# Department of Commerce v. House of Representatives Supreme Court decision of January 25, 1999, holding that the Census Act bars the use of statistical sampling to produce the population figures used to apportion seats in the House of Representatives. The Bureau had planned to use sampling to correct the known differential undercount in the 2000 count. The Court resolved the case on statutory rather than constitutional grounds: 13 U.S.C. § 195 permits sampling throughout Title 13 *except* for the apportionment determination, so the plan failed without reaching whether the Constitution would have permitted it. The practical result is the split that still governs — sampling remains lawful for the redistricting data file and for funding formulas, and is barred for the one number that assigns seats. That split is why the apportionment total is the least correctable statistic the federal government produces. The post-enumeration survey measures [[Census Undercount]] precisely and changes nothing, and [[Utah v. Evans]] (2002) later upheld hot-deck imputation as enumeration rather than sampling — a distinction that moved a House seat. See [[How the Census Apportions Congress]]. > [!note] Stub — fill out > This page was created as a stub during learning-module generation on 2026-09-12. Audit and expand when convenient.