The Conflict in Colombia: A Political Economic Analysis.

Abstract

This research explores the armed conflict in Colombia. It consists of two parts: Part I attempts to provide a theoretical framework for the analysis of the Colombian conflict using a dynamic model. The second part, co-authored with Manuel Pulido Velásquez, uses both difference-in-differences and the synthetic control methods to study the effect of conflict intensity on the country's unemployment rate. The key element connecting both parts is the level of activities run by the agents involved in this conflict. In the theoretical model, the level of each agent's activity represents the control variables determining the state variables and payoff functions. In the empirical model, identification of years with the highest levels of violent activities, helps us divide our period of analysis into pretreatment and treatment periods. Additionally, this identification provides a basis for defining potential control units in constructing our difference-in-differences and synthetic counterfactual. Through the dynamic model we find that an interior solution implies that the sum of the marginal costs imposed by the agents through the activities they conduct on other agents' payoff functions is equal to the sum of marginal benefits generated by those activities. In the empirical model we find that for the difference-in-indifferences approach unemployment rate in Colombia increases 3.7 percentage points relative to the control group. Meanwhile, for the same variable, our synthetic control shows an increase of 4.9 percentage points relative to the synthetic Colombia. Both estimations represent at least one third of the average unemployment rate for the treatment period (1995-2014).