We regularly use outcomes as evidence of discrimination, so I think it’s reasonable to strive to be consistent.

If we look at SAT discrepancies to identify groups harmed by structural issues in the test, we should also do the same with grades.