Census a Touchy Issue for State’s GOP House Members

By opposing the Clinton administration’s blueprint for the 2000 census on the grounds that population sampling would destroy the count’s credibility, Republican lawmakers from California and some other Sunbelt states are risking the possibility that their states will lose out on hundreds of millions of dollars in federal aid.

Republicans who denounce the plans to supplement the traditional door-to-door census with a follow-up survey to correct errors contend that they are protecting the electorate from rank manipulation and upholding the Constitution. Democrats say such scientific sampling is the only way to correct an unacceptably high rate of counting errors in a polyglot and increasingly mobile society.

The partisan debate is getting amplified these days in Washington as Congress and the administration square off over how the first census of the new millennium will be conducted, with just a year until its launch.

Behind the rhetoric lies the plain political reality that the people who are easiest to miss in a traditional census are usually judged more likely to vote for Democrats at a time when Republicans are aiming to hold on to their tenuous majority in the House.

But often overlooked amid the partisan maneuvering are the fiscal stakes for states such as California, Texas and Florida--places with hard-to-count masses of new immigrants and urban poor. Studies suggest that a sampling-enhanced census, yielding higher population figures for those states, could channel hundreds of millions of dollars in federal funding to them over the next decade.

Such aid would benefit programs including indigent health care, child care block grants and vocational education.

A much-debated Supreme Court decision in January barred the use of census sampling for the purpose of dividing congressional seats among the 50 states. The court’s ruling, however, allowed sampling for such purposes as redrawing legislative boundaries within states and calculating how to distribute more than $150 billion a year in federal aid to states. The Clinton administration promptly vowed to use census sampling for those ends.

Despite the strong fiscal evidence, just one of the 24 Republican representatives from California openly supports using sampling in the next census to help divvy up federal money: Tom Campbell of San Jose.

Rep. David Dreier (R-San Dimas), chairman of the House Rules Committee, acknowledged the issue’s sensitivity one recent afternoon as his panel took testimony from Democrats on the importance of census sampling for counting Latinos and African Americans--and for California’s federal take.

Although he declined to spell out his own position in detail, Dreier said afterward: “I do believe we should have a free-flowing debate on this. . . . I want to make sure we count every Californian.”

Democrats say many California Republicans have flip-flopped.

In 1992, a cross-section of the state delegation--including such influential Republicans as Dreier and Rep. William M. Thomas (R-Bakersfield)--petitioned the federal government to adjust California’s 1990 census total of 29.7 million residents upward by using the results from a survey that was in fact a population sample.

“It is not only a fair decision, it is the right decision,” the members wrote, seeking to ensure that the state received a fair share of federal dollars. The state’s undercount in the last census was estimated at more than 800,000, the largest in the nation.

Now, a year before the next census gets underway on April Fool’s Day 2000, such bipartisan sentiments are scarcely heard. Instead, the GOP-led Congress and the Democratic administration of President Clinton are barreling toward a confrontation over the marching orders for the Census Bureau that threatens to cut off the funding for several federal agencies as early as June 15.

Commerce Department officials, who oversee the census, have stepped up their attacks recently on what they call Republican meddling in a task best left to demographic professionals and scientific experts. This week, census administrators are fanning out across the country to mobilize public support for a complete count, particularly in minority communities.

On Monday, Deputy Commerce Secretary Robert L. Mallett said that three census management bills pushed by Republicans in Congress were “simply untenable” and could bust the census deadlines and endanger its accuracy.

The most controversial of the three, HR 472, is known as “post-census local review.” Sponsored by Rep. Dan Miller (R-Fla.), chairman of a House census oversight panel, the bill would expand a program tried in previous counts. But the Clinton administration, which has threatened a veto, says that granting local agencies power to double-check their preliminary census figures failed to make a significant dent in the 1990 undercount.

Republicans reply that the administration’s plan for a “two-number” census will undermine public confidence in an institution that dates back to the nation’s founding. One court-mandated figure the Census Bureau plans to generate for congressional apportionment would use only a direct head count; another, for other purposes, would draw on both a head count and a follow-up sample of 300,000 households.

California’s 52-member House delegation virtually mirrors the national partisan divide.

Rep. Henry A. Waxman (D-Los Angeles), citing a recent study by the General Accounting Office that found California lost $220 million in fiscal 1998 because of the 1990 census undercount, challenged Republicans to explain why they would forgo such sums. “The fact of the matter is, the states are entitled to get money from the federal government to provide these services,” Waxman said.

But GOP lawmakers insist that they do want to count everyone in the state--the traditional way. Although sampling is supported by a number of independent experts and a National Research Council census advisory panel convened by the National Academy of Sciences, those California Republicans who agreed to be interviewed--not all would--derided the method as “polling” and “guesstimates” open to political manipulation.

For Rep. Dana Rohrabacher (R-Huntington Beach), the prospect of a two-number census amounts to “just another case of liberal Democrats wanting to destroy standards.”

Rep. Doug Ose (R-Sacramento) said he opposes census sampling in part because of his own campaign experience last fall. “Sampling,” Ose said, meaning a poll, “showed me up 16 points the Friday before the election. And I only won by eight. I’m not very convinced by the argument of some that sampling enhances the accuracy of actual enumeration. . . . Our objective is to make the census as accurate as possible, and let the money fall where it may.”

But Campbell, a maverick within his party, said he is lobbying for a compromise. Why not bar sampling for the politically sensitive purpose of redistricting, he said, just as the Supreme Court barred it for apportionment? Campbell hopes that would clear the way for legislation to allow sampling for distributing federal funds. “Clearly, that’s a separate issue,” he said, “and it should partake of the very best estimate of the numbers of people.”