Affirmative Action

This seems to me to be the wrong way of looking at it.

Virginia Postrel points out suspicious standards in admissions at UC Berkeley. The San Francisco Chronicle analyzed the school's new admission policy that was supposed to get the university in line with Proposition 209.

According to the article:

More than 400 students -- nearly 90 percent of them minorities -- were admitted to UC Berkeley in 2001 with below average SAT scores under an admissions policy that was to have ended racial preferences at state universities, The Chronicle found in an analysis of admissions data.

UC Berkeley officials developed the policy, which considers grades and SAT scores but includes other factors, such as socioeconomic status, after voters approved Proposition 209 in 1996 to ban affirmative action in admissions.

Bu the analysis of the data shows that of the 422 among the bottom tier of admitted students, 378 were minorities. Seventeen were of unknown race and 27 were white.

The paper seems to be concerned with the results — most bottom-tier students are still minorities, even under the new system. Now, I'm no fan of race-based admissions, but this seems like a ridiculous reaction without further evidence that they are secretly flaunting the law.

Many of the principled objections to affirmative action (including the Supreme Court's prohibition against dictating outcomes) leave room for (or even demand) the use of these other socioeconomic metrics to level the playing field. They just require that they be applied equally to all races. I've argued for additional metrics that take into consideration household income, relative school performance, number of parents, level of education of parents, etc. All of these would embody the (I think legitimate) principle that people who perform equally well, but in a tougher environment, deserve a leg up. And they deserve it regardless of race.

But...., all else being equal, you would expect that any such program would help minorities much more than whites. First, because minorities do worse on the traditional metrics in the first place, you would expect a large percent of "bottom-tier" students (when, I'm assuming, ranked by these traditional metrics) to be minority applicants. Second, because more minority students live in the kinds of situations that you're controlling for with the new system, you'd expect them to be over-represented still more. All without suspicious standards.

Those who are worried about this seem to have lost their principles and suddenly, they're the ones shooting for outcomes. They begin to look more like the charicatures painted by advocates of AA — they just don't want minorities to succeed.

"It is outrageous. They don't have any business going to Berkeley," said [Chair of the UC Board of Regents, John] Moores, who did his own preliminary study of 2002 admissions data recently without looking at race. He was intrigued by the 2001 data and said it appears the students were admitted for "all the wrong reasons."

"I always expect the kid that doesn't test well that turns out to be a whizbang, but there are not hundreds at Berkeley. It can't be," Moores said. "I believe there is a huge element of social justice behind some of the (decisions). I question whether people are really being honest of what the chances are of students being successful."

So, he's not just against race-based admissions, he's against social justice....

Bottom line, if UC Berkeley is really still using race but just obscuring that fact, it's a problem, but other than innuendo I don't see evidence of that. If they've actually switched to a system that is truly race-blind but manages to take into account the fact that the playing field is not level, and this actually helps many minority students, then I'm just ecstatic that someone finally got it right.