High Court Clouds Scenarios for Census
Washington Post Staff Writer
Tuesday, January 26, 1999; Page A8
Rather than resolving a bitter partisan debate, the Supreme Court's ruling yesterday on the federal government's 2000 census plan set in motion a tangled new scenario of how the next national head count will be conducted.
Within hours of the court's decision, Clinton administration officials pointed to language in the opinion as evidence that they have permission to use a controversial counting system for some purposes even if they are barred from using it to distribute congressional seats among the states.
The arcane details of how the census is taken have significant political consequences because the sampling method is likely to add minorities and poor people, groups that traditionally support Democrats, to the overall population estimates.
An administration official said the Census Bureau will likely propose a new design that relies on two sets of numbers: one produced by the traditional questionnaire and door-to-door count that would be used to divide the 435 House seats among the states, and a second set "adjusted" to compensate for those people thought to have been missed by the head count.
Those adjusted numbers could then be used to distribute billions of dollars in federal funds and to redraw political boundaries within each state for congressional, state legislative and local electoral districts.
That suggestion elicited vociferous complaints from Republicans on Capitol Hill, throwing into question the fate of next year's census and once more pushing the esoteric details of statistical sampling to the forefront of national political debate.
"This is a disastrous idea," House Census subcommittee Chairman Dan Miller (R-Fla.) said of the two-number proposal. "It's a nonstarter as far as I'm concerned."
"If those sampled numbers are used for redistricting, the administration is going to find itself in court again," vowed Matthew Glavin, president of the Southeastern Legal Foundation, a conservative organization that joined House Republican leaders in challenging the sampling plan.
The remarkable irony is that the use of these adjusted numbers for redistricting could ultimately carry a much heftier political impact than the use of sampling for apportionment.
Redistricting experts estimate that just three or four seats would be shifted among states if sampling were used to divide congressional seats among the states. But the use of adjusted figures to redraw congressional districts within states could affect many more elections.
In California, for example, adjusted numbers would be likely to add tens of thousands of people, mostly to urban areas where low-income and minority families reside and are often missed in the traditional census. The larger numbers would mean that suburban districts, which are more likely to elect Republicans, would be redrawn stretching closer into the urban center, and that could jeopardize the Republican strength in those redrawn districts.
Clark Bensen, whose firm Polidata conducts redistricting research, mostly for Republicans, estimated that five to six California House seats could shift from Republicans to Democrats if the state legislature uses adjusted numbers to redraw its boundaries.
The same numbers would have even greater impact when used to redraw state legislative boundaries, perhaps shifting political control of some state legislatures between parties.
Whether such a census plan could muster support from a Republican-led Congress remains an open question, although veto threats from President Clinton over past census disagreements have forced Republicans to compromise and made this one of the most contentious issues in Washington in recent years.
Democrats were equally determined not to drop their support for the use of statistical sampling, which uses a survey to project characteristics and numbers for missing households. "We're not going to let the majority [in Congress] prevent those people from being counted," said Rep. Carolyn B. Maloney (N.Y.), the ranking Democrat on the census subcommittee.
Commerce Secretary William Daley yesterday stopped short of saying the administration would pursue a "two-number" census, but an official who asked to remain unnamed said that was the plan unless it was proved to be unfeasible, which he considered unlikely.
Daley and others pointed to the court's majority opinion, which said sampling was required for purposes other than apportionment "if feasible." And, officials said, the recent census dress rehearsal proved that, without sampling, millions of Americans would be missed in the census.
Administration officials said they would probably propose a new plan that would resemble the method used in the 1990 census, when the government used both a head count and a population estimate. But only the head count was used for official purposes.
Staff writer Juliet Eilperin contributed to this report.
© Copyright 1999 The Washington Post Company