A Space-Time Extension of the Lee-Carter Model in a Hierarchical Bayesian Framework: Modelling Provincial Mortality in Italy

The main purpose of this paper is to define a statistical method to model mortality rates by age and sex for provincial areas in Italy. In a preliminary descriptive analysis, we demonstrate the existence of specific spatio-temporal patterns. Thus, we propose a hierarchical Bayesian model that allows area-level estimates to borrow strength from each other by exploiting spatial association of provincial mortality rates and taking into account temporal correlation. As a result, it appears that model based estimates are less variable than direct estimates.