By Larry Jones

In a significant victory for the nation’s mayors, the President signed into law on May 21 legislation that will allow the U.S. Census Bureau to proceed with plans to use statistical sampling to achieve a more accurate census count when it conducts the 2000 census. The legislation extends funding for the Census Bureau, which is a part of the appropriations bill that funds the departments of Commerce, Justice, State and the Judiciary, through September 30. The vehicle used to give sampling the green light was the emergency supplemental appropriations bill, which provides approximately $15 billion mostly for military and humanitarian aid to support efforts underway in Kosovo. It also adds an additional $46 million for the Census Bureau to hire additional enumerators and increase the number of local offices it will use to conduct the 2000 census.

Funding for the Census Bureau and other agencies was set to end on June 15 under a compromise reached last year between the Clinton administration and the Republican leadership. For the past two years Republican leaders have adamantly opposed the Bureau’s plans to use sampling and have tried to include language in appropriations bills to prohibit the use of federal funds for such purposes. But the President, who supports the use of sampling, has been successful in using the power of the veto to prevent such restrictions from becoming enacted.

Last year Republican leaders tried to add language to the appropriations bill to block the Census Bureau from spending funds on sampling for the full fiscal year (October 1, 1998-September 30, 1999). But the Clinton administration refused to agree and threaten to veto any bill with language restricting the use of sampling. However, encouraged by victories last year in two lower courts which declared sampling unlawful, Republican leaders threaten that the Bureau and other agencies would be forced to shut down after June 15 unless the Clinton administration agreed not to use sampling. With strong support from mayors across the nation, the administration and congressional Democrats, led by Representative Carolyn B. Maloney (N.Y.), did not waiver. After the administration appealed the court decisions to the Supreme Court, both sides agreed to revisit the issue in June to give the Supreme Court time to rule. Later, however, the Supreme Court’s ruling and the worsening conflict with Yugoslavia (which made shutting down the State Department an unpopular option), would give the administration the political advantage. As a result, Republican leaders have given up for now on trying to block the Bureau from using statistical sampling.

In the lawsuits filed last year, the Republican leadership claimed that sampling was unlawful under the 1976 Census Act and the Constitution, which they believe requires an actual head count. A district court and a federal appeals court ruled in their favor, that sampling was in fact unlawful under the 1976 law. Although many were expecting the Supreme Court to resolve the issue, the decision added another complication.

In the January 25 decision, the Supreme Court ruled that the Census Act prohibited the use of sampling but only for counting the population for purposes of apportionment (determining how the 435 House seats are allocated among the 50 states). The Court explained however, that the Census Act also requires the administration to use sampling for other purposes "if it deems it feasible." The Clinton administration has interpreted this to mean sampling can be used for purposes of redistricting (re-drawing House districts within states) and for the distribution of federal aid. To comply with the Supreme Court’s decision, the Census Bureau revised its plans for conducting the census next year.

First the Bureau plans to make every effort to count everyone using traditional, person-to-person head counting methods. Figures from the traditional counting methods will be made available to states on December 31, 2000. These figures must be used to apportion House seats among the several states. The Bureau will then conduct a national survey to determine the level of undercounting and over counting that occurred during the census count and then use the data to adjust the census figures for each city and state to provide a more accurate count. Cities and states will be able to use the adjusted figures for purposes of redistricting and receiving federal aid.

In past census counts, many cities with high concentrations of minorities, children and disadvantaged residents have been severely undercounted. As a result, they have lost millions of dollars in both federal and state aid over the last ten years. Scientific sampling, which has long been supported by the Conference, will enable the Bureau to adjust each city’s census numbers for to achieve a more accurate count. By using sampling, the Bureau estimates it will only miss 300,000 people in the 2000 census. Without it, they estimate 5 million will be missed.