Measuring the impact of residential mobility on response: evidence from the Millennium Cohort Study

Abstract

This paper examines the relationship between residential mobility and unit non-response in the first five waves of the UK Millennium Cohort Study (MCS). The objective is to ascertain whether home moves affect the likelihood of response and whether any impact persists over time.  The existing literature is extended by examining the impact of moving home on the likelihood of returning to the survey after dropping out in a previous wave. The findings show that by the fifth wave of MCS more than two thirds of respondents had experienced at least one home move, with most moves happening before wave 2. Residential mobility is found to have a negative impact on subsequent response, even though this impact does not persist over time. Put differently, moving home is circumstantial and movers are likely to come back to the survey after being absent in a previous wave. The findings also shed light on the importance of tracing home movers in order to maintain the sample representativeness in a long-term longitudinal survey.