The Law of the Census: How to Count, What to Count, Whom to Count, and Where to Count Them

Abstract

The Framers of the American Constitution viewed the decennial census as providing a certain rhythm to American politics. Every ten years a state’s tax burdens and representation in the House of Representatives would change to reflect its share of the national population as revealed in the “actual enumeration,” the manner of which Congress “shall by law direct.” Much has changed since the first census, but the rhythm still remains. Perhaps unintended and unimagined by the Framers, however, is the rhythmic and ritualistic dance to the courtroom every ten years to argue over the census numbers themselves and the methods used to construct apportionment totals.

This Article examines the law of the census: specifically, how to count, what to count, whom to count, and where to count them. For the most part this Article draws on my experience and research concerning the use of census data in the redistricting process; however, many of the topics discussed apply to federal funding decisions as well. The Article begins by describing the most recent legal controversies involving census methods, particularly imputation and statistical adjustment. When one thinks of the “law of the census,” these high-profile disputes probably come first to mind. In cases that have arrived at the Supreme Court at the beginning of each of the last three census cycles, undercounted cities and states have argued that census methods were deficient in that the procedures missed some people, double-counted others, or counted people that did not exist.

Second, this Article explores the legal implications of the decisions concerning what to include on the census form, paying particular attention to the topics of race and citizenship. For the second time, the 2010 Census allows respondents to check off more than one race, raising a host of interesting questions concerning the legal implications of alternative methods for categorizing the multiracial population. More significantly for the 2010 Census, the long form, which was previously asked of one sixth of the population, has been replaced by the yearly American Community Survey (ACS), distributed to 2.5% of households. The ACS is the only source for citizenship data from the census, raising questions about the reliability of citizenship estimates for purposes of VRA litigation.

Finally, this Article examines the related issues of who should be included in the census and where they should be counted. The section deals with special populations such as soldiers and other Americans living abroad, college students, the homeless, and prisoners. Prisoners, in particular, present an important case, as some have argued that the wholesale involuntary transfer of convicted criminals away from their communities toward more rural and often whiter areas has allowed for the padding of legislative districts in one part of a state at the expense of districts in other parts of a state. For the first time in 2011, the census will make data available in time for redistricting about the number of prisoners in each census block. Jurisdictions will now be able to subtract out prisoners from the census redistricting data file. Some states have even gone further and have reallocated prisoners to their pre-incarceration address for purposes of redistricting.