A Bayesian Hierarchical Model for Jointly Estimating Subnational Mortality for Multiple Populations

Abstract: Understanding patterns in mortality across subpopulations is essential for local health policy decision making. One of the key challenges of subnational mortality rate estimation is the presence of small populations and near zero death counts. When studying differences between subpopulations, this challenge is compounded as the small populations are further divided along socio-demographic lines. In this paper, we build on principal component based Bayesian hierarchical approaches for subnational mortality rate estimation to model correlations across subpopulations. The principal components identify structural differences between subpopulations, and new coefficient and error models track the correlations between subpopulations over time. We illustrate the use of the model on county-level sex-specific US mortality data and on simulated mortality data with 10 subpopulations. We find that initial results from the model are reasonable and that it successfully extracts meaningful patterns in subgroup-specific mortality.

This project was presented at PAA 2022, the HMD 20th Anniversary Symposium, and EPC 2022.

The final version of this work can be found here.