This paper studies the effect of the Boko Haram conflict on schooling in Northern
Nigeria. We combine individual level data from three rounds of the Nigeria General
Household Survey Panel (GHS-Panel) to detailed geo-referenced conflict data. Using
different econometric strategies, we estimate the effect of conflict on both school
attainment and attendance. We find that a one standard deviation increase in the
number of fatalities in the 20 km radius of each household decreases the number of
completed years of education for the cohort exposed to conflict during primary school
by 0.6 years, compared to the non-exposed cohort, which translates into a 10 percent
drop relative to the average educational achievement. At the same time, controlling
for individual heterogeneity, we find that a one standard deviation increase in conflict
intensity reduces the probability to be enrolled by 0.5 points. Our results are robust
to different econometric specifications, variable definitions, and alternative measures
of conflict intensity. Finally, we provide suggestive evidence of the conflict-induced
reduction in household wealth as a possible mechanism explaining the negative effect
of conflict on schooling.