Here is what I don't get about your argument. You're using one theory of social conditioning to explain why women stay away from CS (the social status of the field), while implicitly rejecting the more obvious bit of social conditioning: the perception that engineering is for socially maladjusted men.
And neither provides a principled reason to maintain the status quo. Why should the socially maladjusted men already in engineering (as you call them) get to define the culture of the field?
What if tech companies tried to actively change the culture of the field. What if, say for 15 years, tech companies put a thumb on the scale by actively trying to recruit say 25% women. You'd definitely have to take slightly less qualified candidates in order to meet the quota, but it would be transient. After 15 years, there would be a critical mass of women in the field, and the perception of the field as being for socially maladjusted men would be greatly diminished. Tech companies would have enough women applicants to maintain their new ratios without putting a thumb on the scale. You can say it's "unfair" to the men who would have gotten those jobs in the interim, but is it any more "fair" for a group to keep another group out of a lucrative profession by defining a particular culture?
How do you change the perceived social status of a job? For example, could you change garbage disposal into a high status job? Or maybe some jobs are simply inherently low status, and being in direct contact with oily machinery (or confusing computer code) will always be among those jobs...
Computer programming is well paying white collar work. It's detail oriented and not social, but neither are accounting and law and those professions have plenty of women.
I'd argue that at face value law is a lot more social. Don't know about accounting - maybe it is not so demanding on the brain - more routine work and less hard thinking? I don't say that to imply women are less good at thinking, just that having to think in general might be unpleasant (the book "Thinking slow and fast" seems to imply that, the brain tries to avoid having to actively think at all costs).
Engineering is generally more social than law. From the minute you get to engineering school, it's all about working in groups. I was shocked to find out how much people in law school hated working in groups. They were humanities majors--they spent all of undergraduate writing papers by themselves instead of working on group projects. When I was an engineer, I had regular team meetings, talked with my team about design challenges every day, kept up correspondence with my customers' teams, etc. As a lawyer I might hole myself up in my office working on a brief for days at a time. My "team" might be three other people on a case who I touch base with once a week.
As for accounting being "less demanding on the brain" I think now you're just grasping at straws.
So you think every accountant could as well be a software developer? Do they earn equal money? Otherwise, why wouldn't they become software developers instead?
With respect to law vs engineering, my point was about the public perception of the jobs, not the actual reality of it. The public perception of lawyers is people wearing stylish and expensive clothes eloquently fighting for what is right. (I am talking of TV and movies).
Absent a large rise in the earnings of engineers relative to lawyers the skills demanded by both professions are such that the average lawyer will always bemore verbally facile than the average engineer. Whoever has better facility with words has better social skills, on average. So lawyers have higher social skills.
Your thought experiment in which companies discriminate against men in hiring might work. It would need to be forced upob all participants in a field/economic sector because if it was voluntary those who were not discriminating would be hiring from a superior talent pool. They'd win, on average. I thonk it would be unfair to women who could have made it on their own merits too, but hey.
Nobody had to force law firms to try and increase their representation of women, at least in any systematic way. Law firms did it because lawyers are a very liberal bunch and demanded it from within.
I don't see why the same couldn't work for engineering. Would engineering companies have to dip into a slightly less talented pool to begin with? Probably. But here is an interesting statistic. At my alma mater, the entering men have about a 30-point higher average SAT math score than the entering women. Of course, the school has gotten more competitive over time, so the women entering today have about the same SAT math score as the men who entered as freshman in my class (about a decade ago). Think about that--the cohort of "affirmative action" women today have about the same level of mathematical aptitude (measured purely by test scores) as the cohort of experienced male engineers who are 5-6 years out from graduation. That's the magnitude of the thumb on the scale, and I'd argue it's very small.
I think the bottom line is this--there is no reason there should be fewer women in engineering than can be explained by any difference in aptitude of women for engineering. In terms of aptitude for engineering, the starkest difference we see is in standardized test scores, particularly the Math SAT. In the range that characterizes engineers from good schools (~700) the pool is ~40% women. That's the upper bound. I honestly think you could get an equilibrium in the engineering field with ~40% women, an equilibrium that would be stable, once established, with no affirmative action.
Your point about possible stable equilibria I don't have any beef with. But honestly I think the “problem” is more on the supply than the demand side. Fewer women find engineering/technology attractive than men. Women have better options given their preferences than tech.
Normally this would be where I linked to the Study of Mathematically Precocious Youth paper showing girls who entered math competitions withr similarly skilled boys to leak out of the tech/science/engineering path much more but I'm on my phone.
And neither provides a principled reason to maintain the status quo. Why should the socially maladjusted men already in engineering (as you call them) get to define the culture of the field?
What if tech companies tried to actively change the culture of the field. What if, say for 15 years, tech companies put a thumb on the scale by actively trying to recruit say 25% women. You'd definitely have to take slightly less qualified candidates in order to meet the quota, but it would be transient. After 15 years, there would be a critical mass of women in the field, and the perception of the field as being for socially maladjusted men would be greatly diminished. Tech companies would have enough women applicants to maintain their new ratios without putting a thumb on the scale. You can say it's "unfair" to the men who would have gotten those jobs in the interim, but is it any more "fair" for a group to keep another group out of a lucrative profession by defining a particular culture?