Birth-Spacing, Infant Mortality, and the Use of Contraceptives: Evidence from Longitudinal Data

Unnati Rani Saha, ICDDR,B and Tilburg University
Arthur van Soest, Tilburg University

Using longitudinal data from rural Bangladesh, this study investigates the effects of contraceptive use after a birth on birth intervals and fertility, as well as the effects of infant death and other factors (such as socio-economic status or gender composition of the household) on subsequent contraceptive use. Our analysis is based upon a model with three parts: an equation explaining infant mortality, a model part explaining whether contraceptives are used after a child is born (and if so, for how long), and an equation explaining birth intervals. Infant mortality is determined by covariates reflecting socio-economic status, etc., but also by the length of the preceding birth interval. The decisions about contraceptives are driven by similar covariates, but also by survival status of the previous child and the family’s gender composition. Birth spacing is driven by contraceptive use and other factors. Each part of the model incorporates unobserved mother specific heterogeneity.