We use geographical variation in the intensity of the genocide, which disproportionately
killed prime-age males during the Khmer Rouge (KR) regime in Cambodia, to study the
effect of violent conflict on the educational and health outcomes of children born years after
the conflict ended. We show that the adverse effects of violent conflict are transmitted from
one generation to the next through its effect on the sex ratio and marriage outcomes of those
who survived the conflict. We find that mortality rates under the KR predict a lower
likelihood of normal grade progression and lower height-for-age Z-scores for children born to
parents who were of prime marriage age (14–29) during the time that the KR was in power.
Using mortality rates during the KR regime as an instrumental variable for the sex ratio, we
find that for every one standard deviation decrease in the sex ratio for the parents’ generation,
the likelihood of children exhibiting normal grade progression reduces by 6.8–7.4 percentage
points and the height-for-age Z-score decreases by 1.5 standard deviations