This paper estimates associations between individual and neighbourhood characteristics and unit nonresponse in a survey of the population aged 50 and over in the Netherlands in 2004. The statistical model includes interviewer fixed effects to control for the non-random distribution of addresses...