It's definitely evidence of bias with a naive prior (i.e. you assume that in other ways a higher SAT-scoring student is as good as a lower SAT-scoring student).

It's definitely evidence of bias with a naive prior (i.e. you assume that in other ways a higher SAT-scoring student is as good as a lower SAT-scoring student). Not proof of bias, but evidence -- in that you'd probably think bias was more likely than if Asian students who got rejected had the same average scores. I think it's likely there is some anti-AA biasing in the admissions process. The alternate explanation, that higher scores on tests are more than cancelled out by lower merit in other all other categories on average, seems really unlikely. The real question is how much of a problem you think that is. This probably depends on your personal system of ethics or morality. Or, more likely, under which of the two alternatives (pure academic merit, or some form of race-based admissions) you personally stand to benefit.