> There tend to be self-reinforcing trends in human behavior. But we also know how quickly human behaviour changes and adapts to new circumstances. > Jobs tha

> There tend to be self-reinforcing trends in human behavior. But we also know how quickly human behaviour changes and adapts to new circumstances. > Jobs that are highly gendered frequently stay that way, making it harder for the categorically-challenged to partake for purely social reasons. I can think of many jobs where this is true (for either sex), but software engineering doesn't seem to fit in this category. In STEM about 30-40 % of students are female; less at trade schools (=lower educational attainment). About 25 % of the IT workforce is female. > One, we should want a society more free from constraints on individual self-realization than less. Absolutely. > Ironically, this is also an argument people use to claim that sexism and racism in the work place can't exist, in that annoying way some economists do. I think the classical Friedmann argument was that there is a cost to biasing against competent people. It doesn't really say anything about what people will actually do (like much of economic theory), since they might be (1) not aware of the cost [and indeed it would be difficult to quantify] (2) valuing their *isms above that that cost. As far as classic economic theory goes, the argument seems to be correct, it's just a poor predictor for what humans will do.