In more or less defending Henry Louis Gates Jr. for bullying the Cambridge police and playing the race card, Michael Kinsley opines:

[G]eneralizations about race don’t lead only to bad things, like an unjustified arrest. They can lead to good things, too. The best example is well known around Harvard Square and other academic communities: affirmative action. Part of the rationale for affirmative action is that African Americans are more likely than whites to have struggled harder, under the burden of greater disadvantages, to reach the point where they are poised to enter Harvard. Therefore, they deserve a break. No doubt this is true on average. And no doubt it is false in many cases. You can easily decide that some generalizations are just too toxic to allow, even if true on average, and race might be a good area to start. But you’d be hard-put to justify forbidding racial generalizations in split-second decisions during tense confrontations between citizens and cops, while allowing them in the relatively leisurely precincts of a college admissions office.

Let’s put aside whether the generalization here was based on race or on Gates’s abusive and unruly behavior, but Kinsley has it wrong. There is no good generalization (he means discrimination) based on race in the academic or any other setting. In the course of trying to defend “good” discrimination, Kinsley goes badly astray.

For starters, affirmation action has rarely been justified and was not instituted on the notion that minority candidates possessed inordinate virtue. Whether as a remedy against past discrimination or as an effort to produce a “diverse” environment, academic institutions generally adopted race preferences to lift up students who lacked the same qualifications and achievements as their white (or Asian) peers. If one wanted to search for those who “struggled” with adversity, we would long ago have adopted class, not race-based, preferences. And we wouldn’t be filling diversity slots with foreign students, children of third-world ambassadors, or the not-so-hard-working upper-middle-class minority applicants whose SAT scores are hundreds of points lower than their peers’.

And of course, in rooting for “good” discrimination in universities, Kinsley neglects the very real victims who are excluded in favor of lesser-achieving affirmative-action candidates. There simply isn’t any basis for equating “struggled harder” with race.

So when Kinsley hints that we shouldn’t be so hard on generalizations in the criminal justice arena if we are going to stick by them in the academic one, he ignores the obvious rejoinder. To paraphrase Chief Justice John Roberts, if we want to stop this generalizing by race, we should stop generalizing by race — in all contexts.