Civil Conflict and Long Term Human Capital Accumulation: Evidence from the Genocide in Rwanda

Abstract

This paper provides empirical evidence of the long-term effects of exposure to civil conflict on human capital accumulation. I use the extreme conflict shock from the 1994 genocide in Rwanda as a natural experiment, employing census and geocoded conflict data to exploit variation in early life exposure to violence by geography and birth cohort to identify the precise effect of a civil war on educational attainment and literacy in the long term. Conditional on being exposed to violence, the average individual accumulates 5% less schooling and 3.8% lower literacy, signifying that both the quantity and quality of the schooling is adversely affected by conflict. Children who were in utero or of primary school age at the time of the conflict saw the most drastic impact. Male children were affected more significantly than female children. These effects persist through the completion of the schooling cycle, indicating that individuals exposed to conflict are unable to catch up to their peers during peaceful times.