I'm not immediately against Twilio's actions here (though the messaging is pretty bad), but not one of these examples strikes me as a bad thing beyond layoffs b

I'm not immediately against Twilio's actions here (though the messaging is pretty bad), but not one of these examples strikes me as a bad thing beyond layoffs being bad in general. Deciding whether or not to close an entire office should probably not take into account the racial makeup of that office. You either have the funds to keep it open if you don't, because if you now keep the office you have all these other expenses and end up having to lay off even more people. If someone doesn't meet the bar to survive a layoff saying "oh but they help us hire people with green skin, keep them on!" is pretty blatantly saying "you're not good enough to do your job, but you're a good enough token to keep your job." It's offensive and discriminatory. Recently hired, early-career/less-skilled, and QA/test are always let go first before mid-career and seniors who aren't at the top of their pay band. If you're at the point of double-digit percentage layoffs, the testing group is a pretty attractive target. I think the issue I have is that I have no problem whatsoever with examining your hiring (and firing) practices to try to eliminate implicit or unintentional biases against any group - disadvantaged, underrepresented, or not. What I have a problem with is saying "here is a completely legitimate, bias-free business decision that is in the best interest of the company (e.g. closing the Atlanta office for cost savings). But it hurts this minority group, so we're not going to do it, which puts the business on less solid ground and has a larger net negative impact." If bias results in you disproportionately impacting one group over another, you should fix it. If business does, it's more racist to try to "fix" it by taking everyone's demographics into account.