Assume all demographics produce the same distribution of bridge quality. Under the historical discrimination regime, we've been taking all the majority-designed bridges simply because they were majority-designed, including the bad ones. Obviously this is bad, and should be rectified.
Parent and I argue that automatically taking a minority-designed bridge is equally bad, because it's equally likely to lead to a faulty bridge. Bridges should be evaluated according to their quality/safety independent of who designed them.
