Long-term impact of mortality on population age structures

Abstract

In this study, we tested the hypothesis that mortality has a greater influence than fertility on shaping population age structures in the long run and that recent mortality rates provide a satisfactory initial approximation for describing observed age structures in most empirical cases. In the theoretical part of this article, we elucidate a potential fallacy in the line of reasoning based on simulations and counterfactuals frequently used to attribute population aging to low fertility rates. The alternative view that we propose leads us to hypothesize that age structures conform, albeit not exclusively, to a standard derived from survival conditions: the age structure of a stationary population within a given period. We tested this hypothesis on all countries, using the data from the United Nations database (1951 – 2021) and specifically on 10 European countries using the data from the Human Mortality Database (1860 – 2019). The empirical results indicate that current survival conditions sufficiently explain a significant portion of the observed age structure across all examined countries and epochs. However, deviations from this underlying, long-term (mortality-driven) path exist, which our approach cannot fully explain. This is where the role of fertility arguably becomes more prominent. Several implications arise from our findings, including the debate on the relative role of fertility and mortality in shaping age structures in the long run, the theoretical meaning and practical use of cross-sectional life tables, and the notion and measure of demographic dividends.