Kieran Healy

False Necessity

The normally sensible Michael Kinsley gives this analysis of the Supreme Court decisions on affirmative action:

Admission to a prestige institution like the University of Michigan or its law school is what computer types call a “binary” decision. It’s yes or no. You’re in, or you’re out. There is no partial or halfway admission. The effect of any factor in that decision is also binary. It either changes the result or it doesn’t. It makes all the difference, or it makes none at all. Those are the only possibilities.

For any individual, the process of turning factors into that yes-or-no decision doesn’t matter. Any factor that changes the result has the same impact as if it were an absolute quota of one. It gets you in, or it keeps you out. And this is either right or it is wrong. The process of turning factors into a result doesn’t matter here, either. In this sense, the moral question is binary, too.

This isn’t right. Doesn’t Kinsley know the difference between necessary and sufficient conditions? He’s writing as though race (and everything else) were both necessary and sufficient—either it makes all the differene or none at all. Last time I looked, though, being a member of a racial minority group wasn’t sufficient to get you admitted to Michigan Law. It wasn’t necessary either, as a glance at most of the law students there will tell you. Of course, minority status can raise your chances of admission, in conjunction with other conditions, and may sometimes be an added variable that pushes a candidate into the ‘Yes’ pile. That’s the whole point of affirmative action. But it’s wrong to argue, as Kinsley does, that “’no role’ and ‘determinative factor’ are in fact the only possible options” and that “There cannot be an infinite variety of effects on a yes-or-no question.” Actually, there can. Either you’re married or you’re not married. Yet there are an enormous number of causal paths to this simple binary outcome.

There is a subtler point here that depends on whether you are looking at things from the point of view of the admissions officer or the point of view of the individual applicant. The admissions officer cares about the overall composition and characteristics of the whole class. Each applicant just cares about whether they get admitted. If Michigan mechanically applied a formula to each applicant, then one’s particular characteristics might directly determine admission or rejection, from your point of view. From the admissions officer’s point of view, though, any particular characteristic would make a quantifiable but not determining difference. Perhaps this is what Kinsley is trying to get at.

Update: Dan Drezner makes the same initial point about necessity and sufficiency.

You’re missing Kinsley’s point. In general, what you are saying is true… race is neither a necessary nor a sufficient condition for admission, but…

because admission is a binary decision, when considering a particular case, either race tipped the scale or it didn’t (when you change the race the admission decision changes or it doesn’t). That is why O’Connor is being incoherent when saying race can play a role but it cannot be a determinative factor – for a binary decision any factor that plays a role will be determinative some of the time.

This is hard to wrap one’s mind around, because assumptions about how the process works have been incorporated into our thinking.

Kinsley is, I think, assuming that all the applicants are somehow linearly ranked, a cutoff point is determined (based on desired class size, perhaps), those above the cutoff point are admitted and those below rejected. He then considers the last few applicants admitted and compares them against the applicants immediately below the cutoff point. If different consideration of race would have caused any or all these people to switch places, then their admission was determined by their race. If not, not.

Healy (and Drezner and probably O’Connor) visualize admissions officers reading an application and marking it “A”, “D” or “?”, without much reference to other applications. Their thought processes are not easily reducible to rule nor amenable to counterfactuals. A different officer reading the same application may mark it differently (in a fair number of cases does, which is why admissions processes need some way of resolving such disagreements). The same officer reading particularly a marginal application at a different time, in a different mood, in a different state of digestion, might have marked it differently. The question, Would the decision change if the race changed? is unanswerable.

It may be true that “for a binary decision any factor that plays a role will be determinative some of the time” but it is not necessarily true that we can recover which factor was determinative in any particular case.

And it’s probably worth emphasising that the actual admission mechanism in Gratz looked more like the model I’m speculating is in Kinsley’s head and the actual admission mechanism in Grutter looked more like the model I’m speculating is in Healy’s. Which may be why the decisions came down the way they did.

Just to note that from the persepective of an applicant, every factor is determinitive. Either you get in or you don’t, and your residence in Long Island or Alabama either got you in or it didn’t.

So Kinsley’s way of thinking (from the presepective of the applicant) really doesn’t add a lot fo value.

I think people are misreading Kinsley. See my extended comments here.

Do you want to see the congress enact a liberal agenda this year?

You can demand that they do, at

http://www.thePetitionSite.com/takeaction/365235275

The George W Bush 2000 Stolen Election Commemorative Coin

http://www.stolenelectioncoin.com

All profits go to charity.

I think you and Drezner misunderstood the type of causality Kinsley meant. Obviously he knows that race is neither a necessary nor sufficient condition. He meant that on the margins it is a but-for condition. His (just) criticism of O’Connor’s ruling is that she didn’t seem to realize that even in a holistic, individualized process, each factor can be a but-for condition, and thus determinative of the outcome. Since the other Michigan case in which O’Connor joined held that race cannot be a determinative factor in the outcome, there is a clear fudge in the reasoning.

I think that Kinsley has a point (based on Kieran’s summary; I don’t read Kinsley). And I think JL Mackie agrees with me in “Cement of the Universe”.

Being black is an insufficient but necessary condition for “being a person who would not have got in based on SAT scores but does so after extra points are awarded to black people”, and being a PWWNHGIBOSSBDSAEPAATBP is an unnecessary but sufficient condition for being awarded a place. So it is permissible under the ordinary language meaning of “cause” to say that being black caused some PWWNHGIBOSSBDSAEPAATBP to get into university.