This project aims to enhance the understanding of the public finance related impacts of the German local-level direct democracy reforms of 1990s. The central focus is on Bavaria, where since the introduction of direct democracy by a state-wide referendum in 1995 around 2500 referendums and citizen initiatives took place. Yet, unlike the extensive literature on Swiss and US direct democratic institutions, there exists no empirical evidence on how these referendums shape and interact with local-level fiscal decisions. The main research goal in this project is to quantify the effect of these direct democracy reforms i) on municipal spending levels, ii) on the level and structure of local taxes, and iii) on efficiency aspects of local-level public good provision. The project has a particularly strong identification strategy, which in the initial phase relies on a “selection-on-observables” approach by using the observed variance in the intensity of initiatives and referenda across municipalities and years. To verify the results against endogeneity and causality concerns, the project benefits from a quasi-experimental type of approach called “difference-in-discontinuity” design which exploits the variation in direct democracy institutions. The results of the project will contribute to the ongoing debate on the political economy aspects of fiscal policy decisions with a particular focus of the role of direct democratic institutions in the context of German cooperative federalism.