We exploit the exogenous variations of arsenic contamination across regions in Cambodia and apply two-period difference-in-differences (DID) to examine the impact of arsenic-contaminated groundwater on school participation among children aged 5-17. We find that conditional on the DID assumption, a standard deviation increase in the arsenic reduced the probability of having ever been enrolled for children aged 5-17 by approximately 0.017. The effects for girls are slightly larger than those for boys. Our results are robust to adding parental education, province fixed effects and zone trends. We also show that children of all specific age groups experienced the reduction in their probabilities of having ever been enrolled, while the children, who were between 12 and 17 years old, currently enrolled and active in the labor market, experienced more reduction in school enrollment relative to their peers, who were not active in the labor market.
Arsenic Cambodia Difference-in-Differences School participation