Saturday, 19 September 2020

Covid-19 deaths per million population (Europe)

Stat's from Worldometers, chart created using

Make of it what you will. There seems to be a gradient between north-east (we can put Sweden to one side, they only did the mildest of lock downs) and south-west (reasons?).

What baffles me is the range - 7 deaths per million in Slovakia up to 857 deaths per million in Belgium? I accept that some countries fared better than others, but a ratio of hundred-to-one? Are these numbers even credible?

Click to enlarge:


Piotr Wasik said...

Making sense of covid numbers is very difficult, for example Belgium is an outlier as they count suspected Covid deaths, not only confirmed: "For example: if you have one or two confirmed cases (in a care home - PW), then the week after you have 10 deaths in the same home based on similar symptoms." from

I gave up on counting in the UK, I only rely on week-on-week 7 day rolling average changes, assuming that from week to week testing methodology is not changing much. So now, I believe the number of cases is going up (from one week to another), but I don't dare to compare current numbers with April ones - too much has changed and nobody is normalising these numbers.

Ralph Musgrave said...

Slovakia was the first EU country to insist on wearing face masks on public transport. Source:

George Carty said...

Northern Italy was the very first region in Europe to be hit by the virus, because Milan's fashion houses have close links with manufacturers in China: the city has a direct flight from Wuhan.

Much of the rest of Europe was seeded with the virus by holidaymakers returning from February half-term skiing trips to the Italian Alps: this explains the low prevalence of the disease in eastern Europe (where people are mostly too poor for Italian skiing holidays) as well as in Norway (which is full of snow-covered mountains that allow enthusiasts to enjoy skiing in their own country).

Mark Wadsworth said...

PW, yes, there was over- and under-counting, partly accidental and partly deliberate.

RM, thanks, makes sense.

GC, thanks, also makes sense... BUT what about Spain? Do Spanish people go ski-ing in Italy in winter?

Mark Wadsworth said...

GC, also, my impression is that Germans like going to Italy on holiday, being a few hours drive away, and their death rates are quite low.

benj said...

Easy to get statistical noise with such small numbers. We are in all likelihood looking at a disease that is no more fatal than flu for the under 50s.

How many countries failed to identify care homes as such as risk? Not all apparently.

Mark Wadsworth said...

B, that might be part of the answer. The March/April excess deaths were real and observable. Nowadays it's not even rounding errors.

Agreed on care homes, that was and is an outrage.

benj said...

@ MW

talking about outrage, have you seen this?

George Carty said...

Don't know about how Spain was initially infected, but in France the main clusters were in Paris (a global city) and in Mulhouse (where the super-spreader event was an evangelical church gathering, with hymn-singing almost surely being the direct culprit).

Germany wasn't as badly affected both because its skiing enthusiasts were younger (and thus both had more natural immunity and were less likely to have parents in care homes) and because (as you point out) they tended to travel by car rather than by plane.

Mark Wadsworth said...

B, that's hilarious! Do these people not realise how stupid they look?

GC, yes, all good points. They sound perfectly plausible, but did you make up the bit about German skiers being younger?

Sobers said...

The council banning home workers from smoking thing will be more of an insurance arse covering exercise than one of heavy handed employer ideological overreach. The lawyers will be worried if they don't formally ban smoking in 'the workplace' (which just happens to be the employees spare bedroom now) they would be liable for employees who got cancer from smoking claiming on them.

ontheotherhand said...

The Law of Small Numbers explains Belgium compared to bigger countries, or BBC news about a City outbreak in the UK. This is concept of Probability that says you will see the most variation in small sample sizes. In the US for example they found that small schools were over represented in the best results. Bill Gates backed charities and government ventures to break big schools into smaller schools. However, if they had checked, they would find that small schools are also over represented in the worst providers.

A small remote hospital might have one day when all of the 8 babies born are boys, but a city with 1000 births a day will never get all boys. Belgium is a small sample that flipped a lot of tails.

Mark Wadsworth said...

OTOH, that logic certainly applies to micro-states like Vatican, Faroe Islands, San Marino. Those countries all have very high or very low covid-19 rates, with no discernible pattern.

But Belgium is not a small country, it's 11 million people, only a bit less than average for an EU member state.