African Census Analysis Project (ACAP)

The study uses the data from the Kenya 1989 and Uganda 1991 population censuses to examine regional differences in infant and child mortality. Using the Brass indirect methods of estimation, the findings suggest that under-five mortality rates were substantially higher in Uganda compared to Kenya. However, the rates in Kenya were found to be increasing in the most recent period, since the mid-1980s. Using the number of children dead for each woman as the dependent variable, generalized Poisson regression equations are used to explore regional child mortality differences. Results suggest that regional differences in under-five mortality are quite substantial even after controlling for other variables in the model. Among other findings, this study reinforces the fact that education of the mother relates inversely to under-five mortality.