Mobilising Immigrant-Origin Voters into Survey Participation: The Effect of In-Person Canvassing Versus Postal Reminders on Response Rates in a Longitudinal Election Survey

Abstract

This paper experimentally analyses the effects of in-person canvassing and postal reminders on the mobilization of non-respondents for a telephone panel survey in Germany. We compare the response rates among randomly selected individuals who were visited by canvassers and those reminded by an invitation letter. We further conduct multivariate analyses to control for different contexts in the survey area, characteristics of canvassers and information about (non-)respondents available from the sampling frame. The sample was drawn from the population registry of the city of Duisburg, Germany, and fieldwork started during the peak of the COVID-19 pandemic with most shops and restaurants being closed. Target persons in the sample were eligible voters from the two largest immigrant-origin groups in Germany who had not responded to an initial postal invitation to the survey. In-person canvassing significantly increases the probability of reacting to the survey invitation (18.3 % diff., 95 % c.i., 16.2 – 20.3 %) and to participate in the first wave of a three-wave longitudinal telephone survey (2.1 % diff., 95 % c.i., 0.7 – 3.5 %). Within the canvassed group, direct personal contact to a target person has a significantly higher effect on participation rates (14.0 % diff., 95 % c.i., 18.2 – 9.7). As a collateral plus, the fieldwork documentation helped the research team to gain more knowledge about non-respondents. However, panel attrition rises among those being canvassed (17.9 % diff., 95 % c.i., 9.7 – 26.0 %) and the cost per successfully mobilized individual is by far lower with postal reminders.

Keywords: survey response, response rate, immigrant origin, immigrant, reminder, canvassing, COVID-19, Duisburg, Germany