Access to and utilization of health services are concerns in poor countries. If implemented correctly,
health insurance may help solve this challenge. This paper examines how participation in the
National Health Insurance Scheme (NHIS) affects health care utilization in Ghana, using
representative data for the whole country and an instrumental variable approach. This methodology
overcomes selection and omitted variable bias. The exogenous variation in enrollment comes from
unexpected membership rules variations in the District Mutual Health Insurance Schemes
(DMHIS). Using the 2008 DHS and a census of all DMHIS, we find insurance membership
increases the probability of: i) seeking higher quality but no greater quantity of maternal services and
ii) parents becoming more active users of child curative care. IV estimates are larger than OLS ones
indicating adverse selection and underlying heterogeneity in returns to NHIS participation. The
instruments and results are robust to several validity checks.