There's DEI in rhetoric, DEI in actual practice, and DEI as it could be practiced. These aren't necessarily the same when it comes to hiring.

The ideal is of course not to do the above. But the ideal isn't ever what actually happens. So, yay!, nice ideal, and let's see what we can actually do.

If you look at how the success of or need for DEI initiatives are measured, it's usually something like "does the percentage of this group match the general population".

Are you actually hiring the general population for a Managing Editor position? Heck no! Expertise matters. Are you hiring the general population for AI finetuning jobs? Heck no! Hiring the general population as speech pathologists for schools? Heck no!

But what's the bottom line that people most often use to judge companies? Why, it's: "do your group demographics match the general population?"

Speech pathologists turn out to be over 95% women. If you're running a large school district and you want gender diversity among speech pathologists, and you expect to be scored based on demographics of the general population, of course you are going to have to hire every man you can get your hands on, no matter how incompetent, and keep them on staff, no matter how bad their performance, and you still aren't going to meet your goals.

And you have to do the best you can to avoid a Title VII lawsuit while keeping these incompetent men on staff.

Speech pathology has a pipeline disparity. You can't fix it during hiring. You can shift it a little bit, you can make sure your outreach includes places to get men, but you're not getting 50% male speech pathologists.

If you judge your DEI by "less than 50% is a failure" and people do, then you bet that the in-effect pressure is to hire incompetent people.

In the speech pathologist case, to hire every man regardless of competence.

Now, companies do actually hire people to do things, and they have quality controls, so more often what happens in situations like these is the companies spend a bunch of money on DEI, ratios don't change appreciably (because they already were hiring mostly based on the available talent), the strongly pro-DEI crowd yells about how the companies don't care about DEI and are horribly biased white patriarchal colonial whatever whatever, and...the companies spend more money on DEI and nothing changes, and so on.

The problem is that DEI is being asked to fix society, and it's not big enough to do that; and people aren't sophisticated enough or mature enough to reliably come up with reasonable goals for DEI efforts.

Sensible DEI efforts that keep firmly in mind what can and cannot reasonably be achieved can be a boon to both companies and society as a whole. Companies can hire good people and make sure people of different backgrounds feel empowered. Colleges can admit people with the prerequisites needed to take advantage of the level of education they offer and make sure people of different backgrounds feel empowered.

But by asking for far too much, the strongly pro-DEI crowd threatens to warp the efforts into what the strongly anti-DEI crowd charges that it already is. If you look at the reality of what DEI has accomplished, it's plain that mostly it isn't warped in that way. But it would have to be to make the strong pro-DEI crowd happy with the outcome.

Meanwhile, this all distracts from fixing the problems at the source, most of which are disparities in childhood educational opportunities (taking all things fully into account, including food insecurity, emotionally abusive living conditions, etc. etc. etc.). It's not everything, but if you ask "which factors contribute how much to the unfair disparities we observe", then in most cases it's education first, wealth second, and everything else together less important than either of those two.