Sunday, January 10, 2021

The importance of Sweden

The importance of Sweden

by George J. Dance

"One baffling feature of 2020," British Tory MP Neil O'Brien recently wrote in the New Statesman, "was that so much energy was wasted puffing up Sweden." Over at The Guardian, OpenDemocracy editor Peter Geoghegan echoed: "why, as the UK experienced one of the world’s worst Covid outbreaks, did so many prominent public figures spend so much of 2020 talking about Sweden?" 

I cannot say for sure why they talked about it (and I suspect neither writer wants an answer anyway), but I can explain why I keep writing about it. (Full disclosure: I have written seven articles on the country.) 

First, Sweden notoriously did not have a lockdown in 2020. Lockdowns were an unprecedented experiment in totalitarianism, justifiable only if they were the only way to prevent mass death. Whether lockdowns can be justified, then, depends on the factual question of whether they prevented mass death; which can only be answered by comparing real-world outcomes between countries that locked down and those that did not. Outlier nations like Sweden that remained open act as controls in the real-world lockdown experiment; the question to be asked about them being, how well did they do versus the closed societies? Without those real-world examples to give us real facts, we would have only guesses embodied in models. 

The original lockdowner prediction, from a model, was that Sweden's health care system would collapse, and it would suffer ~96,000 deaths by May. Contrary to an internet rumor, that modelling was not done by Neil Ferguson and Imperial College; it did, though, borrow its assumptions from Ferguson's infamous March 16 Imperial death model report, which forecast 200,000 British deaths under mitigation (also by May) versus just 20,000 with suppression. All three of the above forecasts were wildly wrong – one would say, comically wrong, if there were anything comical about human death – hence the need for real-world comparisons like the one Sweden offered to the locked-down countries of Europe. If every government had been pushed into locking down, there would be no factual basis for comparison, and we would have to rely on the guesswork of specialists. 

The British government opted in March for a Swedish-style mitigation strategy, but was forced to pull a U-Turn a week later in the face of widespread media and public criticism (including a nasty Twitter campaign accusing Boris Johnson of genocide). Britain was locked down on March 23; the same day, a similar media campaign began against Sweden. The past 40 weeks since then have seen a steady stream of articles – more than five every week – pronouncing the failure of the Swedish "experiment" [sic], like a stopped clock. (See my bibliography, "Sweden-bashing in the media".) The intent may be to convince the world that mass death has occurred in Sweden, or to panic the Swedes into their own U-Turn; either could be used as a victory for lockdowners in their propaganda war.

That brings me to my second reason for writing on the topic: many of those articles are pure propaganda, full of out-of-context claims, cherry-picked facts and outright misrepresentations that call out for rebuttal. On the other hand, even a stopped clock can be right; such articles should be checked not only for what may be wrong in them, but for what may be right.      

Admittedly, I sometimes get tired of writing in the genre. Over and over I find myself reading the same assertions I have already dealt with, and writing the same things in reply. This writing is not a creative act, but simple drudge work I sometimes wish someone else would do. Balancing that off, though, is my third reason for writing them: my articles have been somewhat popular, one getting a satisfactory 2,000 reads. So it is once more into the breach, to deal with these two latest examples. 

Geohegan's article is actually about a shadowy conspiracy involving "backbench Tory MPs, anonymously funded thinktanks and ubiquitous media commentators" that oppose lockdowns, so he deals with the alleged failure almost in passing: 

We now know with certainty what public health experts have long predicted: a light-touch coronavirus approach does not work. Sweden has recorded far higher death rates than its Nordic neighbours, while suffering a similar economic hit. Even the country’s king thinks it has 'failed'."

The first sentence is recycled claims I have dealt with before (and deal with again, later on), but the second is new. Geohegan seems to be implying that the King of Sweden was calling the country's 'light-touch' strategy a failure. Where did he get that idea? Probably from an earlier Guardian story, "King of Sweden blasts country's 'failed' coronavirus response," which declared in the subhead that the royal "Criticism of anti-lockdown stance comes as hospitals struggle to cope with surge in cases". Reading down, though, the story clarifies that that was not the king's message: 

The comments were initially taken as a criticism of Sweden’s controversial anti-lockdown strategy, but the royal court later said the king was referring 'to the whole of Sweden and the whole society. He is showing empathy for all those affected.' 

I am glad The Guardian clarified that point, though it is a pity that clarification did not make it into the headline. 'Tis a pity, as well, that Guardian reporters do not read the articles in their paper.

Sweden did have obvious failures, and the king deserves credit for acknowledging them. The most glaring were the failure early on to shield senior citizens, (particularly those in care homes) and immigrants, two groups that suffered disproportionately. That has little if anything to do with lockdowns, though, as the same failures occurred in countries that did close down. So much for his majesty's comments. 

O'Brien's article, on the other hand, makes several criticisms that need a deeper dive. So let me turn to those, giving his points followed by my responses.  

Recent days have seen Sweden’s Nordic neighbours Finland and Norway offering emergency medical assistance as Stockholm’s hospitals have been overwhelmed, infections and deaths have spiked dramatically upward, and the King of Sweden has made an unprecedented criticism of the government’s bungled strategy.

None of which is an argument for Sweden bungling into a lockdown strategy strategy instead. It would be if the European countries that locked down had managed to avoid a second wave, or at least had  less deaths from it; but neither of those are the case. Infections and deaths have been "spiking" since November all over Europe, and barring a vaccine miracle they will continue to do so until spring. In the process, one locked-down country after another has been recording a worse outcome, in terms of deaths per capita, than Sweden. None of which matters to O'Brien, who recycles just one cherry-picked comparison (that I last saw reused by Dominic Lawson in a recent Sunday Times): 

Unprecedented, but hardly surprising: Sweden has suffered a death rate that is roughly ten times that of neighbouring Norway and nine times that of Finland. 

As with Lawson, O'Brien makes no attempt to look for reasons to explain the difference; the reader is left to infer that those countries' lower death rates must be due to lockdowns (although neither Norway nor Finland had a fall lockdown, and Norway arguably had no spring lockdown, either). O'Brien also makes no attempt to deal with, explain, or even mention the larger European context.

At the beginning of autumn, Sweden had the fifth highest death rate per capita in Europe, behind only Belgium, Italy, the UK, and Spain; all of the last four had suffered more deaths while the country was locked down. By November, Sweden had fallen from #5 to #10, being passed in deaths per capita by France, North Macedonia, Bosnia and Hercegovina, Montenegro, and Czechia. In the following months it has since slipped further to #15, being overtaken in the body count by Bulgaria, Slovenia, Hungary, Croatia, and Switzerland. Switzerland has lockdowns on the state level only; the other 13 have all relied on national lockdowns, most of them at least twice.    

Graph courtesy CTV News
A searing government report concluded the state had failed to protect the vulnerable. Mats Persson, a former UK government adviser, said of his home country: 'For a social model largely designed around the state levelling the odds and caring for the vulnerable, this will leave a very difficult moral legacy.'

Yes, Sweden's biggest failure came in care homes, which were locked down (just as they were in Britain and elsewhere) with disastrous results. Perhaps the UK government and its 'advisers' should take the beams out of their own eyes before preaching 'morality' to countries with less deaths. 

O'Brien does seem aware that Sweden has had less deaths than the UK, though he never mentions it, as he comes up with a couple of reasons why he thinks that should be. 

To understand the magnitude of what’s gone wrong, it’s worth noting that Sweden started the pandemic with several huge advantages. First, it’s a far less urban nation than the UK, for example, and the virus spreads much more rapidly in dense, built-up areas. While the UK has 273 people per square kilometre, Sweden has just 25.

Perhaps O'Brien is not being misleading, but has simply been misled from seeing that "25" figure (as I have) out of context on the internet. In any case the figure is misleading, as Sweden's population is not spread out more or less uniformly across the whole country but concentrated in a few urban areas.  Norrland, which covers more than half the country, is almost uninhabited, with less than 5 people per square kilometre. On the other hand, the population density in urban areas, where most of the people live, is more than 200 times greater. Wikipedia, the likely original source for O'Brien's "25" figure, puts it in context by noting that "87% of the population live in urban areas, which cover 1.5% of the entire land area." That works out to an urban density of just over 1,330 people per square kilometre; but Wikipedia gives a higher figure for the settled area: "The average population density is just over 25 people per km2 (65 per square mile), with 1,437 persons per km2 in localities (continuous settlement with at least 200 inhabitants)." 

Second, Sweden has the highest rate of people living alone in the world: 42.5 per cent of households are single people, compared to just 29.9 per cent in the UK. Obviously, it’s much easier for the virus to spread within the home, and places with large, multigenerational households suffer most.

This statistic, that also appeared in Lawson, is not wrong but is rrelevant. The places in Sweden where infections occurred most were care homes, then hospitals, and then new immigrants households (which tend to be larger and more intergenerational than the national average). None of those could conceivably have been prevented by closing restaurants and bookstores. Confining the non-essential to their households by lockdown might have even boosted household deaths among the last group. 

O'Brien then compares Britain with an imaginary Britain – another comparison used by Lawson, and a favorite one for lockdown aficionados – which he combines with his earlier comparison into a logical pretzel:    

To form an idea of the consequences that would have followed if the UK had followed the Swedish model, you would need to compare Sweden’s outcomes to its similar neighbours. Given the country’s death rate is ten times higher, imagine the chaos we’d have seen if we had multiplied the UK death rate by a factor of ten.

This is not a sound argument, but a non sequitur. I suppose the best way to show the non-sequitur is to substitute terms: 

"To form an idea of the consequences that would have followed if Sweden had followed the British model, you would need to compare the UK's outcomes to its neighbous. Given the UK’s death rate is more than 2.5 times that of its neighbor Ireland, more than 3 times that of its neighbors the Channel Islands, and four times that of its neighbor the Isle of Man, imagine the chaos the Swedes would have seen if they had multiplied their death rate by a factor of 2.5 to 4." 

Nor has there been an economic upside for Sweden: in fact, they saw a bigger hit to their economy than their neighbours, as well as much worse health outcomes. 

O'Brien is relying on economic data from August, not to mention from a notorious Sweden-bashing site. In October, Danske Bank forecast that Sweden's economy would fall 3.3% in 2020 (versus 3.6% for Norway, 4.5% for Finland, and 8.3% for the locked-down euro zone).   In November, SEB Bank estimated Sweden's GDP hit at 3.1% (versus 1.9% for Norway, 4.0% for Finland, and 7.6% for the euro zone). In any case, actually calculating economic damage will require more than just looking at GDP, but also looking at things like bankruptcies, unemployment, crime, and suicide. The countries that locked down will be totting up the damage for years. 

Swedish virologist Lena Einhorn concluded: “Sweden’s strategy has proven to be a dramatic failure.” This matters, not only because health and lives are in danger, but also because the Swedish experiment reveals the failures of the underlying theories of Covid-sceptics. 

The Swedish experiment of mitigation has to date resulted in less loss of life than the British experiment of continual interventions, roughly 20% less. Sweden has had more confirmed cases – over 4% of the population versus 3.5% – but it is unclear what that means in terms of health outcomes, other than a lower case fatality rate. Pace Dr. Einhorn, there is no actual evidence here of failure; cases that result in recoveries rather than deaths are not failures.  

Sweden’s controversial state epidemiologist Anders Tegnell predicted in the summer that because the country had a high rate of infection in the spring, it would have “a high level of immunity and the number of cases [in the winter] will probably be quite low”.

We now know he was disastrously wrong. But he wasn’t alone; his theory was exactly the same as that still relied upon by the UK’s Covid-sceptics.

Tegnell admitted he was wrong about that prediction a month ago. It is a stretch to call that wrong prediction a "failure" – it was not a failure of the theory, but a matter of not having adequate data – and jumping the shark to call it "disastrous." (If O'Brien's implication is that Sweden did not prepare for a second wave, that implication is false.) Sweden's consequences to date have been less "disastrous" than Britain's, or those of 14 other European countries that have had lockdowns. Of course, there are also European countries that have locked down and had less deaths as well; but the 14 counterexamples to date are reason enough to not attribute the latter countries' fewer deaths to their lockdowns.  

Sunetra Gupta, a lead author of the “Great Barrington Declaration”, promised in May that, in the UK, “the epidemic has largely come and is on its way out in this country… due to the build-up of immunity”. One leading Covid-sceptic, Michael Yeadon, wrote that thanks to “prior immunity”, “the pandemic is effectively over.”

It’s these same failed theories that still lead Covid-sceptics to argue it is safe to let the virus rip 

What exactly are "Covid-sceptics"? Does O'Brien mean "lockdown sceptics", as in his title, or "Covid deniers"? If the latter, he is attacking a strawman, and a rather dimwitted one at that: someone who doubted there was a virus would have a hard time believing that a person could develop antibodies specific to it. If the former, O'Brien is still wrong. The sceptic case against lockdowns is that their economic damage has been underestimated, while their value in saving lives has not been demonstrated. All that "immunity" adds to the discussion, for me, is the observation that the coronavirus does not "rip" (a word that sounds like what a Covid Cultist would say); on the contrary, community outbreaks can be expected to slow down with the buildup of immunity as people in the community catch it and recover. 

and attempt to shelter the old and vulnerable. 

Those are the groups that have taken the most casualties. Anyone who is serious about reducing deaths from Covid-19 should be concerned with trying to shield them; even politicians who locked down their states have paid lip service to the concept. I do not see any reason for it to be controversial at all.  

One baffling feature of 2020 was that so much energy was wasted puffing up Sweden, while countries such as Australia, Japan, South Korea, Taiwan and New Zealand revealed to us genuinely successful models based on hard suppression of the virus with decisive action: all suffered a fraction of the European average rate of coronavirus cases.

Fine; let us briefly look those countries. All but Japan are effectively small, isolated islands (South Korea's only land border is impassable), which were able to eliminate the virus by using what I've called a 'NIMBY strategy': quarantining everyone entering the country and the relatively few cases found or traced in the community. Iceland, Norway, and Finland are using the same strategy successfully in Europe. In no case was a lockdown needed (although Australia and New Zealand did have a couple, possibly for the political theatre). It is hard to see how that strategy could be made to work in the European Union. I do not think anyone has yet figured out exactly what Japan did, but we do know that it has not included wholesale lockdowns or business shutdowns either. 

There is, of course, no scope for triumphalism here. 

Of course not; nor any grounds for it, either. If Sweden does end up with far more deaths than any of the locked-down countries - say, the 66,000 that Imperial College estimated would have happened by May without NPIs - that would be something for lockdown champions to gloat about. But Sweden sitting in the middle of a pack of lockdown countries, the latter all markedly different in their mortality rates (by as much as 0.1% of the whole population) does not even indicate a correlation between locking down and reducing deaths. 

Things can always change. The pandemic looks likely to continue until spring; Sweden could always have an even greater winter surge, followed by overwhelmed hospitals and mass deaths. On the other hand, that can happen in any of the locked-down countries as well. One can only wait and see. I am sure this is not the last time I will be writing about the subject. 

No comments:

Post a Comment