U.S. Supreme Court Rejects Statistical Sampling

The U.S. Supreme Court has ruled that the government may not use statistical sampling in the upcoming year 2000 census for the purposes of apportionment. The Court did not rule on whether statistical estimates could be used in calculations for other purposes.

Census figures are of great political importance because they used in apportionment (defining Congressional districts and hence the number of Congressional seats per state), provide an estimate of the U.S. population, and determine the amount of federal funds granted to a given area, based on its population. Given that the high court’s ruling applied only to those numbers used for apportionment, the U.S. Census will likely provide two separate sets of figures, one for use in apportionment and another for other uses.

The Clinton Administration has advocated the use of statistical sampling in addition to the U.S. Census Bureau’s regular counting procedures in order to provide a more accurate count of the U.S. population. Advocates of statistical sampling argue that traditional counting methods fail to measure many of the poor and homeless, and under-represent racial minorities as well as those who do not speak English.

The Clinton Administration reported that it would proceed with plans to provide states with population estimates that could be used for all projects other than apportionment.

In a 5-4 ruling, Justices Sandra Day O’Connor, William H. Rehnquist, Antonin Scalia, Anthony M. Kennedy and Clarence Thomas ruled that a 1976 Amendment granting permitting the use of sampling in some situations did not apply to apportionment. They argued that federal law has banned the use of sampling for congressional apportionment for over 200 years.

Justices John Paul Stevens, David H. Souter, Ruth Bader Ginsburg and Stephen G. Breyer dissented, arguing that the 1976 Amendment does authorize the use of sampling for the purposes of apportionment.


Washington Post - January 26, 1999