In this paper I draw on the study by Saideman et al. (2002) which has contrasted the effect of federalism with respect to the degree of groups spatial concentration. The study found that the intensity of rebellion decreased with federalism, but this was true only for spatially concentrated groups.
I undertake a similar approach for fiscal decentralization and consider that beside groups spatial concentration, the local majority status of groups is crucial to understand why some groups increase their level of violence and why some others decrease it when decentralization goes up. In a nutshell the rationale is as follows.