In 2010, I posted a six-part series on the U.S. census and race (1, 2, 3, 4, 5, 6). In it, I pointed out numerous changes in race categories and sub-categories over the 23 censuses, and multiple contradictions between scientific knowledge about human variation and the census race categories. I also offered a simple solution that would allow the government to collect the information it needs without contradicting science and offending or perplexing many citizens.
Because race is a cultural concept, beliefs about race vary dramatically from one culture to another. In this regard, America and Brazil are amazingly different in the categories they use. The United States has a small number of racial categories, based overwhelmingly on ancestry. Thus, it is possible for an American who "looks white" to "really be black" because he or she has "black blood."
In contrast, Brazilians classify people according to what they look like, using a large number of different terms. For example, one study in the Brazilian northeast conducted by the Instituto Brasileiro de Geografia e Estatística (IBGE)—the entity responsible for the census—asked people what color (cor) they were and received 134 different answers! (Other studies have found even larger numbers, and the results vary regionally, with much fewer categories used in the south of the country.) In many Brazilian families, different racial terms are used to refer to different children, while such distinctions are not possible in the United States, because all the children—no matter what they look like—have the same ancestry.
Thus, I was fascinated to read that "For the first time, non-white people make up the majority of Brazil's population, according to preliminary results of the 2010 census."
Slavery was much more widespread in Brazil than in the United States (and ended only in 1888), with the number of Africans always outnumbering the number of Portuguese, and with Portuguese men probably fathering more offspring with African slaves than with Portuguese women. In fact, there is a saying in Brazil that everyone has "one foot in the kitchen"—meaning an ancestor who was an African slave. Thus, if Brazilians thought about race the way Americans do, it would always have been true that "non-white people make up the majority of Brazil's population." To put it more strongly, using American racial categories, Brazil has always had a majority black population.
Of course, Brazilians do not use American racial categories, and are critical of Americans for "calling people black who are not black." Put differently, Brazilians would say that the American census over-counts the number of blacks, while Americans would say that the Brazilian census over-counts the number of whites. Specifically, the 2010 Brazilian census lists 47.7 percent of the population as white, and only 7.6 percent as black—numbers that would seem unreal to visitors from the United States. (43.1 percent were classified as mixed.)
Just as American census categories of race are unscientific and do not correspond to the cultural categories that Americans use to think about race, Brazilian census categories of race are also unscientific and also do not correspond to the cultural categories that Brazilians use to think about race. For example, the largest number of non-white Brazilians would be classified as pardo, a census term that Americans can think of as roughly meaning mixed. However, pardo is a term that is rarely used in everyday speech. So the census categorizes tens of millions of Brazilians by a term they would not use to describe themselves or others.
The biological sciences tell us that the human species has no biological races—all that exists is gradual variation, with more distant populations differing more from one another than closer ones. The social sciences tell us that different cultures use different concepts of race to categorize people. And governments tell us what "racial" categories to use to count people.