This dissertation examines the evolution in thinking about public policy devoted to social protection by drawing on welfare theory and its extensions, state-civil society frameworks, and transnational networks. It does this through case studies of the Ismaili Imamat, Ismaili community institutions, and AKDN in Canada and Tanzania. The key question posed is: How are we to understand the role in welfare provision of an evolving trans-state -or perhaps international non-state -cluster of institutions that have their origins and owe their existence to a religious faith? The importance of understanding hybridity, intermediary roles, scale, and regional configurations are highlighted.The dissertation comes to the following conclusions on the AKDN and welfare provision. First, regarding the AKDN, although the Imamat level may be occupying that intermediary space and is involved in the orchestration and management of welfare production, the capacity for the mechanisms of the Imamat to also engage in a fully realized potential of a logic of hybridity is a more difficult proposition. The AKDN is a culmination of elements that began in the Ismaili community, and it is still evolving. So the capacity to seamlessly engage in a logic of hybridity or not be confined to a particular sector or scale is an ongoing process. Second, regarding welfare provision, the focus on self-reliance and independence are important factors in meeting welfare needs. However, context is important. For example, in Canada, the Ismaili community institutions are more robust, highly professional, and are operating within the broader context of Canadian state welfare provision and other external services. This in not the case in Tanzania where community institutions are resource strapped and government or external iii services are unavailable. As well, the extent of collaboration between the more comprehensive AKDN entities and community institutions in Tanzania are unresolved.Although there were pockets of collaboration between AKDN entities and government, it could be described as strategic cooperation rather than co-production. Finally, the level of Imamat engagement in welfare production demonstrated that innovation in this area is rooted in the need to break away from sector confinements and re-evaluate thinking about how and where welfare production can take place.iv