Aligning with post-functionalist approaches to the territorial dispersion of authority, we argue that the why and when of these reforms can only be explained by an actor-based approach. We embed territorial reforms in multidimensional party competition and demonstrate that medium term office seeking calculations explain approximately 90 percent of symmetric shifts of authority between nation-state and regions. We test our argument with a newly compiled dataset of election results and party positions on the regional level in 22 countries. Diverging from existing approaches, we use national election data and aggregate party support from the district level to regions. This allows us to include cases where regions have not yet evolved. That way we can extend our explanatory model to the first territorial reforms and the initial endowment of authority for regions in the early state of democracies. We conclude by discussing characteristics of counterfactual cases, where reforms are theoretically expected but fail to appear.
