Supreme Court to rule on census undercount

September 28, 1995|By Lyle Denniston | Lyle Denniston,SUN NATIONAL STAFF

WASHINGTON -- The Supreme Court stepped yesterday into the historic but highly partisan fight over how the nation counts its population every ten years -- a dispute that might lead, in one stroke, to the addition of 30,000 or more people to Baltimore's size.

At issue is the census undercount: the millions of citizens, mainly minorities, who get left out of the national canvas made at the start of each decade. Citizens have been getting missed in large numbers for at least a half-century.

When the Supreme Court reaches a final ruling next year on a key constitutional issue surrounding the undercount, that might mean the shift among the states of a few seats in the House of Representatives, even before the next census in the year 2000.

And it could mean the shuffling of hundreds of millions of dollars in federal funds among various states.

Baltimore stands to get millions more, although city and state officials have no firm estimates of how much is at stake.

There have been estimates that Baltimore's undercount might translate into $15 million or more per year in federal money, although City Hall officials have doubts about the accuracy of those numbers.

Many federal aid programs are keyed directly to population figures, and thus a rise in the population count -- by making up for an undercount -- has dollars-and-cents meaning. Baltimore's population in 1990 was counted at 736,014.

Dr. Jai Ryu, the mayor's census coordinator that year, said yesterday that at the time of the 1990 count, "we were talking about an undercount of over 30,000." If that was the size of the undercount, Baltimore's population would have been bigger in 1990 than the Census Bureau calculated in a 1992 estimate: 759,127, or 23,113 more.

Nationwide, the magnitude of the undercount dispute now before the court is clear from the numbers: in 1990, the final Census Bureau count was 249,632,692. Had the Bureau adjusted for the undercount, the final tally would have been 254,902,609 -- meaning that more than 5.2 million Americans did not get counted.

Although the court will be facing only a legal issue, the undercount controversy also involves a mix of science -- particularly, the science of statistical estimating -- and politics -- particularly, which party would gain the most if more minority citizens get into the census count?

The dispute the justices will confront grows out of a July 1991 decision by a Republican administration -- President George Bush's -- not to adjust the final 1990 count to make up for the undercount. The decision ran into heavy criticism from Democrats, who accused then-Commerce Secretary Robert A. Mosbacher of doing the Republicans a favor by leaving the count uncorrected.

Mr. Mosbacher and others who defended the decision not to make the adjustment contended that it had been the Census Bureau's tradition to rely upon actual counts of people -- the first mail count, plus follow-up, even door-to-door personal contact. Doing a purely statistical adjustment to cover the undercount, it was said, would compromise the integrity of the census itself.

The 1991 decision by Mr. Mosbacher was challenged in federal court, by a group of states and cities that believed they had a stake in the undercount, and by various private and public organizations.

That led to a decision by the 2nd U.S. Circuit Court of Appeals in New York City in August saying that the failure to adjust for an undercount is unconstitutional unless the government made a genuine effort to count everyone equally.

A refusal to adjust for the undercount, the appeals court ruled, must be justified with very strong reasons, to make the refusal valid under the Constitution's guarantee of legal equality.

The Clinton administration, as well as the states of Oklahoma and Wisconsin, then took the dispute to the Supreme Court. Yesterday, taking its first actions on cases that reached it while it was in summer recess, the court agreed to hear all three appeals. A final decision is expected by next July.

It has been calculated that, if the 1990 census is adjusted for the undercount, Wisconsin and Pennsylvania would each lose a seat in the House, while Arizona and California each would gain one.

Usually, seats in the House are divided up after a ten-year census. If the high court rules that the 1990 undercount must be corrected, an earlier adjustment might have to be made. But even if that reshuffling is not done, a decision in favor of an adjustment would translate quickly into a shift of funds, well before the new census.

The Census Bureau is currently working on new statistical formulas for adjusting for the undercount in the 2000 census, but the Republican-led Congress may seek to head that off.

Should the Supreme Court rule that the Constitution dictates an adjustment, that could take the matter entirely out of Congress' and the Census Bureau's hands.

Supreme Court cases

In other cases, the court:

Baltimore Sun Articles
|
|
|
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.