Demography, Diversity, and Democracy: The 2000 Census Story

Kenneth Prewitt
Kenneth Prewitt Director, The United States Census Bureau; Former President, Social Science Research Council; Senior Vice-President, Rockefeller Foundation; Director, National Opinion Research Center, University of Chicago

December 1, 2002

On December 28, 2000, the Census Bureau released the first results of the decennial census completed a few months earlier, announcing, “Never have we been so diverse; never have we been so many; never have we been so carefully measured.”

The “never so many” and “never so carefully measured” were front-page news immediately. The national count of 281.4 million persons was the first big story. The total national population had grown more rapidly than expected, and the numerical growth—32.7 million since the 1990 census—was the largest recorded in any decennial census. The related census story was how these 281.4 million people were distributed across the 50 states. As required by the Constitution, that distribution is the basis on which the House of Representatives is apportioned. The winners and losers were the political story. The national count and its political implications were what mattered as the year 2000 came to an end.

In the months that followed, with the release of detailed census data, a story implicit in those early numbers steadily grew more prominent. “Never have we been so diverse” became the big story, and it shows every sign of staying on page one for some time to come.

Falling Birth Rates and Immigration

Each of us telling the story of America’s diversity will enter at a different point. I start with the perhaps unfamiliar term “replacement migration.” As defined in a March 2000 United Nations report, replacement migration is “the international migration that would be needed to offset declines in the size of the population, the declines in the population of working age, as well as to offset the overall aging of a population.”

The term draws attention to below-replacement birth rates, a problem that is especially acute in Europe. The UN projects negative population growth for 31 European nations, with especially dramatic declines for the working-age cohort. For example, under current (median) UN projections, in the next half-century Italy’s population will drop from 57 million to 41 million and the Russian federation’s from 147 million to 121 million—respectively, 28 percent and 18 percent. The replacement migration numbers get very high. To maintain Italy’s working-age population, for example, would require some 370,000 new migrants each year; Germany’s, just short of a half-million (Germany is now debating whether to grant 30,000 or only 20,000 temporary visas for immigrant workers). Population is declining not only in Europe but also in the advanced economies of the Far East, such as South Korea, Hong Kong, and Japan. Japan, for example, expects to drop from 127 million people to 105 million in the first half of the 21st century.

As in other industrialized countries, the native stock in the United States has stopped reproducing at replacement level. But the U.S. adjustment in immigration policy was swift. During the past three decades the U.S. foreign-born population grew at rates not seen since early in the century. Immigration plus higher-than-replacement fertility among the foreign-born added nearly 33 million people to the U.S. population in the past decade, with especially high growth among the working-age cohort. Both low-end workers for service and farm jobs and high-end workers for information technology have flocked to the United States.

Two Centuries of Migration

The foreign-born are now 10 percent of the U.S. population. The proportion itself is not remarkable; numbers that high were common in the late 19th and early 20th centuries. What is new is the demographic composition, a point best made by starting in 1790, when the first census recorded about 1 percent native Indians, 19 percent African slaves, and the remainder European. The Europeans were overwhelmingly from a northwestern corner of Europe—70 percent of the population were English, Welsh, Scottish, and Irish, and 10 percent primarily Dutch and northern German. More to the point, they were overwhelmingly Protestant. This was about to change.

Starting early in the 19th century, Roman Catholics began to dominate immigration flows as the Irish escaped economic deprivation and religious persecution. By 1850 the Roman Catholic Church was the largest denomination in the country (though still less numerous than all Protestant denominations combined). The next immigrant surge in the 1880s added more Roman Catholics, primarily Italian. Also arriving, for the first time, were large numbers of Eastern Orthodox and Jews when southern and eastern Europe displaced northern Europe as the primary source of immigrants. The change unleashed decades of aggressive populist nativism drawing on deep currents of anti-Catholicism and anti-Semitism.

But the nation worked its way through and around this darker side of a new religious diversity, in the process establishing what perhaps can be called history’s first “pan-European nation,” a century before the European Union. The First Amendment’s careful separation of church and state was critical. It presupposed substantial equality of religious association and allowed for a religious pluralism initially exercised within American Protestantism and then grudgingly extended to Catholics and Jews. This religious inclusion eased the way for the civic embrace of non-Protestant but still European nationalities.

The nation became pan-European, however, at the expense of non-European groups, of whom there was quite an array even in the 19th century. The pre-19th-century groups—Native Indians and African Americans—saw their claims to citizenship and inclusion rejected out-of-hand, the Civil War Amendments notwithstanding. Other non-Europeans were added through conquest: the southwestern Mexican population, Caribbean Islanders, Native Hawaiians and Pacific Islanders, and Native Alaskans. “Coolie” labor brought Asian immigrants to the West Coast, to work the mines and build the railroads. As a narrative about demographic diversity, the 19th century is about more than voluntary immigration; it is about slavery, indentured servitude, imperial wars, and territorial expansion. Civic membership, of course, was largely denied to the non-European groups.

The Census and Racial Classification

Handmaiden to this denial of rights was a racial classification system that grew with the census. In 1790 the census divided the population into three racial groups: free whites, slaves, all other free persons (the American Indians). In every census thereafter, the population was racially classified, though the categories changed with the times. By 1820 it was necessary to add “free colored persons” to the classification scheme. After the Civil War, interest in shades of color led the census to classify people as mulatto, quadroon, and octoroon. Asians began to appear in census racial categories around the same time. Often this involved confusing nationality with race or ethnicity. Chinese and Japanese were separately counted in 1890. Hindus, here confusing a religion with a race, were counted in 1920. Mexican first appeared in 1930, but was quickly dropped when the government of Mexico complained Mexican was not a race. Mexicans were then counted as white until 1970, when Hispanic origin appears on the census form as an ethnic rather than racial category. Hawaiian and Part Hawaiian appear on the 1960 census form, as do Aleut and Eskimo.

Classification separated those entitled to civic participation from those whose national origin, race, or ethnicity was cause for exclusion. The slave system, the relocation of Indians to reservations, implementation of the separate but equal doctrine, denial of citizenship to Asians, racist immigration quotas in the 1920s, and Japanese-American internment are familiar chapters in this story. From 1790 until the Civil Rights movement, policy designed to protect the numerical and political supremacy of Americans of European ancestry used a classification system that assigned everyone to a discrete racial group.

This racial (and racist) story took an important turn in the 1960s. The Civil Rights movement was built on the promise to end policy based on racial groupings. It celebrated unity, not diversity. But discrimination did not easily give way. And soon the country was enmeshed in the politics of statistical proportionality—equal opportunity became proportionate representation, individual rights became group rights, nondiscrimination became affirmative action. Institutional racism entered the political vocabulary. Statistical patterns were offered as evidence. A racial classification that had perpetuated discrimination would now end it.

This brings us to Census 2000, which challenges the nation to struggle again with the tension between the universalism promised by the American enlightenment and the realities of differences and diversity.

From Pan-European to Pan-World

The United States, having first painfully become pan-European, will now have to become pan-world.

The labor pool that moved voluntarily from the Old World to the New World during the 19th century was European. Today, it is Asian, Latin American, Caribbean, Middle Eastern, and African—all demographically young regions of the world, with low wage-rates, a combination that generates emigration pressures. Demographically older nations, especially those with high wage-rates, are where the migrants want to go. If restrictive immigration policy tries to overrule these demographic realities, illegal border crossings increase.

A mix of legal and illegal immigration to the United States has transformed our sense of ourselves as “a diverse nation.” In 1900, nearly nine of every ten foreign-born in the United States were from Europe; as late as 1960, the share was still three of four. The drop-off thereafter was sharp, as Hispanics and Asians arrived in large numbers. Of the foreign-born today more than half are from Latin America, more than a quarter from Asia. These patterns show no signs of reversal. Lesser but still significant numbers arrive from the Middle East, from Eastern Europe, and from Africa.

These trends have given rise to heavy media description of a more diverse America, with stories foreseeing a majority-minority nation. Setting aside the meaningless of the phrase, and the fact that these stories misinterpret the data when they exclude from the white count the Hispanics who identify themselves as white, the picture of a nation increasingly diverse in its ethnic and national origin composition is correct. The young are more diverse than the old, giving weight to the assertion that there is no reversal in sight.

Excepting advanced European economies, every region in the world is sending emigrants to the United States. A nation that started as European Protestant, and became Protestant-Catholic-Jewish but still European, is now constructing Hindu temples for Indians, Buddhist temples for Chinese, and Islamic mosques for Arabs, West Africans, and Southeast Asians. Between 1995 and 2000 the number of mosques increased 25 percent, to more than 1,200.

“Mark One or More”: A Quiet Revolution

However we define demographic diversity—linguistically, culturally, religiously, ethnically—the United States today is the most demographically diverse nation in world history. How will we manage that diversity? Probably not with a policy tool heavily used for 200 years—the racial taxonomy. The taxonomy frequently changed its categories and was used both to discriminate and then to undo discrimination, but it had one constant. It sorted the population into a small number of discrete summary categories: white, black, Native Indian-Native Alaskan, to which more recently were added Asian and Native Hawaiian-Pacific Islander and, starting in 1970, Hispanic as an ethnic category.

Census 2000 revolutionized matters by making multiple-race an option, an option that will now be used across the federal statistical system. The census instruction, “mark one or more,” is the early tremor of an earthquake in political and social life. What for 200 years had been a racial classification based on a handful of discrete groupings suddenly became 63 categories, and, when cross-classified by Hispanic-Not Hispanic, 126 race-ethnic groups. A limited, closed classification became, politically speaking, wide open.

In racial classification, proliferation begets proliferation. On what grounds does the government declare that enough is enough? No scientific grounds, no political grounds exist. The government will accommodate the demand from groups wanting a separate racial identity—Arab Americans, for example, who only reluctantly accepted the 2000 census form. With continued proliferation will come growing doubts about any racial measurement. Neither biologists nor anthropologists believe race has scientific merit. The frequent changes in how race has been measured across 200 years make the point eloquently. A scientifically grounded classification would not be continually changed depending on prevailing political sentiment.

What, then, is the future of racial measurement and the social policies to which it is anchored? There is no turning back to the simpler, discrete classifications of the 19th and 20th centuries. The current classification has too many categories for race-based policies using statistical proportionality yet too few to accommodate the pressures of identity politics and the desire for separate recognition. The situation is unstable.

Census 2000 puts two issues in play: public awareness of a new, unprecedented diversity and an unstable racial taxonomy. What will most matter is how and how quickly the two become intertwined. Policymakers and statisticians who design future censuses will be pressed by ethnic lobbies, demographers, and indeed common sense to provide data that allow for meaningful generalization about America’s diverse groups—but also to accept that the country’s growing, blended diversity makes such distinctions less administratively useful or sociologically meaningful.

Measurement decisions will not be made in a political vacuum. They will be contested, with outcomes hard to predict. Perhaps the greatest political uncertainty is whether new immigrant groups will want to be separately counted, so as to separately matter, or will choose to make their way in a manner that blurs boundaries between national origin groups. The hyphenated American was a product of early 20th-century politics, when the idea of subnationalities that were not racially but ethnically defined won broad acceptance. The prominent role of identity politics then came to the fore when statistical proportionality was introduced as an administrative and political tool. Is this our future, now expanded to include the many new nationalities and ethnicities of late 20th-century immigration? Or will the nation return to an earlier politics focused less on recognition than on redistribution and based more on individual merit, effort, and educational mobility?

The answer will determine how we will measure and sort in censuses to come. My sense is that Census 2000 stumbled upon a reasonable though probably short-term accommodation given the current uncertainties. In today’s America, people are what they say they are. If they say they are black or white or black and white, who is to judge otherwise? If the newly arrived Nigerians or Somalians insist that they do not belong in the same “race category” as seventh-generation slave descendants, will we force them into that category or allow them their own “African” category? Caught between many crosscurrents, the census kept traditional categories but allowed individuals to chose all that apply (or none). This is a fresh start toward telling us who we think we are. Given that the nature and degree of diversity is changing as well as our understanding of what diversity even is, this is a reasonable short-term strategy. It recognizes the present political realities of race and keeps intact the government’s obligation to police and prevent racial discrimination. And in the longer term, “mark one or more” opens the door to a postracial future.