Monday, January 09, 2006

Affirmative action and current bias

From the pro-affirmative action side comes a great new paper trying to defeat what I think is one of the two or three most persuasive claims of anti-AA folks: that there's currently not much individual or institutional bias, and so current open-minded people shouldn't be punished for their jackass forebears. Even though there are individual inequalities resulting from past discrimination, they'll be washed away eventually because the system is structurally fair.

In comes Implicit Cognitive Science, claiming that there is bias, and we know this because we can look behind people's considered answers and go to measure their snap judgments. This is actually kind of frightening. I'm not sure I want to know what I know--I'm sure I'm not even as wholesome as I like to think I am, and that's not necessarily that high a bar. (Sadly, though I try.) The bias also undermines the structural system, because it means our measures for merit might be systematically skewed against the other. [Description of ICS here. Two online tests available.]

A few considerations: first, people have huge incentives to try to not only cover but correct for their implicit biases. Can they successfully get outside themselves to the bigger picture of neutrality? In some ways, it is amazing how neutral we are, given how biased ICS tells us we're supposed to be. We regularly refuse to apply stereotypes that have evidentiary value. Heroic, almost, if we do it well enough. But if we are really so good at consciously overriding our unconscious, shouldn't it show up in our snap reflexes eventually? Is ICS a refutation of the success of these attempts?

Will O'Connor's vague words in Grutter v. Bollinger about an end to affirmative action ("in a generation or two") be replaced by some national test of debiasing? After the Social Contact Hypothesis has had a chance to de-other the other, will we be a measurably happy family? Can that ever happen without an end to the socio-economic gap, though? As long as minorities are poor, they'll be distinguishable and we'll have less contact with them, and so affirmative action might have to last as long as inequality does--which goes expressly against O'Connor's formulation.


Post a Comment

Links to this post:

Create a Link

<< Home