1) We've been hearing that #Sweden is supposedly doing well without a substantial lockdown. This depends on how we define "well". The graph shows weekly deaths for all ages in 2020 compared to the average deaths in 2015-2019. By week 14/April 5, deaths in 2020 were up ~20%. 
2) This excess mortality is driven by the older and more vulnerable age groups 65-79 and 80+ years. The graphs show the weekly deaths in comparison to 2019. Deaths in the age group 65-79 are up by ~30% in 2020 compared to 2019 in the most recent calendar week.
3) Note that 2019 and 2015-2019 had been less lethal than 2020 for several early weeks before #coronavirus became a thing. These data are not from hotspots, but country-wide. The original file can be found at @StatsSweden: scb.se/en/finding-sta…
4) Do we know the excess mortality is entirely due to #COVID19? No. Do we know it would have been lower with a severe lockdown? No. But the excess mortality has been there in recent weeks and it is substantial. Looking forward to an update of the date.
Thread: Not sure Sweden is doing all that great in terms of excess mortality, @wwwojtekk @Noahpinion @tylercowen @willchamberlain @BachmannRudi @EconTalker.
5) And, of course, February 29 in 2020 is a nuisance that I didn't properly account for. 😉 Below are weekly averages of the daily (!) deaths that are not influenced by the missing 29th in the other years. Same tendency in the most recent weeks.
6) And once again, now with 2020 having the same color in all diagrams and the same line pattern (for the color-blind). I should have a checklist for doing graphs right from the get-go...
