Declining public cooperation with the census (Miskura, 1992) and surveys has been attributed in part to increasing public concerns about privacy and confidentiality (Singer, Hippler, and Schwartz, 1992). Such concerns appear to be increasingly important barriers to public cooperation in recent censuses. Concern about privacy and mistrust of confidentiality predicted lower mail response in the 1990 Census, although it had not in 1980, controlling for demographic correlates (Fay, Bates, and Moore, 1991; Kulka et a., 1991; Signer, Mathiowetz, and Couper, 1993). Receipt of a long form increases concern about privacy (Fay, Bates, and Moore, 1991).

This paper examines trends in privacy and confidentiality attitudes during the course of Census 2000, analyzes concerns about long form questions, and explores what information might allay those concerns.