A Bayesian Model for Estimating Population Size and Demographic Parameters with Uncertainty

Mark C. Wheldon, University of Washington
Adrian Raftery, University of Washington
Patrick Gerland, United Nations Population Division

High quality estimates of vital rates and population counts are essential for successful policy development. These should include probabilistic statements about measurement uncertainty. In the developing world, where data collection is often infrequent and/or limited, this is particularly challenging. A Bayesian model is proposed which combines information on fertility and mortality rates of the past, and baseline population counts, with independently measured population counts to produce joint posterior estimates of age-sex-specific vital rates and counts. The model is tested on synthetic data for which the true underlying vital rates and counts are known as well as real data from Burkina Faso. Eighty percent posterior predictive intervals for all parameters capture most of true values in the simulation study while results from the real data suggest that the method could be used in practice.