Population aging and public health expenditure: a dynamic macro-panel decomposition of mortality-linked and morbidity channels

This paper examines whether the conditional association between population age structure and public health expenditure operates more strongly through a mortality-linked pathway or a morbidity-duration pathway. Using macro-panel data for 100 high- and upper-middle-income countries over 1995–2024, the study estimates a dynamic pathway decomposition within a CS-ARDL framework using the Dynamic Common Correlated Effects mean-group estimator. The mortality-linked pathway is represented by the mortality rate among adults aged 65 and older, while morbidity duration is measured by the gap between life expectancy and healthy life expectancy at age 60. These aggregate measures support an associational, not causal, decomposition. The mortality-linked pathway component is positive and statistically distinguishable from zero in both the short run and the long run. By contrast, the morbidity-duration pathway component is small and statistically imprecise in the measured series. A remaining conditional population-age-structure–expenditure association is also positive in both horizons, statistically distinguishable from zero in the long run but not in the short run, indicating that the two specified pathways do not account for the complete association. The results provide dynamic associational evidence consistent with a mortality-linked expenditure channel, not an identified individual time-to-death effect. Projection models based exclusively on age composition may provide an incomplete account of expenditure processes, but this study does not estimate forecasting bias. In the short term, the mortality-linked component predominates within high-income nations, whereas the remaining conditional association prevails long term among upper-middle-income countries; imprecise morbidity-duration estimates do not imply structural irrelevance.