[bzbarsky]

The point when you transition from "X and Y are different" to "X is better than Y".

It commonly takes the even more pernicious form of going from "X and Y are different on average" to "any given X is better than any given Y". If you're lucky, this last statement at least has a "until proved otherwise" caveat.

Unfortunately, history is rife with this sort of reasoning. People slip into it _really_ easily. It doesn't help that there is a natural tendency to perceive your in-group as better than out-groups, so to the extent that X above ends up feeling like someone's in-group and Y ends up feeling like an out-group, the "X is better than Y" conclusion is very hard to avoid.