Reconstructing Past Populations With Uncertainty From Fragmentary\nData

Abstract

Current methods for reconstructing human populations of the past by age and sex\nare deterministic or do not formally account for measurement error. We propose a\nmethod for simultaneously estimating age-specific population counts, fertility\nrates, mortality rates, and net international migration flows from fragmentary\ndata that incorporates measurement error. Inference is based on joint posterior\nprobability distributions that yield fully probabilistic interval estimates. It\nis designed for the kind of data commonly collected in modern demographic\nsurveys and censuses. Population dynamics over the period of reconstruction are\nmodeled by embedding formal demographic accounting relationships in a Bayesian\nhierarchical model. Informative priors are specified for vital rates, migration\nrates, population counts at baseline, and their respective measurement error\nvariances. We investigate calibration of central posterior marginal probability\nintervals by simulation and demonstrate the method by reconstructing the female\npopulation of Burkina Faso from 1960 to 2005. Supplementary materials for this\narticle are available online and the method is implemented in the R package\n“popReconstruct.”

Keywords

Affiliated Institutions

• Agincourt Health and Socio-Demographic Surveillance System ZA
• University of the Witwatersrand ZA
• University of Washington US
• United Nations Department of Economic and Social Affairs US
• University of Colorado Boulder US

Related Publications