The effects of conflict in education outcomes: Evidence from Colombia

Conflicts have a negative effect on human capital accumulation and hinder the growth potential of affected countries. Widespread violence from a conflict can disrupt education through a series of channels. First, households modify their behavior in response to the risk of violence; second, students must interrupt schooling when schools are damaged or destroyed; third, people might fear to visit public places, such as schools. The Colombian conflict has opposed the military against rebellious groups for over sixty years. However, its intensity peaked in the final two decades of the 20th century as other non-state armed actors began disputing strategic regions in rural zones of the country.

The purpose of this paper is to estimate the effects of violence on human capital accumulation. To do this I estimate how educational outcomes of Colombian children are affected by violent shocks. I use a rich dataset from the Longitudinal Survey of Universidad de los Andes (ELCA by its Spanish acronym). This survey has tracked 10,000 rural and urban households and contains a specific module aimed at children. Two different waves (2010 and 2013) are available and the attrition rate is low (6%). I construct two measures of violence exposure: one is the presence of rebellious groups and the intensity of their actions from 2000 to 2012. The other is the number of massacres and its casualties by municipality in Colombia since 1993.

The nature of the database allows me to control for unobserved individual effects in the model using panel data methods. Early results show that the presence of rebellious groups and the intensity of the conflict result in fewer years of education and lower net enrollment ratios. This impact is even stronger for children in rural areas and from smaller towns.