A populist and a liberal are standing outside a burning house. A man shouts from a third-floor window: ‘I’m trapped in my bedroom, smoke is coming under the door.’ The populist shouts ‘Don’t worry, people don’t get killed in fires, that’s just fake news spread by the fire safety industry.’ The liberal screams up: ‘For God’s sake, don’t listen to that evil person. The fire is going to kill you now. Get out, just run downstairs, just do anything to get out of the house.’ The man in the burning house listens to the liberal, goes out of his bedroom door and is promptly burned to death. Ten minutes later a sensible person arrives and explains that if they had called the fire brigade and told the man to stay put, then by now the firefighters would have arrived, and they would have got the man out of the building unharmed. The moral of the story is that even when someone’s house is on fire, telling people to panic does not make you better than the opinionated idiot.
The unprecedented nature of the global COVID-19 lock-down is not ultimately the result of ‘the science’ but the direct result of conflicts in the West between different interest groups who have coalesced around the rival banners of the left-liberal and the conservative-populist. What may make sense in terms of political rivalry in the West makes hardly any sense to those in the developing world who face very different conditions. To illustrate the point, we need to look at the work of Neil Ferguson and his team at Imperial College London who have been hugely influential in promoting the need for a world COVID-19 lock-down.
THE ROLE OF IMPERIAL COLLEGE
Most of the world first became aware of this team’s work with their paper of the 16th March. This gave frightening figures of up to 500,000 COVID-19 deaths in the UK and 2 million in the US (1). In the developing world, the figures get even worse with Ferguson’s team calculating a figure for 7,687,000 deaths in South Asia if nothing is done to prevent the spread of COVID-19 (2). The only way to prevent vast death tolls is the ‘suppression’ or lock-down strategy, according to Ferguson’s team. The lesser strategy of ‘mitigation’ will still lead to huge death tolls. Ferguson suggests in a 25/04/2020 interview with ‘Unherd’ that once the number of new cases is low enough in a given country the lock-downs could be alleviated a little bit. This depends on the use of the South Korean style tracing and isolating of the contacts of anyone who tests positive for COVID-19, however (3). His team admits that many developing countries will not have the infrastructure to do this (4).
Ferguson is probably motivated by a desire to push reluctant populists like Trump and Boris Johnson into action. The promotion of world panic, however, will be deeply harmful to the developing world. Even richer developing countries tend to have fairly limited welfare coverage for the unemployed. The rise in unemployment in Brazil up to 13.7% in 2017 led to 31,415 deaths according to one study (5). It is less clear if recessions in the West cause higher overall mortality (6).
Ferguson’s work suggests that if developing countries cannot trace and isolate they must lock-down totally until a vaccine has been found or let economic activity carry on and suffer a huge death toll. Given, no vaccine for a coronavirus ever has been found his sounds like he is prescribing economic suicide for the Third World. The most likely outcome, though, is that developing countries cripple their economies with lock-downs which end because they are unsustainable but only after they have done fairly severe economic and social damage.
SWEDEN-THE VALUE OF A GOOD EXAMPLE
The fact is that Ferguson’s figures are almost certainly wrong, and because they are wrong they are leading to the wrong policy decisions. Ferguson is totally incorrect about likely death tolls, and this means that the draconian solutions to Covid-19 he recommends do not need to be adopted. This is illustrated very well by the experience of Sweden. According to Ferguson’s model, Sweden should be in the middle of a huge mortality crisis at the moment, but this clearly is not the case. Now if the figures for Sweden are wildly inaccurate, then the same will be true for other countries. This is because, as we shall see, Ferguson uses the same evidence base for calculating the predicted death toll for every country in the world.
The example of Sweden is useful purely because it shows that Ferguson’s figures indicating millions and millions of COVID-19 deaths are exaggerated and that governments in the developing world should adopt less draconian measures than he suggests.
Sweden has not insisted on closing schools (at least up to age 16), shops, bars and so on. However, the Swedish government has recommended such measures as working from home where possible and avoiding non-essential travel.
Ferguson, in turn, has insisted that this means Sweden will have a higher death rate than countries that lockdown. He argues that the reproduction number for COVID-19 in Sweden will be higher than in countries like the UK. The reproduction number is the number of people that each infected person will infect in turn. Thus a reproduction number of 2 means each infected person infects two more people. The Ro or ‘r nought’ number is the basic reproduction number of the disease before any vaccination, or any immunity has developed or suppression measures have started. According to the graph in the Ferguson team’s March 30th paper, Sweden’s basic Ro was about 3.9. The paper goes onto show that the reproduction number came down to 2.64 from 3.9 by March 28th. Ferguson’s team states that his reproduction number estimate for Sweden of 2.64 is a lot higher than his estimates for countries like the UK (about 1.2) because of Sweden’s less restrictive approach (7) .
A reproduction number of 2.64 is a frightening number, implying that every infected person infects 2.64 other people. As so many COVID-19 cases are unreported, it makes sense to calculate reproduction rates from death rates, not reported cases. Although there is a time lag between death and infection, once the deaths begin, they should also be increasing by 2.64 times every serial interval (note that Ferguson tends to assume death rates remain constant.) The serial interval is the time between one person showing symptoms and the person they infected showing symptoms. Ferguson in his March 30th paper, assumes this is 6.5 days. Although Ferguson gives a 23.9 day time lag between infection and death it would be clear that past infections would be leading to deaths now that would be increasing by the amount of the reproduction number in the past. This would have to be either the current reproduction number of 2.64 or somewhere between the Ro number of 3.9 he gives and the later reproduction number of 2.64. .
In order to see whether this is the case, I used the daily death rate from the Worldometer site to calculate death rates for 6.5 day long time periods for March 30th to 1st May (8). This is done by a crude method as a ‘reality check’ on Ferguson, it is not meant as a precise calculation. I simply count 6 days deaths from the entry for Sweden on the Worldometer Coronavirus website. I then divide the next day’s toll in half and add one half to the first period’s figure and the other half to the next. Thus the end day of one period also appears in the next period with half of the death toll for the day counted in each. I simply divide the figure for 1st May in half and add it to the total for the other days in the period.
The figures are:
30th March-April 5th: 277
5th April-11th April 500: (1.8 times higher than previous period)
12th April-18th April: 568.5 (1.137 times higher than previous period)
18th April-24th April: 696.5 (1.225 times higher than previous period)
25th April-1st May: 467.5 (0.67 times the figure for previous period)
Clearly, we are nowhere near a 2.64 reproduction figure here.
Ferguson in his 25th April interview with ‘Unherd’ seems to recognise this. He claims that in actual fact the Swedes are pursuing a suppression strategy but only a half-hearted one and then claims that their reproduction number has come down to 1.4 or 1.5 as a result. It is difficult to see how a country where schools and bars are open is pursuing any kind of suppression strategy like that of the UK or France. In addition, there have not been any massive changes in the restrictions in Sweden between the Ferguson team paper of 30th March and his interview of 25th April. Really, there is no ‘suppression’ or lock-down in Sweden. The fact is the death rate, or the reproduction number he calculated in his 30th March paper must be wrong. Either that or his beliefs about the prolonged severity of restrictions necessary to bring down the reproduction number must be incorrect. Ferguson is trying to claim that Sweden does not show there is any need for him to revisit his original model whereas in actual fact the experience of Sweden clearly provides good evidence that his model lacks good predictive power.
Also, the reproduction number of 1.4 or 1.5 that he now suggests is looking wrong too. By 25th April there was no sign that Sweden’s reproduction number was so high and this is with death rate figures with a more than three-week time delay when used to calculate the infection rate.
Despite the fact this is only a rough reality check, it should be clear that the reproduction number in Sweden is not anything like 2.64 or even 1.5 since the period ending 12th April. Assuming last week is not an anomaly it may even have gone below 1. Ferguson’s 25th April argument that deaths in Sweden will just go on and on with this reproduction rate and they will never reach herd immunity frankly seems to defy arithmetic and evidence and is very likely to prove false.
There are all sorts of caveats about the figures such as the gap between death due to COVID-19 and the death being recorded/announced by the state. It must be remembered that we are just looking at the rough trend in deaths here as a basic reality check on Ferguson’s figures for Sweden. Gaps between death and the announcement of death may mean the death figures here reflect infections slightly more than 23.9 days old. I anticipate the objection that Sweden is not counting large numbers of COVID-19 deaths by pointing out that excess death statistics are pretty close to the number of official Covid-19 deaths (9).
THE ERRORS OF NEIL FERGUSON
The obvious question is why Ferguson’s predictions are so wildly out. Ferguson calculated his original reproduction number for Sweden and the other countries in his March 30th paper by using the infection fatality rate for COVID-19. He takes figures for the number of deaths in each country and uses the infection fatality rate (IFR) to work out the number of infections. Ferguson’s team calculated the IFR by using infection figures for people repatriated from China during the outbreak who were tested on arrival in their destination country. They then compared these figures with known numbers of deaths in China to arrive at an infection fatality ratio of 0.66% (10). This ratio of 0.66% is then used to calculate fatality rates in other countries, including Sweden and the UK, by adjusting for the age profiles in the different countries. I would suggest that this one sample of international travellers is made to do a lot of work in the various papers Ferguson works on. It is used to come up with the figures for deaths and reproduction numbers across the world mentioned above, and a total figure of 40 million deaths worldwide if nothing is done (11).
The experience of Sweden falsifies all of Ferguson’s gloomy predictions based on this figure. Clearly, other models using other population samples need to be looked at rather than just relying on the sample estimate for China of 0.66%.
One important point is that herd immunity is calculated as 1-1/Ro. This means that if Ferguson has overestimated Ro, then he is also wrong in his gloomy predictions about a huge percentage of the population needing to be infected before herd immunity is reached. The Ro number for the 1957 flu epidemic in the UK was 1-5-1.7 (12). The Ro for the 2009 swine flu epidemic was 1.2-1.6 (13). A Ro number of 3.9, which is approximately Ferguson’s estimate for Sweden, means that 74% must be infected or vaccinated before herd immunity takes place. Let us say the actual Ro number for Sweden was 1.6. This means that only 37.5% must be infected or vaccinated before herd immunity takes place.
WHAT SHOULD BE DONE?
Once we lose our absolute fear of massive death tolls, we can think of appropriate ways to manage the risk of Covid-19. The action we take has to be proportionate to the risk. Locking down the economies of the developing countries for 2 years while they wait for a vaccine that may never come is clearly grossly disproportionate and damaging. The leaders of countries like India and South Africa should not have let themselves get panicked by Ferguson’s figures in this way.
The example of Sweden, however, shows that developing countries may have been able to avoid the need for starting a total lock-down that they will not be able to sustain. This is not to say that every developing country can and should pursue exactly the same policies as Sweden. Sweden has much higher per capita health spending than the nations of the developing world, so different policies will have to be adopted by these nations.
In addition, not everything in Sweden has been done in a marvellous way. Sweden has let COVID-19 cause a lot of deaths in its care homes. Neil Ferguson in his interview for ‘Unherd’ said that no country had been able to protect its elderly, which he thought proved the point that only a total, and in the case of countries that cannot contact trace effectively never-ending, lock-down would work. Again we see how Ferguson’s useless advice all seems designed to push governments in the direction of unsustainable long-term lock-down when other policies are available. China has been able to implement better policies to protect its elderly. For example, in care homes it quarantines all residents entering or returning to the care home from hospital. It also sends any care home resident showing symptoms of COVID-19 to hospital (14). In contrast in the UK patients with COVID-19 are being sent from hospital to care home and there is no quarantine of care home residents unless they show symptoms (15).
In general, China has not adopted the same laissez-faire attitude to COVID-19 as Sweden has. In addition, it has used technology to contact trace and monitor people in a way that might be difficult for some developing countries. We have to ask how much of a barrier lack of technology really is, however. Not all the measures China used to stop COVID-19 so successfully were to do with technology. First, it put a cordon around Wuhan and some other affected places. It then sent officials door to door to find people with symptoms of COVID-19 and then quarantined them.
This type of short, sharp shock approach in designated areas of a country may be a good way to control an infection reasonably quickly in order to get the economy moving again as soon as possible. Where this is not feasible other measures might be used. For example, families living with elderly relatives could be requested to remain at home until the pandemic peaked. They could have been provided with government financial assistance and food for this period to cover the loss of employment, with local officials checking they did not claim the money and carry on working anyway. This may be an expensive prospect for the government of a developing country but much cheaper than a total country closure.
Striving to infect sections of the population to achieve herd immunity is an anti-people strategy. The correct path for a people’s government would be to give public health advice to try and reduce transmission as much as possible while protecting the elderly and vulnerable as above. Events and venues which are likely to lead to a major spread of the disease such as weddings, night clubs, church services and so on can be cancelled until better times without shutting the whole economy down. The wearing of face coverings on public transport may also be very useful in preventing disease spread.
THE QUESTION OF BRAZIL
It might be objected that we are relying too much on one country for our analysis. The problem is that virtually all other affected countries seem to have gone for some sort of lock-down. The only possible exception is Brazil, where the President and the regional governors are giving contradictory advice. No-one could really recommend Brazil’s messy and chaotic approach to COVID-19 as a model. Based on a very rough calculation, the reproduction number seems to be around 1.6 or 1.7. Again, this is calculated from death rates as many cases of COVID-19 do not get registered, and the death rates are assumed to be more reliable. This reproduction number suggests that Brazil should adopt a different approach to COVID-19, but it is nowhere near the very high reproduction numbers and death rates that Ferguson claims would occur in the absence of draconian policies.
LIBERAL AGAINST POPULIST AND BOTH AGAINST CHINA
The COVID-19 panic appears to be part of a long-running social divide in the First World countries of the West (and Australia, New Zealand, Japan etc.) The relative decline of the old First World compared to the Third World of developing nations has given rise to a sort of nationalist revolt among the people’s of the First World with the focus often on ‘unfair competition’ from ‘cheap labour’ whether this is from immigrants or from imports from low wage countries. Once upon a time, anxiety about wages might have led to trade union struggle by European workers, but now this is a distant memory outside the public sector. Instead, this revolt has led to the election of completely incompetent populist leaders who are unable to handle the COVID-19 issue.
Middle class liberals launched a counter-blow with ‘lives before money’ type slogans, and weak populist leaders were forced to reverse their initial anti-lockdown positions. The whole populist pro-western worker stance and the focus on jobs and wages are undermined, which was always the intention of the liberals. Ultimately the middle class usually find themselves on the same side as the capitalists and are in fact the intellectual foot soldiers of the ruling class. The capitalists will endure short-term pain but also see the lockdown as an opportunity for a shakeout of redundant workers as the economy moves away from retail outlets in high streets and shopping malls and office work to internet shopping and working from home. Furlough payments and loans to small businesses are really only short-term measures to ease the transition. It is rather like the mass factory closures the UK endured in the 1980s where redundancy payments were used to prevent mass resistance to change among most if not all workers. Long term COVID-19 will allow the capitalists to reduce their payrolls, keep wages down, impose greater labour discipline and end the pro-worker rhetoric that has taken over political discourse.
The motivations of these western actors have no counterpart in the Third World. The countries of the Third World cannot afford the luxury of the lockdown and must not swallow western policy prescriptions. Moreover, the populists are likely to try and survive by diverting the blame for their handling of coronavirus to China. They will use this as an excuse to try and undermine the Chinese economy. The support of western liberals and their capitalist overlords for globalisation is based on their enthusiasm for the exploitation of underpaid labour. It does not extend to China or other countries becoming high-tech challengers to western capitalists or taking away the dominant global position of the West. They may well find themselves united with the populists in a crusade against China. An overblown panic over COVID-19 will only fuel such an anti-Third World movement, and overall the developing countries would do well to avoid this tendency.
(1)Ferguson, Neil et al. (2020) ‘Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand.’
(2) Walker, Patrick GT et al. (2020) ‘The Global Impact of COVID-19 and Strategies for Mitigation and Suppression.’ Imperial College London (2020),doi:https://doi.org/10.25561/77735
(3) ‘Imperial’s Neil Ferguson Defends Lockdown Strategy’ https://unherd.com/thepost/imperials-prof-neil-ferguson-responds-to-the-swedish-critique/
(4) Walker et al (2020)
(5) Thomas Hone et al. ‘ Effect of economic recession and impact of health and social protection expenditures on adult mortality: a longitudinal analysis of 5565 Brazilian municipalities’ The Lancet 7, 11, e1575-e1583, November 01, 2019
(6) Lynne Peebles (2019) ‘How the Next Recession Could Save Lives’, Nature 565, 412-415 , doi: 10.1038/d41586-019-00210-
(7) Flaxman, Seth et al. (2020) m’Estimating the number of infections and the impact of non-pharmaceutical interventions on COVID-19 in 11 European countries.’
(10) Verity et al. (2020) Estimates of the severity of coronavirus disease 2019:
a model-based analysis Published Online in The Lancet March 30, https://doi.org/10.1016/S1473-3099(20)30243-7
(11) Walker et al. (2020)
(12) Ferguson, Neil M et al. (2006) “Strategies for mitigating an influenza pandemic.” Nature vol. 442,7101 : 448-52. doi:10.1038/nature04795
(13) Fraser, Christopher (2009) ‘Pandemic Potential of a Strain of Influenza A (H1N1): Early Findings’ Science 19 Jun 2009:Vol. 324, Issue 5934, pp. 1557-1561
(14) International Long-Term Care Policy Network ‘Report from Mainland China: Policies to Support Long-Te r m Care During the COVID-19 Outbreak’ https://ltccovid.org/wp-content/uploads/2020/04/Report-from-Mainland-China-18-April-final.pdf
(15) ‘Admission and Care of Residents During COVID-19 Incident in a Care Home’ Department of Health and Social Care ‘ https://www.gov.uk/government/publications/coronavirus-covid-19-admission-and-care-of-people-in-care-homes