709
Margo Anderson, Beth Osborne Daponte, Stephen E. Fienberg, Joseph B. Kadane, Bruce D. Spencer and Duane L. Steffey
Long before the issue of adjustment for differential undercount of
blacks and other minorities arose in connection with the 1980 and
1990 censuses, public controversy had attended the publication of
census results. Recently, Brown, et al. critiqued the results of
sampling-based adjustments to the 1980 and 1990 censuses. They
concluded that the proposed design for Census 2000, which includes an
updated version of such adjustment, is too risky. Their account of
past practices contains many errors and misleading statements,
however, and is deeply flawed. Here we attempt to provide a more
balanced perspective of past practices and the fascinating statistical
and legal issues that Census 2000 raises.