True Colors: The Confusion Over Who We Are

By STEVEN A. HOLMES

     WASHINGTON — EVER since the framers of the Constitution reached a compromise to count black
     slaves as three-fifths of a person, racial numbers have been about political power. In contemporary
America, the apportionment of political power on the basis of population, the use of race in the drawing
of legislative districts and racial bloc voting have prompted people to inflate their numbers. Or at least
fear their dilution. That was why several civil rights groups and the Congressional Black Caucus opposed
allowing people to choose more than one race on the 2000 census.

But when the politics of racial numbers ran headlong into the desire of interracial couples to not have
their children be forced to choose one race, the result was a Solomonic decision that may give a distorted
picture of the nation's racial landscape. Responding in part to political lobbying by civil rights
organizations, the Clinton administration decided that people who designate themselves as both white and
a member of a racial minority — the overwhelming bulk of the 6.8 million people who listed themselves
as multiracial — will be counted as members of the minority. For example, people who say they are white
and American Indian are counted by the federal government as American Indian.

The dizzying uncertainty over the best way to count people who list themselves as more than one race
points out two of the starkest facts now emerging about America's racial and ethnic profile: never have
Americans been so diverse; never have they been so confused.

The expansion of the racial categories and the sharp increase during the 1990's in the number of
Hispanics, who can be of any race, has scrambled the racial landscape. It has also challenged the way the
country looks at race. How, for example, should a people be counted? The way a government sees them or
the way they see themselves?

If people can list themselves as an amalgam of race, should they then be asked how they think of
themselves primarily? If the government can expand the officially designated racial categories from 5 to
63, why not expand them further? Is it the government's role to ask people about their race at all? If not,
how does it measure racial progress except by anecdote?

Since the census allocates political power as much as it sketches a portrait of the country, there is another
abiding truth that flows out of the 2000 census: the answers to these questions will be worked out by
politicians and interest groups, rather than scientists and demographers. Numbers are power, too
important to be left to technocrats.

Expect creative solutions, some of which may create other problems. Modern black leaders may cringe at
the comparison, but the compromise reached over how to count multiracial people is little more than a
continuation of a tradition of political deals involving the census that began with the three- fifths
compromise.

In this latest case, the compromise may have assuaged officials of traditional civil rights groups and
parents of multiracial children, but at least one expert, Roderick J. Harrison, former head of the Census
Bureau's racial statistics branch, believes the solution may have made racial and ethnic statistics far less
reliable.

Mr. Harrison, who helped design the racial question on the 2000 census form, now says that because
people were given no guidance on how to answer it, many who listed themselves as more than one race
may have done so not because their parents are of different races, but because they have an ancestor,
possibly generations back, of another race.

Indeed, while the census does not ask people what they consider their primary race, the National Health
Interview Survey, an annual survey that also allows people to list themselves as more than one race, has
been doing just that. What it found, according to Mr. Harrison, was that in recent years 25.2 percent of the
people who described themselves as both black and white considered themselves white; 46.9 percent
who said they were white and Asian thought of themselves as white, and 80.9 percent who designated
themselves as white and Indian believed themselves to be white.

Mr. Harrison argues that the gap between the number of people the government lists as a particular race
and the number of people who think of themselves as being of that race not only adds to the confusion, but
leaves the government figures open to legal challenge.

"What one could easily foresee, if I was called in as an expert witness and asked if this was a reasonable
and fair way to count people, I would have to say no," Mr. Harrison said.

How to count those who opt for the multiracial categories is not the only confusing issue flowing out of the
census. Recently Orlando Patterson, the Harvard sociologist, has criticized the tendency of many
newspapers, including The New York Times, to describe Hispanics who list themselves as whites as
different from non-Hispanic whites. After all, 48 percent of Latinos list themselves as white. Drawing a
distinction between these two groups has allowed the erroneous view to become popularized that by the
middle of this century, "whites" will be a minority in America, he says.

But Robert B. Hill, a statistician who is the vice chairman of a citizens advisory committee to the census,
argues that while many Hispanic people may describe themselves as white, they may not see themselves
that way. "Many of those Hispanics who checked white feel they didn't have much of a choice," he said.
"They didn't want to check black. They're not going to check Asian. But they sure don't see themselves as
Anglos."

While Professor Patterson and Mr. Hill would insist they disagree, they are in accord on a more
fundamental question. They are both asking whether people should be classified with a particular group
using the old criteria of skin color or language, or whether they should be classified based on behavior or
attitudes.

Layered over the new, often muddled demographic landscape is the country's love of numbers, and the
political, social and psychological weight that the data carry. Liberal civil rights groups look at the
expanding number of minorities and see a justification for policies like affirmative action. Conservatives
look at the same numbers, and the growth of multiracial categories, and say race has become too fluid a
concept to maintain race-based programs like affirmative action. Some groups, like Arab-Americans and
people from the Indian subcontinent, look at the racial designations and seek categories of their own.
Nativists see the numbers, particularly the shrinking proportion of non-Hispanic whites, and fan the flames
of racial hatred.

"Especially in the Beltway, or in policy circles, numbers matter for good and bad reasons," said Charles
Kamasaki, senior vice president of the National Council of La Raza, a Hispanic advocacy group. "You
would hope that policy makers would take into account information like statistics and base policy
judgments on them. But they become more than that."

--------------

June 3, 2001

Who Are You? When Perception Is Reality

By GREGORY RODRIGUEZ

   LOS ANGELES — THE growing national crusade against racial profiling is being called the civil rights
   movement of this generation. Polls have shown that a majority of Americans believe profiling is
widespread, and an even greater majority disapprove of it. Yet throughout the country, the evidence for its
existence is largely anecdotal. For that reason, the first wave of legislation and public policy initiatives
designed to address the problem focus on the collection of data intended to substantiate allegations of racial
bias in policing.

Two years ago, no police department in the country collected sufficiently detailed records of traffic stops to
show whether officers were targeting minority drivers. Today, roughly 400 law enforcement agencies are
compiling such data. The Justice Department is encouraging more agencies to follow suit as a way to overcome
public distrust.

But the gathering of this information and the methods used are on a collision course with another rising national
trend. Just as America's insatiable need for information has engendered a growing concern for privacy, the
government's desire to collect more racial data conflicts with a rising public reluctance to answer questions
about racial or ethnic backgrounds.

Perhaps as early as this week, a federal judge will officially place the Los Angeles Police Department under a
sweeping federal consent degree, a portion of which requires officers to record what they perceive is the race,
ethnicity or national origin of each driver they stop — without asking a direct question.

Arguing that Los Angeles was too ethnically and racially diverse to ever make sense of such information, both
the police department and the city strongly resisted the data collection section of the decree. "How are officers
going to guess what background people like me come from?" asks Raquelle de la Rocha, president of the Los
Angeles Board of Police Commissioners. Ms. de la Rocha, the child of a Mexican father and a Filipino mother,
is particularly offended by the notion that police officers will be guessing the racial and ethnic origins of each
citizen they stop.

"With all the racially mixed people in L.A., and Latinos coming in all shades, the data will be garbage in,
garbage out," she says.

But the Department of Justice — as well as the many police departments that have voluntarily begun to collect
racial data — all agree that having officers explicitly asking people their racial background would only
worsen tensions between the police and the communities they serve. Furthermore, say experts on racial
monitoring, the purpose of such data gathering is not to find out the drivers' actual backgrounds, but to record
the police officers' perceptions of them. "We're not trying to get at truth, we're trying to get at bias," says Margo
J. Schlanger, who teaches at Harvard Law School.

TODAY, two days before Angelenos go to the polls to elect a new mayor, the L.A.P.D. is to inaugurate a pilot
program in which 40 officers from two police divisions will begin testing two handheld computer devices into
which they will register the newly required data. Back at the station house, officers will upload the information
into a mainframe computer. But in Los Angeles, as in other cities, it is still unclear how the data will be
analyzed and what benchmarks will be used to measure it. "That's the $10 million question," says Professor
Schlanger, who studies racial profiling data collection. "Comparing the number of people stopped to the
number of people there are is tricky."

But other academics, including John Lamberth, an associate professor of psychology at Temple University, say
appropriate benchmarks can be devised. Because drivers are often just passing through a neighborhood, it
would not be statistically valid to compare the number of stops in one section of a street or highway to the
area's residential population. Professor Lamberth has proposed local sampling, whereby enumerators sitting in
stationary cars or standing on street corners would guess the racial and ethnic breakdown of the commuter
population in a given sector of the highway at various times of the day and night.

In North Carolina, one professor has suggested that researchers could estimate the racial and ethnic makeup of
the "violator population" by having troopers move with traffic and record both the speed and relevant
demographics of the drivers from behind tinted windshields.

In New Jersey, all 2,700 state troopers were required to receive instruction on how to classify a motorist's
race by judging "skin color" and "facial characteristics." While the multiracial option on the 2000 Census now
allows Americans to identify themselves as any one of up to 126 potential racial and ethnic combinations,
troopers are asked to place motorists into one of six racial- ethnic categories.

When Supreme Court Justice Harry A. Blackmun wrote his famous dictum, in 1978, that to get beyond racism,
we first must consider race, most Americans could understand his logic. A generation later, however, racial
monitoring efforts that make the public even more race conscious than it already is may strike many Americans
as feeding into a vicious circle.

So, despite the noble purposes for this monitoring, the prospect of the most visible arm of the government, the
police, guessing the racial backgrounds of citizens is problematic to many observers. "We paint ourselves into
a corner," says Ruben G. Rumbaut, a sociology professor at Michigan State University. "In order to combat
discrimination, we monitor race, which, in turn, only solidifies and hardens these racial categories."

In addition, the Census Bureau has allowed Americans to self-identify their backgrounds on questionnaires
since 1960. In 1995, the Office of Management and Budget, which coordinates the activities of all federal
statistical agencies, issued a directive stating that self-identification should be encouraged.

"There's a real inconsistency in a system based on the right to self-identification also allowing people to be
identified on the bases of the perception of others," says Kenneth W. Prewitt, a dean at New School University
and the former head of the Census Bureau.

COMPLICATING matters is the fact that over the past several decades, a growing number of multiracial
Americans have demanded the right to identify themselves outside of the four racial — and one ethnic —
categories normally provided by government agencies. Last year's decision by the Census Bureau to allow
Americans to mark more than one racial box on their questionnaire was only the first salvo in a war that an
increasing number of Americans are waging against the old mutually exclusive racial categories.

"Race in America is increasingly becoming a matter of subjectivity," says Peter Skerry, a fellow at the
Brookings Institute, who recently published a book on the racial politics of the census. "At the same time, the
government is relying more and more on this morass of data to make public policy."

For example, in 1998, the first year the University of California abolished affirmative action in admissions, the
number of applicants who declined to state their race shot up by 213 percent. Citing the need to get a full
picture of the impact of race-blind policies, admissions directors at the University of California at Los Angeles
and the University of California at San Diego later admitted they peeked into SAT data in an attempt to
ascertain those students' racial backgrounds.

Americans' distaste of the idea of the government assigning racial or ethnic categories to individuals comes, in
part, from the nation's past failure to live up to its individualistic values when it came to members of various
minority groups.

Civil rights advocates and racial data experts reject the charge that increasing the collection of racial
information will only make race more salient in everyday life. If race were not already a pervasive factor in
policing, they argue, there would not be a national uproar over racial profiling.

"I think it's a good thing to make officers more conscious of race," says Deborah A. Ramirez, a professor at
Northeastern University Law School and a consultant to the Department of Justice.

Yet recording the apparent background of every citizen involved in a traffic stop could engender even more
types of racial score-keeping. In mid- May, seeking to combat charges that the New York Police Department
targets minorities, Mayor Rudolph W. Giuliani and Police Commissioner Bernard B. Kerik announced that the
city would begin releasing monthly statistics showing the race of crime victims, suspects, people stopped and
frisked and those arrested.

As with all other racially charged issues over the past decade, the first major battle over racial data collecting
is likely to take place in California over a ballot initiative. Ward Connerly, the Sacramento businessman who
sponsored the successful anti-affirmative action measure in 1996, is now gathering signatures to place his
"Racial Privacy Initiative" on the state ballot in March of 2002. Ironically, Connerly has exempted law
enforcement data from the measure, along with information collected to aid medical research. Aimed at barring
public agencies from identifying people by race or ethnicity, the initiative, according to Mr. Connerly, will
help "California government to stop obsessing about race."

But the abolition of racial data is as likely to eradicate the national obsession with race as widespread racial
monitoring will make Americans less race-conscious. A frontal assault on race statistics could also provoke
the opposite reaction, reinforcing the very demographic categories that racial privacy advocates believe
already hold too much sway over American life.

Gregory Rodriguez is a senior fellow at New America Foundation, a nonpartisan public policy institute.