Hella more than you think. Frank Underwood, certainly.
For most of its history, the stereotype of the racial line in the United States was the one-drop rule. You had any African ancestry and you were considered black.
Of course, the stereotype of the American caste system never really held. The United States was never organized enough to universally enforce the one-drop rule. If you looked white enough to convince strangers, then it was possible for you to “pass” for white. We know that happened a lot. What we don’t know is how common it was. Consider Daniel Sharfstein: “According to just about anyone who has considered the question, the migration [from black to white] is impossible to reconstruct... At best, such evidence is scattered across local archives and county courthouses, in library stacks and microfilm reels. Beyond the isolated anecdotes, there seems to be only silence.”
Except that’s wrong! The U.S. census identified people by race; the census-taker did the identification. So we can, in theory, find out how many people moved from black to white and what happened to them.
Using the full population of de-identified historical Censuses for ten Southern states during 1900-1940, we document that over one-fifth of black males “passed” for white at some point during their lifetime, a tenth of whom later reverse-passed to being black; passing was almost always accompanied by geographic relocation to communities with a higher percentage of whites; occurred at all ages; and was positively associated with political-economic opportunities for whites relative to blacks, such as schooling, earnings and voting rights.
The figures in their paper are even more surprising than that. The one-fifth number (21%, to be more exact) is a lower bound: it refers to the individuals that they were able to identify across censuses with perfect confidence. Their most likely estimates are much higher. Consider Table 5: of Southern black children aged 5-14 in 1900, 43% of them were identified as white in 1940. (The absolute lower bound for that number is 28%.)
They also find, less surprisingly, 95-97% of “passers” moved counties and 54% switched states. In addition, passing was associated with crappier conditions for black people in the county of origin. They are going to code up lynching in future work but I doubt that the results will surprise.
But there is still a lot we want to know! How did passers do relative to non-passers, adjusting for socioeconomic status? Did they marry whites or blacks and how did their children identify? What determined the counties that they moved to, besides having a higher percentage of whites? How many moved to the North, and where? What determined reverse-passing, e.g. going back to identification as black?
The authors note that according to data from 23andme, the number of self-identified whites who would be black under the one-drop rule is about 20%. (The 23andme sample is far from random; I suspect the white people skew towards the descendents of Ellis Islanders.) If the Nix-Qian numbers are correct, however, then 20% is likely an underestimate unless one of three things is true: (1) The children of passers re-identified as black in significant numbers; (2) Passers showed significantly below-average fertility; or (3) Passers showed high levels of endogamy, that is, marrying other passers.
Even in Jim Crow America, the color line is turning out to have been extremely fuzzy.