Department of Commerce v. United States House of Representatives

1999-01-25

Headline: Census sampling ban affirmed: Court strikes down the Census Bureau’s plan to use statistical sampling for apportionment, blocking adjustments that would have changed how states’ House seats and local redistricting counts are calculated.

Holding: The Court held that the Census Act prohibits using statistical sampling to determine population totals for congressional apportionment purposes, so the Bureau may not use the challenged sampling methods to assign House seats.

Real World Impact:

Prevents the Census Bureau from using sampling to adjust population totals for House seat apportionment.
Keeps apportionment based on the unadjusted decennial headcount, affecting states' seat counts.
Means state and local redistricting will rely on traditional census totals, at least for now.

Topics: census counting, congressional apportionment, statistical sampling, redistricting

Summary

Background

The dispute arose after the Census Bureau announced plans to use two statistical methods in the 2000 census to address a persistent undercount of certain groups: a Nonresponse Followup program (NRFU) and an Integrated Coverage Measurement (ICM) using dual-system estimation. Two sets of challengers sued — four counties and many residents in multiple States, and the United States House of Representatives — and separate three-judge federal trial courts enjoined the Bureau from using the challenged sampling procedures for apportionment.

Reasoning

The main question was whether the Census Act allows the Bureau to use statistical sampling to determine state population totals for assigning House seats. The Court examined the Act’s text and history and concluded that 13 U.S.C. §195, read in historical context, prohibits the use of sampling to determine population for congressional apportionment. The Court therefore affirmed the Virginia district court’s injunction and dismissed the appeal from the D.C. court, finding it unnecessary to decide the constitutional issues.

Real world impact

As a result, the Bureau may not adjust the decennial headcount by the challenged sampling methods when calculating state populations for apportioning Representatives. That outcome preserves the traditional unadjusted headcount for apportionment and affects which States may gain or lose House seats and how states conduct redistricting that depends on federal census totals. The decision addresses only apportionment use; other statistical uses of the census were left intact or not reached here.

Dissents or concurrances

Several Justices disagreed. Justice Stevens (joined by others) argued §141(a) clearly authorizes sampling for the decennial census and would have reversed. Justice Breyer viewed the challenged methods as permissible supplements, not substitutes, for a headcount. Justice Scalia emphasized historical tradition and raised constitutional doubt about replacing an "actual Enumeration" with estimates.