Results of the 1980 US Census Challenged

Abstract

In the United States, population policymaking has often taken place in the courtroom-viz court decisions on abortion, contraception, immigration. Whatever the outcome, such a process necessarily entails policy formulation on the basis of legal merits as opposed to technological, medical, or other scientific considerations. A current example of such policymaking in process is provided by litigation between the US Bureau of the Census and several state and city governments over the validity of the 1980 US census results as a basis for apportionment of political representation and for federal funding formulas. That the 1980 census produced a nonrandom undercount is undisputed. An extensive evaluation of the size and sources of underenumeration in the decennial census of the United States, undertaken by the Census Bureau in the 1970s, led to the conclusion that the 1970 census undercounted the national population by 2.5 percent. More detailed analysis revealed, not surprisingly, that the undercount varied greatly by social stratum and by race. It was 7.7 percent for blacks and 1.9 percent for whites. Further studies highlighted additional correlates of undercount by age, education, employment status, region of the country, and rural/urban residence. Although the Census Bureau took a number of steps (including a National Academy of Sciences study of the 1970 undercount and a second study of plans for the 1980 census) to minimize the undercount problem, its position with respect to the 1980 census is that no statistically defensible method is available to adjust the reported census count and, thus, the reported count should continue to be used for apportionment and federal disbursement formulas. As has happened in numerous other countries in the past, however, the undercount issue came to take on a life of its own as a political issue. Ethnic groups with a high undercount, most notably blacks and Hispanics, and various states and localities that had high minority concentrations and that were losing population began to call for remedies to the undercount. Even before the 1980 census results were in hand, a number of localities (about 15 by the end of 1980), including the State and City of