Wednesday, April 19, 2006

Diversity in Law Schools

On his law school blog, Brian Leiter posts a note from a staffer at a top-10 law school about the lack of socio-economic diversity in the elite law schools. You can read it here. One interesting thing the staffer says is:

Despite all of the focus on diversity, it seems that there is a great deal of homogeneity among the students when it comes to the socio-economic makeup of their families--most come from affluent backgrounds--at least at the school I work at.

Now, the staff member is not implying that somehow, the admissions systems are stacked unfairly against those with under-privileged backgrounds. Rather, diversity is the staffer's concern because he/she comes from a socio-economically disadvantaged background.

But I've often seen this sort of argumentation as evidence of a deliberate bias, for example with the SAT's (but you can apply this argument to LSAT's as well, since they tend to measure intelligence more directly). Often the argument goes like this: "SAT scores are highly correlated with socio-economic background. This is prima facie evidence that the SAT's are unfairly biased towards those who are socio-economically advantaged."

This is the kind of reasoning that I would like to challenge here. It essentially says that the fact that wealthy kids tend to do better and poor kids tend to do worse, is evidence that the SAT's themselves (rather than just being proxies for some other underlying problem like bad schooling in rural and inner-city areas, where there tend to be more poor people) are biased against those who are socio-economically disadvantaged. But this argument implicitly assumes one of two things. Either, children's intelligence and work ethic is not at all correlated with parents' intelligence and work ethic or in our society, wealth is not at all correlated with intelligence and work ethic. These are highly suspect claims.

First, most of the evidence seems to point to correlation between parents' IQ and that of children, either because of genetics or due to environmental factors (e.g. intelligent parents tend to speak well at home, use proper grammar, read to their children, etc). I do not know about studies on work ethic, but just as a casual thought, I would imagine children at least learn by example from their parents.

Second, the assumption that wealth is distributed more or less randomly (as far as intelligence and hard work is concerned) seems almost economically-determinist in a way. It would have to mean that intelligent and hard-working people would not be able to obtain wealth very easily because it is in some sort of immutable possession of the Old Guard. This Old Guard may or may not be intelligent and hard-working, but in any case, obtained the wealth through family connections, inheritance, etc. First, the entire concept of "New Money" in our society seems to indicate that this is not true. Jews in this country are a primary example of people who, through hard work and possibly natural intelligence, have been able to rise up above their humble beginnings. On the other extreme, if a person is not at all intelligent, certain lucrative careers (e.g. banking, law, medicine, architecture) are closed off to him, so that leaves him with few lucrative options and most likely reduces him to a lower class or middle-lower class life.

So my point is that it does not seem all that surprising that SAT scores would be correlated with socio-economic status, since it seems highly plausible (and probable) that 1) wealth is correlated with intelligence and good work ethic in our society and 2) intelligence and good work ethic tend to run in the family and this is why children of the socio-economically privileged tend to do better. Even if we take work-ethic out of the equation since it's much more difficult to quantify and thus study in "scientific" settings, it seems as though the correlation with intelligence alone would be quite high.

I am not claiming that all those who are wealthy "deserve" their wealth because they are smart and hard-working or that all those who are poor must deserve to be poor because they are either dumb or lazy. I am merely claiming that one would expect, on face, a pretty substantial correlation. I have not read any studies attempting to correlate intelligence with wealth, but if anyone has any relevant material, I'd be interested.

P.S. This entirely leaves aside the substantial body of evidence that points to a reasonably good correlation between SAT's and IQ's.