Decentralization of public health systems has been a worldwide trend to improve accountability, despite it can also increase inequality and coordination costs (Jiménez-Rubio et al., 2008;Rodríguez-Pose & Ezcurra, 2010). After the multiplication of programs aiming to combat specific diseases in the past decades, the efforts turned to the strengthening of health systems (Hafner & Shiffman, 2013). Thus, unraveling the externalities of health spending among bordering jurisdictions can help to improve public healthcare coordination. 1 This paper aims to identify the potential existence of health spending spillovers among bordering municipalities in Brazil. 2 We take advantage of the fact that Brazilian local governments have taken responsibility for primary health care in the past decades. Still, subnational entities need technical and financial cooperation from the federal government, as seen in many countries (Arzaghi & Henderson, 2005;Goodspeed, 2002).In this context, the main federal grant to Brazilian municipalities, which is called Municipalities' Participation Fund ("Fundo de Participação dos Municípios" in Portuguese [FPM] from now on), is distributed monthly according to an estimation of the local population size. The absolute value of the grant changes abruptly at pre-determined population thresholds, allowing for a "fuzzy" Regression Discontinuity Design (RDD). However, our identification strategy relies on the change of population brackets induced by local population growth, which is calculated by an algorithm that accounts for the national

No abstract

This article documents fiscal spillovers after an exogenous increase in the main federal transfer to Brazilian municipalities. We explore that Municipalities’ Participation Fund is distributed according to the local population and abruptly changes at some thresholds. We disentangle spillovers using bordering municipalities near different cutoffs, showing that the flypaper effect in local economies can be partially explained by bordering municipalities’ grants—roughly 20 percent. Fiscal spillovers are generally positive, except for spending in public health and sanitation for some population ranges.

No abstract

No abstract