Projections Life Table AG2026

For the publications of Projections Life Table AG2022 and AG2024, it was decided to base the long-term trend on the European data up to and including 2019 before the appearance of COVID-19. The mortality that followed from this pre-covid projection was corrected with an excess mortality term that reflected excess mortality in the Netherlands. This excess mortality term was modelled separately with assumptions about the future course of excess mortality. A similar approach has been chosen in the model for Projections Life Table AG2026.

With regard to the European data underlying the pre-COVID projection in Projections Life Table AG2026, a data update for the years up to and including 2019 applies, because the source data (Human Mortality Database) have been updated retroactively. Furthermore, the pre-covid model specification has been adjusted on one point. In order to improve the reconciliation of the projection with the mortality observations in the years before 2020, it was decided to change the method used to determine the starting point of the pre-COVID projection. The prognosis now starts from mortality probabilities derived directly from observed mortality. This adjustment of the starting point will eventually converge to the precovid projection without this adjustment.

Since the last publication in September 2024, two additional years of new mortality data for the Dutch population have become available. These data show that excess mortality is still present, more so for females than for males, and that this excess mortality is higher than what was assumed two years ago. To estimate excess mortality in Projections Life Table AG2026 the years since 2022 were examined, similarly as in AG2024. This effectively means that the years 2024 and 2025 have been added to the data. In addition, in the Projections Life Table AG2026, the excess mortality terms are applied to all ages, whereas in AG2024 this was only the case for ages 55 and upwards. This adjustment does justice to the observed excess mortality at lower ages.

In AG2024, the excess mortality was assumed to decrease by about 25% per year. As a result of the current mortality developments, combined with an evaluation of various scenarios, the CSO has decided to adjust the future development of excess mortality. In contrast to previous publications, where it was assumed that excess mortality will eventually fade out completely, the Projections Life Table AG2026 assumes that 25% of excess mortality is permanent and also that it will take a longer time to reach this level by reducing the run-off factor of excess mortality from 25% to 5%.

These changes affect both mortality rates in the short term and the development of life expectancy in the medium to long term. Life expectancy is still rising, but less rapidly than expected in the AG2024 prognosis. If we compare the projection of AG2024 with that of AG2026, the cohort life expectancy decreases by about 4 months for someone born in 2027. This applies to both males and females. The cohort life expectancy at age 65 decreases by 3 months for males and 7 months for females. As a result of these new life projections, an average purchasing factor for pension and partner's pension around retirement age decreases by about 1.0% for males and 1.5% for females. The provision of an average pension portfolio decreases by about 0.9% for males and 1.4% for females.