Many years ago I read an article that said something along the lines of minority populations stop being "noticed" when they cross a threshold of something like 20%. Below 20% or so they are perceived as being fairly rare, and above 20% or thereabouts they stop being noticed as different. The example I remember our prof using at the time was the number of women CS students in the class. Below 20% women CS was perceived as being mostly a man's discipline, but as the numbers crossed 20% or so it stopped being perceived as primarily associated with one gender or another.
I have been looking for citations or references to this effect for some time, but haven't been able to find any. Does this ring a bell with anyone?