Integrating Subnational Peripheries: State Building and Violent Actors in Colombia

Modern states are selective and calculating in their decision to establish their rule throughout the territory. Within the same country, state officials may be willing and capable of making major efforts to deliver public goods and provide security to the population in some areas, while choosing to neglect others and delegate their control to different actors. This dissertation studies why and how political elites differentially build state capacity in peripheral and marginalized areas of a country and in the midst of violent conflicts. What drives incumbents to increase state capacity in peripheral and marginalized areas within a country? What type of state expansion is likely to take place in such areas?

My central argument is that exogenous shocks that suddenly increase the political and economic value of peripheral areas may prompt incumbents to invest in state capacity, but whether this capacity increases or not depends on the preexisting configuration of violent actors and local rural elites. On the one hand, the type of violent actor---threatening and non-threatening---exerts a differential effect on state capacity by (i) shaping subnational politicians' incentives to cooperate with the central state and (ii) establishing collusive agreements with violent actors. In addition, because greater state capacity tends to undermine rural elites' economic and political power, they will have incentives to oppose to greater state presence and---in contexts of insecurity and violence---establish alliances with non-threatening groups. In short, state capacity is likely to be higher in municipalities with weak rural elites and facing threatening violent groups. In contrast, state capacity will be more difficult to attain in areas with stronger rural elites and where violent groups do not pose a major threat to state authority.

This dissertation studies the case of contemporary Colombia and, using three empirical strategies, leverages municipal-level data, survey evidence, and in-depth interviews of bureaucrats and politicians. First, I exploit the international price of oil palm, a crop mainly produced in rural peripheries, to understand state expansion in these areas. I find that the oil palm boom had a positive impact on state capacity in areas where rebel groups had been a predominant actor, where land concentration was lower, and were mayors were more financially dependent from the center. Second, I exploit mayoral close elections to understand how collusion between non-threatening groups and mayors hinder state authority. I find that the election of `paramilitary-friendly’ parties in 2007 had a negative effect on property taxes and positive effect on homicide rates in subsequent years. Lastly, I conduct a case study of the peripheral Macarena region and the state’s attempt to increase its presence there. I show that the state’s approach to consolidate its presence---mostly driven by insurgent threats---produced important transformations by promoting cooperation between central authorities and mayors, but it failed to incentivize local authorities to develop capacity on their own and to strengthen tax morale among the population.

While my dissertation focuses on Colombia---a case of democracy and violent conflict---it provides a more general framework to understand the differential impact of violent groups on the state, the relationship between democracy and the state in violent contexts, and the important role of subnational politicians in facilitating or hindering state consolidation.