This paper investigates the causal effect of armed conflict on the educational achievement of Colombian high school graduates, proxied by their scores on a national standardized test. Given the possible endogeneity of conflict, we use as an instrument the exogenous occurrence of natural disasters in neighboring municipalities. Using this strategy and based on information at the student level, we find that an increase of one standard deviation in our conflict measure decreases scholastic achievement by 0.75 standard deviations. The results suggest that this effect is not driven by supply-side factors but rather by conflict affecting students and families directly. Moreover, the results are robust to different specifications and alternative identification strategies.

ACCESS PAPER