Imputation, Apportionment, and Statistical Methods in the U.S. Census: Issues Surrounding Utah v. Evans
Patrick J. Cantwell, Howard Hogan, and Kathleen M. Styles
KEY WORDS: Count imputation, probability sampling, legal issues, census coverage
As in all recent censuses, the U.S. Census Bureau used statistical methods in the 2000 Census to account for missing or contradictory information concerning the number of people living in some identified housing units. These statistically corrected counts were used for Congressional apportionment and redistricting. In 2002, the Supreme Court ruled that this limited use of statistics was both lawful and constitutional. This paper provides a context for that decision by tracing the evolution of statistical methods in the U.S. Census, and the accompanying litigation. It then summarizes the statistical and legal arguments and issues raised in Utah v. Evans.">
CITATION: Much of the material in this report can be found in Cantwell, Hogan, and Styles (2004), "The Use of Statistical Methods in the U.S. Census: Utah v. Evans," The American Statistician, August 2004, Vol. 58, No. 3, 203-212. This report provides a more extensive description of many of the statistical procedures applied in the 2000 Census.
Source: U.S. Census Bureau, Statistical Research Division
Created: September 30, 2004
Last revised: March 7, 2005.