The 5 dubious pillars of the Covid-19 panic narrative

Daniil Gorbatenko
12 min readApr 5, 2020
Photo by cottonbro from Pexels

The Covid-19 pandemic and the apparent surge in cases and deaths worldwide has engendered a global panic and a stringent response up to total lockdowns in most affected countries. Here, I will try to summarize the evidence that all five key pillars of both the panic and the harsh and extremely costly response are highly dubious.

  1. The Covid-19 response is justified because the mortality rate is too high compared to other endemic respiratory diseases like flu

I will save you the discussion of why the published crude case fatality rates (CFRs) should just be disregarded. A lot has been written on that already.

Here, I would like to summarize the best evidence that the number of infected people is vastly undercounted everywhere, even probably in places like South Korea.

Let us start with the evidence from Iceland. Already more than 3 weeks ago the company Decode Genetics started testing thousands of Icelanders in an almost random fashion. Already on March 15, the data made it clear that around 1% of the country’s population were infected. In parallel, a lot of symptomatic cases have been detected by the public testing institution.

Overall, according to the data through April 4 (source), there are 1486 confirmed infections, 38 people are hospitalized,12 are in the ICU and 4 have died but actually only 3 should be counted because the fourth is a foreign tourist who probably died with Covid-19 but not from it. Given that around 3,000 Icelanders were probably already infected weeks ago, we can safely assume that there have been at least 10,000 infections in Iceland. This gives us the current estimate for the infection fatality rate (IFR) of 0.03%. To get even to the IFR of 0.1% (estimated for seasonal flu), many more deaths should take place because the number of infected will also keep growing.

Now, Iceland can be an outlier but there is suggestive evidence from the United States and Italy telling us that the infection is way more widespread than the detected cases suggest. Silverman and Washburne used publicly available CDC influenza-like illness outpatient surveillance data to conclude that the infections in the US are undercounted by a factor of 100–1000. This would mean that the current US CFR of 2.83% turns into an IFR of 0.03–0.3%, and the US begins to look a lot like Iceland.

Finally, what about the most difficult example to reconcile with a more optimistic view of Covid-19’s fatality? How about Italy in general and Lombardy, in particular? The IFR there will probably end up to be higher than everywhere else (because of a whole range of potential factors, including air pollution, the complete unpreparedness of hospitals, the share of the elderly, the high rate of cohabitation of young adults with parents, the average age of the small communes that were first hit, even the potential role of the obscure TMPRSS2 gene). But it will not be an order of magnitude higher.

A tantalizing recent piece of evidence that is pointing at just that has come from the commune from which the first Lombardian detected case originated, Castiglione d’Adda. This commune has essentially been under non-stop lockdown (local, regional and national) since February 22, and yet a study of blood of 60 asymptomatic donors has shown that 40 of them have been infected. This is almost 70%.

Now, as I discuss in the next section, there are probably not 70% of people infected in Lombardy but the infection is probably very widespread because a detailed study of the outbreak by Tirani et al. shows that it started in early January and had already been widespread by the time the first case was detected on the night of February 20 at the Codogno hospital. If 30% of the Lombardian population have been infected, this will mean 8,656 deaths per 3 million infected, and an IFR of 0.29%. Of course, there will be more deaths in Lombardy but the new deaths have been going down and the number of people in the ICU has essentially stopped growing several days ago (see the detailed data source here). The disease may have hit most of the vulnerable people in that unfortunate region, and the lockdowns may be an important reason for this, as I discuss below.

2. Without harsh interventions, the epidemic will just grow exponentially until more than half of the population become infected

Another cornerstone of the Covid-19 panic is the idea (enshrined in the form of an assumption in the most-widely used epidemiological models) that if left to its own devices, a Covid-19 outbreak will rapidly grow exponentially until more than 50% of the population have been infected. Put differently, every epidemic is considered to be characterized at its unchecked stage by its basic reproductive number, or R0.

It was already clear weeks ago from the Lombardian death data that even at the unchecked stage, Covid-19 did not have a fixed reproduction number as death growth rates had been declining before any intervention effects could have had an impact. The aforementioned much more detailed study of the Lombardian outbreak shows exactly the same thing. It suggests that the epidemic initially grew slowly, then briefly became exponential and then slowed. All that before the first interventions. Finally, Wodarz and Komarova demonstrate the brief exponential growth replaced by growth described by power laws for a range of countries. They write:

Alternatively, it is possible that the local epidemics characterized by power laws demonstrate an intrinsic power law spread that is independent of interventions. Computational models suggest that infection spread across networks or in spatially structured populations can lead to dynamics that follow a power law rather than an exponential trajectory, see below.

If some countries truly show infection spread that is governed by a power law, the question arises why other countries show clear exponential spread, and why yet other countries are more difficult to classify. We hypothesize that different countries are at different stages of epidemic development, but they all roughly follow the same trajectory, where an initial exponential growth is gradually replaced by a more power like behavior.

Why is the assumption of R0 implausible? Because of the further assumption needed to sustain it, namely that societies are well-mixed. Translating this into clearer language, every person in society is assumed to be equally likely to contact every other person.

This is extremely dubious. It has long been established in the literature on social networks that human societies are characterized by significant degrees of clustering and regular contacts. Human contacts are also locally clustered, as Wodarz and Komarova rightly emphasize in the work just cited. Furthermore, not all contacts matter remotely to the same degree. And then, there are also the hotbeds of Covid-19 transmission like nursing homes and hospitals that do not exist in epidemiological models and to which we will return below.

Let me just illustrate the potential effects of clustering based on a toy model of a society as a clustered network.

Here, the toy society consists of groups with a lot of connections inside them but only one or two connections to other clusters. Suppose in the initial period the red node is infected and infects another node in its cluster. In the second period, the second node infects the node that is connected to the second cluster. But it is entirely possible that in the third period, the infection is not transmitted to the subsequent cluster. The infection then dies out even though most members of society are still susceptible.

Obviously, real societies are much more complex but this toy illustration is only for explaining the basic idea.

3. Targeted measures cannot work when the epidemic is widespread

Already, in February, South Korea gave a mass-testing masterclass and demonstrated that this idea is totally false. Countries like Germany, Iceland and Switzerland have followed suit, and they now have vastly smaller numbers of deaths per million as a result. But this is not just about testing.

Two enormous elephants, mammoths in the Covid-19 room are nursing homes and hospitals. In France, as of April 4, just the deaths that happened inside nursing homes stand at 2028. In Spain, around 2000 people probably died in nursing homes just in Madrid (sic). 40% of fatalities in the region of Castilla y Leon are from nursing homes, as are 511 of the ones in Catalonia. And everywhere you look there is a massive spread of the disease to them — with predictable results. Hopefully, we will soon know which part of all deaths from Covid-19 is attributable to nursing homes but it could be a third or perhaps even close to half.

However, crucially the massive spread to nursing homes is not an inevitability. The testing of all the nursing home personnel, earlier closures of nursing homes and other possible measures would likely have been the most life-saving measure in the fight against Covid-19. A key reason why the number of deaths is still low in Norway may be because of the strict measures to curb the spread to nursing homes. If only the resources and attention had been focused on them instead of trying to police whole societies…

And outbreaks in nursing homes do not just cause a lot of deaths, they probably also inundated many hospitals with extremely sick patients and helped spread Covid-19 there. Which brings us to hospitals. There is also a wealth of evidence that hospitals have been a hotbed of Covid-19 transmission, especially in the crucial earlier phase of the epidemic.

Given the probable dependence between the viral dose and severity of symptoms and the capacity of severely sick patients to spread disease, this should not be surprising. This author with a background in virology gives a glimpse of the potential creative measures that could be taken in this regard:

Think, for a moment, of how we monitor those who work with radiation. Using radiation dosimetry, we quantify someone’s total exposure, and we set limits on it. We already know how critical it is for doctors and nurses to limit exposure to the coronavirus by using protective equipment (masks, gloves, gowns). But for health-care workers on the front lines of the covid-19 pandemic, especially in places where protective equipment is scarce, we might also keep track of total exposure, and put in place viral-dosimetry controls, so that one individual can avoid repeated interactions with some set of highly contagious patients.

This same approach could also be translatable to handling patients who have not yet developed severe symptoms:

Establishing a relationship between dose and disease severity could, in turn, affect patient care. If we could identify pre-symptomatic patients who were likely exposed to the highest doses of viruses — someone cohabitating or socializing with multiple sick family members (as with the close-knit Fusco family of Freehold, New Jersey, which has had four deaths), or a nurse exposed to a set of patients shedding large amounts of the virus — we might predict a more severe experience of the disease, and give them priority when it came to limited medical resources, so that they could be treated faster, earlier, or more intensively.

And, finally, the care of covid-19 patients could change if we began to track virus counts. These parameters could be gauged using fairly inexpensive and easily available laboratory methods. Imagine a two-step process: first, identifying infected patients, and then quantifying viral loads in nasal or respiratory secretions, particularly in patients who are likely to require the highest level of treatment. Correlating virus counts and therapeutic measures with outcomes might result in different strategies of care or isolation.

4. There is no way to significantly boost the healthcare system’s capacity to handle the sick and especially the critical cases. And maintaining the ICU capacity is an imperative that trumps everything else

The idea that the capacity of the medical system to handle the influx of Covid-19 patients (especially the severe one) is very limited was and is the key rationale for the extremely harsh measures to suppress the spread.

This is perhaps the simplest claim to rebut. Examples like the NHS Nightingale hospital in the UK, the field hospital in Stockholm, countless companies ramping up the production of ventilators, the relocation of patients from heavily-hit areas in Netherlands and France are just devastating for it.

But even if it were true that the ICU capacity were very limited, it would still be unclear why it trumps all other considerations. Evidence seems to strongly suggest that once they are placed on ventilators, Covid-19 patients are unlikely to survive. The question then becomes why absolute priority must be given to ICUs even at the expense of risking a huge toll for the society at large? Is it to maximize the lives saved or make people feel good because we are trying to save those most of whom may not be savable?

To this it is worth adding that most of the Covid-19 patients in ICUs are not on ventilators but are just being monitored and (or) are receiving oxygen support through non-invasive means. This does not have to be done in the ICU context for all or perhaps most of them.

5. Keeping families inside as much as possible is a good idea

This seems a very sensible approach but only because most people assume that every contact is equally likely to infect someone, and once infected, someone randomly becomes either mildly or severely ill. But longstanding line of research in virology suggests that this is not remotely the case. The severity of disease has been shown to depend dramatically on the viral dose and the capacity to spread disease further may depend on the severity of one’s viral load. I will not summarize the evidence in detail here because Robin Hanson has already done it better than I could.

In addition, to this there is a mountain of evidence showing that Covid-19 spreads vastly better in closed spaces and through close contacts. There is not a single superspreader event that took place outside. An analysis of 11 clusters of spread in Japan showed that they all happened in confined, poorly ventilated spaces. Nursing homes and hospitals where Covid-19 spreads like wildfire are confined spaces, and so on.

This is starting to make the confinement of families (whether enforced through a direct order or a panic campaign and the closure of all public spaces) look a pretty dubious strategy. As a result of it, optimal conditions are created for the virus to spread in large doses to family members of those infected prior to the lockdown, and if the rate of cohabitation of the older and younger adults is high, you probably get… the badly-hit regions of Italy, where some of the worst-hit communes have been under lockdown since February 22.

It is important to note here that mass staying inside may reduce infections but this is not the point if it leads to higher rates of severe disease and more deaths. Add to it the social, economic, medical and psychological toll of keeping the whole populations essentially under house arrests for months,

And this is not just armchair theorizing. The authorities of one significantly affected country, Sweden, has so far resisted the pressure, including from outside, and refused to force their citizens indoors or even close all the public spaces. In this paper, I used this fact to compare deaths per million for Sweden and 3 Italian regions (Piedmont, Liguria and Marche) to which Covid-19 spread significantly relatively late and in which the majority of days with deaths occurred under lockdowns. The data show vastly higher deaths per million for the first 21 days with deaths for the 3 Italian regions compared to Sweden.

Of course, there are probably factors that make the situation in those Italian regions worse. But we can also compare the evolution of Covid-19 deaths per million in Sweden to that in Denmark. Denmark has not yet imposed a total lockdown but it has banned gatherings of more than 10 people, closed almost all public places and encouraged as many people as possible to work from home. This is almost bound to have made most families to stay inside most of the time, so what is the result? So far, there have been 21 days with deaths in Denmark, and the comparison of deaths per million with Sweden should give pause to anyone who advocates the mass shelter-in-place approach.

That said, it should be reiterated that Denmark is not under a total lockdown like Italy, and that in Sweden, a lot of people have probably been staying at home, too. So, the comparison between Sweden and the Italian regions is probably cleaner.

I am also aware of the low death counts in Norway but Norway appears to have done a lot of testing. Already as of March 25, 73,000 tests had been done.


The evidence is mounting with every passing day that neither the panic, nor the prevailing response to the Covid-19 pandemic are remotely justified or productive. All the draconian measures aimed at making people stay inside should be discontinued immediately, and the resources should be allocated to targeted measures such as serological testing, testing suspected cases and screening of contacts, shielding the nursing homes to which Covid-19 has not yet spread, boosting ventilation at workplaces to get them to reopen and so on. Some restrictions like those on mass gatherings and bars and restaurants can be useful but their goal should not be to keep families sheltered at homes.



Daniil Gorbatenko

PhD, economics (2018) from Aix-Marseille University, independent blockchain adoption consultant based in Aix-en-Provence, France, Email: