“This would all be some feel-good diversity policy if the ten percent students failed to produce results. In fact, they earn higher freshmen grades and stay in school and graduate at higher rates than students accepted by all other methods, even ones with higher SAT scores. In other words, the ten percent admissions policy does a better job of screening applicants than the university’s own admissions office.”

Can you provide the empirical basis for this? I’m interested in seeing the numbers themselves.

It seems to me that your last statement — that the ten percent admissions policy does a better job than the admissions office — is almost surely an misreading of the evidence. No one is disputing that many of the best college prospects are in the top 10% of their classes; many of these people would be accepted anyway, under any system. It’s unreasonable to compare everyone in the top 10% to people outside the top 10% who had good SAT scores, and declare that since the former does better than the latter, a crude cutoff is more predictive than the admissions department’s evaluation. To get a real sense, you’d have to limit your sample to those in the top 10% who would have been rejected if not for the 10% policy — and I suspect that the outcome for that group is a lot different.