Gender differences in infant and child mortality: Estimation and identification of countries with outlying levels or trends

Under natural circumstances, the ratio of the male to female under-five mortality rate is greater than one. However, deprivation of girls’ access to health care or proper nutrition could lead to distorted ratios of under-five mortality. Monitoring of mortality by sex is challenging because of issues with data availability and quality. Moreover, the sex ratio is expected to vary with under-five mortality, which makes it challenging to define “expected levels”. We present a Bayesian model to estimate the sex ratio of under-five mortality for all countries. In addition, we estimate the relative difference between national sex ratios and expected sex ratios based on the global relation between mortality and sex ratios. All estimates include an uncertainty assessment to enable assessments of whether differences between countries or within countries over time are significant or highly uncertain.