Hans Lugtigheid
December 28. Connie looks at the coffin with her grandmother. ‘Terrible that covid’, she thinks, ‘without it granny would have made it easily to the summer. And she loved spring so much.’ Her inner statistician muses on: ‘Without covid she would have died next year. Hmm, then we have to lower the expected mortality for next year with one.’ Connie startles herself. ‘Come on girl. This is not the time for statistics.’ she tells herself.
Indeed, a funeral is not a good moment for statistics. Connie makes a valid point though. I will show that in this article.
In the report ‘Kernprognose 2019-2026’ (KP19) from December 2019 the CBS (Dutch Central Bureau of Statistics) estimates the expected mortality for 2021 to be 155.232. This was pre-pandemic. These are usually reliable estimates by the CBS. The CBS adjusts these estimates yearly. Without grave disturbances these adjustments are marginal. I use this estimate as base.
In 2020 approximately 21.000 people died from covid. I estimate that 8.000 of these 21.000 would have lived for another year and would have died in 2021*. Hence for 2021 I lower the estimated mortality from KP19 for 2021 with 8.000 to 147.232. This then is the expected mortality with covid as the only disturbing factor.
In the report ‘Bevolkingsprognose 2020-2070’ (BP20) from December 2020 the CBS estimates the expected mortality for 2021 at 154.887. This is 345 less then the estimate of 155.232 from KP19 and 7.655 more than the expected mortality with covid as the only disturbing factor. According to the CBS the change from covid (here: 8.000) is included in this estimate. Hence the estimate from the CBS contains a second disturbing factor of the size 7.655. From this we get table 1.
It is the second disturbance of 7.655 in BP20 that gives an overestimation of the expected mortality by the CBS. Excess mortality is calculated as real mortality minus expected mortality. The estimate from the CBS then underestimates the excess mortality with this 7.655. Mortality in 2021 was 170.972 and this gives table 2.
According to the CBS without the pandemic the estimate of the expected mortality in BP20 would increase accordingly. This would be equal to 154.887 + 8.000 = 162.887. Compared to the estimate of 155.232 from KP19 this is a remarkable increase. The CBS has not (yet) explained this increase**.
* : In 2020 approximately 21.000 people died from covid. These included many very old people who would have died in 2021 without covid. I estimate this group to be 8.000. With another estimate, say 6.000, the reasoning remains the same. With so many older people who have died an estimate of 1.000 or 2.000 is not realistic. Hence the estimate, although not precise, has to be significant. Therefore it is relevant.
** : With determining the expected mortality as of January 1st of a year the CBS bases its estimate on the population as of January 1st of that year. According to the CBS everything from the past is included in their estimate and the differences are mainly due to immigration and changed population. This is to general. According to the CBS the estimate of 8.000 for 2021, or at least the underlying effect, is included in their new adjustment of the estimate by 345 for 2021. This gives a difference of 7.655 between the two estimates. If this is explained by, for example, immigration then the question is what the CBS sees as the cause for the extra mortality. Say the extra immigration is 20.000, then which 7.665, or part thereof, of these will die in the same year of immigration? Are all immigrants older than 87 so that a third of them will die within the year? Or does the CBS expects the new young immigrant to die from a new disease in the Netherlands from which they have no resistance? And so on. It is not enough to say that the higher than expected mortality is explained by higher immigration and changes in population. A further explanation for the increase in the expected mortality as the CBS calculates is needed.
Note: For more on this subject and an explanation of the used estimates see:
https://www.hanslugtigheid.nl/excess-mortality-is-strongly-underestimated/
And for the years 2020-2030 see:
https://www.hanslugtigheid.nl/pandemic-adjusted-excess-mortality-paem/
This article is licensed under
Creative Commons Attribution 4.0 International