Social capital is a contentious and multifaceted topic. A broad consensus has been reached, however, that norms such as trust, networks of association membership, and practices of volunteering and socializing are essential to its makeup. It is also increasingly recognized that such elements fall into two distinct types of social capital – bonding and bridging. Social structural influences, such as welfare, have an effect upon social capital. A common conservative conception is that welfare induces dependency and thereby erodes social capital; this can be called the ‘dependency hypothesis’. I suggest this is largely limited to bonding capital only, however. I suggest an alternative, that welfare cutbacks or contingencies upon mutual obligation or status preservation is socially divisive. I call this the ‘division hypothesis’, and it is relevant to the more definitively positive bridging capital. This paper shows trends in bridging social capital in nine OECD countries of differing welfare regime type from 1981 to 2000. It uses data from the most recent versions of the World Values Survey and Multinational Time Use Study. It suggests that welfare regime type, and importantly, welfare regime re-structuring, bears strong relations to national levels of bridging social capital.