Decomposing Mortality Improvement Rates into Cause-Specific Contributions

The aim of our research is to explain changes in all-cause mortality by identifying the contributions of changes in the mortality from specific causes. We measure changes in mortality using the commonly applied improvement rate and present a simple and intuitive method to identify cause-specific contributions. In that way, we can attribute observed mortality improvements to specific causes of death.

The contribution of individual causes to changes in all-cause mortality and life expectancy has been studied by others. For example, Yiu et al. (2022), consider a collection of log-linear models to identify changes in mortality and life expectancy caused by trend changes for specific causes, and Villegas et al. (2023) apply a period–cohort improvement model to identify key drivers affecting developments in US mortality.

In this paper, we present an approach that allows us to decompose all-cause improvements at a specific age (or age group) that does not rely on a specific mortality model and ensures that cause-specific contributions add up to all-cause improvement rates. Empirical results, both for pre-pandemic and Covid-19 years are presented, and we discuss the consequences of this decomposition for log-linear mortality models.