Census nightmares: the more we know, the less we trust it?

In 2006 elevated levels of non-response to the census were recorded in affluent, innerurban areas characterised by high levels of socioeconomic status. This finding challenged the view, supported over decades by studies in Australia and elsewhere, that people of higher socio-economic status are more likely than others to participate in civic activities such as the census. Measures of socio-economic status rely heavily on indices of education and occupation, which may be taken as proxy measures for being 'well-informed'. The pattern identified in the 2006 census raises what would be for the Australian Bureau of Statistics the unwelcome possibility that the more we know about the census, the less likely we are to complete it. This paper reports on a qualitative study concerned with investigating reasons for non-response to the census. It begins by defining some key concepts and outlining a model for thinking about non-participation in the census, then discusses some ideas about non-participation as it relates to questions of trust. The paper concludes by setting out what remains to be done to investigate the possibility that, regarding the census, 'the more we know, the less we trust it'.