It's not surprising that deaths should drop in 2023. A cull in 2022 leaves the strong and the lucky behind. But the fact that 2023 is still historically high suggests an underlying cause of death that has not previously existed. And these high death rates occurred during the summer semester, when death rates are typically low. I'm describing the pattern in NZ. Note also, that death rates had been declining steadily over the past 30 years, until 2021 when the bounced back from the isolationist lows of 2020, even though the country was still isolated.
Thank you Raphael for pointing out that the provisional mortality statistics have gently moved us into the land of the 'new normal'. The ABS says the Provisional Mortality Statistics are not going to compare against a baseline in 2024 so they are now useless as a measure of excess deaths. Their only use will be in giving us actual deaths figures every two months. Since July 2023, the ABS has published new 'official' excess mortality statistics, released 6 monthly. https://www.abs.gov.au/articles/measuring-australias-excess-mortality-during-covid-19-pandemic-until-august-2023
This new model has halved the numbers of excess deaths over the pandemic admitted to by the ABS, but it's still a heck of a lot of deaths! The history of how the ABS has reported excess deaths over the pandemic is here: https://clarityonhealth.substack.com/p/2022-excess-deaths-think-66-plane
'Modelling was used to compare the number of people who died with or from COVID between August 2021 and July 2022 (which captures the Delta wave and the onset of Omicron) against a hypothetical scenario in which vaccines never made it to Australian shores.'
'...............The PLOS One study also found that unvaccinated people had 7.7 times the death rate of people who were double vaccinated and 11.2 times the death rate of those who got a booster shot.'
What do mathematician readers think of the conclusions of this paper?
The point is that the "expected mortality" curve is a man made estimate.
While you can pick an impossible curve, within certain limits all the estimates are neither right nor wrong.
Eliminating outliers is standard practice, but it's up to the researcher to decide whether doing that and what constitutes an outlier.
If the mortality data are correct - and it's hard to claim that the dead count in the developed countries isn't substantially correct - anyone can make one's own "expected mortality" curve: there are several methods. And from there, one can make his/her own analysis and draw his/her own conclusions.
This isn't to say that the Australian conclusions are correct. It's to say that unless they made some really weird estimate, you can't say that the estimate is wrong.
But you can say that it's an estimate among hundreds.
It's not surprising that deaths should drop in 2023. A cull in 2022 leaves the strong and the lucky behind. But the fact that 2023 is still historically high suggests an underlying cause of death that has not previously existed. And these high death rates occurred during the summer semester, when death rates are typically low. I'm describing the pattern in NZ. Note also, that death rates had been declining steadily over the past 30 years, until 2021 when the bounced back from the isolationist lows of 2020, even though the country was still isolated.
Yes! Thank you for stating the government statistics problem so clearly and concisely! Keep up your great work, Dr Lataster!
Thank you Raphael for pointing out that the provisional mortality statistics have gently moved us into the land of the 'new normal'. The ABS says the Provisional Mortality Statistics are not going to compare against a baseline in 2024 so they are now useless as a measure of excess deaths. Their only use will be in giving us actual deaths figures every two months. Since July 2023, the ABS has published new 'official' excess mortality statistics, released 6 monthly. https://www.abs.gov.au/articles/measuring-australias-excess-mortality-during-covid-19-pandemic-until-august-2023
This new model has halved the numbers of excess deaths over the pandemic admitted to by the ABS, but it's still a heck of a lot of deaths! The history of how the ABS has reported excess deaths over the pandemic is here: https://clarityonhealth.substack.com/p/2022-excess-deaths-think-66-plane
The ABC has just released journalist Angus Dalton's review of this April 16 publication.
'Assessing the impact of Australia’s mass vaccination campaigns over the Delta and Omicron outbreaks'
https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0299844&utm_content=INTRO
'Modelling was used to compare the number of people who died with or from COVID between August 2021 and July 2022 (which captures the Delta wave and the onset of Omicron) against a hypothetical scenario in which vaccines never made it to Australian shores.'
'...............The PLOS One study also found that unvaccinated people had 7.7 times the death rate of people who were double vaccinated and 11.2 times the death rate of those who got a booster shot.'
What do mathematician readers think of the conclusions of this paper?
The point is that the "expected mortality" curve is a man made estimate.
While you can pick an impossible curve, within certain limits all the estimates are neither right nor wrong.
Eliminating outliers is standard practice, but it's up to the researcher to decide whether doing that and what constitutes an outlier.
If the mortality data are correct - and it's hard to claim that the dead count in the developed countries isn't substantially correct - anyone can make one's own "expected mortality" curve: there are several methods. And from there, one can make his/her own analysis and draw his/her own conclusions.
This isn't to say that the Australian conclusions are correct. It's to say that unless they made some really weird estimate, you can't say that the estimate is wrong.
But you can say that it's an estimate among hundreds.