Bayesian Probabilistic Projections of Mortality

Jennifer L. Chunn, University of Washington
Adrian Raftery, University of Washington

Every two years, the United Nations Population Division (UN) publishes projections for populations of over 200 countries through the next 50 years. However, UN projections do not report uncertainty in mortality projections. We present a probabilistic projection model of life expectancy at birth. This is done by a random walk with drift model where the drift is determined by a Bayesian hierarchical model (BHM). We assess the performance of our model by fitting it to UN estimates from 1950 through 1995 and projecting life expectancy for males from 1995 to 2005 for 158 countries without a generalized HIV/AIDS epidemic. Our country-specific predictive distributions were well calibrated with 80% Prediction Intervals capturing the “truth” 81% of the time.