What Is Up with the Swedish Model?

Is this the Swedish model we want?




Well,

It's been a long time since I wrote.

I started the blog because I felt the news coverage I saw wasn't showing the real story, namely that the stats presented were done very poorly. Virtually nowhere did I see anyone provided context for the numbers, specifically what the per capita infection rates were as opposed to absolute numbers of infections. The blog showed me (and anyone else who was interested) that the outbreaks were bad in some areas not being covered, and that they were not so bad in other areas that got a lot of attention.

But eventually I stopped writing, because of two reasons. One was that some news services started doing a much better job of putting together the data in more meaningful ways. Another is that things seemed to be going pretty well, nationwide. Almost every state had some sort of official "lockdown" orders, and overall the rate of new case increase was dropping pretty much everywhere in the US. I didn't feel I had an important story left to tell.

That did not mean I stopped watching. And since I last wrote (April 6) a lot of things have happened. The first is that we've seen a political reaction that has completely surprised me (though I guess it shouldn't have). The brief sanity where (most in) the US agreed what needed to be done has dissolved, and many states have started opening up despite not meeting federal government guidelines for doing so. And there's a lot more going on internationally, too. So I will start writing again, pointing out things I find interesting in the US and around the world.

Sweden

The question that has most sparked my determination to start writing again is "what's going on in Sweden?" Sweden has been cited by many who wish to get the lockdown orders reversed (in the name of opening up the economy) as a model for how best to handle the crisis.

But is it? Sweden's number of deaths is FAR higher than it's Scandinavian neighbors, but that's not necessarily a problem. Because Sweden's policy was essentially to not wait until vaccines were available, but rather to achieve herd immunity as quickly as safely possible, while protecting the most vulnerable. But please note this does not mean social distancing did not happen. Sweden only outright banned a few things, such as large gatherings. And while kindergarten and primary schools were kept open, high schools and colleges were closed. Businesses, including restaurants, were not closed; though Swedes were asked to work from home if possible, and to practice social distancing where possible.

In the US, this has been hailed by anti-lockdown advocates as a way to save the economy. So in evaluating the success of this policy, we have to look at both the direct effects of the pandemic, as well as how Sweden's economy is doing.

Looking at the pandemic is not a straightforward matter of counting the infection rate, or the number of deaths, though. Because in some ways, Sweden wants more infections than it's social-distancing neighbors do. Sweden needs the population to be infected in order to eventually have enough of the populace infected that herd immunity has developed. But they want those infections to happen to the right people (those most likely to not suffer serious ill effects from the infection), and at the right pace (fast enough to achieve herd immunity relatively quickly, but no so fast as to overwhelm the medical resources available to treat those infected who require hospitalization).

So let's take a look at some facts, starting with this chart, which shows the number of new cases per day, averaged out over the previous 7 days (I create this average in order to smooth out the data a bit, so that we don't see the wild day-to-day swings; the downside is that it can trail the current numbers a bit):

Sweden's numbers are far higher than those of the other Scandinavian nations. Now, to a certain extent, this is to be expected. Sweden, with a population of 10.3 million, is almost twice as large as each of its fellow Scandinavian countries, who average around 5.5 million each.

So what if we look at the same chart, but this time account for the population differences? This chart's Y-axis is the number of new cases per day (7 day average), per 100,000 people. And, just for fun, I've thrown the US numbers in as well.

In this one, you will notice that the peak rate of Sweden is comparable to those of the other Scandinavian countries (and, unfortunately, that the peak rate of the US was 67% higher than that of any of the Scandinavian countries). But Sweden's rate hasn't really dropped from its peak levels. Even the disastrous US situation (and yes, I think the US's numbers are disastrous, and we'll talk more about those later), while still showing a higher rate of new infections per capita than even Sweden, is at least improving, But it looks as though Sweden's infection rate is holding quite steady.

But is that bad? Remember, Sweden was supposedly essentially wanting people, some people, to get infected in an effort to achieve herd immunity*. So maybe these people being infected is a good thing, in that Sweden might achieve its herd immunity, emerge from its more limited social distancing, and keep its economy humming along.
* - maybe they are trying to achieve herd immunity, or maybe not, as will be discussed below.

The Cost of a Herd Immunity Strategy

What, then, has the price been? If Sweden had peak infection rates similar to the peaks of other Scandinavian countries, though its peak has lasted essentially a month now, compared to 1-2 weeks for Denmark, Finland, and Norway, one would expect the number of deaths to be higher. Also remember that Sweden is twice the size of the other countries, again a reason for deaths to be higher. Counting deaths may not be the best way to determine if Sweden is doing well, but we can at least look and see that the total number of deaths in Sweden to date is, well... let's let the numbers speak for themselves.

We can then compare the deaths on a per 100,000 basis in order to take population out as a factor, and once we do that we can add the US in as a point of comparison. It's not good for Sweden.

As you can see, not only are Sweden's numbers MUCH higher (4-9 times higher) than its fellow Nordic countries, they're even worse than those of the US (more than 25% higher).

And one more interesting fact about what's going on in Sweden. Let's take a look at the Case Fatality Rate (CFR). This is, quite simply, the percentage of those known to have been infected who have died. This is NOT the mortality rate, which would require us to know how many people overall have been infected. I don't believe any country currently has a good picture on the number of people currently infected, so this number is not known anywhere in the world. We will use a CFR calculation based on the death rate of a 7-day average to help smooth out any wild swings, and once again we will include the US for comparison.

I'm not sure why there is the weird bump at the beginning of the US's numbers. It may have something to do with the first deaths occurring in the Seattle-area nursing home outbreak. Regardless, if you ignore that one anomaly you see some remarkably similar patterns, of a steady increase and then a plateau-ing of the rate. This indicates two things to me: first, that it takes a while to die of Covid-19, so there is a lag between infections and the "natural" death rate; and second, that none of these countries have seen a spike due to their healthcare facilities being overwhelmed. But why is Sweden's CFR 2.5 times that of Denmark and Finland, and 4 times that of Norway? This truly has to concern the Swedish authorities.

The Benefits of a Herd Immunity Strategy

Were there any benefits, or at least any ameliorating factors? One thing I'm hearing from Sweden is that while the number of deaths is very tragic, it is almost entirely confined to the elderly. It is treated as some strange outcome that they tried but failed to prevent. "'Once we know how the virus got into our elderly care facilities, the government can make recommendations and take measures to try to stop that, because that is the biggest tragedy of all this, that it has gotten into the nursing homes,' [Swedish Ambassador] Olofsdotter told NPR." (from an article in late April). At roughly the same time, Dr. Anders Tegnell, chief epidemiologist for Sweden's Public Health Agency was telling CNBC "Unfortunately the mortality rate is high due to the introduction (of the virus) in elderly care homes and we are investigating the cause of that." These comments feel like a weird "oops" statement to me. The article from National Review I cited extensively in my earlier post on Sweden April 6th states that "By not requiring social isolation, Sweden’s young people spread the virus, mostly asymptomatically, as is supposed to happen in a normal flu season. They will generate protective antibodies that make it harder and harder for the Wuhan virus to reach and infect the frail and elderly who have serious underlying conditions." If that was the goal, well... they fucked up big time. They were supposed to keep the old and vulnerable safe, and they didn't. But another question is whether this ratio of deaths for the elderly is even in any way a unique situation for Sweden. Swedish numbers as of May 8th are that just over 90.4% of the Covid-19 deaths have been 70 or older. But in Norway as of May 20th 88% of the Covid-19 deaths have been 70 or older. In the US the stats are not broken down quite the same way, but even there 80.6% of Covid-19 deaths have been 65 or older. In other words, I don't think it reasonable to somehow claim that the only reason Sweden has such shockingly high CFR and high number of deaths per 100,000 is because they had an "oopsie" with the old-age homes. But it may be even worse than that; it may not be an error, but intentional. This BBC article reports that many Swedish elder care facilities have been actively denying care to elderly Covid victims.

What about the economy? Will Sweden's economy do better than its neighbors? One would assume that with a less stringent social distancing policy and restaurants able to stay open that the economic forecast for Sweden would be relatively bright. Unfortunately, that doesn't seem to be the case. The EU has just released their semi-annual forecasts this month. This chart shows the projected GDP growth and unemployment for 2020 and 2021 for the four Scandinavian nations, and for the EU as a whole:

Note: Norway is not a member of the EU, and the data does not come from the EU, nor from the Norwegian government, but a different source. I almost didn't include it because direct comparisons are impossible, but it actually still communicates the most important fact, which is that Sweden's more lax social-distancing policies have not bought it an outlook any brighter than its peers, or the EU as a whole. Even if you argue that the EU's forecasts are absurdly optimistic, which wouldn't be surprising given they are forecasting for themselves, it seems unlikely they are using different assumptions for Sweden than the others.

Testing

Let's circle back to this whole question of herd immunity. If one were seeking to implement this kind of policy, I'd think it would be important to measure one's progress by testing as many people as possible to know when success had been achieved. But Sweden's testing numbers are the lowest in Scandinavia.

This chart compares testing in Sweden and Norway. The columns represent the total tests performed per week (week 1 ends Feb 23rd, week 13 ends May 18th), with the scale on the left. The lines represent tests performed per 100,000 population, with the scale on the right.

(I was not able to get good data over time for Denmark and Finland, but Finland's current per capita testing numbers are about 20% higher than Sweden's, while Denmark leads Scandinavia with per capita testing more than 4 times that of Sweden. For comparison, the current per capita numbers of tests performed in the United States are similar to Norway's, ie twice that of Sweden.)

As you can see from the chart, Norway reacted quickly, launching a fairly aggressive testing campaign at the start of their response to the pandemic, but is now tapering off its testing (a reasonable response to a pandemic situation that seems under control). Meanwhile Sweden has progressed its testing numbers steadily upwards. Their current testing levels are similar to Norway's, but remember that Sweden's infection levels are FAR higher than Norway's, and that if developing herd immunity is the strategy, it would seem that testing would be an important part of that strategy as the only way to determine how close to that goal you were. Even Swedish officials acknowledge a need for more testing, with Dr. Tegnell telling CNBC the were "testing around 20,000 people a week and hopes to increase that to 100,000 in a few weeks’ time". Now "a few weeks' time" is a pretty vague target, but I would argue that since that statement was published April 22nd, and Sweden had only tested 32,000 people the week of May 18th, that they have completely failed to meet that goal (they wanted a 400% increase in 2 to 3 weeks, but got a 60% increase in 4 weeks).

What Herd Immunity Strategy?

But wait, because now there's a bigger question. Is Sweden even trying for herd immunity? Officials are saying, no, that is not their goal. From a CNN story dated April 28th, "Lena Hallengren, Swedish Minister for Health and Social Affairs, told CNN: 'There is no strategy to create herd immunity in response to Covid-19 in Sweden.'" If that is the case, then two questions arise:

1. Why are so many Swedish health officials talking about herd immunity if that is not the strategy?
2. If herd immunity is not the strategy, then what is the strategy?

While Ms. Hallengren denies that herd immunity is the strategy, somebody seems to have forgotten to tell that to the Swedish Ambassador to the United States, who told NPR on April 26th that Stockholm already had 30% of its population immune to the disease (presumably meaning already infected), saying "We could reach herd immunity in the capital as early as next month." Now granted, Ambassador Olofsdotter is (we assume) not a health expert, and you could argue that she may have gotten the messaging wrong. But that would be a difficult argument to make given that on April 22nd, Dr. Tegnell told CNBC "In major parts of Sweden, around Stockholm, we have reached a plateau (in new cases) and we’re already seeing the effect of herd immunity and in a few weeks’ time we’ll see even more of the effects of that. And in the rest of the country, the situation is stable," Remember, Dr. Tegnell is Sweden's chief epidemiologist. Now I'm going to show one last graph. This is the raw numbers of new cases per day in Sweden (not the per day average of the previous 7 days' numbers I showed earlier), and the only dates I'm showing are April 21st, 25th, and 27th (that is one day before each of the quotes I've mentioned here was published). I'm also making the vertical scale logarithmic because I think it's easier to read that way.

Assuming these officials had access to these numbers, would you declare that they had "reached a plateau"? I see no sign of it. And even if they indeed had, as my first graph shows, that plateau was at a very high level. If herd immunity were beginning to make an impact, as Dr. Tegnell said on the 22nd, you would expect new case numbers to drop. Herd immunity should have the same appearance on the chart as strict social-distancing--slower spread due to fewer new targets to infect, either because they're already immune or they're just not around.

So, what is Sweden's strategy? I can't state one, because I don't see one (that's not herd immunity). I have seen some talk about the psychological benefits to Swedish children of being able to play with their friends and attend classes. I don't want to minimize that, but I wonder what that benefit would be compared to those who spend the rest of their lives thinking that playing with friends or going to school killed their grandparents? Frankly, I don't believe that herd immunity wasn't the strategy. It seems that the denial of the strategy is more of a reaction to the failure of the strategy. An awful lot of Swedes are dying. Even though 90% of them are 70 or older, in a nation where the average life expectancy is 83 that seems like a lot of years are cut short. By May 5th, the Financial Times' reporter John Burn-Murdoch has shown that Sweden had endured 3600 more deaths than average, a 30% increase for the time period (you can view Murdoch's fascinating visualizations on Covid-19 here). I'm not trying to pick on Sweden. A lot of countries have been horrible in their response to this pandemic. The US, the UK, Brazil, and many others. But Sweden is curious for their strategy, for their certainty that they are the ones who have figured this out while others have it all wrong. But when the numbers tell the opposite story, something smells fishy.

Something is indeed rotten, it's just not in Denmark.