School of Information

This paper uses the results of a natural experiment to assess consumer demand for a particular kind of privacy—protection from unwanted telemarketing calls. We obtained the phone numbers placed on the Federal Trade Commission's do-not-call (DNC) registry, redacted for privacy. After matching those numbers to demographic and other information based on geographic location, we regressed the observed signup frequencies on individual demographic variables to profile the DNC registrants. Grouped logits of county averages of signup frequencies were run on explanatory variables suggested by a simple choice model of the DNC registration decision and the calling pattern of telemarketers. We find that a parsimonious specification accounts for most of the variance explained by the full set of variables. Signup frequencies are larger in areas of higher incomes and greater educational attainment. They are lower for areas with greater incidence of Latino origin and linguistic isolation. The results show that a state DNC list acts as a substitute for the FTC registry unless the two are merged. Finally, we make some crude estimates of the value of the FTC's DNC list, and also use the signup patterns, and the apparent independence from Internet access, to predict the popularity of a hypothetical do-not-spam registry.