"The stakes are a bit higher for us"

The NAACP's Washington bureau chief takes the Census Bureau to task for its new multiracial categories.

Published February 16, 2000 5:00PM (EST)

Tiger Woods could be the poster boy for the 2000 Census, as it attempts to count multiracial Americans for the first time. Although the champion putter won't find "Cablinasian," the term Woods famously told Oprah he preferred to African-American, he will this year be able to check all four racial categories he belongs to: African-American, Asian, white and Native American.

But while the new data may please Woods, as well as demographers and historians keen to trace multiracial population trends, civil rights groups are worried. Ever mindful of the basic principle of strength in numbers, organizations such as the NAACP and the National Asian Pacific American Legal Consortium are concerned that the myriad combinations resulting from the new categories could dilute estimates of racial populations, and cost some groups political clout. They say the new formulas could curtail the enforcement of equal-rights laws and the allocation of funding for race-targeted government programs such as health care, education and public transportation.

But their solution -- let Americans of mixed parentage identify all the races in their family tree, but then have the government assign them to a designated racial group for the purposes of civil rights enforcement -- will no doubt bother some advocates for the multiracial, who don't want the government, or the NAACP, deciding which race they legally belong to.

Only weeks before the release of its latest survey, the U.S. Census Bureau still hasn't decided how it will process the new data -- an example of bureaucratic disorganization that has only compounded the fears of all parties to the debate. The new form could result in as many as 63 racial combinations, or 126 if Hispanics -- an ethnic group, not a race -- are factored in.

In a recent interview, Hilary Shelton, director of the NAACP's Washington bureau, talked about how the good intentions that led to the introduction of multiracial categories in Census 2000 could backfire.

What are the NAACP's arguments against the proposed "mark all boxes that apply" multiracial census?

We're not opposed to the "mark all boxes" -- we're probably the last group that would opposed something like that. We went through the process where the country referred to us as all sorts of things, and we wanted to identify for ourselves what we are. We decided that we're African-Americans. It's simply a way of allowing people to self-identify instead of simply being locked into an area.

How, then, have you been critical of the census revisions?

Our biggest criticism is how long it's taking for the federal government to decide how they're going to process this information. We've had a series of meetings with the various government agencies, including the White House, the Office of Management and Budget, the Census Bureau and others, to discuss how to guarantee that our traditional civil rights are protected as we allow people to self-identify.

And how do you think it's progressing now?

It still hasn't. There are still things that need to be worked out -- they have not announced how exactly they're going to process the information. After the information is gathered, it really has to be configured in a way that allows us to enforce our existing civil rights laws. And those laws aren't written based on 63 multiple-choice categories. A decision is going to have to be made to make sure that we can continue, even as we self-identify, to process the information.

What alternatives do you propose?

I would strongly recommend a tabulation process that would look at the various combinations and then place them in enforcement categories. For example, if you have someone who is half African-American and half white, and if you look at the historic discrimination problems that people have had in our society -- whether it's in employment, admissions into colleges and universities, securing government contracts or living in communities where they've had redlining or whatever -- the conclusion you would draw is that somebody who is half white and half black would be discriminated against if they applied for a bank loan, because of (being) half black. You want to make sure that, as we look at the historic need for enforcement against discrimination, you recognize that that person is part African-American.

What about someone who is Native American and African-American?

When you get into more of the historically discriminated-against ethnic minority groups, you would want to be able to look at where they are and how they are and use that as a premise for calculation. If someone were African-American and Native American and was recognized at an Indian reservation as being a member of that tribe, then the best thing for them to be recognized as is Native American. That provides the greatest civil rights protection for them, in so many different ways.

But won't the new mixed categorizations be more reflective of American diversity?

Yes, I think it's a good thing. We are a very diverse country, and that's our strength.

So what most concerns you?

The census figures are used for the reapportionment and redistricting process that determines our representation in Washington. The census figures are also used to figure out whether that representation is adequate -- that the representation is reflective of the diversity of people who live in that state and in that newly cut congressional district. You want to make sure that the districts are cut in a way that's consistent with the people who live there, whether they be working-class people, agricultural people or otherwise reflective of the interests of the folks who live there.

Probably the biggest danger is making sure that nobody gets left out. For the sake of argument, let's say a particular congressional district has been cut this way: 15 percent of the population is African-American, 4 percent of the population is Asian-American, 8 percent is Hispanic and so forth. You have a very large employer, maybe an automobile manufacturer. Arguments are being made that the automobile manufacturer is discriminating against people of color, and they say, "No, our workforce is reflective of the community that our plant is located in." Then as they begin to make their case for why they are representative, they say, "Well, let's look at the census figures" for this particular area that we live in. They may take a look and say, "No, 95 percent of the working population here is Caucasian." If it's inaccurate, then we've got a problem here, and let's see what we can do to work it out.

If people choose three or four different categories, won't that present problems for a strict definition of races, like African-American or American Indian?

Sure. As with anything else, what you're looking for is to see just how reflective it is, whether it's African-Americans or Native Americans or Hispanics or people who are Native American and Asian, whatever the case might be. But it's important that there are some numbers. With the history our nation has for discriminating against people across racial lines, there needs to be something you can point to, to say that these numbers are reflective of the population; and that's what we count on the census figures to be.

With all the new data and new racial categories, how will researchers compare the data to the 1990 census levels? Will it complicate analysis?

That's a part of the challenge. That's why we also argue for what's called a historic arc -- that is, where the compilation of the multicheckoffs find their way back to a historic category, for the purposes of the very evaluation you're talking about.

If an acceptable solution isn't found for analyzing the revised census data, do you think African-Americans will be disproportionately affected?

The historic problems that ... ethnic minority groups have had in this country as a result of their ethnic minority status -- like having lower wages -- affect us more severely. If you look at the number of people who are unemployed, you will see that the unemployment rate has dropped significantly, but the group that has had the lowest drop is African-Americans. If you look at even things like violence, every group has experienced a decline in homicides and gun-related incidents, except African-American males between the ages of 14 and 17. Because of our experiences, the stakes are a bit higher for us.

Census statistics are used for the allocation of programmatic dollars -- everything from education and health care to transportation. In the 1990 census, 11,000 school-age children in Richmond, Va., were not counted because they were in the areas where they weren't able to actually do the enumeration, for one reason or another. Even though they knew those people lived there because of indicators like mortgages, rent or income tax, but because they weren't actually able to enumerate them, they didn't get counted. As a result, the highest undercount among children were children of color and children in the high-poverty areas. Because they were in high-poverty areas, they were the children who could have most benefited from the federal programs that give dollars to these communities. About $640 per child is given to the poorest communities to help offset a strapped tax base, since (about) 90 percent of school funding is local and 10 percent comes from the federal government. In Richmond alone, that's over $7 million per year that did not go into the school system, that it should rightfully have gotten. The people who lost out were disproportionately ethnic minorities and African-American.

George Bush was the president when the figures came down in 1990. When he saw what a fiasco the census had become, he sent a challenge to the Congress to come back to him with a recommendation that would ensure that this would never happen again. The Congress then allocated funding to bring on some of the top scientific associations to do some research and decide the best approach to it. The National Academy of Sciences and a number of others rose to the challenge and came back and made a recommendation: scientific sampling. The same scientific sampling that was struck down by the Supreme Court because of a lawsuit brought on by Newt Gingrich.

And you support scientific sampling?

Absolutely. What it does is simply say that we work as hard as we can to collect the data using the traditional enumeration process, mail out the data with the hope that people will mail it back and send out enumerators to go door to door. If we get to at least 90 percent of the population and we still know that there's a 10 percent gap that needs to be closed, then they use a basic scientific process that's used by both Wall Street and Madison Avenue to fill in those gaps. Nothing about it is partisan in any way -- it's used by scientists, pollsters and economists all over our country and the world to process data. We could then crunch out the figures and have a 100-percent count. Then, when we start to look at everything from where to put bus stops to new hospitals, we would have a completed strand of data to work with.

But the Supreme Court specifically barred the Census Bureau from using sampling in its 2000 survey.

Well, even though the Supreme Court decision that was handed down doesn't allow for the use of scientific sampling for reapportionment, it doesn't speak to the issue of redistricting. It would be much more accurate if we could use scientific sampling figures to do the redistricting in the states. Unfortunately, however, we have a number of states that have decided that it's not beneficial to them to do this. Arizona, Alaska and others have already passed legislation outlawing scientific sampling for redistricting.

We're a nonpartisan organization and we don't lean to one party or the other, but we were miffed to say the least to see that the chairman of the Republican Party would send a memorandum out to all of his state chairs telling them that scientific sampling wasn't helpful to the Republican Party because the people that would be counted were not people they saw as being traditionally Republican. That's why they wanted to fight the use of scientific sampling. In this person's interpretation, they put party interests above broader national interests. Instead of wanting to make sure that every single American was accounted for, they thought it would be just as well to maintain the status quo and not count a large group of Americans that is predominantly poor and predominantly people of color. Because not counting them would make it easier for them to get reelected. And that's unfair.

Your position is: People should have the right to self-identify as multiracial; then the NAACP wants the Census Bureau to revise those figures and put people back in their historical categories in their final reports. How can you have it both ways?

The tabulation at the end is extremely important. The emphasis on the tabulation at the end is to see to it that we can most fully and consistently enforce our existing civil rights laws. Civil rights laws are there to protect ethnic minority groups that have a history of being discriminated against. When you do the analysis of who's being discriminated against, there has to be some tool to see who is self-identifying and which ethnic minority groups they would have historically been placed in for the purposes of civil rights enforcement -- so the full weight of the law can be used to provide these people with protection.

But how can the NAACP take on the role of determining the ethnic identity of someone who would historically be considered African-American but is now self-identifying as a mixture of black and other races? How is it qualified to make such personal decisions for people?

The NAACP wouldn't be involved in making the determination -- it would be involved in helping to steer the government agencies that are responsible for that enforcement to make sure that they're sensitive to what the multiple checkoff is going to mean for historic discrimination and present-day enforcement of civil rights laws. Does that make sense?

The fundamental fear is that you're going to have people who look African-American, who have grown up in communities that are African-American, who are going to experience the discrimination that African-Americans have historically experienced, but not have even the existing civil rights protections because they're being identified now in this census in a way that's inconsistent with how it's been monitored in the past and how law enforcement is tooled to enforce it now. The numbers aren't important. The enforcement of civil rights are. No matter what you're calling yourself, we want to make sure you don't get lost or left on the periphery of civil rights protection.

By Daryl Lindsey

Daryl Lindsey is associate editor of Salon News and an Arthur Burns fellow. He currently lives in Berlin and writes for Salon and Die Welt.

MORE FROM Daryl Lindsey

Related Topics ------------------------------------------