> So a very, very rough transitionary solution is to _accept_ that your hiring practices have biases, and introduce targeted fixes for those biases. This looks like discrimination, but is actually _countering_ existing discrimination.
I mostly agree with you until this paragraph. I can quibble about specific points, but this is the core dispute. Multiple attempts at double-blind hiring practices have more or less reproduced the status quo. So you're arguing that our best known scientific techniques for eliminating bias can't completely eliminate this specific bias. That's a big claim, but possible.
In principle, I think one's beliefs should be falsifiable, so from your side, what empirical data would convince you that this systemic bias either doesn't exist in STEM, or isn't a significant factor in hiring outcomes? I don't know if such data exists, nor will I necessarily look for it, but I think it's a question we should always ask of ourselves so I want to know if you're acting in good faith.
For myself, a double-blind trial demonstrating differences in hiring outcomes would have been pretty convincing. That said, I can still see a few possible sources of bias in such trials since I've done a lot of hiring myself. For instance, a hiring process might highly value open source contributions or publicly available projects/source code to evaluate skills. Since minorities are more economically disadvantaged on average than Caucasians or Asians, they might have to work more to pay for school and so have less time to do that kind of work. As such, they might be less appealing on paper.
But would you really call a hiring quota a "targeted fix" for this bias? What sort of unbiased or less biased metric could replace this sort of evaluation? We all know coding tests are terrible.
Finally, "transitionary solution" seems like a polite way of acknowledging that this is an unethical practice intended to move us towards a more equitable society more quickly than an ethical approach, like studying the problem to find actual targeted fixes. But:
a) It then seems disingenuous to be offended when people question the ethics of compromising one's ethics, even if it's (allegedly) a shortcut to a greater good. Not everyone is a utilitarian, and calling these people evil/racist/sexist for disagreeing with compromising one's ethics is dishonest at best.
b) This is then subject to an empirical evaluation as whether this shortcut actually works. There's considerable debate about whether past attempts at affirmative action were actually successful, so there's some reason to be skeptical.
For myself, I'm no expert on all discrimination issues, but I've read quite a bit on the gender disparities in STEM, the stats, papers and so on. I'm quite confident that hiring quotas won't fix gender disparities one bit. I don't want to get into that debate, but I've received quite a bit of pushback for arguing that politics should be subordinate to such evidence, and not vice versa.
> It's easy to look at it and go "this person is unqualified, which is why they needed a "special" path", but there's little evidence that this is actually true, and most suggestions that minorities are simply worse overall than white men at a given task are usually pretty empty.
I agree, there's probably very little average difference in ability between all human populations for STEM-like work, given sufficiently similar experience. But what can we use to objectively evaluate experience/claims of qualifications?