Excess Deaths Continue to Climb With No Sign of Slowing Down

It has been just over two years since I last wrote an article in the Daily Sceptic about excess deaths. In the meantime I have continued to compile data from the weekly ONS mortality reports in the hope that an obvious fall in excess deaths would eventually appear in the mortality statistics, but unfortunately this has not yet happened to any significant degree.

Five full years have now elapsed since the year that the first two most deadly Covid waves hit in 2020, so it is a good time to summarise where we are in terms of aggregate data gathered thus far.

The table below shows the data that I will present in this article. The upper half of the table shows the years from 2015 through 2019, which for the purposes of this article represents the baseline situation before the six-year span of the Covid years shown in the lower half of the table. The average for the 2015-2019 data appears in between the lower and upper halves of the table.

The figures have been totalled for each year and ‘smoothed’ to account for the fact that 2015 and 2020 were both 53 week years statistically speaking, and so for example the five-year 2015-2019 average figures are actually for a 52.2-week elapsed period. The meanings of each column are explained in the description below the table itself.

Column 1 shows the total deaths for each year for all respiratory disease (ARD) for the years from 2015 through 2025. Note that this is deaths ‘due to’ respiratory disease as reported weekly by the ONS, as opposed to deaths ‘with’ respiratory disease where the cause of death was something else and respiratory disease was just a comorbidity at the time of death and thus not the direct cause of death. The lower half of the table includes deaths ‘due to’ Covid in ARD as well as all the other respiratory diseases. 

It is obvious that, particularly in 2020 but also to a lesser extent in 2021, there were very high numbers of deaths from ARD. Column 4 shows the excess figures relative to the baseline. However the situation was normalising already in 2022 and 2023, and had completely returned to baseline levels again through 2024 and 2025.  Bear in mind that several recent studies have appeared that cast doubt on the numbers of deaths ascribed to Covid, particularly in the early phases of the pandemic, so it is possible that the ARD situation normalised even more quickly than appears from the table. 

Deaths due to Covid have fallen from a high of over 80,000 in 2020 to just under 3,000 reported in 2025 (out of 71,000 of total ARD), so any artificial inflation of the Covid numbers would now be little noticed in the total figure anyway. Column 14 shows the actual deaths due to Covid in each year just for reference purposes.

Column 2 shows the total deaths for all cause mortality (ACM) for each year. Again 2020 and 2021 were the worst years in this regard, however every year since 2020 also had significantly higher excess death than the pre-Covid baseline (see column 5), and the total cumulative excess since 2019 now stands at just under 300,000.

Column 3 shows the total deaths for non-respiratory mortality (NRM) for each year. This figure encompasses deaths from all other causes than respiratory disease, so it is calculated by subtracting column 1 from column 2. The figure for 2020 is within normal baseline values, and although in 2021 it does rise slightly, it is not until 2022 and each subsequent year thereafter where we can see the significant and sustained rise in the level of this figure over baseline. Columns 6 and 7 show the excess over baseline expressed in absolute numbers and as a percentage respectively.

We can hypothesise from these NRM numbers that there must be some factor operating throughout the whole period from 2022 through 2025, other than Covid itself (which was falling rapidly over that time), which is contributing to an ongoing excess NRM year on year. We may have our suspicions as to the nature of that factor, but the lack of an excess in 2021 weakens the hypothesis.

So this brings me to the invisible factor, which explains the reason why we only can see the tip of the NRM excess death iceberg in 2021, but then do notice it increasing significantly from 2022 onwards. For this factor to become more visible though, we need to rely on modelling.

Now I am as sceptical about modelling as any Daily Sceptic reader, but nonetheless models have been used to justify all kinds of non-medical interventions applied during the pandemic, including the necessity for extreme measures taken on individual citizens, such as the Covid vaccine mandates and travel restrictions, so it should be fair game to make use of models that argue against the narrative. Besides which, this is a very simple model that simply aims to take into account the impact of high mortality years on subsequent years.

This article discusses the phenomenon of mortality displacement (MD), which was to be expected in 2021 following the high levels of excess mortality in 2020. The authors use a model to calculate MD, which has been derived from following a selected subgroup of actual vulnerable people across a given time period, and then applying the model across the general population.

What we learn from this article about periods of excess mortality is that there is a displacement across time for the recorded excess deaths, and that this is most visible in short term statistics when the excess mortality has largely happened to the most vulnerable sections of society, whose deaths would most likely have occurred in the near future anyway. 

In other words, a period of excess mortality amongst the most vulnerable portion of a population will be rapidly followed by a period of reduced mortality in that population, which ought to show up in the mortality statistics a relatively short time afterwards.

We know that the average age of death during the Covid waves was over 80. This elderly cohort already has a heavy weighting in mortality statistics, accounting for more than 80% of the all-cause mortality during normal years. It is no surprise then, that the study calculated a high MD factor (8.9%) for the following year or so as a result of the excess deaths from the first two deadliest Covid waves. This result is in agreement with the earlier study that I commented on back in 2022.

Although the new paper describes its model as a “counterfactual approach”, mortality displacement is directly observable in statistics and is a very real phenomenon. The issue however with the non-respiratory mortality figures in 2021 is that even though the first few months do show a negative excess NRM as predicted by MD, this then rapidly goes into reverse around May, such that the cumulative amount of excess NRM ends the year as a positive figure.

We can surmise then that something else must begin to operate during 2021 to again begin to increase excess NRM to the extent that it eventually outweighs the negative effects on NRM of the MD from the 2020 Covid shocks. Looking at figure 6 in the study paper it can be seen that almost 90% of the MD effects on statistics for the vulnerable cohort studied would be completed and thus visible in the data, within one year from the time of any given spike in death rates.

Looking at figure 8 in the study paper we can see graphically how this plays out during 2021 (starts just before week 40), with the gap between recorded excess mortality and adjusted excess mortality gradually increasing over the course of 2021, up until October 8th.

Fundamentally then, in order to get an accurate picture of excess mortality after a recent spike in death rates, one has to add the appropriate figures for MD onto the officially reported figures for total annual deaths. This is achieved by lowering the expected deaths baseline figure for ACM and is eloquently referred to as “harvesting” in the paper’s abstract.

Column 8 shows the calculated MD percentage for the year, which is based on the excess death figure for the preceding year. Column 9 shows the lowered expected deaths from the 2015-2019 baseline as a result of the MD from the previous year. For the purposes of this piece I have ignored any MD operating beyond one year from the excess deaths accumulated during the previous year. Column 10 shows how this translates to the additional MD excess deaths for the year in question, and Column 11 represents the new adjusted NRM total after the MD excess has been added in. Column 12 then shows the percentage increase on the 2015-2019 baseline figure. Finally column 13 shows the adjusted NRM excess numbers for that year.

The reason that I have added the MD onto the NRM figure, rather than the ACM figure, is simply because all the excess deaths that the MD represents are a statement about what didn’t happen but could have happened – the counterfactual. They are deaths pulled forward from the future and all of them will be non-respiratory because they would have occurred due to the poor health of the individuals concerned, which made them vulnerable to dying in the near future from the long term morbidities that they were carrying at the time of the Covid shock.

The authors of the paper on the MD model write that “their general approach can be applied to any adverse event-based outcome (death being just one example) that follows any ‘at-risk’ defining event (here it just happens to be developing COVID-19), to investigate impacts by time or by cause”.

I hypothesise then, that some other factor than Covid has caused a new wave of excess death in the population, albeit temporally more spread out than from the spikes from the obvious at-risk defining event caused by Covid itself. After the adjustments made in accordance with the MD model, the picture suddenly becomes much clearer with respect to when this new traumatic factor began to operate – namely from 2021 and every year thereafter. 

It is likely however that the MD figures derived from using the model are an underestimate for 2021, and an overestimate for the following years. This is because, as previously mentioned, a portion of the MD for 2020 will not appear in the statistics until 2022, whereas for subsequent years the MD factor will definitely be less than the 8.9% calculated from the excess deaths that happened suddenly during 2020. This is because the excesses from 2022 and onwards are no longer concentrated in the most vulnerable cohorts where MD effects are visible shortly after their deaths (and thus are caused by something else than what happened in 2020), so the MD effects will be spread out over several more years. A recent post from Dr Clare Craig illustrates the point that excess death after 2021 is no longer restricted to the most vulnerable cohorts.

Furthermore, I have used the ratio of ACM excess from the years after 2020 to the ACM 2020 excess as a proxy adjustment to the 8.9% ACM baseline adjustment figure, and this will be less accurate than the calculations made in the model from the very specific concentrated excess deaths arising from the 2020 Covid spikes that they used.

It may be argued though that an ever-increasing population size from 2020 through to the present could account for many of these excess deaths, and thus using a 2015-2019 baseline annual mortality rate can no longer be applicable. However a brief look at net migration figures show that there were no substantial increases in net migration until the mid-90s (see the first interactive chart shown on this ONS page).

These younger people will have very little impact on ACM for many years to come. The reason is that, just like in many other kinds of statistical data, there is an 80-20 rule operating. That is to say that 80% of deaths are found amongst just 20% of the population, which unsurprisingly is the older subset, comprising those aged between 70 and 95 years of age.

The vast majority (92%) of all non-UK born residents had not yet reached this death zone (DZ) in terms of their age because new net migration is overwhelmingly composed of waves of younger individuals hoping for a better economic future in the UK. The 2015-2019 baseline already includes mortality data from the bulk of more longer term foreign born residents, and their aging over 2020-2025 is not a significant factor in overall mortality terms due to their position on the death rate curve well before the DZ, where the rate changes rapidly (see figure 9 from the ONS 2021 population census).

Meanwhile when one looks at the UK population who are currently in the DZ, it is possible to estimate the relative sizes of the five-year cohorts within it by looking back at the historical birth records.

Plotting these against a standardised ‘deaths by year of age’ graph, it can be seen that in 2020 a peak of births in the cohort born after WW2 (those born between 1946-1950) had already entered into the death zone. Fast forward to 2025 and the subsequent five-year cohort born between 1951-1956 had entered the DZ over the intervening five years. This cohort was smaller than the previous one.

Looking at the other end of the DZ, the very oldest cohort born between 1921 and 1925 exited the DZ over the next five years, to be followed by the cohort born between 1926 and 1930, which was a considerably smaller cohort. The DZ is marked with blue arrows in both charts.

Furthermore if one looks at actual death rates for each five-year cohort and plots this against the years of birth of each cohort, it can be seen that the cohort most affected by the very highest death rates (cohorts to the right of the red line) is of course the oldest cohort. As the 1921-1925 cohort dwindles rapidly to nothing after it moves beyond the DZ, the 1926 to 1930 cohort that replaces it will experience far fewer deaths overall just because it is so much smaller.

In terms of total numbers, there were 235,000 fewer people in the DZ in 2025 (according to birth cohort ratios), when compared to 2020 (–1%), and also fewer of them were at the oldest ages with the highest death rates. This all points to a falling number of expected deaths from 2020 through 2025, rather than an increasing number.

In conclusion then, I believe that the baseline mortality used to calculate excess deaths in 2020 (the 2015-2019 average) is still a valid baseline for use up to and including 2025. When this is used there has been excess all-cause mortality in every year from 2020 through to 2025, totalling almost 300,000 people. Of note, the EU is still using the 2016-2019 baseline when compiling its mortality statistics. RD has normalised to the 2015-2019 average over this period, which further validates the usage of the 2015-2019 baseline. Nearly all the excess mortality is therefore now concentrated in the non-respiratory mortality.

When mortality displacement is also taken into account – as it ought to be, because the number of expected deaths should have been below the baseline figure in all years after 2020 – then the excess death count is in reality higher still, to the tune of a further 170,000 excess deaths thus far since 2021, so around 470,000 deaths. This is of course a very crude estimation and a more sophisticated study would no doubt whittle this figure significantly downwards. But it will still amount to a considerably larger figure than current official statistics show for excess deaths over the period.

It should be remembered, though, that every excess death represents a mortality displacement to some future year when the death would have otherwise been expected to occur. As younger cohorts now appear to be experiencing increasing death rates in proportion to all other age cohorts from 2021 onwards, much of this younger cohort displacement will be many years from now, and cannot therefore be counted in the figures for 2020-2025 which I am presenting here. But the baseline of expected deaths ought to be reduced for many years going forward in order to accurately gauge future excess deaths.

As Covid is a negligible factor in overall mortality statistics now, something else has caused these excess deaths. Furthermore, whatever it is appears to still be working its insidious effects as we begin 2026. The only way to describe such a situation is carnage on a scale that would be considered unacceptable even in a war situation. It is quite extraordinary that this matter has not been properly investigated by the UK authorities, who seem neither to notice nor to care.