Red,As much as you'd love for me to be some bumpkin bigot, I'm not. What you're looking for in my statement is just not there. I'm clearly saying that I think that alternate measures of diversity will succeed in creating a suitably diverse class -- and in some ways more diverse than just a check in a box as a measure of diversity. I'm NOT saying that black people can't play the cello and Hispanics can't paint well.
Maybe URM and non-URM candidates a separated by more than LSAT. But I think what separates them is pure speculation on anyone's part without seeing the applications. And in the cases when a URM is also socioeconomically disadvantaged, there are likely to be gaps in other areas of the application as well but I would attribute that to $$ rather than race. Which is why I favor socioeconomic AA and not race-based URM AA.
Red - I was wondering specifically which data YOU were referring to.