Anton Johansson: Can we increase contact rates and reduce costs in a longitudinal survey by including an SMS in the contact protocol? Results from an embedded experiment

In social surveys where respondents are contacted by telephone, a large part of the nonresponse often consists of noncontacts. We have tested SMS text messages as pre-notification with the aim to increase contact rates and reduce costs. We evaluate this intervention using an embedded experiment in a longitudinal survey conducted by Statistics Sweden. In a longitudinal survey we have knowledge about the result in the previous round of data collection. Results from the experiment suggest that the pre-notification has different effects depending on the previous survey round (i.e. if the respondent was coded as an interview, a refusal or a noncontact).