Here's a quick quiz. According to federal statistics, does Sacramento, Calif., have:
A) A slight majority of whites and nearly 138,000 blacks and Asians?
B) A minority of whites and 123,000 blacks and Asians?
The answer is both. And if this seems confusing, it's only the beginning of a new era of racial counting. The new counts will more accurately reflect the United States' increasing diversity, especially the rapidly growing number of biracial and multiethnic children. But they will also so blur racial and ethnic lines that judges, policymakers, and everyday citizens use as they try to draw voting districts, dole out federal block grants, and figure out who goes to which school.
No one knows how the new counting will shape the way Americans see themselves. The first step comes next year with the 2000 census.
"It's going to be a real shock - a real flowering will come out of the census - when people understand the complexity of ethnicity," says Michael Dear, director of the Southern California Studies Center at the University of Southern California in Los Angeles. "What we're trying to do is invent a new way of seeing."
The old method, still used in federal data gathering, forces people to describe their racial and ethnic background by checking a single box: black, white, Asian, and so on. (The one exception is Hispanic, a mixed group to begin with, which is typically asked as a separate question.)
The ethnic pigeonholing works for most people. But it doesn't fit the rapidly growing population of children from mixed marriages. Between 1970 and 1990, their numbers have grown from less than 500,000 to about 2 million. And their numbers are likely to skyrocket in coming decades as racial and ethnic intermarriage becomes commonplace.
That's why the federal government has begun changing its statistical methods. Next year, for the first time, the decennial census will allow people to check as many ethnic and racial categories as they want. By Jan. 1, 2003, all departments of the federal government must make the switch.
A more-complex count
Unfortunately, boosting accuracy also means a far more complex picture of what the US looks like. "It's going to be an accounting nightmare," says Dowell Myers, a demographer at USC.
Even the simple statistics look hard. For example, how do you count black-white people? Do you tally them with blacks? Do you include them in a separate category with all other multiethnic people?
Groups representing the rising multiethnic population have fought hard for the second option, arguing that people of mixed-race heritage are neither one nor the other but a separate group with its own agenda. Yet traditional minority groups, fearful their ranks will shrink, want anyone who checks their box to be included in their totals - even if they check other boxes as well.
The Census Bureau isn't clarifying matters. It plans to report both ways: One estimate will count multiethnic people in a separate category, a second will put them in whatever category they check - never mind that they're double-counted and that the totals add up to more than 100 percent. That's why the numbers for Sacramento, which participated last year in a dress rehearsal for the 2000 census, are so confusing.
Some demographers praise the new method; others detest it.
"The whole situation is untenable," argues Peter Skerry, author of a forthcoming book on the politics of the census. "It baffles me that the Census Bureau thinks it's going to get away with that."
Another challenge: If minorities no longer get counted as a single number but as a range of numbers the totals can easily be manipulated.
"Anybody can play the numbers," says Peter Morrison, a demographer with the RAND Corp., a Santa Monica, Calif., think tank. On one issue, for example, a minority group will be able to argue that it represents 13 percent of a city's population but only gets 6 percent of city jobs. On another, the group can claim it represents only 10 percent of the population, but its members make up 25 percent of the prison population. The discrepancies are small but will grow as the multiethnic population increases.
The biggest immediate bugaboo is redistricting. Every decade, state legislators take the latest census figures to carve up their states into congressional voting districts. The emergence of multiethnic neighborhoods will make this difficult process harder.
"If you get more and more of these numbers and more variety, each of these groups will be clamoring for representation," says Bobby Bowers, director of South Carolina's data center for statistics and a consultant who has helped draw district lines in six states. "I would not want the job in California."
Indeed, California represents the prototype for other US urban centers in the decades ahead, demographers say. Already in last year's rehearsal, the Census Bureau found that 5.4 percent of Sacramento indicated a multiethnic origin compared with 0.8 percent in Columbia, S.C., and 1.2 percent in Wisconsin's Menominee County.
California leads the nation because it contains several large ethnic groups that have lived there for generations. And since intermarriage really doesn't take off until the grandchildren of immigrants reach adulthood, multiethnic children are only now appearing.
Around Los Angeles, for example, black-white and white-Chinese families are beginning to congregate in particular suburbs, says Mr. Myers. And the diversity is fragmenting the region's politics.
"You can't any longer work without seriously talking about coalitions," says Mr. Dear. "There's no single power elite."
Nationally, more than half of the grandchildren of Hispanic and Asian immigrants marry someone outside their own group, estimates Jeffrey Passel, a demographer at the Urban Institute, a Washington think tank. And their children, by their sheer numbers, are beginning to challenge the sharply drawn identity politics practiced today. In 1995, for example, 1 in 15 Americans had multiple ancestries; by 2050, the figure could be 1 in 5.
Two trends remain unclear, however. First: Will these new groups join the mainstream or will they blur the mainstream so much that Americans will have to exchange their ethnic and racial lenses for some other way of defining themselves? Dear thinks it will be the latter.
"We're not asking people anymore to become good average American white people," he says. "What we are moving toward is this multicultural paradigm in which the hybrid becomes the rule: not the white, not the brown, not the black, but the hybrid."
The second question concerns blacks. They intermarry far less than either third-generation Asians or Hispanics (10 percent versus more than 50 percent). On the other hand, the number of children from black-white marriages is rising rapidly from a small base. In 1968, the census counted 8,800 such births; in 1994, it counted 63,000.
Will they cross the barrier that Hispanics and Asians seem ready to tear down?
"It may be that five or six decades from now we won't recognize African-Americans any more than we recognize German-Americans," says Mr. Morrison of RAND. "But my sense is it may take that long."
(c) Copyright 1999. The Christian Science Publishing Society