As for the correct percentage of gender/race/whatever in programming, I don't think that really matters. I think a certain amount of people in general have the mindset to be a programmer and think the way a programmer does, and I don't really see anything like gender or race playing a part in someones ability to be a programmer.
Whatever numer of people that, barring societal interference, in an isolated vacuum with neutral exposure to all professions, choose the profession based only on personality and aptitude.
Maybe some market pressure towards in-demand professions, but that should be mostly extrinsic to someone in their developmental years until after they are exposed to most possible career paths.
Like the OP mentions, it isn't something that the software industry alone can solve, and IMO, all the affirmative action only does the inverse draw that doesn't actually attract those with a passion, but those for the easy access or money. It require society collectively to get over itself and stop thinking you can't do STEM without a Y chromosome, and it needs to stop demonizing the intellectually curious girls. But that is hard, because the parents of those girls were raised the same way.
I think that teaches us two things:
1) It's foolish to jump to conclusions about the "natural number" of women in programming. If you equalize the ratio and it sticks when you remove the affirmative action, that's the natural number.
2) Affirmative action can work when it comes to gender issues, even if it has been less successful for race issues. I think that's because gender doesn't have heritable socioeconomic status like race. A girl is equally likely as a boy to be born into a family that can say afford college, but that's not true of say blacks and whites.
And it's not that I don't care about women's issues, but I feel annoyed when I get grouped with other men just because I'm a man, despite the fact that race could also play a role in opportunities.
While there are few women in programming, there are a lot more women than Hispanics or people of color. In my opinion this is even worse than the gender disparity. Are the reasons the same? I suspect so, especially #3.
I didn't mention race in the original post, because I'm white. I have no credibility there.
Thank you for bringing it up.
From my (poor) understanding of neurophysiology, it seems to me that women should actually generally be better than men at jobs like software development and engineering. So the fact that women are generally less interested in these things is alarming to me.
If that's so, then we have a major social stigma around technical interests that is preventing all but white and Asian men from gaining even a basic interest in these fields. And this is a travesty of the highest order.
Suppose that 200 years ago you took a poll and most women "simply preferred" not to vote. What does that tell you about whether the womens' suffrage movement was right or wrong?
A better comparison would be if most women didn't vote even though they could. (Because of thugs at the booth perhaps?)
>>When I go to Kindergarten Moms Night and someone asks what I do, "computer programming" usually ends the conversation
I don't this is specific to women - I run into exactly the same thing talking about my tech job with friends, family, strangers, etc.
There's also always an underlying question that gets avoided in these types of articles - what's the "correct" percentage of people of a particular gender/race/whatever in a profession? Obviously we should aim for a society where no woman is pushed away from programming if that's what she wants to do, but is there any reason to think every job should match society demographically?