Incentives in Random Digit Dial Telephone Surveys: A Replication and Extension

Singer, Van Hoewyk, and Maher (2000) found that a $5 prepaid incentive markedly increased response rates during the period 1996 to 1998 on the University of Michigan’s monthly RDD Survey of Consumer Attitudes (SCA). Although this led the study to adopt that incentive, SCA’s response rate has declined sharply since then. We examine (a) whether the effect on response rates of the $5 prepaid incentive has diminished; and (b) whether prepaid incentives of $10 would, as predicted from earlier research, show a larger effect. In addition, we look at the effect of incentives on number of calls to obtain an interview, item nonresponse, response distributions, and sample composition, thus replicating analyses of the earlier article. We also examine the cost-effectiveness of prepaid initial incentives versus promised refusal conversion payments and the effect of differences between cases for which addresses can and cannot be obtained, topics neglected in the earlier article. Finally, we discuss implications of incentives for nonresponse bias.