Identity Crisis

When Tiger Woods is polled in the 2000 census, he will no longer be forced into a racial pigeonhole. The young golf champ doesn't want to be described as an African-American, as he usually is, but as a multiracial "Cablinasian"--a person of caucasian, black, American Indian and Asian ancestry.

The Office of Management and Budget (OMB) has decided, after years of agonizing, to shift the way the government collects racial information. This promises to be a fundamental change that goes to the heart of the emotional, political and legal wrangles over how Americans identify--and view--themselves.

No longer will federal forms demand that you pick just one of four racial categories (white; black; American Indian or Alaskan Native; or Asian or Pacific Islander) and declare whether you are of Hispanic origin. Starting with the next decennial census, you will be able to check as many boxes as apply, from a slightly altered menu of five racial categories (including the new category of Native Hawaiian or other Pacific Islander, which has been split off from Asian), in addition to saying whether you are Hispanic.

At first, the change won't affect all that many Americans. Estimates of what proportion of the population is of multiple ancestry range from just over 1 per cent to almost 7 per cent. But the proportion of interracial marriages is increasing dramatically, and the numbers could be much bigger within a few decades. Already, in some large cities, the concentration of mixed-race populations in some neighborhoods could make a difference in city council and school board districts.

OMB, which oversees all federal paperwork, arrived at its decision after spending four years conducting test runs and other research, holding public hearings, creating a 30-member federal interagency committee and collecting the conflicting opinions of civil rights leaders, state and local officials, business executives, social scientists and a gaggle of other interested parties.

That was the easy part. Now, OMB and the federal agencies that use racial data must answer even harder questions: How will they tabulate the results? How will they report them? And there's not much time to figure it all out.

"It took a couple of years to decide . . . to allow [people to check] more than one [box]," Isabelle Katz Pinzler, acting assistant attorney general for civil rights, acknowledged at a Nov. 6 meeting of a National Conference of State Legislatures task force on redistricting. "Now all the rest of these issues are going to have to be decided in a couple of months. We are in a real crunch."

A lot is riding on the tabulations. They'll have far-reaching implications for redrawing congressional, state legislative and local council districts; for enforcing civil rights laws; and for planning and targeting a host of health, social and economic programs.

Say, for example, Tiger Woods or someone like him checks the boxes for white, black, American Indian and Asian. How would he be counted under the Voting Rights Act for the purpose of congressional redistricting? If he's placed in a combined category, would that offer the same legal protections as a minority category--such as black or Hispanic--covered by the statute? Would he be put in one category? If so, which one? Or would he be apportioned among four categories? How?

"We can't make any decisions yet on how to handle something like that," said Ellen Tewes, deputy legislative attorney in the Tennessee legislature's Office of Legislation. She reflects the views of many of her counterparts in other states: "I don't have any answers now. And it probably won't be resolved until there is litigation."

And that's only one side of a multifaceted problem. For instance, since many casual observers perceive Woods to be African-American, would the Justice Department consider him black when enforcing employment discrimination laws? That's the direction in which Justice is leaning, according to Pinzler, but the courts will probably be asked to decide that question as well.

Or, if the Centers for Disease Control wanted to target a health problem that disproportionately affects Asian-Americans, how would a multiple response like Woods's be factored into the equation?

In deciding to allow multiple racial responses, OMB policy makers struck a compromise. Old-line civil rights groups wanted to keep the traditional race categories, and a handful of new groups representing people of mixed race lobbied to add a single, "multiracial" box. Officially, the agency rejected the multiracial category on statistical grounds. Someone of mixed white-and-Asian ancestry, for example, wouldn't have enough in common with someone of mixed black-and-Indian ancestry to make the catch-all classification statistically useful, by OMB's explanation.

The process that led to the compromise was "comprehensive, thorough and done in a professional manner," insisted Sally Katzen, who supervised the undertaking as the administrator of OMB's Office of Information and Regulatory Affairs. The next phase, too--deciding on how to tabulate the data--"is being driven by career statistical professionals who bring to the process their professionalism and not some political agenda."

Politics, however, is lurking. The traditional civil rights groups fear that a multiracial category would result in a loss of constituents. Many supporters of a single, multiracial category who also oppose race-based policies such as affirmative action hope to fragment the longstanding, rigid racial identities.

"My own take is that this is not a statistical issue, it's a political issue," said Edward J. Spar, executive director of the Council of Professional Associations on Federal Statistics, in Alexandria, Va. "The politics are in the details and the devil is in the politics, not in the statistics."

Mix and Match

Like any compromise, the decision to allow multiple responses allowed each side of the debate to claim victory--initially, at least. "Although Asian-Americans and American Indians have expressed some concerns, I think many people felt it was a rather elegant response to a difficult problem," said Wade J. Henderson, the executive director of the Washington-based Leadership Conference on Civil Rights. "You allowed people the freedom to choose, without making a judgment on what that represented."

But, as with many compromises, the decision has also sparked mistrust all around, as the interested parties contemplate how to put the new rule into effect.

Civil rights groups worry that government agencies tabulating the data, when confronted with a plethora of racial combinations, are likely to lump the smaller ones into a single category labeled "multiple response." Indeed, OMB's background paper on its decision suggested that course as an agency option. If that happens, "as far as I'm concerned, you've just back-doored yourself into a multiracial category," said Karen Narasaki, the executive director of the Washington-based Asian Pacific American Legal Consortium.

But the advocates of a single multiracial category are no less suspicious that government statisticians might stick some people who claim more than one race into a single-race category. Susan Graham, the executive director of Project RACE in Roswell, Ga., among the most vocal of those advocates, expressed her concerns in an editorial on the group's Web site: "Will [people who check more than one box] be retrofitted back in to ONE box? We are certain the NAACP will have a hand in this, as well. As one NAACP official said, 'Let those mixed people check all the boxes they want--but COUNT them as black.' Is this fair and accurate? No. Is this racist and discriminatory? Yes."

The intent, as OMB's Katzen described it, is for federal statistical agencies to present as much detail as possible by listing the many racial combinations and then mixing and matching them to suit particular statistical or administrative needs.

It will be complex and confusing, she conceded. "This is an increasingly sophisticated world we're dealing with and simplistic, one-size-fits-all solutions, I think, are less in favor. . . . Life is complicated. People are complicated."

With that complexity could come both practical and legal problems. "Most likely, you'll get more combinations than you've got colors on the screen, if you want to map [racial identity]," statistician Spar commented.

Perhaps more important are the concerns about confidentiality. Federal law requires the Census Bureau to keep the facts and figures about any individual strictly private. Providing statistics about people of rare racial mixes could help others guess who they are. So federal agencies are certain to lump those groups together, whatever they do with groups of more-common racial mixes.

Another side issue relates to self-identification. The whole racial classification system is predicated on people declaring their own race, no matter what they look like or who their parents are. Except when they can't. OMB's directive allows for a designated "observer" to determine a person's race--"if necessary." The example that OMB director Franklin D. Raines often cites is the need to specify a person's race on a death certificate, for which self-identification is clearly impossible. But these observers--employers, school officials or the like--may make the judgment in other circumstances, although no one seems sure just which ones or how many. This troubles interest groups on both sides of the debate.

"We call this 'eyeballing,' " Project RACE's Graham wrote. "A school administrator, federal agency employee, employer, etc., can LOOK AT YOU AND DETERMINE YOUR RACE. . . . This is one of the most unfair and prejudiced of all racial injustices."

The National Council of La Raza, a Washington-based Hispanic advocacy group, is worried about the practice for a different reason. On most federal forms, the question about Hispanic origin is asked separately, before the question about race. But when an observer does the identifying, the two questions are combined. Since Hispanics--people whose ancestors' native tongue was Spanish--are of various races, "there could be major distinctions between those who self-identify and those who are observed," said Eric Rodriguez, a La Raza policy analyst. Studies in predominantly Hispanic communities in California show that people of mixed Hispanic and Anglo ancestry more often identify themselves as Hispanic. An observer might not.

Academics who take a more detached view tend to play down the hand-wringing. "This is a tempest in a little teapot," said University of Wisconsin (Milwaukee) history professor Margo Anderson, an expert on the census. "The cultural and identity and symbolic politics are what get people's juices flowing. This isn't going to make that much difference," at least right away.

Trial runs the Census Bureau and other agencies conducted over the past four years found that given the option, fewer than 2 per cent of Americans identified themselves as multiracial. Allowing multiple responses did not significantly change the number of people who identified themselves as black or white. But it did reduce the number of people who called themselves Asian-American or American Indian. Compared to people of other races, a much higher proportion of Asians and Indians marry outside their group.

Groups that represent Asian-Americans and American Indians naturally find these implications troubling. At the same time, black and Hispanic groups don't find them comforting. African-American activists, in particular, point to the continuing stigma of being black in America. Merely allowing people to check more than one racial category, they contend, will encourage many African-Americans to do so.

"There is an unspoken fear that by giving people the option to claim multiple ancestry, you will play into a desire that people have, not so much for cultural self-determination, but for being something other than black," a longtime civil rights advocate said. "By giving people the option, you will in fact diminish the number of people who claim to be African-American--not because the definition has changed, but because people have an option to escape. It's not that they think their lives will improve, but they'll be able to say to themselves, 'This is not me. I'm different.' "

First Stop: The Courts

When it comes to allowing multiple responses, there's consensus on one thing. "Everyone knows for sure that there's going to be litigation," La Raza's Rodriguez said. "It's sort of like, you're damned if you do and damned if you don't--and there's not much you can do."

Civil rights groups, exhausted from their battles over affirmative action--most recently the thwarting of the nomination of Bill Lann Lee as assistant attorney general for civil rights--aren't eager for a new round of judicial fights.

"This is basically going to reopen a lot of civil rights enforcement issues," Narasaki said. "The litigators are telling me that they're going to be spending their time in court now, relitigating a lot of issues. There couldn't be a worse time to be trying to do that."

Indeed, Pinzler prefaced her remarks to state legislators recently by acknowledging, "The Department of Justice, of course, is not in a position to dictate how the courts will apply the data that result from the changes."

Those involved in redistricting, particularly in the 18 (mainly southern) states that are covered by the Voting Rights Act, fret that allowing multiple responses could complicate their task after the 2000 census. Language in the 1965 law prohibits "retrogression"--the adoption of a state law that diminishes the voting power of any covered minority group. Results of the 2000 census may do that. Say a congressional district had a 51 per cent black majority after the 1990 census, and say the majority shrinks--because of new tabulation methods--to 49 per cent. State legislators are nervous about a legal challenge.

Katzen insists that statistical "crosswalks" will let numbers crunchers make historical comparisons between the 1990 census and the 2000 census. But legislators express concern that these statistical connections may not stand up in court. If there's not "a good bridge from the 1990 data to the new 2000 data using this different way of collecting," asked Tim Storey, the resident expert on redistricting at the National Conference of State Legislatures, "how do you measure retrogression? How do you tell if the states are drawing more or less districts [with a majority of minority voters] than they should have vis e vis the 1990 maps?"

Timing is also a problem. Under another provision of the Voting Rights Act, Justice or a federal court must approve districting changes in advance if they bring a decline in so-called majority-minority districts.

"There can be a fight on that," explained Jeffrey M. Wice, a Washington attorney who consults with Democrats on redistricting matters. "Different folks in the the legislative redistricting bodies might try to . . . maintain, strengthen or dismantle these districts.

"So, we were hoping that the Justice Department would have provided some early guidance to the OMB decision makers," he continued. "But that hasn't happened yet."

Based on the research, officials at OMB and Justice predict that African-American identification won't change enough to undermine existing black-majority congressional districts. Any changes the 2000 census brings, they say, will happen in the handful of districts with pockets of Asian-Americans or Indians.

"There were no statistically significant differences [in the test runs] in the numbers reported for blacks, for example, which is the area in which we're hearing questions about voting rights," Katzen said. "That leads me to believe that, proceeding as we are, we'll be getting very comparable information and not changing the relationships."

But nobody knows for sure.

Redistricting probably won't be affected this time around, historian Anderson said. But, she conceded, many imponderables on racial identification remain. "Is there such a thing as a multiracial community?" she said. "Or, are there multiracial people living inside the white, Spanish, Asian and black communities? We don't know yet because we haven't been able to literally see it."

On this quandary, civil rights groups and redistricting practitioners are worried that OMB's tests don't really predict how people will mark their census forms on April 1, 2000.

"I don't believe we're going to have a very good sense of what could happen, because we won't have any tests to show where there could be pitfalls [in the tabulations]," Rodriguez said. "We don't have a lot of evidence. What we have is theories. We'll have to think hypothetically about how this is going to look--and that could make things very difficult" when state legislatures redraw the district boundaries.

Or when civil rights groups go to court to enforce laws that ban discrimination in housing or employment. "I am very sympathetic to the whole issue of multirace," Narasaki said. "I have nieces of different races, and so I understand the emotional issue about self-identification.

"But the point is, you collect the racial and ethnic data for a reason," she continued. "If you can't tabulate it, you've undermined the ability of the federal government to provide information that will help set policy and help ensure that the civil rights laws are effectively enforced. That's the bottom line."

To reinforce that bottom line, the lawyers on all sides of the debate are sharpening their brief-writing pencils and pulling out their yellow legal pads. To back them up, there's another cadre of professionals who are breathlessly warming up their computers.

"Just the way changes in the [tax code] keep accountants in business, this is going to keep the statistical and social-science analysts in business for the next 10 years, at least," Spar said. "Now is the time to go to graduate school in social science."