Sweden did not achieve herd immunity and is a cautionary tale to avoid, not an example to emulate.
Back in March and April, when the first wave of COVID-19 was running rampant through Europe, most nations did the responsible thing and imposed mandatory nationwide lockdowns in order to stop the spread of the SAR-CoV-2 virus. And while it largely worked in the countries that took their lockdowns seriously, Sweden decided to go another route and try for what was referred to as “herd immunity.” Essentially, Sweden refused to implement a lockdown or even mandatory infection controls and tried to keep the virus from causing outbreaks among the most susceptible to the virus. The so-called logic was that the virus was going to burn its way through the community no matter what the country did, so the sooner it finished, the less damage to people and the economy three would be over the long run
As a result, Sweden’s COVID-19 cases exploded. While most of the rest of Europe was recovering from the first wave over the summer, Sweden suffered for an additional three months of high cases. Sweden’s deaths also exploded, with more nearly 10x Swedes dying from COVID-19 by July 31, 2020 than any of the neighboring countries of Denmark, Finland, and Norway. But until the second major wave of COVID-19 hit Europe, it was possible that Sweden’s pain and suffering had been worth it. The argument was this: if COVID-19 cases and deaths stayed low in Sweden while the pandemic erupted in neighboring countries, then Sweden’s experiment with “herd immunity” could be deemed a success.
Europe is now in the midst of the second wave of COVID-19, so we can look at Sweden and its neighbors to assess whether Sweden’s painful experiment was successful or not. The data is undeniable – herd immunity was not achieved and Sweden’s experiment was a deadly failure.
Looking strictly at new cases of COVID-19, we can see that Sweden’s cases are skyrocketing. Denmark’s cases are up as well, as are Finland’s and Norway’s, albeit less dramatically. There’s clearly a new wave of COVID-19 hitting Scandanavia, but that’s hardly a surprise given COVID-19 is on the rise globally at the moment.
Sweden’s new cases are also up even when we adjust the data for population (lower figure). Here we can see that there was an early spike in Denmark that subsided before the big wave that is currently affecting everyone, and Finland is doing the best of these four countries.
Before I proceed, let me make a comment here about increased testing. It’s likely that part of the recent increase in new cases is at least partly due to increased testing for COVID-19. However, all four countries show an increase in testing (Finland dropped off during the summer but has since increased again), and so we can still compare the results from one country to another.
If we look at just the last 30 days, we can see from the table below that Sweden is outpacing all of its neighbors on total number of cases and in number of cases per 100k population. If Sweden had developed anything resembling herd immunity this would not be the case.
|New cases, last 30 days||Sweden||Denmark||Finland||Norway|
|per 100k population||724.1||466.5||111.1||212.7|
Let’s look at deaths from COVID-19 as well, as the deaths from the disease are largely independent of testing. We can see that deaths in Sweden are up, but when we look at the overall number of deaths since the start of the pandemic, it’s hard to see the details. The problem is that Sweden’s refusal to lock down and impose restrictions resulted in a massive spike in deaths that topped out at 115 in mid-April. We can also see that Sweden’s initial spike in deaths was not only much larger, but lasted much longer than its neighbors’.
And adjusting the data for population doesn’t make Sweden look any better. Sweden has by far the largest population of all the four countries, but when adjusted for population Sweden still had 3.4x more deaths than Denmark did. Denmark’s population density is nearly 6x higher than that of Sweden, so you would expect Denmark to have the higher number of deaths, but because Denmark instituted significant infection controls, they were able to save lives.
If we look at just the last 30 days, we can see from the table below that Sweden is again outpacing all of its neighbors on total number of deaths and number of deaths per 100k population. Norway, Finland, and Denmark all reimposed significant infection controls early in this wave of COVID-19 – Sweden, again, did not.
|New deaths, last 30 days||Sweden||Denmark||Finland||Norway|
|per 100k population||2.55||1.38||0.35||0.30|
Finally, let’s look at the 7-day moving average number of deaths per day as adjusted for population since August 1. We can see that all four countries had pretty low deaths through most of the summer, with Denmark, Finland, and Norway all being slightly lower than Finland. Denmark’s deaths started rising first, in mid-September, followed by Sweden 2-3 weeks later, and lastly by Finland and Norway two weeks after Sweden. But except for a couple of short periods where Denmark outpaced Sweden (that can easily be explained by the fact Denmark was 2-3 weeks ahead of Sweden in the initial rise), Sweden has again had deaths climb faster than its neighbors.
If Sweden had achieved herd immunity, they would have fewer deaths per capita in the second wave than their neighbors who locked down. Sweden has more deaths per capita, therefore they did not achieve herd immunity.
And Sweden didn’t even save their economy in the process, which was a large part of why they didn’t impose the kinds of restrictions that their neighbors did. As you can see from the table below, the IMF expects Sweden’s economy to shrink the most out of the four countries we’ve been looking at.
|Real GDP growth, 2019 to 2020||-4.7%||-4.5%||-4.0%||-2.8%|
Let’s be clear about this – Sweden did not achieve herd immunity and their approach led to thousands of unnecessary deaths that could have been prevented. Sweden is not an example to emulate, it is a cautionary tale of what not to do.
We follow Sweden’s lead at our own peril.
All graphs and analyses were done using data from here, accessed on 11/15/2020 at around 11 AM, Mountain time (1800 UTC).