> Diversity programs are not about lowering the bar, they are about outreach.
At colleges, lowering the bar is in fact how they have often been implemented. This article has some decent numbers:
https://www.nytimes.com/2017/01/30/opinion/white-students-un...
Lowering the bar is also the easiest way for a mid-level manager to achieve a target percentage. It's probably also the first method that may come to mind for many people.
Also, if you spend 3x the resources to find 3x the number of candidates (is it linear? I dunno), conduct normal interview processes, and then randomly throw out 2/3 of the male candidates that passed, while keeping all the women that passed, that is one way to triple the number of women you hire while having the same bar for men and women. However, the step where you throw out a random 2/3 of the male candidates might strike someone as highly irrational. Why not throw out the bottom 2/3 of male candidates? But then you end up with male candidates that passed a higher bar. You're not lowering your standards for women, but you're raising your standards for men, and then employees might notice and form the impression that the average male employee is better than the average female employee, which would be unfortunate. But it would probably be difficult to convince the people involved that avoiding the "unfortunate" outcome is worth throwing away 2/3 of the very best male candidates.
It might be possible to spend 3x the resources in such a way that all the new candidates you find are women. That would be a way to avoid the above problems. How would this be accomplished? Perhaps you send a recruiter to a university, and have them only talk to the female students. Hmm, that might be a little weird. How about a female-only university? That would be great. Likewise any female-only programming meetup groups. And of course any qualified female engineers that anyone knows—they'll never be wanting for a job.
How well will a company perform that spends significant resources recruiting from female-only groups, compared to one that recruits from groups that simply take the best candidates? How significant a fraction of company resources are normally spent on recruiting? And I did see a study indicating that demographic diversity (measured as age and sex) improved morale. I guess we'll see.
What would really help is if universities published the CS accomplishments of all their female students, but not of their male students. I wonder if they'd be willing to do it.
In any case, my original point was, even if a company does in fact implement its diversity program with one of the methods that does not systematically lead to higher-quality male hires than female hires, I think it is natural for someone to assume until proven otherwise that it is implemented by shifting the bar.