> If a company does not figure out how to hire, retain, and promote people who are not white men

Agreed. Not hiring the best talent is always bad for a company. If you are biased against your best talent, you've selected worse talent.

However, diversity goals keep moving - they are well past bias issues. Note that tech companies aren't hiring mostly white men at this point - they are in fact outnumbered by Asian men (Google was at 30% and 39% in tech hiring respectively last year.). And yet, we are still talking about the lack of diversity in tech (women remain underepresented, but in terms of ethnicity it's a very diverse place)

Remember this research was done across the globe including countries like Brazil and Japan.

But to focus on the U.S. most really competitive institutions in the U.S. like medical institutions are not running into the problem of having too many white people. Medical institutions have too many Asians, and not enough Black and Hispanic people to have a representative graduating class.

https://www.kff.org/other/state-indicator/allopathic-distrib...

> I would expect a company with more diverse leadership to outperform, because they are able to access more talent.

You implicitly assume that the performance of a leadership team is the sum of the talent of its members.

I'm sure that's important, but the more direct variable to study (somehow) would be how similar people's backgrounds are. If they are all precisely alike, then they'll miss perspectives and be blindsided. But if they are totally different, then they will struggle to communicate their assumptions, and also won't perform well.

So I'd expect honest research to show a U-shaped loss curve. Or more realistically, to have gathered cautionary tales of companies that fell apart for both of these reasons.