Demographics and Race
Demographics and Race
The United States Census has collected information about race ever since the first census was taken in 1790. Indeed, the Census and the collection of information about race were originally mandated in the U.S. Constitution. In Article One, Section Two, the founders of the United States set forth the language for collecting information about race in the decennial census:
Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons. The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.
This language acknowledges several racial designations. In almost all circumstances “free persons” and persons “bound to service for a term of years” meant European immigrants and their descendants. Very clearly, American Indians who resided outside the jurisdiction of Anglo-American communities were excluded, insofar as they could not be taxed. The notorious “three-fifths compromise” was an obvious allusion to African slaves and their descendants, who were born into slavery for the duration of their lives.
Despite the obvious racial overtones of this language, it is important to understand that the Constitution was a product of Enlightenment philosophy. The Enlightenment had an enormous influence on the framing of the U.S. government, but, ironically, this intellectual tradition offered little insight into the nature of race and racial distinctions. As a result, the incorporation of race into the Constitution was shaped more by considerations of conquest and oppression than by abstract philosophical principles. Race, for the founders of the United States, was mostly a matter of civil status.
The articulation of racial thinking that would justify the collection of information about race awaited the emergence of biology as a full-fledged body of scientific knowledge in the late eighteenth century. The Swedish biologist Carl Linnaeus deserves credit for devising the first racial classification for Homo sapiens, published in 1735. This classification consisted of four groups—Americanus, Africanus, Asiaticus, and Europeanus—and each was associated with race-specific behavioral characteristics. In 1775 the German scientist Johann Blumenbach elaborated Linnaeus’s scheme and introduced a vocabulary for describing race that still more or less remains in use. His categories were Caucasian, Malayan, Mongolian, Negro, and American. Blumenbach also assigned particular behavioral traits to each of these groups. The work of Linnaeus, Blumenbach, and the French biologist Jean-Baptiste Lamarck had a profound influence on scientific thinking about race in the United States, and on the way racial data were collected in nineteenth-century censuses.
From 1790 to 1810, there were few changes in the U.S. Census. Households were enumerated according to the number of persons, both free and slave, and most American Indians were excluded from the count. The first significant modification of the census with respect to race occurred in the 1820 enumeration, in which a distinction was made for “colored” persons, free or slave. At the very least, this signified that there were sufficient numbers of free colored persons to merit enumeration, as well as a tacit acknowledgement that race was a physical quality in addition to being a condition of civil status.
The distinction between free and slave colored persons remained in the next several censuses, but the 1840 census was particularly noteworthy; it was the first to precipitate a controversy over the accuracy of racial data. In particular, the 1840 enumeration resulted in an unusually large number of “colored insane” living in northern states. This, of course, became fodder in the slavery debates, with the proponents of slavery using this finding to argue for the deleterious consequences of emancipation (Anderson 1988).
The debates over the accuracy of the 1840 census proved to be long lasting and especially bitter. They also set the stage for a larger role for scientific opinion in the 1850 census. Congress appointed a special Census Board consisting of the secretaries of state and commerce and the Postmaster General. This board received considerable authority to conduct the census, and it appointed Joseph C. G. Kennedy as the superintendent of the census. Kennedy proved to be a highly effective leader and an especially important superintendent because he actively sought scientific advice about how to conduct the census. He sought, for the first time, the advice of leaders in the American Statistical Association and the American Geographical and Statistical Society, as well as leading European statisticians such as Charles Babbage.
The 1850 census was taken in an era when the country was not yet at war, but the sectional conflict over slavery was growing in intensity. It was also a period when eugenics and ethnology were reaching a pinnacle of influence within the scientific community. Among the innovations introduced by Kennedy for the 1850 census was a special schedule for the enumeration of slaves—they were identified by a number along with the name of their owner. The 1850 census also admonished enumerators to carefully record the racial heritage of household members. One noteworthy innovation in this census was the introduction of the term “Mulatto.” It was the first tacit acknowledgement that sexual relations transcended racial barriers and that the offspring of these unions existed in sufficient numbers to merit enumeration.
In the decades following the Civil War, the race question was modified in ways that reflected the social and political concerns of the era. The preoccupations of scientific racism played a very large role in the development of the census. The enumerator instructions in the 1870 census epitomized the obsession with racial purity that characterized this period. These instructions admonished enumerators to “Be particularly careful in reporting the class Mulatto. … Important scientific results depend on the correct determination of this class” (Snipp 2003, p. 567).
The 1890 census did not stress the scientific importance of its racial classification or the gravity of accurately assessing racial heritage. However, it was significant because it introduced an even finer classification for persons with African heritage by identifying persons as Octoroons, Quadroons, or Mulattos. It took careful stock of the racial composition of the American Indian population, particularly the numbers of mixed-race persons. Another significant innovation reflected the racist and nativist concerns about immigration from China and Japan. Earlier censuses had enumerated the presence of Chinese and Japanese immigrants in California. Their numbers had grown sufficiently large in other parts of the country causing an outpouring of alarm that culminated in the 1882 Chinese Exclusion Act. The 1890 census was the first effort to monitor the spread of Chinese and Japanese immigrants to locations outside California.
At the dawn of the twentieth century, ideas about race that first appeared in the early and mid-nineteenth century were still influential. Jim Crow racism institutionalized these ideas and made them the scientific underpinnings for laws and public policy governing race relations in the early twentieth century. For example, the so-called one-drop rule stipulated that even the smallest amount of African heritage was sufficient to be designated as black. Consequently, for the 1900 census, efforts to divide the black population into subgroups such as Octoroon were abandoned. By 1920, at the peak of Jim Crow, even the category of Mulatto was jettisoned.
Changing patterns of immigration and continuing anti-Asian public sentiment led to the addition of several new categories in the 1930 census. Specifically, the government added categories for Mexican, Filipino, Hindu, and Korean. With the exception of “Mexican,” these categories were retained in a virtually identical format for the 1940 census. In response to lobbying by the Mexican government, the Census Bureau deleted “Mexican” as a separate designation and counted Mexican Americans as “White.” One other change connected with the 1940 census was that data for ethnic minorities ceased to be published as part of the decennial reports for states and localities. Instead, a special subject report was issued titled Characteristics of the Nonwhite Population by Race.
Racial measurement in post–World War II America became an urgent matter for at least two reasons. First, a significant undercount of black men was discovered to exist in the 1940 census, and a similar undercount was documented after the 1950 census. Second, and more important, the civil rights movement gathered momentum and the census, as the principal tool for allocating political power and resources, became recognized as essential for ensuring the enfranchisement of blacks and other minorities. Given the unmatched importance of the census for obtaining fair political representation, any undercount of racial or ethnic minorities was simply unacceptable.
In every census since World War II, the U.S. government, particularly the U.S. Census Bureau, has struggled to minimize the undercount of racial and ethnic minorities while also confronting other challenges, such as containing the costs of the census. One cost-cutting measure that proved to have profound consequences for the way the census measures race was introduced in the 1960 census. Prior to 1960, the decennial census was taken by enumerators conducting door-to-door interviews. Race was a characteristic that was visually determined and recorded by the interviewer. To save the expense of hiring thousands of enumerators for the 1960 census, the Census Bureau mailed census forms to households and allowed individuals to self-identify their race. This led to a subtle, though fundamentally profound, shift in the measurement of race in the census. The race that individuals reported in the census was determined less by the phenotypical cues that might influence an enumerator’s judgment, and more by the personal beliefs, feelings, and attitudes that shape the racial identity of individuals. Race was no longer a matter of observable physical traits; it was now a matter of personal subjectivity.
The shift to self-identification had little or no effect on the enumeration of some groups, such as blacks or whites. However, for one group—American Indians and Alaskan Natives—the effect was profoundly significant. The shift to self-identification resulted in a significant increase in the number of American Indians, rising from 357,000 in 1950 to 524,000 in 1960—an increase of 47 percent. This increase was followed by similar or even larger increases in subsequent decades. Some of this increase was due to an excess of births over deaths. However, an even larger amount was due to the fact that many persons who had once been identified as white or black chose to report their heritage as American Indian.
The Census Bureau continued to use the same question format for obtaining information about race in 1970, 1980, and 1990, albeit with small modifications. The 1990 census proved to be a pivotal event that triggered a storm of protests because it instructed respondents to choose only one race for their heritage. Even more infuriating was the fact that if respondents chose to mark more than one race for their background, the Census Bureau had a complex procedure in place for assigning one and only one race to each individual for whom more than one race was reported. The groups who objected most strongly to this measure were mixed-race couples that were expected to choose one over another in designating a race for their children. Why, they demanded, should a child with a black mother and a white father be forced to choose one race or another, when in fact they were genuinely multiracial?
These complaints eventually led Congress and the Office of Management and Budget to undertake a thorough review of the procedures the federal government uses for the collection of information about race. In October 1997 the U.S. government issued new guidelines for collecting this information. These guidelines were mandatory for all federal agencies, their contractors, and their grantees, effective January 2003. The new guidelines made two significant changes. One was the creation of a new category: “Native Hawaiians and other Pacific Islanders.” Formerly, people in this population had been counted among “Asians and Pacific Islanders.” The second and most profound change was a mandate that whenever the federal government (or its contractors or grantees) collects information about race, the instruction to “mark one or more” or “choose one or more” must appear on the questionnaire.
The Census Bureau was the first government agency to adopt this language. Thus, the 2000 census included language on the race question that instructed respondents to “Mark one or more races to indicate what this person considers himself/herself to be.” In response to this question, 6.8 million persons were identified as having more than one race— about 2.4 percent of the total U.S. population. Relatively little is known about this population except that it is an extraordinarily diverse group. For example, persons of black and American Indian heritage are very different from persons of white and Asian descent in terms of where and how they live. Another notable quality is that about 22 percent of the multiracial population is composed of Hispanic persons who identify themselves as “White” and “Some Other Race.” Studies conducted by the U.S. Census Bureau indicate that multiracial persons frequently change the way they identify their race, perhaps in ways to meet the social expectations of other persons around them (Bentley et al. 2003).
The racial data produced by the 2000 census is an extraordinarily complex body of information, but in many respects it mirrors the racial complexity of American society. Once it seemed possible to frame race and ethnicity in American society entirely as a matter of black and white. The changing nature of American society, however, now demands an approach that allows for myriad forms of racial identification, and data will undoubtedly be even more complex in the future.
Bentley, Michael, Tracy Mattingly, Christine Hough, and Claudette Bennett. 2003. Census Quality Survey to Evaluate Responses to the Census 2000 Question on Race: An Introduction to the Data. Washington, DC: U.S. Bureau of the Census. Available from http://www2.census.gov/census_2000/datasets/CQS/B.3.pdf.
Fredrickson, George M. 2002. Racism: A Short History. Princeton, NJ: Princeton University Press.
Snipp, C. Matthew. 1989. American Indians: The First of This Land. New York: Russell Sage Foundation.
———. 2003. “Racial Measurement in the American Census: Past Practices and Implications for the Future.” Annual Review of Sociology 29: 563–588.
C. Matthew Snipp