Fighting Fire with Aid: Development Assistance as Counterinsurency Tool - Evidence for Colombia

Abstract

I study the causal effect of the foreign aid for development assistance —implemented by the U.S. Agency for International Development (USAID)— on the intensity of municipality-level armed conflict in Colombia, for the period 2009-2013. To address potential endogeneity biases, I use a Bartik-like instrument which exploits the spatial persistence of aid from USAID in Colombia. Specifically, I instrument foreign aid with the interaction between the United States GDP and municipality-level intent-to-treat indicators for the Malaria Eradication Campaigns (circa 1957). The results point out that foreign aid reduces the insurgency associated with left-wing guerrillas, especially FARC. However, foreign aid does not affect the violence associated with criminal gangs from right-wing paramilitary origins (BACRIM). I provide both quantitative and anecdotic evidence on two potential mechanisms behind my results: (i) Development programs raise the opportunity cost of fighting and; (ii) foreign aid improves the trust, and the information flows between civilians and the government. Finally, I provide empirical evidence that casts doubt on two alternative channels whose predictions cannot be reconciled with the results: (i) that foreign aid increases the potentially- looted rents by the insurgents and; (ii) that development programs rise targeted assassinations committed by insurgents to sabotage and reestablish bargaining power.