Getting the Facts on the Census Correct: A Response to Brunell

Extract

As the initial results of the 2000 Census were made public, the December 2000 issue of PS was published. In that issue, articles by Lynne Billard (2000), Thomas Brunell (2000a), and by us (Anderson and Fienberg 2000a) explored the technical innovations for adjusting the census to ameliorate the effect of the differential undercount and the political controversies surrounding the implementation of some of those innovations in the 1990 census. We commented on Brunell (2000a) in Anderson and Fienberg (2000b). Brunell (2000b) responded to our article on mythmaking, which focused on the implication for apportionment of a computer error in the proposed adjustment. This is our response to Brunell's comments on our original article about census adjustment and apportionment in 1990. In particular we point out that Brunell misread both out tables and our discussion about the computer error. In his comments, Brunell (2000b, 2000c) also raises new issues about adjustment. We respond to these, albeit briefly.