Monday, September 30, 2002
Court Rules U.S. Need Not Use Sampling Data for Redistricting
By KENNETH OFGANG, Staff Writer/Appellate Courts
The Bush administration did not abuse its authority by deciding not to adopt statistically adjusted census data for redistricting purposes, the Ninth U.S. Circuit Court of Appeals ruled Friday.
Secretary of Commerce Donald Evans was entitled to rely on expert opinions that the methodology used for sampling of data to account for individuals not located through the traditional “head-counting” method in 2000 was flawed, Senior Judge Warren Ferguson wrote for the court.
Judge Susan Graber concurred in the opinion. Judge Stephen Reinhardt dissented, saying Evans’ conclusion was unreasonable and thus contrary to Sec. 195 of the Census Act, which says that the secretary “shall” approve sampling “if he considers it feasible.”
All three panelists were appointed by Democratic presidents—Ferguson and Reinhardt by Jimmy Carter and Graber by Bill Clinton.
The court affirmed the grant of summary judgment in favor of Evans by U.S. District Judge Gary Feess of the Central District of California, a Clinton appointee. The suit was brought by the cities of Los Angeles; Inglewood; San Francisco; San Antonio, Tex.; New York; Chicago; Albuquerque, N.M.; and Toledo, Ohio; Santa Clara County; Cook County, Ill.; and a number of individual public officials including Lt. Gov. Cruz M. Bustamante.
Funding at Stake
State and local governments have a great deal at stake when it comes to use of census data. The final population figures published by the Census Bureau are used not only to apportion the House of Representatives, but to draw legislative districts throughout the country and to distribute some $185 billion in federal grants.
Ferguson, writing for the court, explained that the Census Bureau has used sampling techniques since 1940 in order to supplement the information it gains from the headcounts. Legislation specifying that the secretary of commerce “may” use sampling “where he deems it appropriate” was enacted in 1957, and was amended in 1976, when the present “shall...if he considers it feasible” language was adopted.
Consideration has been given over the years, Ferguson explained, to using sampling to correct the census undercount, which is believed to result in disproportionately low numbers for racial and ethnic minorities, children, and renters. But the Carter administration, and that of President George Bush, rejected proposals to incorporate sampling into the official counts for 1980 and 1990, the judge noted.
In 1997, however, the Clinton administration submitted a plan to Congress that would have used sampling to correct the initial headcount. That plan precipitated legal challenges, and three years ago the Supreme Court ruled that—for purposes of apportioning representatives among the states—the use of sampling was prohibited by the Census Act.
The Census Bureau followed up on that ruling by adopting procedures for the use of sampling for purposes other than apportionment, and appointing a committee of senior bureau officials and statisticians—the Executive Steering Committee for Accuracy and Coverage Evaluation Policy, or ESCAP—to evaluate the data.
In October 2000, then-Secretary of Commerce William Daley adopted a rule delegating authority to the census director to make the final decision regarding adjustment of the census data, but most of the rule was repealed by Evans last year. The current secretary did, however, leave intact a provision that ESCAP’s report be released to the public.
Two weeks after Evans revoked the bulk of the Daley rule, ESCAP reported that after weeks of “examining voluminous evidence,” it could not recommend the use of the sampled data to adjust the census count because it was not fully convinced of its reliability. The conclusion was unanimous.
The acting census director concurred, and Evans directed that only the unadjusted data be released for redistricting purposes.
The various cities and officials brought their suit in district court in February of last year, arguing that Evans’ decision violated the Census Act because the use of sampling was “feasible.” But Feess said the secretary’s administrative determination was entitled to deference, and that both ESCAP’s concerns about the data and Evans’ interpretation of the statute were reasonable.
Ferguson agreed with the district judge on all issues.
“[Census Act] Section 195 confers broad discretion on the Secretary to decide if and when sampling is feasible in the taking of the decennial census,” he wrote.
Congress’ use of the term “feasible,” the judge went on to say, was ambiguous, enabling the secretary to rely upon his own definition of the term as long as it was reasonable.
“We find that Secretary Evans’ interpretation of the statute, as permitting him to consider accuracy as a component of feasibility, was a permissible construction of the statute,” Ferguson said. “Further, his decision not to statistically adjust Census 2000 was neither arbitrary nor capricious.”
Reinhardt, however, argued in dissent that any reasonable definition of “feasible” would have resulted in the use of sampling.
He noted that ESCAP had found that “a majority of the evidence indicates both the continued existence of a differential undercount of the population and the superior accuracy of the adjusted numbers,” but had recommended against using those numbers because of “[t]he potential for a reversal of these findings.”
That “potential,” the dissenting jurist declared, “does not constitute a determination as to feasibility, or even a consideration of that question,” and does not justify ignoring the unadjusted headcount’s “profound and unjust effect on polities and communities in which members of undercounted groups disproportionately reside.”
The case was argued in the Ninth Circuit by Thomas M. Riordan of O’Melveny & Myers for the plaintiffs and Michael S. Raab of the Department of Justice for the government.
The case is City of Los Angeles v. U.S. Department of Commerce, 01-55986.
Copyright 2002, Metropolitan News Company