Abstract:
This paper analyzes the effect of being born to a teen mother on child health outcomes in South Africa using propensity score reweighting. Exploiting the longitudinal nature of the Cape Area Panel Study, we estimate the probability of being a teen mother conditional on pre-childbirth characteristics. We use this score to construct a weighted counterfactual group of children born to mothers over nineteen whose pre-childbirth characteristics are very similar to the teen mother sample except for their age at the birth of their first child. Our reweighted regressions indicate that being born to a teen mother has some significant adverse effects on child health, especially among Coloured children. In particular, children born to teens are more likely to be underweight at birth and to be stunted with the negative effect being double the size for Coloureds than Africans. No negative impact of teenage childbearing is found on head circumference at birth or the incidence of incomplete first year immunizations. These results remain robust even when we simulate influential unobservable effects in both the reweighting equation and the outcome equation.
