"Gender gap" is a fancy way of saying "not 50%". More to the point, assumptions are often unstated, as I'm sure you know.
I agree with you insofar as the mechanisms you point out are likely contributers to this so-called gap. Rather, I'm highlighting that -- unpopular though this notion may be in certain circles -- part of this effect is likely to be driven by bottom-up biological factors.
How big is this part? I haven't got a clue.
I'm copying my "tl;dr" from the above comment because it applies here as well: if we're going to have an honest debate, we have to entertain the notion that at least part of the gender gap is the result of "nature", rather than "nurture".
Addressing disparity in the pipeline is a separate issue.
No. Gender gap is underrepresentation relative to the qualified labor market. Right now the gender ratio in entry-level software engineering jobs does not represent the segment that would be interested in and qualified for them (CS grads).
The 10% difference between grads and junior developers is on us to fix.
You say elsewhere that while 20 percent (the actual number is around 18) of CS grads are women, only 10 percent of junior devs are in fact female. Others have mentioned that maybe the women went on to graduate school or are working some CS related job that doesn't fit the definition of "junior dev".
The latter points seem reasonable to me but you keep bringing it up though so here's my question, where do you think 50 percent of female grads are ending up? The unemployment line? That would seem like a very large number of people, such a large number that it would have been recorded somewhere. Do you have any actual statistical data beyond the raw numbers you keep bringing up?
But that's not the point. A supply of women with both the qualifications and the interest exists. We ought to be seeing a representative sample of that supply, and we should do what we can to avert bias within our organizations.
According to your source, the unemployment rate for women in Computer and mathematical occupations in May was 2.6 vs 1.8 percent for men. That's actually a 31 percent higher amount. Still high but not 50 as you claimed in your previous post. It also says the rate for May of 2015 was exactly the same for both genders at 1.5 percent. This issue has gotten a lot of press in the last 12 months so it is very unlikely that suddenly the numbers are disparate as a result of discrimination after they were the same a year ago.
And in other sectors on your chart like in Management, professional, and related occupations, the unemployment figures are the opposite. Women unemployment went from higher to lower than the men's rate. I mean, in Architecture and engineering occupations, the unemployment rate of men is literally 100 percent higher than women at 2.2 vs 1.1 percent. It just looks like natural statistical gyrations of numbers and not conclusive or even particularly indicative of anything.
And the point is, if you are saying that huge numbers of women are graduating with these degrees and then being shut out of the job market, you can't just stop there, you have to back that up. Even ignoring the fact that just because someone has a degree doesn't mean they're qualified, quoting the percent of females in entry level dev jobs doesn't say anything about where the other women went or what they are doing. I'm pointing out that if they're out there looking for these jobs then the unemployment statistics should reflect that and via your source, they don't. Other posters have suggested graduate school and other jobs not described as entry level dev and so far that explanation seems more plausible. Ask yourself, what are CS graduates qualified to do besides be junior devs and you may be surprised with how many answers there are.
I think where your confusion lies is in order for there to be discrimination, there has to actually be discrimination. Point out the discrimination, who's doing it, which companies, what policy allows this, and we can all rally and do something about it. But just throwing out cherry picked employment numbers and claiming bias is somehow self evident which is what your argument implies isn't good enough. It actually comes across as a kind of weird p-hacking[0]. Even worse is when you introduce a credibility problem by conflating different employment sectors and inflating your numbers like you have here.
What if some percent of those women simply went into a slightly different career?
There are structural problems that make the available talent pool unequal per-capita anyway (women being socially pressured out of software engineering, resulting in a lack of role models, etc. I'm sure there are plenty of other factors).
Part of changing the dynamic will take decades because _currently_ a smaller percentage of all women in the world are software engineers vs men. No one wants to hire incompetent women engineers just to fulfill a quota and all the people I've talked to about this seem to recognize and understand it. Mostly what seems to piss people off (and rightly so) is when people try to claim there isn't any bias, there isn't any problem, and everything is just fine as-is.
If all else were equal, what total percentage of women would want to be software engineers and would be good at it? I have no idea, but it does seem clear that we are far away from that number. As long as we can all agree that the structural problems should be addressed then I think we also understand that it isn't going to be 50/50 overnight and it might never be 50/50. But 90/10 is clearly unbalanced in a way that is not correlated with women's aptitude or ability to write software.
I think we should be a welcoming community and make reasonable efforts to make anyone who wants to join it feel comfortable. I don't understand the constant desire to rules-lawyer or dismiss such a simple and straightforward idea: don't be an asshole, be considerate of others, encourage those who are interested to learn.
In the end I have to think it is partially driven by fear and self-doubt, like maybe if we let black people or more women into the club it will all be over. Global demand for software engineering talent outstrips the supply and software is eating the world. My talents have never been in higher demand. My life is not diminished by nursing mother facilities, unisex bathrooms, coding events in black neighborhoods, attempts to reduce bias in interviews, et al.