Measuring Inequality in Early-Life Mortality Within and Between-Groups Over Time: A Bayesian Approach with Application to India
Antonio Pedro Ramos. PhD
Department of Biostatistics,
Fielding School of Public Health, UCLA
California Center for Population Research, UCLA
Most studies on early-life mortality compare mortality rates between large groups of births, such as across countries, income groups, ethnicities, or times. These studies do not measure within-group disparities. The few studies that have looked at mortality across the entire population of births, however, have used tools from the income inequality literature. Using a large data set from India, we estimate mortality risk for over 400,000 births using a Bayesian hierarchical model. We show that while measures based on the income inequality literature, such as Gini indices, are not appropriate for mortality risk, most of the variance in mortality risk occur within-groups of births. We developed a novel approach to investigate inequality in mortality risk. Our approach uncovers several important patterns in the dynamics of inequality in infant mortality and has broader applicability to other health outcomes.