Introduction Both policy and academic debates create new terms and concepts, either to describe new--or newly discovered--realities or to develop new ways of understanding those realities. These terms and concepts are often both politically loaded--as the recent debates over `welfare reform' and `welfare dependency' demonstrate--and poorly defined. However, because the terms and concepts of policy debate so insistently shape the way we frame and analyse social problems and their potential remedies, engaging with new policy languages is essential to properly understand and address contemporary policy concerns. The need to understand new terms of policy development, debate, and analysis has already generated some very useful explorations of the development and usage of key terms. Work by Nancy Fraser and Linda Gordon (1994) on `dependency', Richard Mulgan (2000) on `accountability' and more recently in this journal, Brendon O'Connor's examination of the origins of `welfare dependency' (2001), for example, has given insight into the ways earlier theoretical traditions and debates inform contemporary policy discussions. In the same vein, this article sets out to explore the origins of the concept of `social capital', particularly within rational choice debates over the role of the state. Social capital's rise came largely after the publication of Robert Putnam's Making Democracy Work (1993). It has since been taken up by a number of academics, policy makers and activists in Australia. However, in this country, emphasis has been placed on the concept's ability to challenge the tenets of traditional rational choice theory and the agenda of economic liberalism that is associated with it. Writers such as Eva Cox (1995), Mark Lyons (1997) and Jenny Onyx (1996) make clear a general aversion for what has been termed `economic rationalism'. Even those associated with a more economically orthodox agenda, such as Mark Latham, Peter Botsman and other advocates of the `Third Way' have been careful to use `social capital' as a way of distancing themselves from more extreme forms of economic liberalism, while at the same time rejecting much of the Left critique of the free-market agenda (see Botsman & Latham 2001). Yet, social capital theory has been intimately linked with developments in rational choice theory. This link is well noted by a range of authors in the international literature (e.g. Ostrom 1990). Indeed some definitions of social capital rely entirely on a game theoretic conceptualisation of the social sciences (Paldam 2000). Recently, this rational choice lineage has been used as the basis for a critique both of social capital as a useful social science concept, and of the politics of the Third Way which is so closely connected to this new literature. Ben Fine has claimed that social capital is taking up the role of sociology and political economy in explaining the importance of social relations to economic and political outcomes. However, social capital theory cannot deliver the goods because it is embedded in a rational choice framework that necessarily reduces the social to a small number of variables and therefore loses the complexity and contingency necessary for social analysis, in short that social capital is a rational choice `Trojan Horse' (Fine 2001: Fine & Green 2000). This article investigates the evolution of the rational choice debates around collective action which became one source of the current social capital debate. In particular the article draws attention to the crucial role played by rational choice debates in developing the currently popular conception of social capital as a collective resource that affects the macro political and economic outcomes of communities. The claim made here is that social capital developed as a solution to internal methodological debates within rational choice theory, rather than as a sociological critique, and that these rational choice debates were intimately linked to an anti-state ideological agenda. …