#coviddeaths

Casual neglect /

Of the one hundred thousand /

The Covid response was not an error, and it was not the result of rushing to address a crisis due to an unknown pathogen.

It was a lot of people, mostly professionals in the field, systematically and collectively doing what they knew was wrong, David Bell writes and systematically lays out the facts.

“When laid bare by maths and statistics rather than sponsored modelling, the covid response looks horribly like incompetence that was not completely unintentional,” he says.

The Covid Response Was Not a Mistake – It Was Just Wrong

By David Bell, as published by the Brownstone Institute

Early in 2025, some statisticians from Scotland and Switzerland wrote a discussion paper with a characteristically (for Scots and Swiss) understated, even boring, title: ‘Some statistical aspects of the Covid-19 response’. Good science is stated clearly without fanfare, while “bombshell” announcements, or similar rants, indicate a need to embellish. Good data speaks for itself. However, it only speaks widely if people read it.

The paper, by Wood and co-authors, was written for presentation at a meeting of the Royal Statistical Society in April 2025 in London. It remains one of the best reviews of the early response to covid - in this case with a United Kingdom focus but relevant globally. However, some people don’t avidly read the Journal of the Royal Statistical Society – Series A: Statistics in Society, or attend their London meetings. A pity, as London is nice for three days in summer, and this particular Royal Society seems to have a grasp of reality lacking in some of its siblings.

The paper provides simple statistical truths, as statisticians should. Truths are particularly valuable when applied to subjects where fallacies are more profitable. This is why, in public health, they have become so rare and, therefore, so worth reading. Stating truths dispassionately regarding covid helps to grasp how bad the public health response actually was. 

Covid and the Economy

Public health has always been highly dependent on economic health, so the authors set the scene by stating the obvious of the economics of the response of Western governments that decided in early 2020 that printing money was simpler than making people work to generate taxes:

Creating money while reducing real economic activity is obviously inflationary.

And consequently:

The subsequent sharp increase in inflation is one path by which the disruption has contributed to increased economic deprivation … of the sort clearly linked to substantially reduced life expectancy and quality of life.

This is important because we knew this long before 2020 (the Romans knew it), and we also knew that the resultant economic deprivation would shorten life expectancy. This is Public Health 101, and every public health physician knew it when covid started.

In public health, we recognise that there is a trade-off between spending money to save one person or allocating it elsewhere to save many more. If we just spend without limit, we all get poor and then we cannot really fund healthcare at all. This is not complicated; people understand it. It is why we don’t have MRI scanners in every village. We therefore make estimates of how much can save a life without overly impoverishing society and then losing more. Wood and colleagues looked at the UK standard for this compared to the costs of lockdowns:

…any reasonable estimate of the cost per life year saved from covid by non-pharmaceutical interventions substantially exceeds the £30K per life year threshold usually applied by NICE (the UK National Institute for Health and Care Excellence) when approving introduction of a pharmaceutical Intervention……  

 gives a cost per life year saved over 10 times the NICE threshold. 

Again, this is basic public health. Allocating health resources is a complicated issue as it is (rightly) tied to ethics and emotion but on a societal scale, it is how we manage our health budgets. In this case, the numbers predicted to be saved through the enormous costs of lockdowns never remotely made sense. 

However, the UK government, like governments elsewhere under the same apparent media-pharmaceutical yoke, simply ignored costs and benefits calculations and ploughed on regardless. Guided by its Scientific Pandemic Influenza Group on Behaviour (“SPI-B”), the UK government embarked on a campaign to mislead the public into taking actions they could reasonably expect to be massively harmful on an individual and national level. They knew the campaign to instil fear was unjustified; a campaign of misinformation aimed at the same public who paid them. Wood and colleagues provide “one of the milder examples”:

 … a widely displayed government poster picturing a healthy woman in her mid-twenties in a mask with the slogan “I wear this to protect you. Please wear yours to protect me.”

The actual risk profile that the UK government and SPI-B had at that time is shown in the Figure below, provided in the paper.

This is where statisticians are useful - to provide context in place of anecdote and fear. They provide a good one:

…the current best estimate for the return time of a super-volcanic eruption of the civilisation-ending magnitude that city dwellers are unlikely to survive is 17 thousand years (Rougier et al., 2018). Even only considering the two years of the pandemic this is likely larger than the covid risk to the woman pictured.

So logically, if they were being logical about covid, the UK government should now be gutting their economy to prepare for the aftermath of a super-volcano. But let’s not suggest that, as they might just do it.

Explaining Covid Burden

The UK government’s efforts to mislead the public regarding covid-19 risk were not a case of dealing with an unknown virus, as many are now claiming:

Risk was known early 2020: Diamond Princess, and e.g. Verity et al., 2020; Wood et al., 2020, from Chinese data.

Case fatality data from Figure 3 (B) in Verity et al. published in March 2020 by Imperial College London, noting minimal risk of Covid mortality among young and middle-aged people (i.e. those removed from work and school).

Irrespective, the UK government maintained that covid was severe and debilitating in young, fit people, potentially (as Wood and co-authors note) using actors and fabricated stories and thereby simply lying to people. The UK Office of National Statistics (“ONS”) did its part by, as the authors demonstrate from various studies, also misrepresenting the frequency of long covid.

SPI-B advice on masks was also strange, being at odds with their own citations, thereby grossly exaggerating their impact. This is a strange one - why would a government convince the public to cover their faces, knowing that they are basing their advice on falsehoods, running against previous advice and that it will not significantly help anyone? This is where bad intent starts to look increasingly part of the approach.

The authors then note:

This type of misleading and selective use of statistical evidence was not limited to the media. For example, in 2021, the official online Scottish government advice on face coverings stated that 

“Scientific evidence and clinical and public health advice is clear that face coverings are an important part of stopping the spread of coronavirus.”

and provided a link for the scientific evidence. This turned out to be a SPI-B/SAGE advice summary18, which cited two pieces of scientific evidence, apparently suggesting transmission reductions from mask wearing of 6-15%, or up to 45%, respectively. The paper cited as evidence for the first figure was in fact an editorial (Cowling and Leung, 2020), which also pointed out that the paper cited for the 45% figure (Mitze et al., 2020) was flawed (the design appears unable to pick up the case in which mask wearing is actually harmful, for example). The editorial’s figure is quoting a properly conducted meta-analysis (Brainard et al., 2020) which actually concluded

“… wearing a mask may slightly reduce the odds of primary infection with by around 6 to 15% This was low-quality evidence.”

Again, this government was unequivocally misleading their own people into a major behavioural change, whilst having evidence that it would not be of use; either negligence or simply lying.

Mortality

The discussion of Wood and colleagues on quantifying mortality becomes really interesting, demonstrating how difficult this actually is. Firstly, when covid hit in 2020, the babies born immediately after the Second World War were just turning 75. There were 31% more babies born in the UK in the year after the war’s end compared to the previous year, and high birth rates continued in subsequent years. There is nothing magic about 75, but the point is: a mass of the British public, born in the few years after the War, were entering the ages of rapidly increasing mortality. 

This is a driver of “excess mortality” not widely discussed. It means there should have been an increasing mortality in 2020, and in subsequent years (i.e. above normal compared to pre-2020, but not really an excess if standardised for age). This is important for understanding total excess, whether claiming it’s from “covid,” vaccination or anything else. It does not, however, account for rising mortality in younger age groups or the rate of death at any age.

The other obvious problem with covid numbers is that, as the authors note, people generally only die once. Thus:

Cumulative excess deaths  much lower than the 212,247 officially considered “covid.” Many covid would have died anyway , or were not covid deaths. The cumulative excess … are much lower than the total deaths recorded with covid (212,247 with covid mentioned on the death certificate by the end of 2022, according to the UK government’s data dashboard). There are a number of mechanisms that are likely to account for this. An obvious one is the fact that only some 17 thousand people had only covid and nothing else recorded on their death certificate.

That was 212,247 with covid on a death certificate - only 17,000 had covid only. But official figures frequently imply that all 212,247 died because of covid. Covid mortality events do not simply add to the mortality caused by the other comorbidities. The viral infection, like other viral infections, often simply hastens the deaths of very sick and dying people.

The equivalent figures for the UK in 2020 was a life expectancy drop of about 1 year and a life loss of about 6 days per head.

This is really important to understand. So, people who died of/with covid lost, on average, a year of life. But the vast majority of the population did not die. So, only 6 days were lost on average across the entire UK population.  

This raises a problem that governments and public health officials knew well before imposing lockdowns: the known impact of poverty and inequality on life expectancy. To quantify, well-accepted UK data from Marmott et al. (2020) show a 5-year gap between life expectancy of the upper decile (rich) and lower decile (poorest) people in the country. Covid caused, in comparison, a 6-day reduction in life expectancy (averaged across the whole population). It is therefore almost inconceivable that an intervention that greatly increases poverty could be less harmful than covid, from a public health viewpoint. 

Modelling

The paper points out the really basic flaws in modelling by Imperial College London and others in supposedly predicting the covid-19 impact. These models drove many governments’ responses, though it was clear at the time, and the modellers would have known, that the models were designed to exaggerate harms. In particular, they failed to adjust for population heterogeneity, which tends to slow spread and reduce harms (the most vulnerable leave the population, leaving a more resilient populace). Failure to account for heterogeneity will overestimate future transmission by design.

Perhaps the most surprising feature of the epidemic models used to justify covid policy was the omission of the fundamental role of person-to-person transmission rate heterogeneity investigated by Novozhilov (2008).

They also ignored the fact that close to half of early infections were hospital-acquired (China, Northern Italy) rather than from the community, leading to falsely high community transmission rates being fed into the models.

The Imperial modelling group, one should remember, was the same group that published in the Lancet in March 2020, showing almost no mortality in young and middle-aged people (second graphic above). They knew, when they pretended that very high mortality was expected, that the true picture was very different.

UK predictions were consequently far above reality - as were predictions of lockdown impact. Lockdown models assumed reproductive rate (R0) would be constant before or after lockdowns without intervention, whereas in reality, it always varies with time, steadily declining from an initial peak as fewer people remain susceptible to being infected per case, as more of the population is immune. Again, this is really, really basic outbreak modelling. Consistent failures (e.g. non-lockdown Sweden having about 6,000 deaths instead of 35,000) failed to stimulate any modification and rectification of these basic errors.

While the actual impact of lockdowns on poverty and economic health is clear, controversy does remain on their impact on covid transmission and mortality. Wood and co-authors address this by noting that nearly all lockdowns started after transmission had already started declining (see figure). It almost looks as if lockdowns were imposed at a time that would make them look effective, rather than with the expectation that they would avert more infections.

Time to stop pretending.

While covid started over 5 years ago, people want to move on, and there are myriad papers arguing one side or the other. However, the paper of Wood and co-authors does stand out. It does not push any advocacy baggage or speculate on political motives, but simply lays out numbers and facts. From the point of view of the pandemic industry, it provides a really strong argument for censoring facts and hammering dogma. When laid bare by maths and statistics rather than sponsored modelling, the covid response looks horribly like incompetence that was not completely unintentional.

Perhaps the modellers whose numbers justified covid hysteria simply did what they were paid for and did not expect politicians and media to take them seriously. Perhaps public health physicians promoting long-term poverty and inequality were just trying to keep their careers on track and mortgages financed. 

Perhaps politicians are just resigned to a reality that they must represent corporate sponsors before their constituencies to survive. Perhaps we are just not as smart, virtuous and moral as we like to pretend that we are. Whatever the underlying issues, it is time everyone stopped pretending the covid response was anything but a mess, or that we did not know it would be. There is still a place for truth. 

About the Author