Expectations (II: Game Theory)

I wrote last week a post called Expectations, about how affirmative action has the effect of undermining the credentials of potential beneficiaries, even if they did not receive any preference. R. Alex Whitlock responded with In Case of a Tie, arguing that quiet racism was still more widespread that I surmised, and might still be in need of correction:

We're all wired to judge people based on appearence and skin tone is one of its most obvious features. We're all inclined to make judgments about people based on incomplete data. An interview is about more than a resume. If it was just a matter of qualified/not-qualified there wouldn't even need to be interviews. During an interview, an employer tries to get a feel for who they are going to hire: Will they fit in? Do they have the personal traits we're looking for? A lot of these subjective judgments come down to impressions. Impressions are subject to our biases. People feel more comfortable with people like themselves. People are more likely to think that those that look more like everyone else will get along better than everyone else.
The point to all of this is that it's not hard to see that the deck might - just might - be stacked pretty heavily in favor of the norm. Racially speaking, the norm is presently white.

(I'm glad Alex makes the point about fitting in, which lets me link to this. Another corrosive consequence of affirmative action is, I think, that it makes it harder for minorities to fit in once hired.)

The exchange I am interested in is in Mr. Whitlock's comments section. I said:

Second, the existence of affirmative action is now the most potent weapon of the quiet racists. It lets them cast doubt on the achievements of any minority student, at any level -- and I can't deny their potential validity.

Mr. Whitlock replied:

The problem with weighing too much into that is that it makes white prejudices ("he only got in because of affirmative action") advantageous to the person that holds them. While on a different scale, it's analogous to using the hardships that blacks faced in the newly integrated military decades ago to keep blacks out. If whites just hold out long enough, their prejudices will prevail.

We need to study two aspects of this argument. First, we must distinguish carefully between struggling against racism and against racists, and also between internal and observable racism. I do not care about internal racism per se, and do not think it a fit subject for government intervention, since people's thoughts are their own. If policy changes would help racists, or be perceived by them as vindication, this is not in itself a significant argument against such changes.

My claim is essentially that affirmative action increases internal racism in such a way as to also increase observable racism, in such a way that it is not beneficial in toto to the groups being "helped". Evaluation of this claim requires a judgement of how prevalent affirmative action is (I believe that it is more widespread in middle-salary white-collar and academic jobs than anywhere else, so perceptions of its impact will depend on one's distance from this hotspot), and on how widespread the quiet racism faced by minority job-seekers is.

I think the idea that affirmative action increases racism is highly defensible. Consider the assaults on Clarence Thomas following Grutter v. Bollinger. The attackers may be quiet racists, but they publicly supported affirmative action -- and their means of defense was to denigrate Mr. Thomas's affirmative-action-aided achievements. Quiet racists would, I suppose, have a quiet chuckle at this. [Yes, it is unfair to use Maureen Dowd's words as a criticism of thinking beings. Blame the yokels.]

I think that Mr. Whitlock's worry, that the end of affirmative action means racist views will "prevail", is overstated. Racism is not gone, and we should try to gently ease it into that good night, but the end of affirmative action is not the same as the rise of racism. Further, if the argument against letting racists prevail is translated into action, you are engaging in a test of strength with a determined enemy, while simultaneously providing ammunition to that enemy. This is unlikely to lead to good in the long run.

[Note: the definitive analysis of the Bolliger cases is by Michael Kinsley.]