From targeted private benefits to public goods: Land, distributive politics and changing political conditions in Colombia

Abstract

• There is a political land cycle (PLC) in Colombia: land allocations are higher in electoral years than in non-electoral ones. • The PLC is dependent on political conditions: its size is reduced by half after the implementation of a deep political reform. • The reduction was stronger in areas where increases in political competition and fiscal and administrative capacity were higher • These changes generate incentive and capacity effects affecting the costs and benefits of distributive politics resources. • This led to a re-composition of the distributive politics portfolio: from private targeted benefits towards public goods. This paper analyzes how changes in political conditions affect distributive politics. We study the case of Colombia, focusing on the strategic allocation of land in relation to the electoral cycle. Relying on over 55.000 municipality-year observations on land allocations, exogenous timing of elections and sociodemographic controls, we show that there is a political land cycle (PLC), and that this cycle is dependent on the local political conditions in place. We analyze the changes in the PLC derived from the implementation of a deep political reform that increased political competition and the fiscal and administrative capacity of the state, doing so heterogeneously across municipalities. We show that the PLC decreased by half after the reform, with this reduction being stronger in municipalities in which political competition and fiscal and administrative capacity increased the most. The heterogeneous reduction in the PLC does not appear to stem from an aggregate weakening of distributive politics, but rather, from a re-composition of the distributive politics portfolio: away from the allocation of private targeted benefits like land, and towards the strategic allocation of public goods. We discuss the incentive and capacity effects explaining this re-composition likely affecting the relative costs and benefits of different types of distributive politics resources. The results evidence the importance of understanding not only the territorial dimension of distributive politics, but also how the specific traits of resources determine distributive politics strategies and their resilience to contextual changes.