Per capita just takes the absolute numbers and divides them by the total population and then multiplies it by 100,000. There is no other magic than that.
The size of the populations doesn't affect this calculation. Out of every 100,000 vaccinated people how many died of a non-Covid cause?
It's not the size of the population, it's the mix of the population. Prior to June 19, the mix of the population was that the unvaccinated was 2.5 times the size of the vaccinated population. They were roughly equal at June 19. By September, the vaccinated population was 2.5 times the unvaccinated population.
So, your question should be: "Out of every 100,000 vaccinated people on March 19 [June 19, etc.], how many died of a non-Covid cause?" And so on. Because the population was not static across all time periods.
What you asked is the difference between the data on table 3 and table 4 from the dataset.
| Per Capita | COVID-19 Deaths | Non-COVID-19 Deaths | ||
|---|---|---|---|---|
| Week Ending | Vaccinated | Unvaccinated | Vaccinated | Unvaccinated |
| 19-Mar-2021 | 0 | 0.2 | 1 | 1.9 |
| 26-Mar-2021 | 0 | 0.2 | 0.6 | 1.8 |
| 02-Apr-2021 | 0 | 0.2 | 1.1 | 1.6 |
| 09-Apr-2021 | 0 | 0.1 | 1.4 | 1.4 |
| 16-Apr-2021 | 0 | 0.1 | 1.5 | 1.5 |
| 23-Apr-2021 | 0 | 0 | 2 | 1.4 |
| 30-Apr-2021 | 0 | 0 | 2.8 | 1.3 |
| 07-May-2021 | 0 | 0 | 2.9 | 1.2 |
| 14-May-2021 | 0 | 0.1 | 3 | 1.1 |
| 21-May-2021 | 0 | 0 | 3.3 | 1.3 |
| 28-May-2021 | 0 | 0 | 3 | 1.1 |
| 04-Jun-2021 | 0 | 0 | 2.7 | 1.1 |
| 11-Jun-2021 | 0 | 0.1 | 2.8 | 1.1 |
| 18-Jun-2021 | 0 | 0 | 2.7 | 1.1 |
| 25-Jun-2021 | 0 | 0.1 | 3.1 | 1.4 |
| 02-Jul-2021 | 0 | 0.2 | 2.8 | 1.3 |
| 09-Jul-2021 | 0.1 | 0.2 | 2.8 | 1.1 |
| 16-Jul-2021 | 0.1 | 0.2 | 2.6 | 1.2 |
| 23-Jul-2021 | 0.1 | 0.3 | 2.9 | 1.5 |
| 30-Jul-2021 | 0.1 | 0.4 | 2.9 | 1.6 |
| 06-Aug-2021 | 0.1 | 0.6 | 2.6 | 1.4 |
| 13-Aug-2021 | 0.2 | 0.4 | 2.6 | 1.2 |
| 20-Aug-2021 | 0.2 | 0.5 | 2.5 | 1.5 |
| 27-Aug-2021 | 0.1 | 0.6 | 2.2 | 1.4 |
| 03-Sep-2021 | 0.2 | 0.4 | 2.3 | 1.4 |
| 10-Sep-2021 | 0.2 | 0.5 | 2.7 | 1.4 |
| 17-Sep-2021 | 0.2 | 0.5 | 2.2 | 1.3 |
| 24-Sep-2021 | 0.2 | 0.4 | 2.2 | 0.9 |
The problem here is one of the two populations is older on average, so naturally their mortality rate is higher.
No. The "older" population (55-59) is changing depending on the time slice.
According to this Wikipedia article on the vaccination roll-out in the United Kingdom, the general population under 60 without comorbidities didn't begin getting vaccinated until March (the start of the data). The youngest groups began vaccination in June.
The "older" population is the unvaccinated population before June 18, and is the vaccinated population after June 18 (from the table in my earlier post). The 65+ group was vaccinated before the start of the data (pre-March).
By the equilibrium point in June 18 when the youngest members began to be vaccinated, supposedly they would bring down the per capita death rate as they enter the vaccinated population. However, this does not seem to be happening according to the data.
The unvaccinated per capita non-COVID death rate falls from a high of 1.9 in March to about 1.1 in June. However, the vaccinated per capita non-COVID death rate rises from 1 in March to a peak of 3.1 in June, and remains at that level.
Why does the vaccinated non-COVID death rate rise faster than the unvaccinated non-COVID death rate falls as the population moves from the latter group to the former?
-PJ
Exactly. And during the study the vaccinated 10-59 population was on average significantly older than the unvaccinated 10-59 population. Because there were no vaccinated people between 10-18!!
In essence they were comparing an 18-59 population to a 10-59 population. Of course the on-average younger unvaccinated population had a lower mortality rate. As anyone would expect.
Prior to June 19, the mix of the population was that the unvaccinated was 2.5 times the size of the vaccinated population. They were roughly equal at June 19. By September, the vaccinated population was 2.5 times the unvaccinated population.
Which has exactly zero bearing on the per/100,000 calculation.
The youngest groups began vaccination in June.
Read your own Wikipedia article. Vaccinations for 12-18 year olds wasn't approved until September.
The "older" population is the unvaccinated population before June 18, and is the vaccinated population after June 18 (from the table in my earlier post). The 65+ group was vaccinated before the start of the data (pre-March).
You're confused about the study. The cohorts older than the 10-59 one are fairly matched in age. The 10-59 one wasn't because there were no vaccinated people between 10-18. It wasn't apples-to-apples. This is why Berenson chose only that cohort and didn't report on the significantly higher all-cause mortality among the unvaccinated in all the other brackets.
He chose the one that the researchers warned was invalid and ignored the ones that were a fair comparison.
Do you think he did that out of ignorance?
Do you think it odd that he only reported on that one cohort?
So, your question should be: "Out of every 100,000 vaccinated people on March 19 [June 19, etc.], how many died of a non-Covid cause?"
I think that's what the study shows us.
When a death was recorded the questions were 1) was this person vaccinated? and; 2) what was the age of the person?
The number of deaths were then compared to the size of the vaccinated or unvaccinated populations at that time and the mortality rate was established.
I'll look at the other questions you have about the study but for now let's focus on Berenson's claims which are the subject of the thread.
His characterization of the data is misleading and he doesn't share the researcher's perfectly reasonable and intuitively obvious explanation with his readers.
Do you think he respects his readers when he does this?