It is a fundamental principle of risk management that the benefits of any decision should be weighed against the costs. What is less often appreciated is that risk is an interactive phenomenon. Perceptions of risk often lead to people adjusting their behaviour in ways that reduce (or increase) the chances of a hazard occurring. Risk does not equal a hazard plus some fixed probability of it occurring.
University College London Professor Emeritus John Adams captured this with his idea of a ‘risk thermostat’, whereby we balance our propensity to take risks, risk perceptions and costs and benefits. Aaron Wildavsky perhaps captured this most parsimoniously in his aphorism that ‘the secret of safety lies in danger.’
But overestimating risk can lead to costly responses. In one famous example the 1976 Swine Flu epidemic in the US resulted in the rushed development of a vaccine that was far worse than the disease, causing many cases of Gullain-Barré Syndrome, a rather nasty immune system disorder. This was described in detail by Robert Formaini in his excellent book The Myth of Scientific Public Policy. Unfortunately, all too often, governments do not understand that public policy cannot be decided solely by scientific experts, because human conduct frequently undermines the intentions of policy-makers by exhibiting behaviour scientists did not anticipate.
The appropriate medical field for responding to pandemics is epidemiology, not immunology. As Nobel Prize winning immunologist Peter Doherty put it in an interview on Sky News on March 26, he’s a ‘lab guy’, not an epidemiologist. Good epidemiology should, of course, include a good measure of social science to cover this eventuality, but I once attended a conference on climate change where a prominent epidemiologist disavowed any knowledge of human behaviour, stating ‘I am not a social scientist.’ He went on to argue that the health of future Aboriginal Australians would be seriously harmed by climate change — ignoring the rather obvious point that the impacts of future climates projected by models rested on emissions scenarios that assumed massive increases in wealth; apparently none of this would find its way to Aboriginal communities, which would continue to wallow in squalor.
Which brings us to the current pandemic with the novel coronavirus, or COVID-19. The government policy response has seen some sensible measures adopted, especially in shutting down borders and requiring a degree of ‘social distancing’ in an effort to flatten the curve of infections sufficiently to allow the health system to cope. But has the National Cabinet gone too far?
The measures announced on the evening of Sunday, March 22, went beyond the sensible maintenance of social distancing at cafes, restaurants, pubs and clubs while throwing many of their employees on the unemployment scrapheap, joining those already put there by the collapse in the tourism sector. Estimates of unemployment running at a million people have been floated.
In announcing these measures, Prime Minister Morrison specifically mentioned the flouting of the restrictions on social conduct by the thousands who flocked to Bondi Beach the day previously. Yet the functioning of a National Cabinet ensured that they were applied equally here in Tasmania, where there were then fewer than 20 cases — all contracted by those travelling overseas. Policy was decided on the basis of the activities of sybaritic Sydney sun worshippers.
The impact on the hospitality sector alone has been enormous, and has a human face. To give one example, a café I frequent is struggling to keep operating by offering a takeaway service, but it usually provides work for dozens of people. These people are long-term unemployed, many with disabilities, who gain the work experience and skills that are vital to getting a foothold on the employment ladder. They now have no work, and if the café folds, that situation will continue.
But, we have been assured by the medical experts, this is all necessary because we are on the same tragic trajectory as Italy, which very quickly overtook China in total fatalities. The Victorian state government has been pushing in the National Cabinet for tougher measure, and has indicated it was willing to go it alone. On March 26, Premier Daniel Andrews warned Victoria could not wait for everyone else to catch up. ‘If we did that . . . we will look like Italy,’ he said.
But is that inevitable? Indeed, is it even so?
The Italian Case
One of the very challenging problems for policy-makers has been the lack of reliable data. Not only did the autocratic Chinese regime assist the speed with which this virus spread from bats in the wet markets of Wuhan, arresting those doctors who tried to sound the alarm and exacerbating the global crisis, but it has inhibited the flow of reliable data. Early on, statisticians pointed to a co-efficient of determination (R2) of 99 per cent between two series of data the Chinese official were reporting. Epidemiologists quickly noted these were not credible, and the data were clearly made up. And in late March the South China Morning Post reported that they had found another 40,000 cases down the back of the sofa.
The failure to provide the rest of the world with reliable information on the genetic characteristics of the virus and data on its spread and mortality is a further stain on the reputation of the Chinese regime.
Italy has provided more accurate data, but it requires close examination, because it suggests Australia is unlikely to follow its trajectory.
The impact on Northern Italy has indeed been alarming, but it needs to analysed carefully. An aging population and substandard health system, with fewer acute care beds than the European Union average could explain the mortality rate — but what explains the infection rate?
In point of fact, those factors don’t explain the mortality rate, and neither do air pollution and a heavy incidence of cigarette smoking, though all these no doubt contribute. Former Israeli Health Minister, Professor Yoram Lass has pointed out that Italy has more than three times the morbidity from respiratory diseases than any other European country. The data indicate that COVID-19 is actually less dangerous than media reports suggest. Italian Professor Walter Riccardi has estimated that only 12 per cent of death certificates in Italy have shown a direct causality from coronavirus, in contrast to media reports where all the people who die in hospitals with the coronavirus are deemed to be dying of the coronavirus. (In fact, it is usually the secondary infection from pneumonia that causes death).
This arises from the relationship between the coronavirus and other ailments. Comorbidities are the key. Italian data show that only 1 per cent of those dying have no other ailments, and almost 50 per cent have three or more. In fact, data show that morbidities in the 65+ cohort have been lower than usual this winter, following two relatively mild influenza seasons which have left more of the vulnerable exposed to COVID-19. Riccardi’s point is that most Italians have died with COVID-19, not because of COVID-19. But why are there so many cases in Italy — an in Northern Italy at that?
The answer lies in the economics of the textile and fashion industry. As one writer for Bloomberg put it, “Today, in the textile towns of northern Italy, ‘made in Italy’ often means made by Chinese workers in Chinese-owned mills.” Northern Italy has a huge Chinese immigrant population, many of whom travelled from Wuhan. Chinese New Year was celebrated from January 25 to February 8, and many of these Chinese immigrants went home to Wuhan to celebrate the festival with their families.
Italy has direct flights from Wuhan. When they returned after their leave, the migrants were not tested in Italian airports and their movement was not restricted in any way. When Prime Minister Conte suggested stronger action, these measures were called ‘fascist’ by the Left. When the public displayed some hostility to Chinese migrants in February, the Mayor of Florence launched a ‘Hug a Chinese’ campaign, egged on by Chinese supported social media campaigns, which featured videos such as that of a man of Chinese extraction holding a sign reading ‘I am not a virus – I am a human being.’ (see the clip above)
The Comorbidity Issue
The hedge fund Rebellion Research has provided excellent research on this, and the Italian National Institute of Health has published extensive data, but it is not clear that it has been appreciated by policy-makers in Australia and elsewhere. Certainly, as far as the United Kingdom government is concerned, their lock-down was provoked by modelling from Imperial College London (based on the Wuhan data) that estimated the best case for the UK was 250,000 deaths and that it could rise to 500,000. What government wouldn’t react to such results? They decided to try to restrict the impact to 20,000 deaths.
Problem was, the modelling failed to account for comorbidity, and the fact that there are around 600,000 deaths each year in the UK. Many of the coronavirus deaths will be those harvested early that would occur this year regardless, those with very short life expectancies. There will probably be few excess deaths, as is the case in Italy. It is the case that they will be concentrated in a shorter time period, and that is the most urgent policy problem.
As noted above, the Wuhan data are also not reliable. Even if we could trust Chinese government statistics, infection rates are hard to determine accurately. About 80 per cent of those tested are asymptomatic, but then there are reports that testing kits can also yield false positives. The extra 40,000 Wuhan cases were apparently asymptomatic, but we have little guidance as to what the denominator is to calculate mortality. This uncertainty has contributed to the costly responses by governments and the fear panicking markets. (Thank you, President Xi and the CCP!)
Some better analysis has been provided by John Ionannidis, an epidemiologist and biostatistician at Stanford University. Ionannidis is something of an iconoclast, having published research showing that most published medical research later proves to be wrong (often because of noble cause or more venal corruption). Ionannidis observed that the Case Fatality Ratio (CFR) among passengers and crew on the Diamond Princess, the only example where an entire, closed population was tested, was 1 per cent. This, however, but was a largely elderly population, much more susceptible to mortality. Extrapolating this to the age structure of the US population, he estimated the death rate among people infected with Covid-19 would be a mere 0.125 per cent. But he acknowledged this was based on a small sample size of about 700 people, and that the real death rate could range from five times lower (0.025 per cent) to five times higher (0.625 per cent). And then there might be delayed deaths, so his ‘reasonable estimates’ for the CFR in the general U.S. population vary from 0.05 per cent to 1 per cent.
Ionannidis’ analysis was supported by the Japanese National Institute of Infectious Disease, which found that despite the high average age of passengers on the Diamond Princess, 48 per cent of all those testing positive remained completely symptom-free. Even among the 80-89 year cohort, 48% remained asymptomatic, while in the 70-79 cohort 60% were asymptomatic.
There are obviously many who are infected but remain unsymptomatic. This is how the virus spreads so effectively, but it inflates morbidity rates based upon confirmed cases.
Is it a case of TINA?
Margaret Thatcher was fond of the saying ‘There is no alternative’ (TINA). Decision-makers certainly seem to have been advised by medical experts that COVID-19 is a case of TINA, and measures that have closed down businesses and thrown people out of work — one hopes only temporarily — have been added to the havoc wreaked by the markets.
Is this really the case? The real problem seems to be that, left unchecked, COVID-19 will overwhelm health care (and funeral) sectors, even if the excess deaths over the year are likely to be modest. There are experts who argue that an effective response could come at lower cost — both economic and social. One such is Dr David L. Katz, founding director of the Yale-Griffin Prevention Research Center.
Writing in the New York Times on 20 March, Katz posed the question few others have done in the media: ‘Is Our Fight Against Coronavirus Worse Than the Disease?’ He goes on to suggest another, less costly approach to flattening the curve. Employing the war analogy he suggests that there are two options: the inevitable carnage and collateral damage of diffuse hostilities, similar to trench warfare and massed attack used in World War I; and a precision ‘surgical strike’ targeting the sources of danger, like a special forces operation. He noted that ‘The latter, when executed well, minimizes resources and unintended consequences alike.’
Noting the characteristics of the disease (perhaps 80 per cent of infections asymptomatic, average age of those dying about 80, comorbidity, low death rate among the young), Katz recommended focusing isolation and resources (including testing) on the elderly and vulnerable. Rather than a lock down of the education system that would send the young (likely to be asymptomatic carriers) to be shut up at home with parents and grandparents at greater risk, he advocated keeping it open, along with business as much as possible. Those infected with mild symptoms would acquire immunity — limiting their ability to pass it to others. Limited resources would thus be devoted to those most at risk, and those who display serious symptoms or develop pneumonia.
Katz is hardly a voice in the wilderness. There are numerous other experts taking a different line to that of the experts advising governments.
German virologist Hendrik Streeck holds that Covid-19 is unlikely to increase total mortality in Germany (normally about 2500 people per day or 912,500 annually). Professor Sucharit Bhakdi, an expert in medical microbiology, says there are other important factors at play, notably pre-existing health conditions and poor air quality in Chinese and Northern Italian cities, describing the imposed measures as ‘grotesque’, ‘useless’, ‘self-destructive’ and a ‘collective suicide’ that will shorten the lifespan of the elderly and should not be accepted by society. Danish researcher Peter Gøtzsche believes that Corona is an ‘epidemic of mass panic’ and that ‘logic was one of the first victims.’ German Professor Karin Moelling has stated that Covid19 is ‘no killer virus’ and that ‘panic must end.’
Pietro Vernazza, a Swiss infectious disease specialist, argues that many of the responses are not based on science and should be reversed. Vernazza considers mass testing makes no sense because most of the population will be asymptomatic, and lockdowns and closing schools are even counterproductive. Like Katz, he recommends focusing on protecting at risk groups while keeping the economy and society at large undisturbed.
Former Israeli Health Minister, Professor Yoram Lass, maintains the new coronavirus is less dangerous than the flu and that lockdown measures are likely to kill more people than the virus. He added that ‘the numbers do not match the panic’ and that ‘psychology is prevailing over science’. Professor Stefan Hockertz, a German immunologist and toxicologist, has stated that COVID-19 is no more dangerous than influenza, and that the fear and panic created by the media and the reaction of many governments were greater dangers. Argentinean virologist and biochemist Pablo Goldschmidt agrees.
Dr John Lee, a retired pathologist, presented an analysis similar to that here in The Spectator, as I was completing this essay, adding the important point that the many varying infection and mortality rates between countries are probably an artefact of the way in which statistics are collected. COVID-19, for example, is a notifiable disease in the UK, but influenza is not and respiratory deaths are usually simply reported as ‘pneumonia’ or ‘old age’. COVID-19 is thus more closely monitored and reported and is more likely to be found in hospitals than among the general public.
While COVID-19 is a notifiable illness and the reason for an extensive lockdown in the UK, however, on 19 March it was removed from the list of High Consequence Infectious Diseases on account of its more information was now available ‘about mortality rates (low)’, greater clinical awareness and wider availability of s specific and sensitive laboratory test. If we want to compare COVID-19 with other recent respiratory illnesses, it is instructive to consider thse disease still on the HCID list: Avian influenza A H7N9 and H5N1; Avian influenza A H5N6 and H7N7; Middle East respiratory syndrome (MERS); and Severe acute respiratory syndrome (SARS).
The Missing Social Science Perspective
As it happens, there were some social scientists involved in the development of advice to the National Cabinet, but their counsel was over-ruled by the natural scientists on the panel convened among Group of Eight universities at the request of the Chief Medical Officer. This panel advocated even more extreme measures, including the closing of schools, but it rejected the suggestion of a regional and staged response, tailored to the circumstances of each state (such as whether there was yet community transmission). The Go8 reported that
Those who argued for a more nuanced approach came at it not as epidemiologists and public health experts fearing an Italian situation but as social scientists deeply concerned about the financially and mentally vulnerable, the lonely, and those about to become much lonelier as the restrictions and lockdowns become speedier and more severe.
Again, the view that epidemiology should not contain social science! And the public health experts were ‘fearing an Italian situation’ that they clearly did not understand fully. And were only the ‘social scientists deeply concerned about the financially and mentally vulnerable’? It is uncharitable to take that interpretation, but that is the way the Go8 put it, and the majority of the group did reject that ‘more nuanced approach.’
It is a brave political leader who fails to follow scientific advice in the present climate, but by failing to consider critically that advice, seemingly coming through single, official channels, and to undertake a proper risk evaluation that includes social and economic factors, they have done more harm than they need have done. They have done well on the basics of hand washing, basic social isolation, and so on, but they have inflicted greater harm than they need have if they had followed the advice of Dr Katz.
In the US Donald Trump is being excoriated for daring to consider the costs by those who consider risk management to be solely a matter for natural scientists, and to be decided without regard to cost. He is one of the few leaders (if not the only one) who is including some consideration of the human and economic costs of the responses dictated by the medical experts who have no expertise in risk management as it should be practiced.
And regardless of our opinion of him, he deserves credit for that.
Aynsley Kellow is Professor Emeritus of Government at the University of Tasmania. He is the author of Science and Public Policy: The Virtuous Corruption of Virtual Environmental Science and Negotiating Climate Change