Current global health policies emphasize institutional deliveries as a pathway to achieving reductions in newborn mortality in developing countries. There is however remarkably little evidence regarding a causal relationship between institutional deliveries and newborn mortality. In this paper we take advantage of a shock to institutional deliveries provided by the randomized rollout of a government performance-based financing program in Rwanda, to provide the first estimates of this causal effect. We construct an instrumental variables estimator that combines interrupted time-series and difference-in-differences approaches. We do not find any statistically significant effect of an institutional birth on either 7- or 30-day mortality in any of our specifications. We conclude that attempting to increase institutional deliveries without addressing supply-side constraints is unlikely to result in the large reductions in mortality that policy makers expect.
- institutional births,
- infant mortality,
- health production
Available at: http://works.bepress.com/edward_okeke/8/