Estimating age-sex-specific adult mortality in the World Population Prospects: A Bayesian modelling approach

Abstract

As part of its efforts to revise population estimates and projections for the biennial World Population Prospects (WPP), the United Nations Population Division generates estimates of age- and sex-specific adult mortality for all countries and regions globally, spanning from 1950 to the present. These estimates draw upon diverse data sources, including civil registration and vital statistics systems, sample registration systems, surveys, national estimates, and population censuses, applying standard demographic techniques. Biases and inconsistencies inherent in the available data are rigorously evaluated and addressed to produce a consistent annual time series of age- and sex-specific mortality estimates. This technical paper presents the Bayesian hierarchical model (BHM) developed by the Population Division to estimate trends and levels of adult mortality by age group and sex for all countries and regions since 1950. The model leverages an extensive database maintained by the Population Division, incorporating data from diverse sources and allowing for the sharing of information across countries and time periods. This approach addresses challenges related to sparse, biased, or unavailable data.

Fengqing Chao

Assistant Professor of Computational Social Science

My research interests include statistical demography, global health, Bayesian modeling, and time series analysis.