People are often hired to fill requirements for diversity, all over the place.
It's not insane, not at first. sometimes the hires are fine, and people fit in. but it's tougher and if they don't have the skillset or experience, it's worse for everyone.
management has to deal with the fallout and repercussions regardless. And i guess it's doubly meaningless since diversity hires that come through pad out their resumes with references and experience so they can get jobs elsewhere, while the fallout hits the rest of the workforce and leaves a bad impression with the others in the company/business as to the management's competence / motives.
The argument people want to make is that the hire process is hindering women because they don't get through the interview process as women candidates, because of stereotyped or behavioral /perceptual norms that women aren't capable, or have other subliminal biases.
That could be true. Maybe. You can pretty quickly see that there's a slight chance, but it relies on the same kind of kitsch argument that if you have lousy management or interviewers, then women can't get past the resume or interview wall.
Just like most people firing off hundreds of resumes these days, if you can't get past their biases, you're not getting in the door, let alone the job.
So hiring and PR companies attempted to correct after bad PR, and in the 2000s, brought more women into the HR pools to review and interview. And, that didn't work. So, they built blind review and interview systems where people did open tests or tried to separate the management or review candidates into test rounds and skipped women through the review stages so they'd be put up against filtered candidates who had passed those tests, blanked out names in candidate pools, etc.
There's also a few infamous examples of candidates who changed their names to be more anglicised to get interviews. many examples exist, even a few included in freakonomics,
The famous symphony curtain example (it must be the only positive one available since i hear it every year), or Fortune,
http://fortune.com/2016/06/08/name-bias-in-hiring/ sic. And UBT, training people out of habits or behaviors through essentially, brainwashing to move the dial 2% or 1% in a more diverse way.
However, 20 years later, the stereotypes are going the other way.
The Australian Public Service made this handy animated video with their free time ... on how actual people in 2017, i.e. Current Year, performed against blind testing.
i.e. It didn't go well for the feminist arguments of subliminal bias.
there was a clear 3% pro-women in the management sample and a 7-8% bias for pro-ethnic names in the candidate pool.
As for Vidya Narayanan, there's very likely an impossible disparity to overcome because women aren't disposed to technology. As much as i'd want to champion or believe in the whole "Women in Tech" parade,
the candidates are predisposed, people will choose women when they apply, teams are willing to train on the job, or even handicap their team in order to fill diversity quotas or extend training out to a whole team so that the new hire can catch up. but it's still going to be 3 people out of 200. or 8 if they relax the standards and let any women in. The numbers are minimal. And telling.
I don't buy the misogyny angle, just because i've gone to 10+ coed schools and colleges, universities, and girls want different things once they get into college/university. This might change, as while programming is stupidly easier than when i was growing up in the 90s, I've been in 5 different university courses with a co-ed group makeup, it was never higher than 4% except in games design where it was 16% depending on who turned up out of ~100 students.
You could setup "preschool programming gulags" and develop "Kindy Kreatives" "Girl Game Jams", it won't push women in. It can help those that are predisposed, but it's not going to keep people around for long.