Abstract:
The academic literature on the impact of decentralization on conflict is not yet fully settled. Some scholars find decentralization reforms contributing towards conflict mitigation while others see them leading to conflict intensification. This has been a major constraint in designing conflict-sensitive decentralization reforms and in measuring their impact. This essay argues that decentralization's impact on conflict is a function of how such reforms affect the relationship between national-level elites and subnational-level (group) elites. By definition, decentralization distributes power spatially. Therefore, such reforms invariably impact the power-parity between inter-level elites. It could reinforce the power of elites at one level vis-à-vis that of elites at another level. Affected elites may then try to leverage the influence they enjoy over their constituents (group members) to respond to the shifting power balance. This would either promote cohesion at the national-level or encourage centrifugal forces. This will have direct implications for subnational conflict. This is a generalizable framework that can be applied to most contexts to interpret the relationship between decentralization and conflict. This essay goes onto formalize this relationship using game theory. Specifically, the essay uses a simple two-level game to capture the interaction between inter-level elites (per Daniel Treisman) while using the intuition of the selectorate theory (per Bueno de Mesquita et al.) to capture the ensuing intra-group dynamics.
