (e.g. vital, but "unpromotable" ones) are delegated to women, and tasks that are associated with femininity are considered low-status. This is a well-documented (https://noidea.dog/glue) and easily-measurable phenomenon. I expect there are many harder-to-measure instances of institutional sexism that might make classes of workplace unpalatable, even if there's no gender bias in desire to do the actual work.
If this (or a similar) effect has been going on for a while, I'd expect that to have significant knock-on effects.
For a counterpoint: a friend of mine works at Google, and as an excellent SRE who also happens to be female she's steadily getting opportunities thrown her way; especially invites to speak at external conferences and other internal events. She's also gotten promotions.
It helps that she's very competent, but she didn't have to work extra to be noticed by the organisation.
Systematic discrimination isn't the same as universal discrimination.
If we're trading second-hand anecdotes, I've got a couple dozen of trans women programmers no longer receiving promotions despite flawless performance reviews, and half a dozen trans men programmers suddenly receiving credit for work they were previously ignored for. That's as close to a controlled test as I can think of – though, obviously, marred by the selection bias of anecdotes.
All this doesn't mean it's the same in mathematics – but I'm not sure how someone can deny that there's institutional sexism in the field of computer programming. It's well-documented. "One person at Google" doesn't refute that.
> Where Summers sees innate differences, Barres sees discrimination. As a young woman […] he said he was discouraged from setting his sights on MIT, where he ended up receiving his bachelor’s degree. Once there, he was told that a boyfriend must have solved a hard math problem that he had answered and that had stumped most men in the class. After he began living as a man in 1997, Barres overheard another scientist say, “Ben Barres gave a great seminar today, but his work is much better than his sister’s work.”
