Occupation-based life expectancy: Towards actuarial fairness of determining statutory retirement age

Actuarial fairness is considered a key aspect of a just pension system and can be understood as “equal treatment for equal risks”. For a pension system to be actuarially fair, the individual’s contributions (plus interests) should equal the individual’s expected benefits. However, life expectancy is strongly socially stratified, the “risk” of longevity is not distributed equally in society and differs across occupational groups. Therefore, equal treatment through a fixed retirement age for every occupational group cannot be actuarially fair. This study aims to develop a fairer method to determine retirement ages by quantifying occupation‐based life expectancies.
We examine quantitative differences in survival time between six sectors selected for their differences in educational and gender composition. Next, we investigate to what extent survival differences can be accounted for by five health variables. We address these research question for the Netherlands, using the 1992‐93 baseline cycle of the Longitudinal Aging Study Amsterdam, ages 55‐85 years, linked to vital status follow‐up until December 31, 2016.
Results show that participants having worked in technical and transport occupations have an approximately 1.5‐2‐year shorter life expectancy than the median of the general population, accounting for compositional characteristics (age, gender). The number of days sick in bed due to health problems accounted for a portion of this difference in survival.
These preliminary results provide a proof of principle regarding the extent of quantitative differences in survival time between occupational groups. When replicated, we can achieve a basis for the calculation of more actuarially fair pensions.