Anti-Climacus

Once again, the question of bias in academia has come up. Jacob Levy speculates that we aren't learning anything new, just reinforcing the old convictions. Megan McArdle tries to adduce every possible explanation for bias. I wonder whether this question is framed in entirely unhelpful terms. The problem is that the number who even begin a graduate program is tiny relative to the undergraduate population as a whole. I think it's considerably harder to adduce an explanation for why certain people end up in academia that is equally able to explain why most people do not.

The question, I gather, is about a particular end-state of affairs (academic political orientation disproportionate to the population as a whole). To explain how this came about, one gives an explanation about why particular students opt into an academic career. But at this point it seems to me like we're dangerously close to selecting on the dependent variable.

Why? Though it's true that most right-of-center students don't go into academia, it also seems to me intuitively obvious that most left-of-center students don't go into academia. The number of people who start graduate school is a small fraction of the undergraduate college population, even at highly-ranked schools. The number who finish is even smaller than that. As one who recently completed this trek, I would be inclined to believe that the ability to finish grad school tracks personality and character traits that are orthogonal to political orientation. That is, controlling for the interest and ability to produce new research, be paid little money, and willingness to have uncertain job prospects, my guess is that there is no change in the make-up of political orientation from matriculation to degree, which is the opposite of what one might expect if politics caused them to fall by the wayside (my anecdata suggests this is true).

So I'm not sure any explanation of the -blank- are more intelligent, inquisitive, etc can work, because the people who go to grad school in general are already on the right tail of whatever the distribution of intelligence amongst undergraduates happens to be. Thus you'd have to make an argument like: intelligent liberals are more likely than intelligent conservatives to be in academia--but even then, I'm not sure that'd be true. More importantly, I'm not sure how one would go about substantiating that thesis in any acceptable social-scientific way.

In short, we have a number of competing explanations, and not very much data. Nor is that data of the kind that could be helpful in adjudicating amongst the various claims being made. One would need, as Levy notes, a longitudinal analysis, and one that would have to be comprehensive at varying levels of academic prestige--and the data for why the vast majority of students who do not go on to academia choose not to.