One of the most hotly debated issues of the past four years has been the method to be used to count the population in Census 2000. Broadly, the debate is over whether to incorporate statistical sampling in the decennial census for the purposes of enumerating the population. Each of the previous 21 U.S. censuses has failed to count all persons. Both George Washington and Thomas Jefferson were certain that the state population totals tallied for the first census in 1790 were wrong (Skerry 1992). To make matters worse, the census must do more than simply tally the number of people in the country; it must also report exactly where each of these people live. By any metric, this is a Herculean task. Indeed, enumerating and fixing geographically 100% of the population without some error is impossible.

In this article I outline the plan for Census 2000, define the political debate, and offer an assessment of the plan itself. I argue that deciding how to conduct the census has been, and will continue to be, an inherently political process, and that there are reasonable, scientific arguments against adjusting the census. I became interested in this issue during my tenure as an APSA Congressional Fellow, when I staffed the Subcommittee on the Census in the House of Representatives. Before I began my fellowship on Capitol Hill, I was suspicious of the argument against adjusting the census statistically. How could anyone possibly be against adjusting the census with statistical methods?

Progress, far from consisting in change, depends on retentiveness …. Those who cannot remember the past are condemned to fulfil it.

—George Santayana, 1905-06

Those who are familiar with the “equal proportions” formula used to reapportion the U.S. House of Representatives each decade are aware how sensitive it is to small changes in states' population data. In 1990, this sensitivity took on new significance as statisticians and politicians debated the merits of adjusting the decennial census data to correct for the differential undercount. The “official” apportionment results from the 1990 Census were published in December 1990, and the reapportionment of Congressional seats among the states proceeded on the basis of those data. Adjusted census results were released in July 1991. These data would have changed the reapportionment of Congress and shifted two seats in the House: one each from the states of Wisconsin and Pennsylvania to the states of California and Arizona.

Adding to the complexity of the debates about which set of data to use, and in the midst of the ensuing litigation between New York City and the U.S. Department of Commerce, the Census Bureau discovered a computer coding error in the adjusted data estimates and issued “revised adjusted data” in January 1992.

In “History, Myth Making, and Statistics,” Anderson and Fienberg purport to have debunked one of the “myths” opponents of statistically adjusting census counts repeatedly invoke. The authors strive to show that critics have misplaced blame on a computer coding error for artificially inflating the national undercount by more than one million people and producing adjusted state population totals that would have provided justifications for shifting one or two House seats inappropriately. Rather, Anderson and Fienberg argue, an ad hoc decision to collapse the number of post-strata from 1392 to 357 led analysts to produce inaccurate numbers.

Far from discrediting the criticism, however, Anderson and Fienberg provide compelling evidence that both the computer error and the adjustment model itself produced numbers that would have led to a malapportionment of House seats. The authors report the 1990 apportionment based on official census numbers in their Table 1, the apportionment based on the incorrectly adjusted census numbers in their Table 5, and the apportionment based on the “corrected” (no computer error and with 1392 poststrata) adjusted counts in their Table 6. Note that the incorrectly adjusted apportionment differs from the actual apportionment by two seats and the correctly adjusted apportionment differs from the actual by two seats.

In his preceding article, Brunell offers some background on Census 2000, the use of sampling and adjustment in a census context, the magnitude of the differential undercount from prior censuses, and the debate over the use of sample-based adjustments to the count. Much of his description of Census 2000 is correct, as far as it goes. It is what he leaves out that is problematic. His omissions lead him to draw the wrong conclusions about the undercount and the methodology for correcting it.

The origins of the peculiar American institution of the decennial census can be traced to the Founding Fathers and the federal Constitution of 1787. When the leaders of the American Revolution met in Philadelphia in the summer of 1787 and decided to apportion seats in the new House of Representatives among the states “according to their respective numbers,” they invented a fundamental new instrument of republican government. The infant U.S. government of the Confederation Era had trouble raising taxes and making decisions, in part because representatives in the Continental Congress voted by states and the states were of very disparate sizes and populations. The framers recognized the need for another policy-making mechanism that took account of the fact that states deserved different numbers of representatives and, hence, votes in the House and Electoral College. The answer was the census, a periodic count of the population and consequent redistribution of House seats and economic resources to reflect the relative sizes of the populations of the states.

With “Partisan Politics at Work,” Anderson and Fienberg continue their spirited, if misleading, defense of the plan to statistically adjust the census count. Unfortunately, their argument in favor of making the adjustments misses the mark and they gloss over some of the most serious of the evidence presented against the methods of adjustment. Rather, they are more interested in trying to remove politics from an inherently political process, which I maintain is an unrealistic goal. I conclude my portion of this exchange by reiterating my position that statistically adjusting the census totals will, at best, provide numbers that are no better than those of a traditional headcount; at worst, doing so may provide very inaccurate data.

At the end of their second article, Anderson and Fienberg plead for “legislation to insulate the Census Bureau from efforts at political manipulation and restore to it the authority to manage the technical details of how to fulfill the constitutional mandate for a census.” I do not share these wishes. Why does the federal bureaucracy need any more protection from congressional oversight than it already enjoys? Certainly, Census Bureau officials have more technical expertise and they are more readily able to judge whether a particular sampling plan is sound than are the collective members of the U.S. House of Representatives. Nevertheless, the Bureau remains the fiduciary for Congress. Erecting barriers between the two groups is not a solution to any identifiable problem.

Legislators in democratic countries find it relatively easy to do certain things. For example, they typically have little problem reducing taxes or increasing benefits for a majority of citizens. However, should they attempt to raise taxes or cut benefits for the majority, they will probably encounter opposition, even if taking such measures is necessary for the public good. How members of democratically elected legislatures manage to accomplish unpopular and politically difficult tasks is worth examining. In this article, I illustrate how members of the U.S. Congress have “done the difficult”–i.e., passed laws or instituted programs, often in the public interest, that were unpopular with the voters or would have been were they known to the public–and consider why the legislators were able to take the actions they did.

Determining how members of Congress make unpopular but necessary policy decisions is difficult because members themselves will not emphasize when they deviate from the public's will and how they are able to do so. After all, most of them are more interested in getting reelected than in being perceived as courageous and resourceful statesmen. Even the most civic-minded legislator may avoid calling attention to unpopular votes after realizing cynical and skeptical voters are unlikely to reward him for making difficult choices in the public's interest. Consequently, to reduce the likelihood that the action will be highlighted, unpopular legislation is often enacted in obscure and technical ways.