Abstract:
Data from a number of regions indicate that childhood deaths are unequally distributed across families. This has been identified, in previous research, with (observed and unobserved) heterogeneity between families. In this paper, we investigate whether, on top of these correlated risks, there is a causal process at work within families, whereby the death of a child elevates the risk of death of the succeeding sibling. Borrowing language from the unemployment literature, the causal process is termed state dependence or scarring. To the extent that scarring exists, a social multiplier comes into play, raising the payoff to policies that reduce infant mortality. Acknowledging scarring effects is also potentially relevant to understanding the relation of mortality and fertility behaviour within families. The analysis is conducted using data for the 15 major states of India. Large scarring effects are observed in 14 of the 15 states.