White Americans Are NOT Close to Minority Status

The idea that white Americans are on the verge of no longer being the majority in this country has taken root in the media and in the minds of an awful lot of people. This theme was once again pushed out to the public a couple of weeks ago by William Frey, a demographer at the Brookings Institution. I have known Bill for a long time and he is a very good demographer, but I hadn't blogged about his report because it troubled me--the data he was using--even though from the U.S. Census Bureau--just didn't quite square with my own impression of what is happening demographically. Thanks to my long-time friend, Rubén Rumbaut at UC, Irvine, for pointing me to an "open letter" from Dowell Myers at the University of Southern California in which Dowell summarizes his very important work on this problem:

Most of us are using the same analysis procedures this year as we did back in the 1990s, even though the Census Bureau totally overhauled their racial definitions and measurements in 2000. Now that we are nearing the end of the second decade of the 21st century, why would demographers still be using racial binaries (white vs. nonwhite) and mutually exclusive categories? At best, the public analysis I see reported only uses half of the available race data, the half that comes closest to the oldest idea of race in America, namely the “one drop” rule, that says any portion of nonwhite blood makes a person nonwhite, no matter what is their mother or father’s race or no matter how they truly identify.

Let me share exactly where I am coming from, because this was reported in two publications recently, one in the Annals of the American Academy of Political and Social Science, the other in the Washington Post (both co-authored with the political scientist Morris Levy). The Annals study, part of the May issue, devoted to “what Census data miss about American diversity,” reviews the changes in Census Bureau race projections since 2000 and tests the impacts of alternative versions of reporting on a randomized sample of white voters. It received favorable coverage in outlets ranging from Vox to Reason.com.

Many data consumers do not know that the Census Bureau actually tracks six definitions of white in their projections. There is a larger, inclusive count of each race and a smaller, more exclusive count, the latter having subtracted out all whites who also identify with another race (people such as Meghan Markle, who has a white father and black mother, and now is a member of the British royal family.)

The Census Bureau knows this problem of racial classification and handles it by reporting both exclusive and inclusive definitions of white. You can see the latest projections comparing these numbers out to 2060 here. [Go to Table 5]. It is up to the users to decide which version of white is best.

Thus, if we follow the Census Bureau's guidelines that Hispanic is an ethnicity, not a race, and if we accept the idea that many people of mixed race also include themselves as white, then we see that the Census Bureau estimated that in 2016, 79% of the U.S. population considered themselves to be white, compared to the non-Hispanic one-race only definition by which 61% were white. Then, using the very restricted definition of white, by 2045 whites would drop to slightly less than 50% of the population. However, by the inclusive definition of whites, by 2060 (the end date for the current Census Bureau projections) whites are still 74% of the population.

To put it another way--the melting pot is working and whites are not on the way out the door in this country.