ABSTRACT -------- To what extent do the effects of pork barrel policies in public policy spill over? Pork barrel greases the wheels of politics in many ways, being a crucial tool for building support in Congress and with the electorate. Due to their discretionary nature, they are considered an inefficient, poorly planned expense with a high possibility of generating corruption and creating clientelistic networks. Despite its lousy reputation, Pork barrel policies influence the electorate’s daily life as they positively affect public policy in the target jurisdiction, even with small and dissipative effects. However, these effects could be underestimated as neither research considers the spatial relationship among municipalities. I argue that pork barrel collective effects are more prominent when considering spatial spillover effects, which permit them to reach neighbouring jurisdictions. Spatial interdependence among local units and the strategic allocation of pork monies draws this overflowing feature. The restriction on the amount of pork barrel policies available for distribution forces politicians to target units that can better absorb the demand of neighbouring areas and better process the benefit due to the accumulation of human and material resources. Consequently, spatial spillover effects increase the pork’s beneficiary network, helping to soften its inefficiency and becoming politically efficient despite increasing subnational unevenness in policy provision. Using Brazil as a case, I leverage the municipal level in an original time-series cross-sectional dataset, gathering data on 33,240 impositional budgetary amendments and health indicators for 5,214 municipalities between 2014 and 2018. I apply the Spatial Durbin Model that accounts for the interdependence among local units and verifies both the direct and indirect effects in target units and their neighbouring municipalities, respectively. The results show that legislators allocate pork barrel resources following a well-defined strategy: concentrating the highest share of pork barrel monies in a single municipality while benefiting its neighbours with lower values. Furthermore, suggest that pork barrel has a positive (albeit small) effect on health policy in the target municipality and neighbouring local units. Nonetheless, these effects are more prominent when analysing policy input indicators (measured as annual spending per capita) while losing their strength when analysed as policy process indicators, measured as BCG vaccine coverage, and outcome indicators, measured through the infant mortality rate. Taken together, these results contribute to understanding the effects of distributive policies on public policies at the subnational level and how the territorial dynamics can extend the network of beneficiaries and, ultimately, soothe the inefficiency of spending on the pork barrel policy.