When you look at the situation in the United States, and try to take all the factors into account to figure out whether or not women are being discriminated against...
... you'll come away with your head spinning because there's just too damned many factors.
Consider the factor of giving generous paid maternity leave. Obviously a benefit. Except that a company won't want to hire or pay someone that may disappear like that. Except they are legally obligated to. Except these things can't really be legally obligated because there's always a way around it. And of course a certain percentage of women really do of their own free will choose not to come back, which is hard to put an economic value on for any interested party, but if you can't put a reasonably concrete objective value on both the value of having that choice, and the values of exercising that choice, how can you tell whether women are being shortchanged? Being guaranteed that you can return to your job after X months is still some sort of economic value, even if most of the X months is itself unpaid.
It isn't even necessarily that hard to find women getting paid more than men by some metrics now: http://www.time.com/time/business/article/0,8599,2015274,00.... Note, I'm explicitly calling your attention to the various caveats and logic in the article, because this is an example of my real point about the complexity of the topic, not that one article proves anything about the "real question"; my point is exactly that nailing down the "real question" is very, very complicated, well beyond simply totalling up who makes what and looking at which number is larger.
At this point, given the very market forces that this article talks about, I am unconvinced there is a major systematic problem that comes from human bias; I think the variance caused by the non-zero-but-small-in-practice effect is dominated by the other various concerns that cause disparities. I'm sure there's sexism afoot but I suspect it's been pushed below the noise threshold. But I can't prove it.
Yes, men's versus women's opportunities are complex in the US. Whether they are exactly equal doesn't matter. They are close enough that the most competent, skill women have good opportunities - which means that there's less of a reserve of the cheap and competent women's labor out there.
This is might be a explanation for the decline of women in computing fields. What could be happening that there are now more other fields where the most competent can go and not get any hassle at all or even experience preferential treatment. Thus the only reserve of the cheap and competent labor in the field would be those who are drawn to computing from early age. For whatever cultural, biological or other reason, these are mostly men and so this would explain why programming field has become more and more male over the last twenty years.
I know that the proportion of blacks who teaching in inner city schools declined tremendously from, say 1965 to 1985 because the barriers to those black who attended college dropped down and many more went into industry instead of teaching.
... you'll come away with your head spinning because there's just too damned many factors.
Consider the factor of giving generous paid maternity leave. Obviously a benefit. Except that a company won't want to hire or pay someone that may disappear like that. Except they are legally obligated to. Except these things can't really be legally obligated because there's always a way around it. And of course a certain percentage of women really do of their own free will choose not to come back, which is hard to put an economic value on for any interested party, but if you can't put a reasonably concrete objective value on both the value of having that choice, and the values of exercising that choice, how can you tell whether women are being shortchanged? Being guaranteed that you can return to your job after X months is still some sort of economic value, even if most of the X months is itself unpaid.
It isn't even necessarily that hard to find women getting paid more than men by some metrics now: http://www.time.com/time/business/article/0,8599,2015274,00.... Note, I'm explicitly calling your attention to the various caveats and logic in the article, because this is an example of my real point about the complexity of the topic, not that one article proves anything about the "real question"; my point is exactly that nailing down the "real question" is very, very complicated, well beyond simply totalling up who makes what and looking at which number is larger.
At this point, given the very market forces that this article talks about, I am unconvinced there is a major systematic problem that comes from human bias; I think the variance caused by the non-zero-but-small-in-practice effect is dominated by the other various concerns that cause disparities. I'm sure there's sexism afoot but I suspect it's been pushed below the noise threshold. But I can't prove it.