According to Cox & McCubbins (1986), parties reward their core constituency in an effort to maintain the coalition of voters and activists that brought them into office. There is also ample evidence that intergovernmental transfers in established democracies accrue disproportionally to politically aligned local constituencies (e.g., Scheiner 2005, Solé-Ollé and Sorribas-Navarro 2008, Tavits 2009, Fouirnaies and Mutlu-Eren 2015, Hessami 2017, Coman 2020). In this paper, we make two novel contributions to this literature. First, we develop a theoretical argument integrating both the strategic choice of national parties about whether to compete in local elections or not, and the resulting effects this choice has on intergovernmental transfers back to the local level. The literature on local party system nationalization (e.g., Kjaer and Elklit 2010; Ennser-Jedenastik and Hansen 2013; Dodeigne, Close and Teuber 2021) remains largely silent on this question. Second, we exploit a national transfer program to poorer municipalities introduced in Sweden in 1917 that included a discretionary component which allowed the national government to withhold or adjust what transfers were granted. In a two-period difference-in-difference model based on data from >1600 municipalities in 1937-1939, we find that the newly installed Social Democratic-Farmers Union coalition government disproportionally rewarded municipalities in which these two parties made significant gains in previous local elections. However, if broken down by party, the effect is minuscule for the Social Democrats but quite sizable for the Farmers. We discuss possible interpretations of this result.