Afro-Latin America by the Numbers: The Politics of the Census

By George Reid Andrews

Where did the all-encompassing term “Afro-Latin America” come from? While “Afro-Cuban,” “Afro-Brazilian” and other national terms were invented in the rst half of the 1900s, the broad regional concept appears to have originated in Brazil in the 1970s. A group of black socialist activists and intellectuals in the city of São Paulo were paying close attention to racial issues and struggles not just in Brazil but in Africa, the United States, and throughout the African diaspora. These activists, many of whom participated in the creation of the Movimento Negro Unificado in 1978, named their movement the Grupo Afro-Latino-América. When offered the opportunity to publish a regular section of articles and commentary—edited by the young journalist Hamilton Cardoso—in the leftist magazine Versus, they called the section “Afro-Latino-América.” Thus was coined a paradigm-shifting concept that would eventually reverberate across the diaspora.

The idea of Afro-Latin America was introduced to the United States by political scientists Anani Dzidzienyo and Pierre-Michel Fontaine, both of whom were doing research in Brazil on black social and political movements. Dzidzienyo published his findings in a 1978 article on “Activity and Inactivity in the Politics of Afro-Latin America”; two years later Fontaine reported on “The Political Economy of Afro-Latin America.” Fontaine defined the term to include “all regions of Latin America where significant groups of people of known African ancestry are found.” But that definition left at least two questions unanswered. First, how do we “know” when people are of African ancestry? And second, how large do groups of those people need to be before we consider them “significant”?

For answers to both questions, we might logically turn to Latin American censuses, an invaluable source of information on national societies. But at the time that Fontaine was writing, only two countries in the region, Brazil and Cuba, were gathering and publishing census data on their African-descent populations. Every other country had either removed questions on African ancestry from the census or, in some cases, had never included any. In the face of that statistical oblivion, how do we document the presence of African-descent populations and the conditions under which they live? Throughout the region, black activists posed precisely this question during the 1980s and 1990s, placing the census at the center of their demands for state action against racial discrimination and inequality.

CENSUSES THEN

Map 1. Afro-Latin America, 1800. Numbers under country names indicate the size of the black and brown (pardo) population, in 000s. (Map by Lena Andrews.)
Map 1. Afro-Latin America, 1800. Numbers under country names indicate the size of the black and brown (pardo) population, in 000s. (Map by Lena Andrews.)

Modern censuses began in Europe in the 1700s and arrived in Latin America in the 1770s and 1780s, when Spain and Portugal both ordered comprehensive population counts of their New World colonies. In both empires, caste laws organized colonial societies into racially defined groups that owed different combinations of taxes and labor obligations to the Crown. In order to verify the numbers of people falling into each group, the colonial censuses all gathered data on race, though using different systems of categorization in different parts of the empire. Officials in Cuba and Puerto Rico paid close attention to distinctions between blacks and mulattoes, noting the numbers of slaves and free people in each category. Census-takers in Colombia, Ecuador, Venezuela and Panama created a “free people of all colors” category that drew no distinctions between blacks and racially mixed pardos (browns). Officials in Honduras, Guatemala and Nicaragua followed a similar approach, counting all free nonwhites (except for indigenous people) as “ladinos.”

Despite these varying approaches, the colonial censuses made clear that by 1800, six present-day countries—Brazil, the Dominican Republic, Panama, Venezuela, Puerto Rico, and Nicaragua—were majority Afrodescendant (black and brown), and another four—Cuba, Colombia, Argentina and Honduras—were between one-third and one-half Afrodescendant. Even Mexico, only ten percent black and mulatto, had the second-largest Afrodescendant population in the region, at an estimated 635,000 people. (Brazil was number one, with a black and brown population of at least 1.3 million; see map 1.)

Following independence for most of the region in the 1820s, the new republics faced the issue of whether they wished to keep counting their populations by race. All of the new nations, including monarchical Brazil, had overturned the colonial racial laws and replaced them with constitutional declarations of full civic and racial equality. Those egalitarian principles seemed to argue against retaining colonial racial labels in national censuses, and some countries (Argentina, Colombia, Costa Rica, Mexico, Uruguay, Venezuela) did drop race from their 19th-century population counts. Others, however, such as Cuba (still a Spanish colony), Honduras, Nicaragua, Peru and Brazil, continued to count by race. They were joined in the early 1900s by countries that were either taking national censuses for the first time (Dominican Republic, Panama) or that had decided to return to gathering racial data (Colombia, Costa Rica). 

Those turn-of-the-century censuses were powerfully influenced by the doctrines of scientific racism that prevailed throughout the Atlantic world. Individual and national destinies were largely determined by racial inheritance, those doctrines argued. Nations seeking to improve and modernize needed to know the obstacles they were up against, hence the need for racial data. Or perhaps the opposite: nations seeking the path to modernity might prefer to ignore their racially mixed composition. This was the case with Brazil, which after documenting a relative decline in the black and brown population from 58 percent of the national total in 1872 to 47 percent in 1890, eliminated race from the censuses of 1900 and 1920. The resulting absence of racial data did not prevent census officials from concluding in 1920 that “the coefficient of the white race is constantly increasing in our population,” accompanied by “a reduction in the coefficient of inferior blood.”

During the next two decades, however, notions of superior and inferior blood fell increasingly out of favor in Brazil and other countries. In the 1940s, in response to the atrocities of German Nazism, scientific racism was roundly repudiated in the world community. Brazil and other Latin American countries now re-imagined themselves as “racial democracies,” thoroughly egalitarian societies in which the gathering of racial data served no useful purpose and might even provoke racial divisions. By the 1950s, only four nations (Brazil, Costa Rica, Cuba and the Dominican Republic) were still collecting data on their Afro-descent populations; by 1980, only Brazil and Cuba were doing so. 

CENSUSES NOW

Map 2. Afro-Latin America, 2010). Num- bers under country names indicate the size of the black and brown (pardo) popu- lation, in 000s. (Map by Lena Andrews.)
Map 2. Afro-Latin America, 2010). Num- bers under country names indicate the size of the black and brown (pardo) popu- lation, in 000s. (Map by Lena Andrews.)

Yet as we have seen, it was precisely in the late 1970s that the concept of Afro-Latin America—and the national black movements that had given rise to that term—were taking form across the region. During the 1980s and 1990s, black movements in Brazil, Colombia, Costa Rica, Honduras, Panama, Peru and Uruguay pressed national governments to acknowledge the disparities between racial democracy in theory and racial democracy in practice, and to take action to close those gaps. And in every country, high on the list of those movements’ policy demands was the restoration (or in some countries, the inclusion for the first time ever) of racial data to the national census.

It was imperative, those movements argued, that Latin American nations document their racial composition, the actual size of their black and brown populations, and the conditions of social and racial inequality under which those populations lived. Further supporting those demands were requests from international nancial and development agencies, which by the 1990s had come to see deeply rooted racial and gender inequality as major obstacles to social and economic development. As a first step toward combating those obstacles, the United Nations, World Bank and Inter-American Development Bank all pushed Latin American countries to provide systematic data on their class, racial and gender composition.

Under pressure from local movements and from international agencies, Latin American governments ultimately agreed to add questions on Afrodescendants to their censuses and national household surveys. Colombia and Uruguay inaugurated those questions in the 1990s, Ecuador and most of the Central American countries in the census round of 2000, and Argentina, Bolivia, Panama and Venezuela in the census round of 2010. Mexico and Peru have both agreed to gather racial data in censuses to be taken in 2017 and 2020, respectively. Chile and the Dominican Republic are currently the only Latin American countries with no concrete plans to canvass their Afrodescendant populations.

As in the colonial period, different nations have framed census questions on race in different ways, with results that are not always easy to compare across national boundaries. Afro-Colombian activists charge that, by failing to include the commonly used racial label “moreno” as a possible response, Colombia’s 2005 census substantially undercounted the nation’s black and brown population. Conversely, Venezuela’s use of the same term has led to intense debate over whether or not racially mixed “morenos,” the single largest category in the 2011 census, should be considered as Afrodescendants.

Given the complexities of racial identities (how one sees oneself ) and racial identifications (how one is seen by others) in Latin America, such debates are unavoidable and will surely continue into the coming decade and beyond. In the meantime, numbers yielded by the 2010 round of censuses do enable a provisional estimate of the region’s black and brown populations. As of 2010, three nations were majority Afrodescendant, down from six in 1800: the Dominican Republic, Venezuela and Brazil. (Dominican Republic gures are taken not from the census, which does not count race, but from the Latin American Public Opinion Project [LAPOP] survey of 2010; in Venezuela, I have considered morenos to be Afrodescendants). Cuba’s black and brown population of cially registered at 36 percent of the national total; in all other countries, Afrodescendants accounted for 12 percent or less of the national population. The total Afrodescendant population for the region as a whole was an estimated 135-140 million, most of whom (97 million) live in Brazil. Another 15 million live in Venezuela, and 8 million in the Dominican Republic (map 2).

Brazil have long been informed by census and national household survey data documenting pervasive racial inequality in that country. Now that similar information is becoming available for almost every country in the region, comparable debates will surely take place in those countries as well. And just as Brazil has taken major steps toward racially compensatory policies in education, public health, and employment, similar policies are already being proposed and, in some cases, adopted in Colombia, Panama, Uruguay and other countries. Without doubt, census data will be a central component of those policy debates.

 

George Reid Andrews is Distinguished Professor of History at the University of Pittsburgh and the author of, most recently, Afro-Latin America: Black Lives, 1600-2000 (Harvard University Press, 2016). For more on Latin American censuses, see Mara Loveman, National Colors: Racial Classification and the State in Latin America (2014). To consult recent censuses, access either https://international.ipums.org/international/ or https://www.cepal.org/es/temas/censos-de-poblacion-y-vivienda/enlaces-in...