Distributive Politics and Regionalisation: Evidence from the Brazilian Unified Health System

Does distributive politics promote the regionalisation of healthcare at subnational level? Regionalisation emerges as an institutional response to correct inequalities in access and equalise the provision of services within countries. However, in a context of great fiscal disparity, funding is key to the success of this strategy. In this paper we focus on the role of distributive politics, which by decentralising federal funding based on the preferences of congresspeople might not align with the coordination of policy goals regarding regionalisation. We argue that legislators can promote regionalisation by allocating resources to the hub cities defined by the regional planning or hinder it by uncoordinatedly benefiting other local units. We explore the health regions implemented by Brazil's Unified Health System, which divides the state's municipalities into geographical nuclei responsible for managing health policies and actions. Empirically, we leverage data on allocation of individual budget amendments from 2009 to 2019 to understand the territorial patterns of distributive resources delivered by legislators within Brazilian states. We use spatial analysis to test three allocation mechanisms. Firstly, the strategic mechanism examines whether legislators target resources to hub cities in the health regions defined by the SUS and whether spatial spillover benefits municipalities within these predefined regions. Second, the neighbourhood mechanism investigates whether allocations disregard the regionalisation project, focusing instead on spillover effects that benefit only adjacent municipalities. Finally, we explore whether allocations are driven purely by political goals, with distributive policies favouring municipalities whose mayors have political ties to the legislators. Using Brazil as a case study strengthens our research design by facilitating comparisons between sub-national units within a consistent institutional framework. This enables a deeper understanding of territorial dynamics and the heterogeneous effects of distributive policy, supported by the extensive availability of local-level data. Additionally, it allows for the development of recommendations applicable to other federal and decentralised countries. By analysing the intersection between the allocation of distributive policies and the regionalisation of health policy, we aim to contribute to understanding the mechanisms that promote (or restrict) equal access to public policies at sub-national level.