Race in Another America: The Significance of Skin Color in Brazil (review)

This is a blockbuster of a book. To a topic—Brazilian race relations—historically fraught with ambiguity, uncertainty, and disagreement, it brings clarity, logic, and lucidity, not to mention several truckloads of data. The result is the most important work on race in Brazil since Gilberto Freyre's seminal The Masters and the Slaves (1933).

Telles begins by confronting the core contradiction in Brazil's racial order: high (by US standards) levels of interracial sociability (expressed in cross-racial social contact, friendships, and even marriage) co-existing with equally high (by any standard) levels of racial inequality in education, earnings, vocational achievement, life expectancy, and other areas. Telles labels these the horizontal (sociability) and vertical (material achievement) dimensions of Brazilian race relations. Previous authors, he argues, have tended to focus on one dimension to the exclusion of the other, and have thus lined up in two opposing camps, one seeing Brazil as a hopeful instance of racial harmony and egalitarianism, the other as a case of extreme inequality and exclusion.

The achievement of this book is to acknowledge both dimensions, fully document them, and then ask how they are related to each other. In answering that question, Telles does not shrink from the multiple complexities he finds along the way, beginning with the vexing question of racial classification: how do Brazilians define who and what they are, racially? He finds that Brazilians do not identify themselves "racially," in the sense of belonging to a collective racial group. Rather, they identify themselves as individuals, and in terms of skin color (whence the book's title). And here things get complicated fast. In exploring Brazilians' notions of color, Telles finds three competing (and somewhat overlapping) conceptual schemes: the "official" three-category (white/brown/black) system used in the national census; the more spectrum-like, multi-category "popular" system used in everyday life; and the two-category (black/white) "black movement" system invoked by Afro-Brazilian activists. The "official" and "black movement" systems both strive for clarity and certainty. The "popular" system, by contrast, allows for flexibility and ambiguity. While its largest category, with 42 percent of the population, is "white," the second-largest (32 percent) is moreno (literally, tan or brunet), a rather vague color label that can be applied to members of any racial group (though in practice mainly to members of the "brown" census category).

Despite these complications, Telles finds that "in a large majority [88 per- cent—see page 90] of cases, there is no doubt about who is negro or white in Brazil." (266) Nor is there any doubt about which group is more advantaged. Two devastating chapters present a wealth of data on racial inequality in earnings, education, vocational achievement, even life expectancy, and the relative role of structural factors and discrimination in producing those inequalities.

The next two chapters consider the horizontal dimension of Brazilian race relations. Here we find Brazilians marrying across racial lines, and living in racially integrated neighborhoods, at rates much higher than in the United States. Still, they do not ignore race entirely. As of 1991, 77 percent of married Brazilians were in racially endogamous unions, showing a clear preference for marriage partners of their own color. And measures of residential segregation, while significantly lower than in the United States, are far from 0. Brazilians do take race into account in deciding who they are going to marry and where they are going to live, even if it weighs less in their decisions than in the United States.

Or less for some Brazilians than for others. As he turns to the question of how relative social inclusion can co-exist with high levels of economic exclusion, Telles notes that indices of intermarriage and residential integration are highest among the poor and working classes. Those indices are much lower in the middle class, which is overwhelmingly white, and are essentially 0 for the elite. Since it is white elites...