| Abstract: | To what extent do concerns about privacy and confidentialityaffect people's participation in surveys, in particular thelargest U.S. survey, the decennial census? This study presentsthree separate estimates of the effect of attitudes on behavior.First, it estimates the effect of privacy and confidentialityconcerns on willingness to provide an address to a Gallup interviewer.Second, based on the Census Bureau's matching of survey responsesto its file of census returns, it estimates the effect of privacyand confidentiality concerns on respondents' return of theircensus form. Finally, it estimates these effects in one-personhouseholds. It concludes that concerns about privacy and confidentialityhave a small but statistically significant effect in all threetests, explaining roughly the same amount of variance in the2000 census as they had in 1990. |
