'More community!' Does the social capital hypothesis offer Hope to untrusting societies?

Academics, politicians, and commentators influenced by the idea of social capital make a general claim that enhancing the social capital of the community can restore trust. Trust is produced, they argue, by group membership and participation, intimate connections, and consultation. But we argue here that the kind of relationships that many social capital researchers describe as trust are quite distinct, and that we should not expect that straightforward ‘more community’ solutions will be effective in building greater trust. We use data from the Middle Australia Project to specify regression models that predict three kinds of trust — trust in neighbours, trust in others, and trust in government. We find definite limits to the explanatory power of measures that could meaningfully be described as social capital indicators for the latter two types of trust. We consider why social capital measures do not confirm the intuitions of social capital-based research and propose some alternative ways of looking at the problem.