That might actually be what Loury was getting at (I was honestly as surprised as OP at that line). However, there does seem to be a third, midway option, which

That might actually be what Loury was getting at (I was honestly as surprised as OP at that line). However, there does seem to be a third, midway option, which the author even alludes to: black people have been socially and economically disadvantaged for centuries (to put it mildly), resulting in overrepresentation in lower socio economic groups. That, in turn, would lead to lower academic performance (regardless of race), because of less support and encouragement from the environment. In broad strokes: Option 3 is a combination of 1 (but societal, not just in academics), leading to 2 (but correlation, not causation). The idea of affirmative action then (assuming #3) being: giving black people more space in the academic race will eventually help break that chain. There are a lot of assumptions here, but I don’t see how this would be incongruous with either the author’s earlier statements or affirmative action as a whole. However I doubt this is actually true, because the author (an expert, unlike me) seems to know something I don’t.