Evaluating the immediate and longer term impact of a refusal conversion strategy in a large scale longitudinal study

Refusal conversion is one of the fieldwork strategies commonly used to minimise non-response in surveys. There is, however, relatively little evidence about the effectiveness of this strategy, particularly for face-to-face longitudinal surveys. Moreover, much of the existing evidence is based on observational studies. This paper evaluates the effectiveness of a fieldwork strategy - intensive re-issuing - to convert refusals using evidence from an intervention on a random sub-sample of refusals implemented in wave four of a large scale longitudinal study in the UK: the Millennium Cohort Study. We show that intensive re-issuing is an effective way of reducing the refusal rate. We also show that refusal conversion led to a modest reduction in non-response bias in the survey estimates for several key variables. The longer term impact of refusal conversion is also a key concern in longitudinal surveys. We demonstrate that, although the majority of converted refusals go on to participate in the subsequent wave of the study, there is no overall effect of intensive re-issuing on sample size at this wave.