Yeah but I do think there is merit in comparing Sweden to neighboring countries such as Denmark.

However, luck is also a factor that shouldn't be disregarded. If anything was learnt from the attempts of modeling the development of the pandemic it should be that the virus does not diffuse through the population in some idealized fashion. It's extremely messy and depends on local circumstances and sheer chance.

This can easily be confirmed by comparing different regions of countries. Despite similar regulations the variation is enormous.

Only if you ignore other factors like demographics and population density of population / population-weighted density.

Countries should be evaluated based on performance relative to doing nothing not relative to other counties in very different situations. Niger has a median age of 15.4, Italy has a median age of 45.5 they simply have vastly different ratios of the elderly etc.

Population density (on a national scale) is not relevant. Sweden is an urbanized country, like most (all?) of the western world. More than 85% of Swedes live in urban environments: https://www.statista.com/statistics/455935/urbanization-in-s...

For disease transmission it’s the average population density of the population not the country. Effectively the population density of each area * population in that area divided by total population.

So, assuming 3 areas 1,000 square miles with 2,000 people, 100 square miles with 4,000 people and 10 square miles with 8,000 people you do : (2,000^2/1,000 + 4,000^2/100 + 8,000^2/10) / (2,000 + 4,000 + 8,000) = 468.8 people per square mile, not (2,000 + 4,000 + 8,000) / (1,000 + 100 + 10) = 12.6 people per square mile.

The reason is adding a huge area that nobody lived in (ex: 1 person across 1 million square miles) doesn’t change the calculation. (1^2/1,000,000 + 2,000^2/1,000 + 4,000^2/100 + 8,000^2/10) / (1+2,000 + 4,000 + 8,000) = 468.8 people per square mile

“Urban” Sweden is much different from other countries. For example, someone else urged us to compare Sweden to France. Yet Paris has a population density of 20000/km^2 versus Stockholm’s 4000/km^2. I’ve lived in places with both densities and one is considerably more “urban” than the other.

If you draw the border of Paris further and further out, I’m sure you can get the density as low as you want. Meaning: the density figure of a city says very little of how dense it is and says more about how far out into sparsely populated farmland the official city limit ends.

Go look at a Google street view of the far ends of Stockholm metro area…

Paris has a population density of 9,800/sq mi across 1,101.7 sq mi and an inner core of 53,000/sq mi across 40.7 sq mi.

Stockholm’s urban area of 11,000/sq mi across 147.35 sq mi, is comparable to Paris, but it’s inner core 73 sq mi is a much lower density 13,000/sq mi. Even if you assumed all those people lived in a 40 square mile “inner inner” core you still don’t get close to 53,000/sq mi.

Is that land area or area including water?

Population density is indeed important, but averaging over whole countries gives the wrong impressions.

Covid was deadlier in urban areas. Compare rural regions, such as the ones bordering Norway and Sweden, and the mortality is pretty much comparable across the border. Slightly on the Swedish side, but well within what can be ascribed to random variability. That doesn't make the border closings look especially effective in retrospect. The virus doesn't care about borders.

A different calculation of density works fine: https://news.ycombinator.com/item?id=30812118

We care about the population density of population, not the population density of land. It’s sometimes call the population-weighted density.