On "Sweden as a COVID success story"

When you look at the best comparison cases, it's unfortunately not so

[Btw, if this topic interests you and you haven’t already seen it, you might be interested in a look at this, which I meant to schedule for earlier in the week and then screwed up and accidentally published it without emailing…]

I feel obligated to start by saying, once again, that I’m not a ForeverMasker, and I recognize full well that overreaching lockdowns caused significant social harm in many societies, including this one. What I am interested in is coming to the most accurate understanding possible of what has worked in our response to COVID, what hasn’t, and the reasons why in both cases.

Here’s a representative version of the widespread effort on parts of the libertarian right to argue that Sweden, which did relatively little in the way of aggressive COVID quarantining and lockdowns, represents a successful national outcome in the pandemic.

Note first that absent any investigation of particulars, this claim alone is hardly indisputable. Sweden’s COVID death rate does not even place it in the top 1/3 of members of the European Union on this variable. To claim, as the above linked article does, that Sweden’s death rate is “among the lowest rates you’ll find in all of Europe” is simply untrue. Their rate places them in the middle third of Europe, neither at the end of those that did worst nor at the end of those that had the best outcomes.

Now let’s get into details a bit. A central problem seemingly not well understood by the folks who make this argument about Sweden’s outcome is that many variables are involved in explaining a country's COVID death rate. Lockdowns and quarantining are an important part of the story of a country’s profile when endeavoring to explain their results, but alone they don’t get us far in the way of a thorough explanation of how effective various policies are or aren't.

In fact, countries that aggressively quarantined and did lockdowns had a wide range of outcomes on this variable. In the European Union, just about every country but Sweden used some combination of the aggressive measures (lockdowns, quarantines, required masking) that the Swedes avoided, and the death rates in the EU range from much higher (e.g., Bulgaria at 520 and Hungary at 461 deaths per 100,000) than Sweden’s (176) to much lower (e.g., Norway at 41).

That much lower figure in Norway leads me to perhaps the single most important element of this analysis that is, not surprisingly, overlooked entirely by the people trying to present Sweden as a clear success story. If we want effective comparison between cases in which the number of variables is very high, as it is in this case, we do well to try to find ways to reduce the variation as much as possible. That is, we want to compare the case we’re interested in to the other cases that are the most similar to it.

As it turns out, there are several countries in close proximity to Sweden which are also very similar along many socio-demographic lines, thus effectively controlling for a lot of the variation. And these cases differ significantly from Sweden on the national response regarding lockdowns and quarantines.

These “excellent comparison” countries are Norway, Finland, and Denmark.

Take a look at these three at the link above on COVID death rates and you’ll find all of them are substantially below Sweden’s. Sweden’s death rate of 176 is nearly twice that of Denmark (93), three and a half times as large as Finland’s (50), and nearly four and a half times as large as Norway’s (41). So Sweden did very poorly on COVID deaths in comparison to the other European countries most demographically like them.

If we extend the comparative set to include two other geographic neighbors with significantly similar social and cultural profiles—Germany and the Netherlands—that were much more aggressive on lockdowns than Sweden, we find that both of these also have lower COVID death rates than Sweden’s (153 and 125, respectively), though not as much lower as the first three in our comparison set.

Note that there is not a word about any of this in the article linked above.

Another important variable to explore if one wants to understand differing death rates is vaccination rates. The “Sweden is a success story” people typically have nothing whatever to say about this either (the words “vaccine” and “vaccination” do not appear in the article linked above). This is in large part because that demographic tends to be libertarian on vaccination as well.

The data show a clear pattern on vaccination rates: those countries in Europe with low rates of vaccination are inevitably among the worst in COVID deaths per capita. Note the countries at the bottom of this graph (or this one) and then look for their COVID death rates on the link given above and you will find no exceptions to that rule.

The “Sweden is a success story” folks would, if they were genuinely looking for the full panoply of causes of Sweden’s high in the middle third in the EU COVID death rate, have to note that Sweden is top third of countries in Europe with the highest overall vaccination rate. That variable undoubtedly did some of the work to counterbalance the lack of governmental action to control transmission, i.e., if they had not vaccinated at that rate, they almost certainly would have had a higher death rate given their lackadaisical attitude on governmental action to limit transmission.

It has to be admitted that high vaccination rates alone do not guarantee a European country a place among those with lowest death rates, and some countries with high vaccination rates fared relatively poorly on COVID death rates, but this is consistent with what I’ve argued above. Many variables are involved in determining a country’s outcome: governmental policies on lockdowns, masking, quarantines, and social distancing; vaccination rates; poverty rates; level of populational access to medical care and quality of that care; overall health levels in the country (very unhealthy countries may do comparatively poorly even with relatively high vaccination rates); density of population (high numbers of people packed very densely together is typically undesirable in an infectious disease pandemic); and many others.

But again, there is little doubt that if Sweden had undertaken the approach to COVID it did with markedly lower vaccination rates, they would have done still worse than they did. That is to say, if they’d followed the libertarian platform across the board, they would have had outcomes still more significantly below their closest European peers than they had even with their high vaccination rates.

It is in my view not an accident that the “Sweden is a success story” people don’t talk about Sweden’s high vaccination rate. They have access to all of those data. But it doesn’t fit their narrative.