University of Washington - Department of Statistics
Advisor: Adrian Raftery
Current methods for reconstructing human population structures of the past are deterministic or do not formally account for measurement error. A method for simultaneously estimating population counts, fertility rates, survival proportions and net migration is proposed that incorporates measurement error. Inference is based on joint posterior probability distributions which yield fully probabilistic interval estimates. It is designed for the kind of data commonly collected in modern demographic surveys and censuses. Population dynamics over the period of reconstruction are modeled by embedding formal demographic accounting relationships in a Bayesian hierarchical model. Informative priors are specified for vital rates, migration rates, population counts at baseline, and the accuracies of their respective measurements. Calibration of central posterior marginal probability intervals is investigated by simulation and the method is demonstrated by reconstructing the female population of Burkina Faso from 1960--2000. Future work will include applying the method to other countries and extending to two-sex populations.