Let the rebels rule? Evidence on the economic effects of rebel governance in Colombia

I study the impact of rebel governance on economic development in rural Colombia. In 1998 the Colombian government created a 42,000 square km demilitarized zone (DMZ) to negotiate with FARC, Colombia's largest and oldest rebel group. Using a spatial regression discontinuity design, I exploit the DMZ's border defined by municipalities' pre-existing administrative boundaries to examine the causal effects of rebel based social order on education, living conditions, and agricultural production. I show that rebel governance increased the years of education by 0.1 standard deviations, access to aqueduct systems by 11 percentage points, and agricultural yield by 16 percent. These findings appear to be driven by public goods provision and less exposure to violence during rebels rule. However, I find that the positive gains from rebel governance did not translate into better living standards.