Admissions at UCLA are supposed to be race-blind according to state law.
Here is the report that UCLA Political Science professor Tim Groseclose submitted upon resigning from the faculty committee overseeing admissions, which argues that UCLA is in violation of state law. I have not read it all but the parts I looked at are pretty blistering.
The fundamental problem at schools like UCLA, Berkeley and Michigan is that the preferences of the citizens for race-blind admissions, as expressed through citizen initiatives, do not coincide with those of the majority of faculty members and administrators at these schools. The resulting principal-agent problem means that the effect of passing such initiatives is that race goes from an explicit criterion to an implicit one, either through the means described in Tim's report, such as students reporting their race in their personal statements, or through the use of variables correlated with race, such as economic disadvantage, that can be justified on other grounds.
Although I think college quality matters, and have written papers (e.g.
here and
here) that provide empirical evidence that this is the case, even if you buy the larger estimates in the literature the amount of fire and brimstone (and money and faculty time and administrator time and media time etc.) devoted to this issue seems to me to be well in excess of the possible net costs and benefits times the number of students at issue. Minority students talented enough to get into Berkeley or UCLA or Michigan under affirmatative action but not under their current systems should, in general, get into very fine private schools and/or public universities of nearly equal stature in states still using affirmative action and/or marginally less selective public schools in their home states, such as the lower ranked UC schools in California or Michigan State in Michigan. The exact same point holds for white and Asian students on the admissions margin in states with affirmative action who get crowded out by students admitted under affirmative action. Each policy surely has winners and losers but, especially if mismatch is an issue for some students, the average treatment effects of either affirmative action or its absence seem to me to likely be quite small.
Of course, more research would be useful here; there is no serious cost-benefit analysis to point to in the literature that I am aware of, a fact that should be an embarrassment to both sides of the debate.
At present, I am tempted to argue that the world would be better off if everyone concerned spent the time they would have spent fighting about this issue doing consulting and then donated the resulting money to fund scholarships for talented kids from disadvantaged backgrounds.
