Honest question: are there actually any long-term longitudinal studies of affirmative action that reasonably demonstrate it does or doesn't work? You seem to take it as a truism, but I'm not convinced that concerted after-the-fact addressing of symptoms is necessarily doomed to failure, pragmatically or generally speaking.

Saying there is an issue with the pipeline is he same as saying “white applicants are just better candidates”.

Now, you may be right. It’s a hypothesis. But the other hypothesis is “standard admissions criteria are not adequately measuring how good candidates are.”

One way or the other, you have to make a presumption.

My presumption is that non-white people, as a group, are just as good as white people. That’s why I support continuing to experiment with admissions criteria until we find ones that give us close to equal representation (i.e. affirmative action.)

It sounds like you disagree, you think the tests are measuring reality accurately, and it’s the non-white applicants who need to be better. Why?

This isn’t a scientific debate, it’s a philosophical one. It’s two ways of interpreting the same data.