How race-conscious policies have failed. By Stephan Thernstrom.
The United States has been a racially and ethnically diverse society from its beginnings. But the conventional wisdom these days is that something radically new is happening—that demographic changes are fundamentally transforming our society in unprecedented ways. Peering into a crystal ball, many observers have claimed that the groups we currently designate as minorities are destined to become the new majority. By the middle of the twenty-first century, they predict, and perhaps even sooner, whites will have been reduced to minority status and "people of color" will have become the majority. This, it is claimed, will have momentous implications for the nation’s political, social, and cultural life.
Such is the argument, for example, of Peter Brimelow’s Alien Nation, a 1995 volume that contended that current population shifts were "so huge and so systematically different from anything that had gone before as to transform—and ultimately, perhaps, even to destroy—the . . . American nation."
Brimelow is a conservative, but many observers on the multicultural left are equally convinced that a profound demographic transformation is under way. They are cheered rather than dismayed by the prospect, however. They welcome the arrival of a majority made up of minorities and see it as evidence of the need for immediate action—for more multicultural education in the schools, continued affirmative action and diversity training programs in higher education and the workplace, and an expanded welfare state.
The idea that "race" is a crucial and immutable division of humankind is a product of the primitive social science of the nineteenth century.
The demographic projections on which both sides of this debate depend are too flawed to be taken seriously, as I shall argue later. But the general public seems to have got the message—so it would appear, at least, from the results of a 1995 poll that asked Americans to estimate what proportion of the population belonged to various racial or ethnic groups. This survey revealed that whites (that is, non-Hispanic whites) thought that the black population was almost twice as large as it was in fact—24 percent in their minds, just 13 percent in reality—and that there were 50 percent more Hispanics and almost three times as many Asians in the country as the Current Population Survey figures revealed there to be. These three minority groups together, whites thought, made up fully half of the total population, when they actually were little more than one-quarter. The "minority majority," in the eyes of whites, was not a possibility in the remote future; whites were already on the brink of losing their traditional majority status.
It is tempting to interpret this misconception as evidence of widespread white paranoia. But the delusion was not confined to whites. Indeed, blacks and Hispanics were even more prone than whites to exaggerate their numbers. They also greatly exaggerated the size of other minority groups: minorities together, they believed, were already a distinct majority of the population, constituting 54 or 55 percent of the total. Asians were a little better informed than other groups, but they also greatly overestimated the size not only of their own group but also of other minorities. Whatever their backgrounds, most Americans tended to have similar misconceptions about the racial-ethnic composition of the nation’s population.
Why should the U.S. government distinguish some citizens from others on a basis that is not scientific and use those distinctions in allocating public resources?
It has long been claimed that nonwhite people are socially invisible in American society and that the minority presence deserves to be given far more attention than it receives on television, in the press, in classrooms and textbooks. President Clinton’s Race Initiative was based on the premise that most white Americans do not pay sufficient attention to their fellow citizens with skins of a different hue. These polling numbers suggest that the opposite may be closer to the truth: Americans have become so attentive to racial divisions and so obsessed with racial matters that they have developed a badly distorted picture of the shape of their society.
Four Boxes: Our Arbitrary and Unscientific Official Racial-Ethnic Categories
The survey referred to above employed four crude categories: white, black, Hispanic, and Asian. Why are these the relevant categories for subdividing the population into cultural groups? Why are these few groups singled out for attention, whereas a great many others with some claim to a distinct identity are not? What about Italian Americans, for example, or Jews? Are divisions among "races" deeper, more fundamental, and more enduring than divisions among "ethnic groups"?
The idea that "race" is a crucial and immutable division of humankind is a product of the primitive social science of the nineteenth century. According to theorists of the day, all the peoples of the world were divided into four distinct races: white or "Caucasian," black or "Negroid," yellow or "Oriental," and red or "Indian." White, black, yellow, and red people were profoundly different from one another, as different as robins from sparrows, trout from salmon, rabbits from squirrels. People who belonged to different races were not only distinct physical types but differed in innate intellectual potential and in cultural development. If they were to mate across racial lines, their offspring would be biological monstrosities.
Americans have become so attentive to racial divisions and so obsessed with racial matters that they have developed a badly distorted picture of the shape of their society.
Since these race theorists were white, it is hardly surprising that they fervently believed that Caucasians were the superior race. Orientals were next in line, with blacks and American Indians at the bottom of the heap. Given this premise, it was only natural that representatives of the "most advanced" race believed that they were entitled to rule over the "lesser breeds."
Such ideas have long been discredited and are now held only by those on the lunatic fringe. Scientists today agree that the genetic differences that distinguish members of supposedly different "races" are small and that the races have become so intermixed that few people can claim to be of racially "pure" origins. The range of biological variation within any one race is far greater than the average differences among races.
And yet the government of the United States, remarkably, still utilizes these antiquated and pernicious categories in compiling statistical information about the American people. If you want to know how many African Americans regularly use the Internet, how many Asians were treated in hospital emergency rooms in the preceding year, how many Hispanics usually eat breakfast, or how many Native Americans were arrested for burglary, the answers are all there. The federal government inundates us with data that convey the unmistakable message that Americans of different "races" differ from each other in many important ways.
The idea that it is meaningful and socially useful to cram us all into one of the four racial boxes constructed by racist thinkers more than a hundred years ago remains unchanged. The previous decennial census, in 1990, still accepted the traditional premise that every American belongs in one and only one of four mutually exclusive racial categories; people of racially mixed ancestry were required to record just one race on the census forms. The census of 2000 broke from this tradition and allowed respondents to give more than one answer to the race question, but for purposes of civil rights enforcement the results were tabulated in the same old crude categories, rendering the change virtually meaningless.
The range of biological variation within any one race is far greater than the average differences among races. And yet the government of the United States, remarkably, still utilizes these antiquated and pernicious racial categories in compiling statistical information about the American people.
The issue is not confined to the U.S. census. Nineteenth-century conceptions of race are also alive and well in the official guidelines that govern the statistical information that all federal agencies must gather. The authoritative statement of current practice is the Office of Management and Budget’s Directive No. 15, "Race and Ethnic Standards for Federal Statistics and Administrative Reporting," first issued in 1977 and still in effect. Directive 15 declared that the population of the United States was divided into four "races" and two "ethnic" groups and required all agencies of the federal government to compile data using these categories in order to assess the impact of their programs.
The "racial" groups identified in Directive 15 were the usual ones: whites, blacks, Asians and Pacific Islanders, and American Indians and Alaskan Natives. Even though the old idea of a racial hierarchy with whites on top had lost all intellectual respectability, the guidelines set forth in Directive 15 were designed to subvert that hierarchy. The rationale for requiring all governmental agencies to subdivide the population into these particular racial categories was that these nonwhite groups had been the targets of prejudice in the past (so had many white immigrant groups, of course, but the guidelines made no mention of that). It was necessary to monitor how the nonwhite races were faring in the present in order to overcome the allegedly lingering remnants of a history of white supremacy. The three minority races were victim groups that had once "suffered discrimination and differential treatment on the basis of their race." As victims, they were—and are—entitled to a variety of special protections and preferential programs not available to whites.
America’s "Two" Ethnic Groups
In addition to the three groups presumed to be disadvantaged because of their race, Directive 15 added a fourth—"persons of Hispanic origin." According to Directive 15, Hispanics were frequently the objects of prejudice and "differential treatment," not because of their "race" but because of their "ethnicity." Federal agencies were required to compile data on Hispanics as well as on the three nonwhite races because "ethnicity" for Hispanics was presumed to be the functional equivalent of race for blacks, Asians, and Native Americans.
The concept of ethnicity had long been an essential analytic tool for understanding American society, but Directive 15 used the term in a novel, indeed bizarre, way. The common understanding of American society was that immigration played a central role in its development and that many distinct "ethnic groups" had emerged out of the immigration experience and then faded away as later generations became more integrated into the larger society. Being a stranger in a strange land was difficult, and newcomers naturally felt the need to associate with other people who spoke their native tongue, liked similar food, worshiped in the same way, and had similar customs and values. The Harvard Encyclopedia of American Ethnic Groups describes more than 100 such ethnic groups, many of them extinct or close to it by now.
It is thus remarkable that the official guidelines employed by the federal government maintain that there are just two ethnic groups in the United States: persons of "Hispanic origin" and those "not of Hispanic origin." Several dozen white ethnic groups with distinct identities were suddenly collapsed into a single group with the awkward label "not of Hispanic origin." All the white ethnic groups had presumably merged into the general population, while Hispanics were taken to be an unassimilable, racelike group that would be as enduring as the "races" that the federal government was so dedicated to enumerating—even though most Hispanics considered themselves, and had always been officially classified as, "white."
The Pseudoscience of Race
Does it make sense at the beginning of the twenty-first century to identify "races" as defined by nineteenth-century supporters of white supremacy? The authors of Directive 15 were careful to say that "these classifications should not be interpreted as being scientific or anthropological in nature." True enough, but the admission only makes their decision to utilize them more dubious. If these categories are not "scientific" or "anthropological," what are they? Why should the U.S. government distinguish some citizens from others on a basis that is not "scientific" or even "anthropological" (whatever that means) and use those distinctions in allocating public resources?
If we continue race-driven policies in the delusion that they will enable us to "get beyond racism," we will only ensure the perpetuation of racial and ethnic divisions far into the future.
Perhaps the answer is that the OMB assumed that Americans today habitually draw these crude distinctions in their daily lives and that recognition of social reality requires the government to do the same. This is a feeble argument. What is the evidence of a societal consensus based precisely on these distinctions?
Even if it could be shown that these unscientific racial categories did correspond at least moderately well to the way in which the general public perceives the racial landscape, it does not follow that it is wise for the government to insist upon the saliency of race. Justice Harry Blackmun argued two decades ago that, "in order to get beyond racism, we must first take account of race. . . . And in order to treat some persons equally, we must treat them differently." But the race-conscious policies that have been pursued in the United States for a generation have plainly not taken us "beyond racism."
President John F. Kennedy was wiser than Justice Blackmun, I believe, when he said that "race has no place in American life or law." To continue to draw racial distinctions in our laws and to compile massive amounts of official statistical data about racial differences among racial groups will not serve to make race less important in "American life."
Dispelling the Myth
The picture of the American people as divided into oppressors and oppressed racial-ethnic groups is an oversimplification and a distortion. The errors it entails are compounded when we attempt to peer into the future and calculate what the racial and ethnic mix of the American population will eventually be. Projections of precisely this kind have attracted considerable public attention, thanks to a credulous press. The cover story in the April 9, 1990, issue of Time featured a Census Bureau projection that concluded that the United States would have a "minority-majority" population by the year 2050. Since then, the official estimates have been revised slightly, with the latest indicating that the "minority" population (blacks, Asians, Hispanics, and American Indians) will be a shade less than a majority in 2050—49.7 percent of the population.
Will this in fact happen? Will it matter if it does? The first thing to notice is that demographers have never been much good at prediction. Demographers project future populations on the basis of currently observable patterns of immigration, fertility, and mortality. The more remote the future, the greater the possibility that these variables will change in unanticipated ways. In fact, the Census Bureau cannot even project with any accuracy the total population of the United States in 2050. (The bureau’s high and low estimates vary by a staggering 236 million!) If there is such great uncertainty about what the total population will be half a century from now, there must be similar uncertainty about the size of the various racial and ethnic subgroups that make up the total.
Why do these projections vary so enormously? Because they necessarily rest on assumptions about the determinants of population growth that may prove mistaken. For example, they require accurate estimates of the level of immigration to the United States 30 or 40 years hence. Obviously, we cannot know those levels with any reliability because our immigration policy may become far more restrictive than it is now. Laws enacted in the 1920s sharply cut back on the number of new arrivals, and we cannot be sure that a similar anti-immigrant backlash will not again close the door to newcomers from abroad.
Nor can we be at all sure about a second variable that determines how the size of a population changes over time—its fertility patterns. At the beginning of the twentieth century, a great many native-born Americans worried about the consequences of their own rapidly declining fertility. Many feared that they were being swamped by huge waves of new immigrants and the large families the new arrivals typically had. Lothrop Stoddard, a leader of the Immigration Restriction League, warned that Anglo-Saxons were committing "race suicide." According to his calculations, after 200 years 1,000 Harvard men would have left only 50 descendants, while 1,000 Romanian immigrants would have produced 100,000!
There was nothing wrong with Stoddard’s math. The problem lay with his straight-line projection of the fertility differentials of his day 200 years into the future. He failed to comprehend that in the second and third generations Romanian Americans would adjust their fertility patterns to the American norm and would produce many fewer children than did the immigrant generation. This process of assimilation to the prevailing national fertility norm continues to operate today.
Stoddard also erred in his implicit assumption that Romanian immigrants and their children would keep marrying within the group, perpetuating the cultural patterns of their country of origin. Quite the opposite happened. Romanians, like most other immigrants, often married non-Romanians, with the probability rising the longer they lived in the United States. Ethnic intermarriage complicates ethnic identification. Are you still a Romanian American if just one of your four grandparents was Romanian? What if two of the four were? The immigrants of the early twentieth century, like their nineteenth-century predecessors, usually chose mates of the same ethnic background, but many of their children and a great many of their grandchildren did not. The population derived from the great waves of European immigration is by now so thoroughly interbred that its ethnic origins are difficult to disentangle and of little consequence.
Assimilation via the "marital melting pot" has also occurred at a rapid pace among the immigrants of the post–World War II era. If intermarriage continues at such high levels, a very large proportion of all Americans in 2050 and even sooner will have some Hispanic, Asian, or African "blood." But it does not follow that all or even most of these individuals will identify more with their one Hispanic, Asian, or African American ancestor than with those who were non-Hispanic whites.
Multiculturalism rejects the idea that a common American culture binds us together and minimizes a common commitment to the Constitution and to the rule of law. So much for the melting pot ideal.
The Census Bureau today has more sophisticated techniques for modeling population change than were available to Stoddard, but it has been no more successful than he was at grappling with the reality of ethnic intermarriage, assimilation, and loss of ethnic identity, a reality that fatally confounds all efforts to extrapolate contemporary ethnic divisions into the remote future. Even if the descendants of Romanian Americans and the other "new immigrants" of the early twentieth century who so worried the Immigration Restriction League are now a majority of the population, as Stoddard feared, who could possibly care? By the time that the groups currently classified as minority become a majority, if that ever happens, it will be equally irrelevant because they will no longer be thought of as minorities.
Moreover, if today’s immigrants assimilate into the American stream as readily as their predecessors did at the turn of the last century, there will not be any minority-majority issue anyway. Whatever their origins, they will have joined the American majority, which is determined not by one’s bloodlines but by one’s commitment to the principles for which this nation stands.
Abandoning the Melting Pot
Still, assimilation cannot be taken for granted. We cannot reliably predict the shape of the ethnic and racial future of the United States. The historical parallels drawn above may not hold because the melting pot ideal that was once so widely accepted has by now been largely displaced by the competing ideal of multiculturalism, which implies that racial and ethnic divisions are and should be permanent.
Multiculturalism rejects the idea that a common American culture binds us together and minimizes a common commitment to the Constitution and to the rule of law. So much for the melting pot ideal. This simplistic view that the American people belong to five separate cultures, with all but non-Hispanic whites the victims of racial oppression, is hardly a recipe likely to make us "one nation, indivisible." If we continue race-driven policies in the delusion that they will enable us to "get beyond racism," we will only ensure the perpetuation of racial and ethnic divisions far into the future.
Stephan Thernstrom is the Winthrop Professor of History at Harvard University and a Manhattan Institute senior fellow.
A longer version of this essay appears as "The Demography of Racial and Ethnic Groups" in the new Hoover Press book Beyond the Color Line: New Perspectives on Race in America, edited by Abigail Thernstrom and Stephan Thernstrom.
The new book Beyond the Color Line: New Perspectives on Race in America, edited by Abigail Thernstrom and Stephan Thernstrom, is now available from the Hoover Press. To order, call 800-935-2882.