Infant Mortality in Rural Bangladesh: State Dependence vs. Unobserved Heterogeneity

Unnati Rani Saha, Tilburg University
Arthur van Soest, Tilburg University

Using data from the Health and Demographic Surveillance System (HDSS) in Matlab, Bangladesh, this paper investigates the dynamic of siblings’ death at infancy. Probit models are estimated incorporating survival status of the previous child of the same mother and unobserved mother-specific heterogeneity. The likelihood of infant death is about 32% more if the older sibling died at infancy; estimates suggest that, in the absence of this “scarring” effect, the infant mortality rate would fall by 7.13% among the second and higher order births in the comparison area. There is no evidence of scarring in the ICDDR,B area, perhaps because learning effects play a larger role with the available extensive health interventions. Reducing average distance to the health clinic in the comparison area can lead to a reduction in infant mortality of 17/1,000 live births of the total gap 25/1,000 live births among firstborns between two areas.