U.S. Supreme Court Rejects Sampling for Census

The U.S. Supreme Court ruled 5-4 on Monday that a scientific method of counting Americans in the 2000 census, known as statistical sampling, cannot be used to enhance the census’ accuracy.

The opinion, written by Justice Sandra Day O’Connor, says that the census law prohibits the use of sampling to adjust figures used to allocate congressional apportionment among the states. According to the decision, sampling may still be used to help draw lines of federal, state and local legislative districts and to distribute federal funds.

Direct marketers use demographic information from the census.

Some 4 million people were missed in the 1990 census, especially children and members of racial and ethnic minority groups, according to the Census Bureau. The bureau proposed the use of sampling to supplement the traditional headcount and increase accuracy, which was “strongly supported by the National Academy of Sciences and the statistical community,” the bureau said in a statement.

The ruling is a victory for the Republicans in the U.S. House of Representatives who had filed the suit challenging the bureau’s plan. The Clinton administration and the Democrats supported sampling, arguing that it would make for a more accurate count, though it would also add groups to the total that traditionally vote Democratic.