by Conor Clarke
Down the rabbit hole we go: My Harvard Law School interlocutor responds to my response about racial diversity at the Supreme Court. He emails:
I think the question might be obscured by framing it as empirical. First, the benefit of x diversity factor isn't something that's readily quantifiable in most cases. Second, even if it were, that's just a convenient way to punt the value choice that underlies a commitment to diversity. If, for example, retention rates went down in schools, the military and businesses after a diversity push, would that "prove" that diversity is a bad thing and should be abandoned? If the answer to that question (or the question relating to any other empirical metric you might use to measure the benefits of diversity) is no, then there is some value that empiricism is allowing you to obscure.
To the first point: I agree that the benefits aren't easily quantifiable, but we can obviously try. To the second point: I think this proves too much! One reason why I prefer justifying affirmative action based on fairness ("we are making up for a lack of opportunity") rather than social utility ("this will give us better classroom discussions") is that the social utility argument can always cut in both directions. Maybe diversity will make those classroom discussions worse! (And as a sidenote, I will add that Robert Putnam has done interesting empirical research on this point.) I am of the feeling that fairness should trump utility.