Numbers running

New Crisis, The, May/Jun 2001 by Harrison, Roderick J

The 2000 Census showed that the U.S. population has grown by 6 million more people than had been predicted. The Hispanic population has increased by more than 60 percent in the last decade. And together, Blacks and Hispanics represent one in four Americans. BY RODERICK J. HARRISON

Census 2000 is certain to stand out as one of the most important in our on's history.

It will be remembered for many reasons, from the failed efforts to use sampling to correct for the undercount of minorities to the successful efforts by thousands of organizations to have their communities complete the census and be counted. Then there are the historic changes that allowed respondents to report more than one race and the startling revelation of a nation more heavily Hispanic than analysts had anticipated

For African Americans, it was a census with positive and potentially problematic results. Civil rights and community groups were decisive in helping to reduce the undercount, and there's little doubt that their efforts spilled over into a remarkable turnout at the polls in November's presidential election. The potential pitfalls, however, lie in the new multiple race category, which may have offered more opportunity to define one's heritage, but also could be detrimental to efforts to monitor and enforce civil rights.

What's already dear is that this census will have a profound impact on the quality of American lives and how we see ourselves as a nation and as a people.

THE RACE QUESTION

Race is central to the question of U.S. national identity, and it's played a key part in the census since its inception.

Article 1, Section 2 of the U.S. Constitution, ratified in 1787, spelled it out. In its mandate that the seats of the House of Representatives be apportioned among the states every 10 years based on a population count, it also specified that slaves should be counted as three-fifths of a person and "Indians, not taxed" (those residing in sovereign tribes) should be excluded. On the other end of the spectrum, the Voting Rights Act of 1965 - which prohibits practices designed to dilute the voting power of minority populations - has kept counts of racial and, since 1970, Hispanic origin populations, integral to the redistricting that occurs with each census.

Now in the aftermath of the 2000 Census, we are faced with the continuing challenge to sort out what our population shifts will mean in terms of political and economic power and fairness in a truly diverse country.

THE COUNT

The Bureau of the Census had planned for the 2000 Census to be the first to adjust for undercounted populations. It planned to use samples of the non-responding households in each census tract to statistically represent all such households in the tract. But in January 1999, the U.S. Supreme Court barred the bureau from using sampling for apportioning congressional seats. The bureau was forced to use its traditional methods, the same methods that had yielded undercounts for decades.

Given that reality, the bureau would have done well to keep the undercount at the 1990 level. There had been significant growth in hard-to-count communities of minorities, immigrants, non-English speakers, migrant workers and the undocumented. But what had remained the same was the deep-seated suspicions of government agencies in many of those communities. Bureau director Kenneth Prewitt, new in the job at the time, noted that "using traditional counting methods, [the Bureau] must run harder to stay in place. It will run harder."

So the bureau went to work. It placed 100,000 more enumerators in hard-to-- count areas, kept them there longer and, perhaps most significantly, worked with more than 140,000 partners, including many minority and civil rights groups, to try to assure a complete count.

The results were striking.

The Accuracy and Coverage Evaluation (ACE) survey that the bureau conducts to estimate the undercount indicates that it was cut from 1.6 percent in 1990 to 1.2 percent in Census 2000; even more dramatic reductions were achieved among African Americans and other minority populations. The net undercount of the Black (non-Hispanic) population in the 2000 Census was estimated at 2.2 percent, less than half of the 4.6 percent undercount in 1990. The net undercount of American Indians on reservations fell from 12.2 percent in 1990 to 4.7 percent; the off-reservation undercount was 3.3 percent. The net undercount of Hispanics was reduced by 40 percent, from 5.0 percent in 1990 to 2.9 percent.

The 2000 Census also succeeded in reducing the undercount of children by 50 percent, from about 3.2 percent in 1990 to about 1.5 in 2000. Overall, the bureau and its partners, including the NAACP, seemed to have won a major victory in the decadeslong struggle to reduce the undercount.

Still, Census 2000 is estimated to have missed 6.4 million people, and double-- counted 3.1 million; the result is a net undercount of 3.3 million. The 1990 census missed 8.4 million people and double-- counted 4.4 million, for a net undercount of 4 million. And while the relative likelihood of undercounting African Americans and other minorities in 2000 remained high, the net undercount of white nonHispanics was about 0.7 percent in both 2000 and 1990.