Run the Numbers, Survey the Folly Peter L. Swan
As of May 22, Australia had suffered 7,088 cases of COVID-19 and 102 people who died with the virus, the majority were males between 70 and 89. A sizeable proportion of those admitted to hospitals’ intensive care unitss were suffering from comorbidity issues such as cardiac disease and diabetes. Australia’s Chief Medical Officer (CMO) has told a Senate Inquiry that the Australian government’s actions in locking down the economy saved 14,000 lives.
In the meantime, Camilla Stoltenberg, director of Norway’s public health agency, has confessed:
Our assessment now, and I find that there is a broad consensus in relation to the reopening, was that one could probably achieve the same effect – and avoid part of the unfortunate repercussions – by not closing. But, instead, staying open with precautions to stop the spread.
The cost to the Australian economy of the global pandemic could be as high as one thousand billion dollars with an additional direct cost to the taxpayer of $260 billion this year alone. It will be a while before our Prime Minister and the premiers admit the lockdown was entirely unnecessary and unjustified.
- Rampant alarmism
With the support of 289 or more top economists, four economists, Edmond, Hamilton, Holden and Preston (2020) deny “that there is a trade-off between the public health and economic aspects of the crisis.” I answered this here. Two of the four, Richard Holden and Bruce Preston estimate that without lockdowns and similar interventions, but presumably with voluntary social distancing, 90 per cent of our population of 25.5 million would have resulted in 225,000 deaths, based on an assumed fatality rate of 1 per cent, and yielding an incredible rate of 882 deaths per 100,000 residents.
The claimed deaths saved are higher than argued by Australia’s CMO by a factor of 16 times. The same methodology yields nearly three million COVID deaths in the U.S. and seventy million deaths globally. But is this simply fanciful alarmism and fearmongering designed to frighten all Australians into acceptance of continuing lockdowns and restrictions on normal life, or is it an astute economic analysis?
If one were to accept the Holden/Preston analysis, with an assumed valuation of A$4.9 million per whole of life, the cost of not locking down is placed at A$1.1 trillion. As Paul Frijters has pointed out, deaths are largely confined to the elderly and infirm with very low life expectancy, reducing this outlandish estimate by 95 per cent to a slightly more plausible A$55 billion. Applying the same methodology to the CMO’s more conservative estimate yields a more modest figure of $3.4 billion in comparison with the Commonwealth’s outlay in excess of $260 billion. This is an expenditure of $76 dollars for every dollar saved!
These current Holden/Preston death rate estimates are extreme even relative to discredited March 16 forecasts made by Neil Ferguson of Imperial College, London, of 510,000 UK deaths and 2.2 million US deaths. These forecasts, with little if any scientific justification, prompted massive lockdowns in both the Britain and America, and in much of the rest of the world including Australia.
As of the date of writing, the UK had experienced 34,600 deaths (54 per 100,000) associated with the virus and approaching 100,000 in the U.S. (30 per 100,000), with the global total 340,876 as of May 22, a very modest rate of 3.7 deaths per 100,000 so far. This is less than 0.5% of the deaths projected by Holden/Preston. Why are these death rates so much higher than Australia’s? The U.K., with tens of millions of new arrivals in the last few months, is only now commencing mild quarantining arrangements and similarly, unlike Australia, quarantining arrangements and bans on all non-resident entry in the U.S. did not take place.
By contrast, the far worse Spanish Flu pandemic of 1918-19 that killed 50 million people worldwide resulted in 12,000 deaths in Australia with only quarantining and social distancing and no lockdowns. This represented a death rate of 236 per 100,000 — 0.59 per cent of the 40 per cent of the population that caught the flu. This death rate from by far the worst pandemic to hit the globe in over 100 years is benign compared to the Holden/Preston figure of 882 COVID-19 deaths per 100,000, a factor 3.7 times higher. By contrast, the Spanish Flu was indiscriminate — killing the young and healthy as well as the vulnerable. By contrast, COVID-19 casualties are largely confined to the elderly and those with severe co-mobility factors.
- International quarantining seems to be effective in the short-term
What this analysis does not recognize is that, like Taiwan, Australia is an island nation. While both nations have similar populations, Taiwan is much closer to the Wuhan origin and epicentre of the virus. Taiwan began testing arrivals from Wuhan for flu symptoms from December 31, 2019, but Australia did not restrict access from China until February 1, with a further ban on all non-resident entry introduced on March 20.
Today Taiwan has 441 cases and seven deaths. Hence their infection rate was 6 per cent of ours and their death rate 7 per cent. Moreover, their far superior result, at least from a short-term perspective, was achieved simply by strict quarantining, banning of cruise ships, testing and tracing of carriers with voluntary social distancing, and handwashing. As a result, their economy is suffering less than many others, with their exports expanding.
The Australian death rate is 0.4 per 100,000 residents, compared with 51 for the UK; France 43; and the US, 29. While it is likely that the 100-fold higher rate in Europe could be partly due to higher population density and more adverse climatic factors inclusive of winter, a more plausible explanation is the ability of Australia to impose quarantining earlier and more effectively than in Europe.
- The ineffectiveness of lockdown
Despite the slowness with which Australia closed its borders and mindful of the Ruby Princess debacle, infections in Australia peaked prior to the introduction of severe lockdowns and domestic border closures that were always completely unnecessary. The latter were never supported by the Australian government on medical or other grounds and seem in contravention of Section 92 of the Constitution. In the United States, the most common comprehensive social distancing policy adopted is a shelter-in-place order (SIPO) which requires residents to remain in their homes for all but essential activities.
Dave, Friedson, Matsuzawa, and Sabia analyse the difference in the infection rate and death rate between the 42 states which imposed SIPO laws and the eight Republican states which did not. The study finds that lockdown was effective in reducing the number of COVID-19 detected cases by 44 per cent but were not able to estimate a statistically significant lower death toll, other than noting that it appeared to be lower.
The study is unable to explain the much better record of some states, such as Florida, that did not embrace lockdowns as opposed to states, such as New York, which did. To date, New York City with lockdown in place has suffered 16,565 confirmed COVID-19 deaths and an additional 4,749 probables, which is a rate of 113 per 100,000 and way exceeds even the rate experienced by Belgium, the country which is worst in the world. Florida has suffered 2,233 deaths, a rate of 11 deaths per 100,000, and while still worse than Australia’s performance it is good by comparison with many similar US states that did lock down.
The differences are in part due to philosophy. New York’s Governor Andrew Cuomo declared a universal policy: “We need everyone to be safe, or no one can be safe,” in locking down all “non-essential” activity, despite not closing either the airports or mass transit. COVID-19 sufferers in New York were forced to return to age-care homes to free-up beds in hospitals. As a result, many thousands of the elderly died in these infected homes.
By contrast, Florida Governor DeSantis prohibited visitors to nursing and age-care homes and urged the most vulnerable to stay home. Florida is both more humid and less congested than New York City, perhaps accounting for some of the vastly lower death rate. But the far older population of Florida retirees makes the state even more vulnerable than that of New York City, where very high proportion of the deaths were confined to the elderly.
In particular, the very nature of lockdowns seems to be designed to promote more deaths in certain areas. People confined in cramped and poorly ventilated indoor spaces with insufficient exercise opportunities and limited access to sunlight are naturally going to become more vulnerable to flu-like viruses. This was made worse in Australia as nearly all hospital beds were emptied for COVID-19 sufferers who never showed up. Over 7,000 respirators have been provided but less than 20 ever used. Was this the precautionary principle in action or simply panic?
Contrast confinement on a cruise ship and being free to enjoy healthy beach and ocean exercise in warm sunlight. Huge resources in the form of police helicopters, multiple police boats, high fencing along much of the coastline and shores lined with police officers, were used in a hugely expensive and futile attempt to drive socially distanced swimmers and surfers out of the water in many parts of Australia.
Perhaps in a sustained effort to protect prevalent dolphins from the threat of COVID-19, socially-distancing surfers were driven from the water at McKenzie’s Bay, near Bondi, on numerous occasions by police presumably animated by the need to protect animal rights.
Another study, by Brzezinski, Deiana, Kecht, and Van Dijcke (2020), using a panel dataset based on 40 million smartphone devices across the US, combined with detailed data on state- and county-level government policies, finds that compulsory lockdowns increased the time spent at home by up to 39 per cent but voluntary social distancing also increased the time spent at home. They find that higher socio-economic groups were more likely to respond to physical separation advice and recommend voluntary distancing as being far more cost-effective than lockdowns. However, they do consider stronger measures in areas where voluntary response is problematic.
Hence, like Norway, the evidence from Taiwan and the US does not indicate that lockdown is either necessary or is more effective than voluntary social distancing in combination with protecting the vulnerable, in reducing COVID-19 morbidity. Two other South-East Asian countries, South Korea and Japan, urged voluntary social distancing and introduced strict tracing and testing but did not enforce lockdown. South Korea has suffered 263 deaths so far, a rate of 0.51 per 100,000 — comparable to Australia but perhaps with a lower economic cost; Japan has suffered 749 deaths, a rate of 0.6 per 100,000, and also comparable to Australia’s with lower social cost.
Edmond, Holden and Preston cite the absence of mandatory lockdown in Sweden as an indication of what might have happened in Australia if we had followed the same path.
On the face of it, the Swedish experience does not look spectacular in the short-term, with 3,925 deaths to date or a rate of 39 per 100,000 residents. While this is far lower than many other European countries that did lock down, such as the UK and France, and less than half that of Belgium with its rate of 81 per 100,000, it is far higher than Australia’s and even its neighbour, Denmark, which not only locked down but also instigated severe quarantine restrictions.
What accounts for the apparently poorer performance of Sweden relative to its near neighbour in the very short term? Most likely, it has little to do with the difference in lock-down policies but could be accounted for by the fact that, unlike its neighbour, it kept its borders open while at the same time, and like Australia, it provides generous public subsidies to displaced workers and firms. Hence it is not surprising that its economic costs are not obviously that much lower than Australia’s, and the initial death rate at least, higher due to the absence of quarantining.
Hence the evidence indicates that Australia’s initially excellent COVID-19 record so far is due to relatively early quarantining of international borders and voluntary social distancing. Compulsory lockdown came late in the piece and has made very little difference other than greatly increasing economic costs and adding to the feeling of social isolation, increased threat of family violence, and the appearance at least of home detention.
It is true that people could venture out for exercise so long as they stayed confined to their local area and did not sit down in the park or anywhere else, while denied access to beaches, oceans, pools, gyms, holiday homes, hotels, pubs, restaurants, community sport, and all forms of live entertainment and travel.
How easily are people convinced that the government has the right to remove most of their freedoms on no more than a whim!
- Restrictive measures not justified on medical grounds seem to be there to frighten us
Most of the unpopular measures such as school closure, closure of beaches and water access, bans on sitting on park benches while still social distancing, bans on golf and fishing, bans on entry to a number of states, etc., were instigated by some state premiers contrary to medical advice provided to the Commonwealth Government. CMO, Dr Jeanette Young, stated: “While evidence showed schools were not a high-risk environment for the spread of the virus, closing them down would help people understand the gravity of the situation”. In other words, people needed to be frightened “for their own good”, a task which Holden/Preston have taken up with alacrity.
Can one argue that this peculiar medical advice to close safe schools was an improvement on the frightening medicine provided to patients in Florence during the plague of 1629 consisting of “oil of crushed scorpions in Greek wine”? Closing schools has contributed to the massive economic burden on the states’ economies and Australia’s rapid decline in educational standards, while the scorpion diet was inflicted on a relatively small number of survivors of a pandemic that occurred four hundred year’s ago.
- Incremental cost of lockdowns over voluntary distancing
Estimating the cost of compulsory lockdown in Australia, state border closures and bans on travel is difficult as voluntary social distancing would have reduced domestic travel and use of venues such as dining in restaurants, clubs, and cinemas, in any case. International travel restrictions would have reduced revenues for the university sector from foreign students and the international tourism industry would still have suffered.
Of course, much of this extreme voluntary social distancing was doubtless excessive and not warranted on medical grounds but was promoted by extreme alarmism and fearmongering from media, academic economists, public servant and all forms of government. Those sections of the community not immediately subject to job loss.
Holden/Preston assume a 10 per cent fall in GDP in just one year, 2020, with no long-term effects and a very conservative cost of the COVID-19 crisis of $180 billion, but they argue, inconsistently, that this grossly overestimates the cost of shutting down, as voluntary social distancing would have been costly even without forced closure of much business activity. Might not voluntary social distancing have saved 225,000 lives even without lockdown?
My Colleague Paul Frijters argues persuasively that the true cost of the global pandemic on Australia is likely to be six times higher than that provided by Holden/Preston, making it $1,080 billion.
Holden/Preston, having already argued that voluntary social distancing was going to cost 225,000 lives without forced closure, it is hardly consistent to argue that half the total cost costs would have been incurred anyway. Hence, according to the Holden/Prescott analysis, we are left with a net cost of $90 billion from the shutdown but no sign of any benefit as voluntary social distancing, international quarantining, testing, and tracing would have done the job anyway at far lower cost. Their entire $1.1 trillion, or more credible $55 billion, benefit from lockdown is vacuous.
- The payment of $260 billion to stop people working is far from costless
Holden/Prescott then argue that a $320 billion (16.4 per cent of GDP) cost of supporting the economy during the compulsory shutdown is not really a cost but simply a transfer of resources from one part of society to another. Since it is paid for initially by borrowing, the immediate impact is on the young and future generations. While this figure comes from Treasurer Josh Frydenberg, a $60 billion overstatement in the Government’s JobKeeper estimate means that it is actually lower, $260 billion, once this error is corrected.
So we learn that this sizeable boost to the social security budget to pay about three and one half million people to quit work and perhaps do nothing at home comes at no cost, other than “the economic distortions coming from raising the revenue to service the spending.” In fact, we learn from Holden/Prescott that this cost is likely to be “tiny” as it will simply raise our international debt obligations at a time when real interest rates are low.
Suppose we have two otherwise identical workers with each creating equal value of $1,500 per fortnight, but one is employed in an “essential service” and the other is not. The latter is placed out of work by shutdown orders and receives a JobKeeper payment of $1,500 per fortnight. Not only does GDP fall by $1,500 per fortnight due to the forced lockdown but the essential worker is now taxed $1,500 per fortnight to support the now unemployed person in idleness.
Since he cannot afford to pay the tax, the essential worker joins the inessential worker in idleness. The cost of the lockdown has now doubled to $3,000 per fortnight. Hence, I cannot agree that a policy of paying people not to work is “costless”.
If adding debt is essentially “costless”, I am puzzled as to why we have taxes at all, since these are highly distorting and come at huge cost. Why not go one step further with this costless funding and pay the entire nation not to work? What is wrong with zero GDP? Surely, everyone would be better off if we exploited unlimited international borrowing capacity. Despite what one might infer from these statements, taxpayers will need to be slugged this additional A$260 billion cost in a year in which the budget was supposed to be in surplus for the first time in decades.
One can identify some benefit from JobKeeper subsidies as they retain a link between employer and employee. It is the lockdowns which have forced about three and one half million people into idleness. According to Australia’s CMO, this outlay has saved 14,000 lives at a cost of a sizeable $19 million per life, but in reality, probably close to zero lives.
- Payments to stop Australians working burden every Australian with a debt of $10,200
Yes, in a world in which resources are essentially free, why worry about who is going to meet the largely unnecessary bill of A$260 billion or $10,200 for every Australian man, women, and child? There is an upside for the non-economists in our midst. If resources are free, then there is no need to employ any economists and nor to train them. I find it surprising that a large number of top economists should argue, in effect, that they are either redundant or perhaps should be made redundant.
- Should we be concerned with the additional 750 to 1,500 suicides each year?
According to the May 7, 2020, statement from the Australian Medical Association, between an extra 750 and 1,500 more suicides may occur annually, in addition to the 3,000 plus lives that are lost to suicide already every year, due to the expected increase in job loss and business failure arising from the shutdown. Could simple fear of the virus itself generate a sizeable loss of life from suicide? Very unlikely. The virus could have been presented simply as yet another form of flu that would affect a small vulnerable portion of the population but would be extremely mild or unnoticeable for the bulk of the population.
Holden/Preston argue that these effects are small relative to the mass deaths of 225,000 people that they envisage but are unable to justify. Moreover, from the perspective of deaths, they envisage economic crises as being beneficial as there are fewer deaths in the workplace and on the roads. Yes, almost certainly the mortality rate has fallen greatly during the COVID-19 pandemic, with millions of people forced out of jobs and travel largely banned. Would they wish to live in a COVID-19 world permanently, or do they favour more Great Depressions with 50 per cent unemployment since fewer die on the roads or in the workplace? Holden/Preston have a simple message: ‘The shutdown wins’ and they believe that there are high risks from relaxation of restrictions in the absence of a vaccine.
To the contrary, Holden/Prescott are simply alarmist, frightening the public with entirely discredited pandemic modelling in which they argue that mortality from COVID-19 is far worse than the devastating Spanish Flu of a century ago and would kill nearly 250,000 Australians were the economy not locked down. Yet the current world COVID-19 virus death toll would have to rise by a factor of 202 times to reach their estimate of seventy million global deaths in the absence of lockdown.
- At great economic cost, Australia has failed to generate herd immunity
It would be a mistake to think that Australia and Taiwan, and all the other countries that have employed similar strategies, have got off lightly from COVID-19 by the simple expedient of banning the entry of non-citizens and early quarantining. Quarantining is exceedingly disruptive and impossible to sustain for any length of time.
It is likely Sweden will emerge from COVID-19 better prepared than many other countries due to the creation of at least some herd immunity, which does not rely on everyone being exposed to the virus. Rather, it relies on enough of the young and healthy being exposed with sufficient antibodies such that the elderly and vulnerable are protected from it. This would be especially important if it proves impossible to develop a vaccine in reasonable time. While Sweden may suffer because of the COVID-induced decline in the global economy, unlike most economies it has not deliberately forced millions into potential penury.
We can be sure that Governor Cuomo in New York and authorities in Belgium have exposed a sizeable portion of their populations to the COVID-19 virus making “second wave” attacks less likely when borders reopen, as they must. However, they will never be “winners” in an economic sense because, unlike Sweden, they locked down the most productive while not taking care of the vulnerable.
- Conclusion
Our actual current death toll of 102 would have been far lower if we had earlier implemented the quarantining, tracing, and testing methods so successfully employed by Taiwan with their voluntary social distancing. Forced lockdown in a highly responsible society such as Australia’s yields no lives saved but huge and devastating economic costs and future deaths arising from economic despair. According to Australia’s CMO, some lives have been saved at a cost of $19 million per life, but this is a gross understatement of the cost per life. The sooner we cease the lockdowns and travel restrictions, while retaining voluntary social distancing and protecting the vulnerable, the better.
Peter L. Swan AO FRSN FASSA is Professor of Finance at the University of New South Wales-Sydney. He wishes to thank Gigi Foster and Paul Frijters for useful comments.
Comments are closed.
