That is a classic, but continuing dilemma: One, the basis of a person's identity (skin color, ethnic basis, national origin, their current perception of how they identify themselves, etc.)?; and Two, the implications of such measures.
You bring up a good point about Americans (North and South) as well as people from various African states, and likely other regions. The question arises whether categories like skin color and ethnic basis or skin color and national origin are really divisible. What about a person from the Masai tribe that was born and raised in Kenya? Can one divide that person's ethnic, their skin color, and their national origin? Given there are policy implications to these categories, this creates challenges. If they're not divisible, as you show, then it's hard to count and thus hard to measure. Until the US government included a mixed race category, the President-elect would be stuck between categories.
How accurate are the categories created? Then, is any category under-reporting or over-reporting its weight? Some work has been done to try and determine ethnicity through determining how a person identifies themself: do they see themself as White, European, French, Parisian, etc. They use the "ethnic kin" the person tends to connect to, relying on religion, language, and (a more controversial one) community.
Interesting! Steve Ackerman