The comparison doesn't have to be with homicides. "Every demographic does X at about the same rate" is such a flat-out wrong assumption that you'd really have to cherry-pick X, with lots of caveats and conditions, for it to hold true.

I think there's an ingrained sense of revulsion, in all of us that live in the cultural era that descended from Enlightenment ideals of equality for all, against the idea that there exist observable group-level differences on nearly any metric you could think of, but that doesn't make it any less true.