Concept explication at it's finest: "White" turns out to be a variable, as of course do other "categories" of the construct ethnicity. Amazing, right? The more you think about all of this, the more confusing and and irritating it is. This is why using categorical variables is always questionable: The implication is that there is more homogeneity within categories of ethnicity than between them. I doubt that this is true today in the US, where waves of immigration mixed the various flavors of "white" (see above) with the so-called "minority" categories of ethnicity. Do you think there was more homogeneity within ethnic categories 100 years ago than today? 50 years ago? And so, is there anything to be gained by using the concept "ethnicity" in social science research today?