Does Peace Build Human Capital? Evidence From Colombia's Peace Agreement

ABSTRACT Does the cessation of armed conflict improve education, and through what pathway? We study Colombia's 2016 peace agreement with the FARC (Revolutionary Armed Forces of Colombia), which sharply and differentially reduced violence in municipalities with a historically high guerrilla presence. We develop a framework distinguishing two margins of human capital—educational participation (whether children are in school) and learning (how much they learn)—and argue that peace acts on participation quickly and broadly, but on learning only where the security improvement is deep enough to restore the conditions for schooling (teachers, infrastructure, and safety). Testing this with a difference‐in‐differences design across 1122 municipalities, combining georeferenced conflict data with standardized test scores (Saber 11) and enrollment records, we find three things. First, participation rose: net enrollment in high‐conflict municipalities increased from 2018 onward, after a transitional dip in 2017. Second, the average effect on learning is a precise null that conceals strong regional heterogeneity: in the remote, formerly FARC‐dominated Amazon, test scores rose significantly, while the urbanized Andean region shows no gain. Third, participation and learning gains move together across regions ( r = 0.86), and the average learning null is not an artifact of a changing exam pool. Peace builds human capital, but the learning dividend is conditional on rebuilding the local capacity to teach. We draw out implications for mediators, donors, and post‐conflict governments.