The Mean Age at Death (MAD): An alternative to life expectancy?

Markus Sauerberg , Federal Institute for Population Research
Marc Luy , Wittgenstein Centre (IIASA, OeAW, Univ. Vienna)

In this paper, we compare and discuss the standardized mean age at death (MAD) with the period life expectancy (LE) indicator. Since the observed MAD is affected by the population's age structure it is not an adequate metric for comparing mortality levels between populations. Both, the standardized MAD and period LE are free of the population's age structure and thus, appropriate for comparing mortality levels across population and over time. The standardized MAD, however, might be conceptually closer to the real population because it standardizes by controlling for fluctuations in births. Period LE, on the other hand, relies on the synthetic cohort approach. We first discuss and compare both measures formally. In the empirical part of the paper, we provide estimates for the observed MAD, standardized MAD, and period LE in Switzerland, Sweden, Denmark, and Portugal from 1990 to 2020. We find that the observed MAD and standardized MAD are currently very similar to each other, while period LE suggests higher values for the average length of life. Yet, period LE is the only measure reacting to the sudden increase in death rates observed in 2020. Thus, our preliminary results suggest that period LE is indispensable for examining period shocks in age-specific mortality rates. Still, the standardized MAD can be a valuable alternative to period LE for measuring the average length of life in the real population.

See extended abstract

 Presented in Session P1. Postercafe