¿Son los cuidados prenatales efectivos? Un enfoque con datos individuales de panel

This paper analyzes the impact of prenatal care on low birth weight and pre-term births using panel data on births that took place between 1995 and 2008 in the largest public university hospital in Uruguay (Perinatal Information System, PAHO). The hospital provides free access to prenatal and obstetric care and serves 15% of deliveries in the country. The use of difference-GMM estimation addresses potential biases due to time invariant unobserved heterogeneity and feedback effects from prior pregnancies to the current demand of prenatal inputs. Our results show that if pregnant women initiated prenatal care in the first trimester and had at least 6 controls (which is the goal of the Uruguayan National Health Insurance System) the likelihood of low birth weight would decrease by 3 percentage points, a 30% decrease. Our estimates underscore the importance of controlling for unobserved heterogeneity and feedback effects. They are also larger than others exploiting health policy changes in 2SLS settings. In this sense, our results are in line with the critique that local average treatment effects identified by 2SLS may fail to consider the bimodality of the pregnancy distribution and underestimate the effects of interest.