This article examines the relationship between sub-state nationalism and the welfare state through the case of Que´bec in Canada. It argues that social policy presents mobilisation and identity-building potential for sub-state nationalism, and that nationalist movements affect the structure of welfare states. Nationalism and the welfare state revolve around the notion of solidarity. Because they often involve transfers of money between citizens, social programmes raise the issue of the specific community whose members should exhibit social and economic solidarity. From this perspective, nationalist movements are likely to seek the congruence between the 'national community' (as conceptualised by their leaders) and the 'social community' (the community where redistributive mechanisms should operate). Moreover, the political discourse of social policy lends itself well to national identity-building because it is typically underpinned by collective values and principles. Finally, pressures stemming from sub-state nationalism tend to reshape the policy agenda at both the state and the substate level while favouring the asymmetrical decentralisation of the welfare state.Nationalist movements in advanced industrial democracies operate in a political environment where decisions about a wide range of policy matters are taken on a regular basis. Overall, little attention has been paid to the relationship between nationalism and public policy as researchers have preferred to focus instead on the constitutional and institutional implications of claims for self-determination. To be fair, specialists of nationalism have not ignored all policy areas. There is a significant literature on nationalism and linguistic policies (McRae 1986;Laponce 2003). Education and broadcasting policies have also been studied for how they are affected by nationalist movements (Erk 2003). The interest in these particular policy areas is not surprising. After all, sub-state nationalism in Western Europe and Canada often features language as a key identity marker (for example, in Flanders, Catalonia and Que´bec), which means that language, education and broadcasting policies are likely to come under nationalist pressures. There has been much less research conducted on nationalism and policy fields not immediately related to language (or culture). 1 This is perhaps because it is not readily expected that day-to-day policy-making in areas not featuring language (or culture) would shape nationalist movements, and that these movements would in turn affect policy-making in these areas.