Expected Child Mortality, Fertility Decisions, and the Demographic Dividend in Low and Middle Income Countries

High child mortality creates replacement and insurance motives that lead to high fertility, and in high fertility settings reductions in child mortality have been followed by substantial reductions in fertility. While... [ view full abstract ]

High child mortality creates replacement and insurance motives that lead to high fertility, and in high fertility settings reductions in child mortality have been followed by substantial reductions in fertility. While fertility declines when child mortality falls, the effect on the number of surviving children is unclear; the number of surviving children only declines if a reduction in child mortality leads to a greater than one-for-one reduction in fertility. This is important because if fertility falls sufficiently to lower the net reproduction rate, family and cohort size will be smaller, allowing higher investments in children, and freeing women from child care to enter the labor market, setting off the possibility of a demographic dividend. Estimating this relationship empirically is complicated by the fact that fertility depends on expected future child mortality as well as experienced mortality, and there is reverse causality between fertility and mortality. We construct a model of expected child mortality and estimate the relationship between expected mortality and fertility using a pooled sample of almost 2 million women in developing countries using nationally representative household surveys over the period 1990 to 2010. Using the mix of vaccine coverage in a cluster as an instrumental variable for death rates, we find that each additional expected child death raises fertility by 1.86 children, implying that reductions in mortality reduce the overall number of surviving children.