ONS Data: 25% Excess Mortality Among the Boosted is Obscured by Undercounting of the Unvaccinated

Oops, the unvaxed were 50% undercounted

Today, on Feb 21, 2023, UK’s ONS released the much-awaited “Deaths by Vaccination Status” data set.

The report is expansive, shows COVID-19 and overall mortality, and conveniently breaks it up by age, sex, or both—a very valuable and interesting resource.

It also shows “deaths per 100,000 person-years”, where the ONS statisticians calculated “person-years” lived by slices of the population they analyzed, month-by-month. They divide the number of deaths in a given vaccination status category by person-years lived by that category to show us the mortality broken down by vaccination status.

That number of “person-years” is the denominator of the calculation.

The data appears soothing and demonstrates that unvaccinated people have somewhat higher mortality.

Be aware that the mortality per 100,000 person-years numbers are fake because ONS undercounts unvaccinated people by approximately one-half.

Let me demonstrate. Open up the latest ONS report in Excel. Go to the tab titled “Table 2” and go to line 2,616. You will see data for non-Covid-19 deaths for March 2022 and age category 50-59.

Circled in red, I calculated the population percentage by vaccination status, computed based on person-years for each vaccination status, divided by total person-years for the entire age category for March 2022.

So, according to the ONS, the unvaccinated were 6.19% of the 50-59 subgroup population in March 2022. Were only 6.19% of 50-59-year-olds unvaccinated in the UK last March?

Let’s look for a better source of information!

In another window, open up UKHSA Week 13 vaccine surveillance report. Go to Page 17.

You will quickly see that for the same month, according to the UKHSA, about 87% of 50-59-year-olds had at least one dose of the vaccine, so 100-87=13% of 50-59-year-olds were unvaccinated last March of 2022!

So, whose number is correct? Were only 6.19% of 50-59-year-olds unvaccinated at the time (per the ONS), or 13% (per the UKHSA)? The answer is that the second number (13%) from the UKHSA is correct and is based on databases of named UK residents eligible for vaccination.

So, the ONS understates the number of unvaccinated people by about 2x (depending on age category). Prof. Norman Fenton and Martin Neil discussed this in January. I also discussed that in November 2021 (yes, 2021!). See “Debunking fact checkers” here:

UK: death protection downward slide accelerates

Corrected Data Suggests Higher Non-Covid Mortality in the Boosted

With our knowledge from comparison to March 2022 UKHSA vaccination rates, let’s look at more recent all-cause death rates. Take a look at lines 999-1005 of the same ONS table. (all-cause deaths for 50-59 for Dec 2022, the most recent month of the report)

If the number of unvaccinated persons is undercounted by two times, then their mortality is conversely OVERcounted by roughly two times. (the precise calculation would require us to know the exact, corrected percentage of unvaccinated persons and corrected percentage of boosted persons, which we do not have, so this is a rough estimate). So we would have corrected data along the lines of:

Unvaccinated mortality: 366/2 = 183
Boosted mortality:      229

These rough estimates suggest that boosted 50-59-year-olds experienced 229/183-1.0 = 25% greater mortality than the same age but unvaccinated people during December 2022.

Such estimates are quite consistent with the excess mortality that the UK experienced last December:

I am very sorry that the ONS used incorrect counts of unvaccinated people. Fortunately, we can correct them using publicly available, official UK data.

Such corrected counts of unvaccinated people lead to a much more realistic comparison of mortality between boosted and unvaccinated people, consistent with our observations of overall excess mortality in the UK.

This is NOT a harmless mistake. Undercounting unvaccinated people will distract the public from looking at the real cause of excess mortality because ONS numbers falsely suggest that the boosted people experience lower excess mortality than the unvaccinated.

The opposite is likely true: as I have shown, the boosted people have higher excess mortality. Furthermore, corrected mortality explains the 20% or so excess mortality the UK experienced in December.

Dear Substack authors: feel free to make similar calculations across other age groups. The results will follow the 50-59 age group. The undercounting of unvaccinated people by the ONS is pervasive across the entire UK dataset.

Check out El Gato’s analysis of the same dataset also.

Do you think that the undercounting was an accident? Do you think that my calculations are flawed? All opinions are welcome!