Reframing Race in the Census

This is a contribution to Prospect Debate: The Illusion of a Minority-Majority America.

Richard Alba's analysis is a service to the country. I write to urge the Census Bureau and its various oversight agencies and committees to take his message seriously.

The current xenophobic, anti-immigrant political movement rests on the flawed premise that America will soon be a majority-minority country. The movement demands restrictive immigration policies to slow and even reverse what its leaders see as a demographic train wreck. But what if these anxieties are misplaced? By the time that becomes apparent, the damage will have occurred: Walls will have been built, families separated, deportations accelerated, and refugees refused.

If this is our future, it will be history repeating itself as tragedy. In the 1880s, members of Congress described recently arrived Europeans as a class of undesirables given to crime, uninterested in becoming citizens, unwilling to learn English, and stealing jobs from deserving Americans-in short, unfit for life in the United States. It took several decades, but in the 1920s the opponents of immigration shut America's gates, which weren't reopened for another 40 years. And what of the sizeable numbers already in the country, with their supposedly slothful habits, un-American tendencies, and alien religions? That story is well known. They became "American," and quite successfully so.

One feature of the grand narrative of the 20th-century melting pot was a redrawing of the color line. To the white Protestants who dominated political, economic, and cultural life, the Irish, Italians, and central European Jews were alien races. But as these aliens merged into a more inclusive European race, they redefined what had been a WASP-only mainstream, greatly enriching the country in the process.

Today there is a lively discussion among demographers and sociologists about how immigration is again redefining America. Many possibilities are in play, including differently drawn color lines, perhaps erasing them altogether, as social class, national origin, and immigrant status edge aside color as the primary cleavage in a multicultural world. It is too soon to know how these various forces will play out, and Alba's analysis adds a major dimension to the discussion.

It is a discussion that census statistics can help or hinder. Currently, the Census Bureau classifies as "minority" all those who claim both European and another heritage and all those who are born of mixed marriages. That practice hinders careful analysis of unfolding trends, and it hinders rational public discourse. More troublesome, it invites policies that are misguided, perhaps even socially damaging in serious ways.

We face an unavoidable tension: Racial classifications distort social realities, yet without classification the nation cannot produce the data necessary for a common conversation about race, however distorted. We discuss race through the prism of numbers-numbers educated, incarcerated, promoted, unemployed. The trade-off is difficult: How much distortion must we accept in order to have a common base of factual information and a common conversation about race and ethnicity in America?

The Census classifications currently assume-uniquely among national statistical systems worldwide-that the United States is home to five primary races. To be sure, the Census now more or less accepts blending across those categories and has added two ethnic groups (Hispanic and non-Hispanic) to the racial classifications. The basic elements of the schema date from mid-19th-century race science, which claimed there were "five races of mankind": white Europeans, black Africans, brown Pacific-Islanders, red indigenous people of the Americas, and yellow east Asians. On top of that foundation, the Census added Hispanic as an ethnicity to take account of the fast-growing Mexican American population. (For an explanation of how this system developed, see my book What is Your Race? The Flawed Effort of the Census to Classify Americans, Princeton University Press, 2012).

But if this schema is not a "true" representation of American society, what is a better schema? Some suggest no classification at all, treating everyone as "American." That introduces its own distortions; color-blind answers wish away racism, bias, discrimination, cumulative disadvantage, disparities, and related features of life in America that need public attention. Such attention rests on statistics. Thus we produce statistical races, the product of a classification that sorts the population into a limited number of categories. For many purposes, statistical races are more real than socially constructed races or identity races or, going back, biological races. They are certainly more real in the policy process, where problems defined and solutions proposed are anchored quantitatively.

During the last half-century, the Census Bureau has adjusted its practice as best it can (the constraints are many), trying to improve on the basic five-race classification. These adjustments reflect efforts to better align the classification to changing policy and demographic realities in the wake of the civil-rights movement and the surge in immigration. Those who know this history appreciate the contributions of the OMB's statistical policy office and the Census Bureau. But we are puzzled to encounter the bureau's continuing application of a "one-drop rule" in its forecasts of the nation's mid-century demographic profile. As Richard Alba shows, the bureau needs to adjust its practices again to provide the country with a more accurate picture of itself.

Next: William Darity Jr., "The Latino Flight to Whiteness"