Abstract
Female education is a crucial input to women's agency and empowerment, and has wide-ranging impacts, from improved labor market outcomes to reducing child mortality. Existing gender-specific evidence on the effect of armed conflict on education is conflict-specific and mixed. We link granular data on conflict events to georeferenced survey data on educational attainment from 28 countries in Africa, and use a regression-based approach to estimate the local effect of conflict exposure on female years of schooling. We find that conflict events occurring within 25 kilometers during a female child's primary school years reduces years of schooling by 0.4 years by adolescence. We do not find the same effect for males. Exposure to only low intensity conflict events with at most two casualties has persistent negative and significant effects. Consecutive years of conflict, however, can have positive effects in later years, which offset earlier negative effects, suggesting a habituation to violence. In the past two decades, we estimate excess child mortality in Africa associated with the indirect channel of women's education to be similar in magnitude to the number of direct child casualties due to conflict.