Sweden
Back in March of 2020, there was a reasonable case for lockdown. A new, highly contagious virus was spreading through the population, and while the death rate for young people was low, the death rate for elderly people was quite high. Early data pointed to an IFR of about 1%. We were told that - in the absence of drastic measures - the virus would continue to spread until about two-thirds of the population had been infected. A simple back-of-the-envelope calculation suggested that, if we did not take drastic measures, the death toll would be enormous.

The UK's population is 66.7 million. Two-thirds of that is about 44.5 million. Applying a 1% IFR yields 445,000 deaths. And that was if the NHS didn't become overwhelmed. If it did become overwhelmed, we were told, the IFR might rise to 2 or even 3%. Hence we were looking at a worst-case scenario of around 1.3 million deaths. (Note: this is about ten times the official death toll, which is itself a slight overestimate.)

Although lockdowns would come with massive costs, I reasoned, it was worth having one to prevent hundreds of thousands - or even a million - people dying. Hence I supported the first lockdown. Though it may have been a reasonable thing to do given the information available at the time, I now believe that locking down was the wrong decision.

There are many elements in the case against lockdowns, as I have outlined in an article on this website. But - outside a few specific countries like Australia and New Zealand - the case for lockdowns basically collapsed in May of 2020, when Sweden's epidemic began to retreat.

Sweden, of course, was the only major Western country that didn't lock down in 2020. And the argument for lockdowns made a clear prediction concerning what would happen there: since the country hadn't taken drastic measures, it would see substantially more deaths (relative to its population) than the countries that had locked down. Using a model "based on work by" Neil Ferguson's team at Imperial College, researchers at Uppsala University predicted there would be 96,000 deaths by July 1st.


Comment: The model based on Neil Ferguson's work was based on pseudo-science and designed to show a very high mortality rate, when in reality the virus is not much more deadly than a bad seasonal flu. Even Ferguson himself didn't believe in his own work, if his personal behavior is any indication.


Fortunately, that isn't what happened. The number of confirmed COVID-19 deaths by July 1st was only 5,370. And up to week 51, the country saw age-adjusted excess mortality of just 1.7% - below the UK and below the European average.

Now of course, Sweden isn't identical to the UK. It's more trusting, less densely populated, and has fewer multi-generational households. However, it isn't dramatically different from the UK in these respects. So even if one might have expected fewer deaths in Sweden than in the UK, given the same policies, the fact that Sweden didn't lock down should have massively increased its death toll. But it didn't.

One reply to the argument I've just made is that Sweden did much worse than its neighbours. This reply has been extensively addressed by other commentators, and in any case the point remains that Sweden did not do catastrophically. Both its first and second epidemics retreated long before the herd immunity threshold was reached, and far less than 1% of the population has died.

The evidence from Sweden does not imply that the correct approach to COVID-19 was "do nothing". As I've argued previously, a focused protection strategy like the one recommended in the Great Barrington Declaration would have been much less costly, and might have saved more lives, than the Government's actual policy of intermittent lockdowns.