Document Type

Working Paper

Publication Date



This Population Council working paper examines the causal structure of the relationship between child mortality events and subsequent fertility during a time of rapid decline in fertility in Bangladesh. Results lend support to the hypothesis of an insurance effect, while demonstrating that its demographic significance is likely to be less prominent than that of replacement behavior. Findings indicate that the insurance motive remains intact even after total fertility declined to fewer than three children per woman. The well-documented role of gender bias as a determinant of child health and survival is also a factor in child-replacement decisionmaking. Although the rapid fertility decline in Bangladesh transformed reproductive behavior in less than a generation, no evidence shows that the effect of son preference on replacement motives has changed as the fertility transition has progressed