I think that the reason people fail to understand why there are so few women in software development is because they frame the issue the way you just did. They think about how to convince a young woman that engineering is better than home economics.
Instead, try this as an exercise: imagine you're a career counselor at a university, and a very bright young woman has just entered your office asking for advice. She's a top student who can handle math, science, and arts/humanities, and she's got the chops to score in the high percentiles on whatever standardized test is required for a grad program.
Imagine a conversation where you try to talk her out of becoming a cardiologist and sell her on a grad program in computer science. Make sure you consider employment prospects, earnings potential, social prestige, job stability at age 48, and so forth.
This conversation will demonstrate why it's not engineering vs staying in the kitchen. Young women are starting to out-achieve young men in the US, and I'd be surprised if this isn't the case in the UK.
People (who are aware that the achievement gap is now starting to favor women) often think of the scarcity of women in engineering as a dark spot in an otherwise bright picture, but one that should still be remedied. I think they're missing the possibility that the evident aversion women have to engineering or CS is actually a positive thing for women - that women have realized that they can achieve more real power and influence in other fields. By "better off" I'm not talking about some condescending "oh, you'll be happier at home with the kids" kind of thing. I mean wealthy, high income, highly influential, positions of power in business and government kind of "better off".
Now for a slightly lame backtrack - I actually do think that science and engineering are amazing fields, and the scarcity of women is a complicated thing - especially at the undergraduate level (after all, the most common academic background at the bachelor's level for fortune 500 CEOs is engineering in the US). However, I do think that the perspective I described above needs to be a big part of the analysis. People seem to view the low participation of women in certain technical fields as a negative for women without adequately understanding the other things women are doing and the possibility that they may be making better choices than men.
really?
if you had a teenage son or younger brother, would you be encouraging them to be a doctor, dentist, or lawyer? or would you be encouraging them to learn about computers, software, networking, mobile phones, social networks, the cloud, the web, and all the other cool things we love that are changing the world?
But if we're concerned about how women aren't going into graduate engineering or CS programs, that means we think something's wrong with their choice, right? It might not be their fault. It could be discrimination, an unfriendly environment, a lack of friendly encouragement (or a lack of pressure, which also works, though often at the expense of the person's happiness).
So think of it this way - if my daughter was interested in following grandpa's steps and going into medicine rather than dad's steps and going into computing, I certainly wouldn't discourage her.
BTW, I've posted about an article from the RAND foundation about the "shortage" of science and engineering PhDs in the US (it's easy to find on their site). RAND found that this is, essentially, a rational reaction to poor prospects and pay relative to the professions. I understand we're not robots, passion matters. But when researchers are RAND are reaching this conclusion, it starts to be very difficult to hold the position that women are somehow losing out by choosing professions over grad programs in CS/Eng.
> it starts to be very difficult to hold the position that women are somehow losing out by choosing professions over grad programs in CS/Eng.
but that's not what this article was about. we're talking choices at 16 years old, long before people have chosen a specialty.
That is a) an unproven assumption and b) irrelevant. The right thing to measure is the integral of P(inventing next big thing | intelligence = g, enters computing) P(intelligence = g) dg. (This also assumes g is the only factor, not ambition or other factors, but whatever).
A not so terrible substitute for that would be measuring the ratios of men and women with intelligence over some cutoff . This could bew considerably lower even if women are every bit as intelligent as men, if for example the variance of women's intelligence was lower.
But just a nit - I'm no longer at Courant, my webpage is stale.
Assuming that women are every bit as intelligent as men, our field is missing out on roughly half the population of brilliant minds that could be inventing the Next Big Thing. No?
Edit: OK. Summary of the answers so far. A) Cannot necessarily make the above assumption. B) Women on average may be as intelligent as men but they will never be geniuses like Einstein (or likely to go to prison). and C) We don't need that many engineers anyways.
Conclusion: they should stay in the kitchen