Census Sampling Challenged by House Committee Report
Despite active opposition by the Conference of Mayors and a coalition of public interest organizations, the House Government Reform and Oversight Committee voted 22-12 on September 18 for a report unfavorable to the use of sampling in the 2000 Census.
Among the recommendations contained in the report, entitled "Sampling and Statistical Adjustment in the Decennial Census: Fundamental Flaws," is that, "The [Census] Bureau should not use sampling methods to complete or adjust the actual enumeration of the 2000 Census which is constitutionally mandated for purposes of apportionment."
The report itself does not have the force of law or in any way prevent the Census Bureau from moving forward with plans to incorporate statistical sampling data in the 2000 Census, but calls on Congress to enact legislation which would prevent sampling.
In moving the report to passage, Committee Chairman William Clinger (PA) said that the Census Bureau, "has failed to prove that sampling will help" produce a more accurate census, adding that the Bureau would, "create more problems than they attempt to resolve." This opinion was strongly debated by a number of committee members, led by Rep. Carrie Meek (FL), and dissenting views were filed after the final vote for passage.
The U.S. Conference of Mayors has long had strong policy supporting sampling_policy which was reaffirmed at the 1996 Annual Conference of Mayors in Cleveland this June.
The issue of sampling came to a head when following the 1990 Census, then-Commerce Secretary Robert Mosbacher decided not to use statistical data which was collected to adjust for the estimated five million persons missed by that census. A number of cities including New York City, Los Angeles and Chicago, joined by The U.S. Conference of Mayors, then brought suit to force an adjustment.
In 1995, the Clinton Administration - despite the urging of the Conference of Mayors - appealed to the Supreme Court a lower court ruling which strongly favored an undercount adjustment, and the Court eventually ruled unanimously against forcing a readjustment.
However, the Court did not prohibit the use of sampling data in a future census, and the Census Bureau has tested and does plan to use such statistical data in the 2000 Census, and to incorporate that data prior to release of any figures, thus providing for a "one number count."
In a statement issued responding to the Committee's action, Conference Executive Director J. Thomas Cochran said, "It would be absolutely unforgivable for the federal government to knowingly undercount millions of citizens of the United States of America when there is an acceptable means of including these persons in the official count of the census, and when there is ample and substantiated support within the statistical community for the use of sampling."
Copyright © 1996, US Conference of Mayors, All rights reserved.