A1819
Title: A Bayesian approach for producing subnational population estimates using a population base statistical register
Authors: Jairo Fuquene - UC Davis (United States) [presenting]
Abstract: Subnational Population Estimates (SPE) in Latin America are useful for implementing new public policies in subnational areas with internal armed conflicts or difficult access. A Bayesian approach combining a Population Base Statistical Register (PBSR) and Official Population Projections (OPP) is proposed to produce SPE. The proposed procedures are useful for computing SPE of Population size or SPE as a percentage SPE (\%). However, the focus is on SPE (\%) due to data restrictions and to ensure data confidentiality. The PBSR is constructed using multiple administrative sources with registers from health, education, vital statistics systems, tax registration, and victim records from the current internal armed conflict in Colombia. New fast Markov Chain Monte Carlo algorithms are proposed to produce SPE (\%) using data augmentation procedures to address complications caused by the resulting joint posterior containing gamma functions. The proposal is implemented to compute SPE (\%) by age and sex groups in the municipality of Jamund in Colombia, which is currently affected by poverty, forced displacement, and internal armed conflict, with accuracy evaluated using a Population Census.
