[icebraining]

In 1984, women earned 37% of the CS bachelor degrees and were 38% of the CS workforce. How does the "sexyness" theory explain this?

I agree with you that home computers where the turning point, but it has nothing to do with being "sexy", but with the gender roles prejudices in families, which heavily biased who got computers for birthday and Christmas, while before it was purely an academic activity, therefore more accessible to women (by that time).

[dozzie]

You got it totally wrong. Programming wasn't sexy until late '90s, when internet bubble happened. Earlier it was just another occupation that was boring, like electronics or mechanical engineering (to non-tinkering people, of course).

By now it should be clear that by "sexy" I mean "desired to profess".

All combined, "the 'sexyness' theory" doesn't need to explain why at the peak 38% of the CS workforce were women. It's not meant to explain anything, it's meant as merely an observation that it's untrue that men only got to IT when it paid well. It was the opposite: IT was already a male-dominated field when it started to pay well.

[icebraining]

Fair enough; I understood you meant women went into it now because it's "sexy".

That said, back in '84 when women still composed a good chunk of the workforce, programmers and system analysts earned $60-$100k/year (2015 dollars), which is close to the current average ($80k in 2013).

So it seems to me that it already paid well when it became male-dominated - even if that wasn't the reason it did so.