Most Nordic countries flattened the curve
I’m not sure the precise definition of “Nordic”, but these countries seem to fit the bill:
Switzerland was hit hardest in its Italian section, so it’s less of an outlier than it appears. The Netherlands is clearly an outlier, although it has also flattened the curve.
In the US, the death rate appears to be falling rapidly. The average death rate has been about 5% of reported cases (although closer to 1% of actual cases.) Now the marginal death rate of newly reported cases is falling very fast, and may soon fall close to 1%. The actual death rate relative to total actual cases will likely fall far below 1%, albeit not as low as in Singapore.
A portion of this may be increased testing or better medical care, but it seems to be mostly due to illnesses shifting from the old to the young. Of course this is exactly what economic theory would predict, as the cost of socializing for the young is far smaller (much less risk of death) and the benefit is probably higher—romance!!
I don’t have any normative claims here, just some data I thought were interesting.
PS. Check out the link above; you can create your own graph.
PPS. The US is a bit above the Netherlands, with a steeper line. Most of the big European countries are similar to Sweden, if not higher. But Sweden is rising faster.
PPPS. Quebec, the northeastern US, France, Italy, Spain and Belgium were all hit very hard. What would happen if an epidemiological study put in a dummy variable for “historically Catholic”?
Tags:
27. June 2020 at 12:35
Sumner, who is white, once again oppresses with his words the brown peoples of Italy. I’ll be the first to note this…
Btw I’m looking for movie recommendations from Dr. Sumner, any suggestions?
27. June 2020 at 12:40
Are there differences between the Catholic and Protestant part of Germany? Your graph above doesn’t seem to support much of a Catholic effect. Austria is doing better than Germany, Lithuania and Latvia (which is orthodox but close enough) better than nearby Lutheran countries. It may more ethno-cultural, e.g. French and Italian places vs. German or Anglosaxon.
Also, do you think 20% of all cases are reported? From what I’ve read most people seem to think more like 10%.
27. June 2020 at 13:53
Mark, I guess you’d want to look at northern vs. southern Belgium to make that distinction. Anyone have that data?
27. June 2020 at 16:36
It is the strangest thing. The economy has been wrecked, and macroeconomists have become amateur epidemiologists.
Maybe it is too painful to watch.
27. June 2020 at 18:27
Scott,
Your theory is interesting and there seems to be some evidence for it: Ireland might fit your theory, so would Portugal and possibly Romania.
Within Germany, your theory might be correct as well, Bavaria is by far the most Catholic state, and behold, they are the worst off, closely followed by North Rhine-Westphalia though, which is more Protestant, and Baden-Württemberg, which is pretty much exactly 50:50, with Baden (Catholic) maybe being a bit less affected than Württemberg (Protestant), as far as I can tell, but I’m not 100% sure.
Also Belgium doesn’t fit, Flanders seems to be a bit worse off than Wallonia, which is perhaps not surprising, because Flanders is basically Dutch, and the Netherlands do not fit into your theory either, and neither does Sweden.
Maybe Sweden could be taken out of the ranking though because it is dominated by just one weird epidemiologist, who simply has too much influence and might change the course of the whole country.
What doesn’t seem to fit either is the UK, although one could maybe check if the “Roman regions” are more affected than the regions that were never conquered by the Romans.
And, as others have said, the old “Austro-Hungarian dominion” doesn’t fit either, namely Austria, Hungary, Poland, the Czech Republic and Croatia. Here one could examine whether there are differences in areas that were conquered by the Ottoman Empire compared to areas that were only conquered for a short time or never, Romania and Macedonia might then also fit into this scheme, but Greece might probably not.
South America would fit your theory as well, and so would Australia and New Zealand.
27. June 2020 at 19:11
Scott,
hilarious how the one throwaway phrase at the end dominates the discourse. I’ll take the bait too of course.
My first thought about the US patterns was, rather than catholic vs non catholic, it’s “socially cold” vs “socially warm”. That explains US South vs US North, even California. And the successes of East Asia vs less success in South and SE Asia. Basic idea, some cultures are more socially distant than others to begin with.
If I’d had to dig deeper as a social scientist I would look at social network structures and relate them to the pandemic. The virus spreads in clusters due to rapid infection of close contacts. It then jumps from cluster to cluster, less rapidly so, influenced by degree of overlap between the small local networks that define the clusters. In other words: the more a region has “small world” networks, the faster the spread. Regions with social networks that are les prominently “small world”, are see a slower spread.
So here is my “real” hypothesis:
– the spread patters can be explained by how much a region’s social networks are “small world”
– this is influenced by two main factors, the size of “local” networks and the degree of overlap between them.
Hope someone will pick up the idea and do the work 🙂 = Facebook and their ilk are natural data mines for kind of work.
27. June 2020 at 19:33
Christian, Thanks for the info on Belgium.
Yes, there are plenty of exceptions. If there weren’t, then everyone would be noticing it. I’m not even sure if “Catholic” is the right way of thinking about the cultural differences.
mbka, Interesting way of looking at it.
The data from Quebec was what really got me thinking. Quebec has almost 2/3rds of deaths and close to 90% of all the active cases in Canada
28. June 2020 at 01:05
The disease isn’t shifting to younger people. The young are just getting tested more. This is evident in the ratio of deaths between age groups. It has not shifted.
28. June 2020 at 08:39
rwperu, I doubt that. The surge in cases is very recent; too recent to be explained just by increased testing.
28. June 2020 at 09:48
Scott,
If you add France, Spain, and Italy to that graph, and then switch from cumulative to 7-day average, you get a very clear picture. From that point you can toggle back and forth between deaths and cases and the picture doesn’t change. The US and Sweden have made zero progress on containing the virus, while the rest of the western world crushed the curve. It’s not even close. The difference seems clear. Countries that try do better than those who do not.
3 or 4 months into this experiment, it is very clear that Sweden was mistaken and the results have been catastrophic. It also seems likely that states in the US that prioritized opening up were wrong.
At this point, can anyone argue that Sweden’s strategy wasn’t a mistake?
28. June 2020 at 18:07
Scott,
My prediction would be that when the people infected today eventually die the age distribution will be the same. If skews younger in the sort run that just means at some point in the future it will skew older. The virus will find the old and vulnerable. The idea that we can somehow “protect” then is ludicrous. We are too connected.
bb,
“Countries that try do better than those who do not.”-This about sums it up, does it not? We have a difficult problem that will take collective action so…let’s not even try.
28. June 2020 at 23:00
For Belgium, the picture is somewhat more complicated. The full data by province is available here: https://epistat.wiv-isp.be/covid/covid-19.html. Flanders was definitely hit considerably harder, but driven a lot by proximity to the Dutch border. As you might imagine, there are a lot of cross-border links with regions of common language. France went into lockdown relatively early, at a point where there was very little virus in Belgium, and Belgium quickly followed. As a result, there wasn’t much transmission from France. By contrast, the Netherlands stayed open for much longer. The news was full of reports of people driving across the border to go to shops and bars.
Brussels mostly got infected by people coming back from ski resorts in Italy. There are a lot of Italians in Brussels because of the European institutions. But the city is actually quite small, only a couple million people in the greater metro area, and not anywhere near the density of most European cities.
Nursing homes, especially in Flanders, were abysmally protected. In April, it started burning through them. At one point sciensano was breaking out those figures and they were 60% of the cases, though that was obviously somewhat biased by more comprehensive testing in those settings.
I suspect a good portion of the differences in death rates between countries comes down to how well protected their nursing homes / elderly were.
29. June 2020 at 03:30
@bb – the silence is deafening to your question. Reminds me of posting an anti-gun stat in a pro-gun forum. Discourse not gonna happen.
29. June 2020 at 03:42
It’s sort of besides the point, but for what its worth “Nordic” usually just means scandinavia + finland + iceland. Sometimes the Baltics are thrown in but it’s a stretch. Have never heard of the Netherlands or Germany or Switzerland in that category.
29. June 2020 at 06:16
Christian List: “Maybe Sweden could be taken out of the ranking though because it is dominated by just one weird epidemiologist, who simply has too much influence and might change the course of the whole country.”
You could actually look this up since you seem to have access to a computer. The Swedish system that opposed a lockdown has been in place for many years and hasn’t been dominated by “just one weird epidemiologist.”
Norway initially was going to follow Sweden’s example but bowed to international pressure to have a lockdown.
29. June 2020 at 06:22
bb wrote: “At this point, can anyone argue that Sweden’s strategy wasn’t a mistake?”
Sweden’s per capita Covid-19 death rate will lower than Spain, Italy, the UK and Belgium and about the same as France and the U.S. despite no lockdown.
Norway has low per capita deaths but an analysis last month concluded Norway would have had the same result without a lockdown.
29. June 2020 at 08:26
I think the major story of this phase of the coronavirus is the spread of the virus inside of controlled borders. It’s raging in the US, Brazil, India and Russia, all large countries. It’s contained in small countries, especially island countries. The outlier among the large countries is authoritarian China which has the will and capacity to cordon off parts of its own country and treat them as essentially different countries.
It’s not all about borders as evidenced by the example of authoritarian China and the contrast between Sweden and its neighbors. The will and ability to enforce public health measures can contain the spread. People in the large countries are now grappling with the realization that one of their greatest strengths–the freedom to travel unrestricted within their borders–has turned into a weakness. I don’t see the other large countries suddenly becoming as authoritarian or, if already authoritarian (e.g. Russia), suddenly becoming competent enough authoritarians to follow China’s lead. Instead I see them yoyo’ing between partial lockdowns and opening up, all the while desperately trying to convince their people to alter their behavior.
29. June 2020 at 08:38
rwperu. I’m old. Are you saying I’ll eventually get the virus? Why do you think that?
Thanks Dan, As far as spillover from the Netherlands, it’s worth pointing out that the Netherlands was not hit as hard as Belgium.
29. June 2020 at 08:40
Rob, I have a fairly expansive definition. 🙂
29. June 2020 at 08:43
Todd, You said:
“Norway would have had the same result without a lockdown.”
That’s quite plausible, but Sweden could have had the same depression with a lockdown.
Carl, You said:
“It’s contained in small countries, especially island countries.”
I’ve wondered about this too. Uruguay and Brazil is an interesting comparison. But there are tricky problems comparing the statistics of small and large countries.
29. June 2020 at 10:00
Scott wrote: “That’s quite plausible, but Sweden could have had the same depression with a lockdown.”
I’m not seeing a Swedish depression. It grew 0.4% annualized in Q1 and unemployment has gone from 7.1% in February to 8.0% in May.
29. June 2020 at 10:54
Todd,
You need to take your miracle supplements, they’re obviously not working again.
All the forecasts say that Sweden’s GDP will drop significantly, even the Swedish government says so.
No matter what nonsense you spin here, all predictions say that Sweden won’t do significantly better than Denmark and Norway.
Some current forecasts even say that they might do worse than Norway and Denmark:
https://www.bloomberg.com/news/articles/2020-06-16/one-economy-stands-out-as-crisis-reveals-striking-differences
We’ll see who’s right, Scott or you, when the second quarter numbers come in. They should be ready sometime in July.
29. June 2020 at 11:14
Oh, sorry Mr. List, I was dealing with reality in the present, not what forecasts predict by the end of the year.
29. June 2020 at 11:29
Carl wrote: “It’s raging in the US, Brazil, India and Russia, all large countries. It’s contained in small countries, especially island countries. The outlier among the large countries is authoritarian China…”
Russia’s population is 145 million and Japan’s is 126 million so also a large country (15% smaller than Russia) and one with 1,000 deaths as almost done for this round.
29. June 2020 at 11:34
One more on Scott’s “depression” in Sweden.
Unemployment may rise by 2 percentage points and GDP has been projected to be down 3.2% for the year before increasing 3.5% in 2021. What economist would call that a depression? I mean, besides Scott.
29. June 2020 at 11:59
Todd,
no you are dealing with the past not the present. The second quarter is almost over and you’re talking about the first quarter.
But maybe you are right, for example the first quarter in Norway and Denmark was worse than in Sweden indeed. On the other hand, Norway and Denmark might have been hit earlier than Sweden. Was Sweden even affected in March?
We’ll see soon, the figures for the second quarter come out in a few days.
However, one can assume from your belief in junk supplements that you like picking data a lot, so I still trust the more serious forcasters a bit more right now.
Sweden might have “restructured” its pension funds a bit more though, so this might be a “plus” from a strictly financial point of view.
29. June 2020 at 12:16
@Todd Kreider – the relevant stat is not death rates but infection rates, and, as such, Sweden has failed. The same population as Greece but 68k infected in Sweden vs 3k in Greece.
Put another way: C-19 possibly causes in young people, black toes, long-term lung damage, sterility in men and death to a certain percentage of young people. Even if you don’t mind that grandma dies, if you’re young, you do not want to catch C-19 and take your chances that you’ll be one of the lucky ones that do not experience symptoms.
29. June 2020 at 12:32
@Todd Kreider
I should have included Japan in my survey of the most populous countries (they’re 11th). They are definitely an outlier that is managing to contain the virus without Chinese levels of lockdown. I suspect it is related to Japan being uniquely law-abiding for so large a country: https://worldpopulationreview.com/country-rankings/crime-rate-by-country. I’d expect a nation that obeys laws in general very well to follow public health guidelines well. Japan’s crime rate is about one third that of the other countries in the top 11 per https://worldpopulationreview.com/country-rankings/crime-rate-by-country. And, I think crime rates are a good window into how well a country will follow public health guidelines. Interestingly, China’s crime rates are on the lower end of the big countries, albeit nowhere near as low as Japan’s. Plus, as I said before, China has shown the ability to force compliance to public health guidelines where people will not willingly comply.
There are a couple of countries who have lower crime rates than their geographical peers but not lower contagion rates (Spain, for example, has a significantly lower crime rate than Italy, France, Belgium and the other hard hit European countries). But the correlation seems to hold up pretty well. Interestingly, Sweden has a much higher crime rate than its neighbors Denmark, Finland and Norway.
29. June 2020 at 12:55
Theres probably a medicine tech part to. I’d say improved treatment my guess is reducing deaths 25-50%.
Younger people of course being willing to take risks.
Third reason I’m guessing theres correlation. People with strong immune systems are less likely to be infected and less likely to die. People infected in the first round of the disease were easiest to infect.
And slight chance virus mutating to less deadly which often happens though I put this at low probability.
29. June 2020 at 14:29
To me, it’s looking like we will need to suppress Covid in the end, because we’re not likely to get a polio-like vaccine that can eradicate it. My expectation is that we will need to use a combination of tools to suppress the disease to include contact tracing, vaccines, masks, social distancing, lock down measures, and global coordination. That last one makes be think we won’t make real progress until after January.
The Asian countries are doing great, but they are still living with serious restrictions: masks, no mass gatherings, older people still isolating… And it’s very possible that they could be hit by a second wave. I think we will need to apply all the tools at once, and we’ll need to do so in a coordinated way. I hope I’m wrong.
29. June 2020 at 16:37
Scott,
Without intervention? Yes. California is growing. That means without further action it is going to continue to grow until 70% of the population has been infected. Maybe closing bars and restaurants with universal masks is the secret formula? As of right now whatever CA is doing isn’t working.
If you were old in AZ it’s even worse. Right now about 2-3% of the population is getting infected every week…and growing. It doesn’t take too long to get to herd immunity at that rate. We just instituted mask mandates on the 22nd, bars just got closed today (June 29th), school pushed back a couple of weeks. That’s still not enough. We will need a full scale shutdown that’s much stricter than the first one (it was flat during those wasted months) or our hospitals will be completely overrun on our way to herd immunity (at least 35,000 deaths).
BTW, we’re also starting to ration care. Already. With a minimum of six weeks of pain in front of us. We are New York. We are Lombardy. This is where every place that is growing is headed without intervention.
29. June 2020 at 18:12
I wonder if developed economies would have had depressions without lockdowns. People seem to crowd into bars and restaurants as soon as it is legal—-so much so, LA County is again shuttering bars.
There were scattered, fleeting shutdowns during the 1957-8 Asian flu in the US, which had a similar death rate to C19, of about 0.6%.
The FRED charts show a small recession that coincides with the 1957-8 Asian flu, but then again that could be monetary policy at work, or other variables.
My guess is that without government lockdowns, as pioneered in mainland China, there would have been a recession in developed nations, but not the depression we are seeing. But perhaps an expansionary fiscal-monetary policy could have beat back a 2020 C19 recession.
I wonder where the free-marketeers and libertarians went?
You read the highly intelligent Tyler Cowen re C19, and he is indistinguishable from a statist martinet. Cowen works at a public institution in a state going Democratic—-is that it?
Perhaps freedom of expression in the US is dead.
I think the lockdowns are marginally effective against C19 in the long-run, but certainly economic disasters. Maybe the cavalry will arrive —-that is, a vaccine.
Good luck everyone.
30. June 2020 at 03:04
“My guess is that without government lockdowns, as pioneered in mainland China, there would have been a recession in developed nations, . . . ”
Here: https://mv-pt.org/wp-content/uploads/2019/12/Monthly-e-mail-1912-Global-money-round-up.pdf
is someone who disagrees?
“Elsewhere the message is closer to ‘steady as she goes’. The Eurozone also has enjoyed rather strong money growth in recent months, but this feature has not been as marked as in the USA. In China and India money growth in 2019 has been steady at high rates appropriate in these economies with strong underlying trend growth in output. (But India suffers from serious ethnic and religious tensions at present, which may have economic consequences.) Japan’s money growth is stable at a very low rate, while the UK has a money growth recovery in conjunction with a clarification of the Brexit process. The overall conclusion is that during 2020 the world economy will see at least trend growth of demand and output, after a lacklustre 2019. “
30. June 2020 at 05:44
44K new cases, 345 new deaths. I am pretty sure we have not had a ratio of new deaths to new cases that low. The next two weeks will be interesting—-although “interesting” seems inappropriate. It is highly likely deaths will rise——but how much? I know my “Feelings” on how we track this virus is that we are missing something significant—-that is why next two weeks are important.
So many essays on what is happening. For example, the belief that we really have close to 30 million cases is consistent with more testing. We are up to 100k tests per million—-that will increase by 50-100% in a month.
I look at New York and the decrease is remarkable. New cases is less than 5% of what they were 6 weeks ago—-deaths even greater decrease. One can either create a logic that fits the facts, or look at the facts and say there is no logic which fits them. As Scott surely knows, we have no testable hypothesis——just anecdotes of what is happening.
I just don’t believe we are looking at this correctly.
30. June 2020 at 05:53
Carl wrote: “I should have included Japan in my survey of the most populous countries (they’re 11th). They are definitely an outlier that is managing to contain the virus without Chinese levels of lockdown. I suspect it is related to Japan being uniquely law-abiding for so large a country”
Yet that doesn’t begin to explain per capita Covid-19 deaths in Japan being 50 times lower than in the U.S. or Western Europe. The Japanese government asked people to stay home to the extent possible in the big cities on April 7 and for the entire country on April 16 – weeks after coronavirus had been in the country. The South Korean government asked people to stay home in late February and supposedly did “a lot” of testing, whatever that means, whereas Japan tested less than the United States but got almost the same number of deaths per capita as the lightening fast Koreans.
As I’ve said many times, it might be a major clue that Japan and South Korea had 100 and 200 H1N1 deaths respectively during the 2009/10 pandemic that killed 8,000 to 18,000 Americans.
In Australia, 100 to 200 died of H1N1 according to two different health organizations there. Some think there were really closer to 1,400 deaths then if using the U.S. CDC model, which would bring Australia to the high end of deaths in America. Maybe the same was true of South Korea and Japan. No idea.
30. June 2020 at 06:09
Michael Rulle wrote: “I look at New York and the decrease is remarkable. New cases is less than 5% of what they were 6 weeks ago—-deaths even greater decrease.”
This isn’t at all surprising for typical epidemics. Back in February, I didn’t know much about pandemics except that they almost always creep into an area, exponentially spread and hit hard and then go into decline. I thought in early March that the last part of March and the first two weeks of April would be quite bad and then the worst would be over by Earth Day toward the end of April and the country would open up in May. There was no way 300 million Americans were tolerate a lockdown until a virus was found as Obama advisor Ezekiel Emmanuel and epidemiologist Michael Osterholm insisted needed to happen.
Despite lockdowns being only somewhat effective at slowing down a virus, they have become religious entities in this mass hysteria.l (Of course, there has been very good reason to be concerned if elderly, have a heart condtion, etc.)
30. June 2020 at 06:16
Also, what ever happened to “flatten the curve”? Wasn’t there a presumption that until a vaccine or cure was found, the key was to not have hospitals over crowded. Now, it seems to be about eliminate deaths. But we know there is a trade off. We cannot eliminate deaths except by quarantine. And the more extreme the quarantine, the worse it is economically.
We are obsessed. Trump may be a cat, but his instinct on opening the economy is correct.
30. June 2020 at 07:16
Michael,
Well, countries like Brazil, USA, and Sweden seem to have a problem with the easy part already, i.e. the flattening aspect of the strategy:
https://upload.wikimedia.org/wikipedia/commons/1/1d/COVID-19_Active_Cases_per_100_000_population.png
30. June 2020 at 07:24
@Todd Kreider
50x is a big difference. And sure, the Japanese might have some prior immunity to SARS-COV2 that explains the 50x difference( although I haven’t read of any clinical evidence of it), but it also may simply be that 50x is not as big as it sounds when talking about something that can spread exponentially. It doesn’t take that many drunken maskless wahoos in bars or shouting maskless protesters packed in tight quarters to kick off case spikes around a country.
30. June 2020 at 07:30
The interesting variable we are all missing is “number of people an average person gets in close enough contact to spread the virus”. This is why Spain’s numbers are so bad: Ignore population density, which counts a lot of empty land, and instead look at how many people live in close quarters and use public transportation. Madrid looks like New York, because everyone spends a lot of time in very high risk areas. In that kind of environment, the number of cases you need before the spread goes out of control is very small.
What is interesting now is American regional differences. Why Florida? Why Georgia? Why Houston?
30. June 2020 at 08:36
Carl wrote: “I should have included Japan in my survey of the most populous countries (they’re 11th). They are definitely an outlier that is managing to contain the virus without Chinese levels of lockdown. I suspect it is related to Japan being uniquely law-abiding for so large a country”
But Japan’s voluntary distancing after coronavirus had been in the country for weeks doesn’t explain how there were 50 times fewer deaths than in the U.S. and Western Europe. The South Korean government asked people to stay home to the extent possible in late February and apparently tested “a lot” – whatever that means, so the story has been that all that testing led to deaths per capita similar to Japan’s, which did little testing and less than the U.S.
Why did 100 Japanese and 200 South Koreans die in the H1N1 pandemic with no extra distancing or testing while the U.S. had 8,000 to 18,000 deaths?
30. June 2020 at 08:40
@Carl – Sorry, that was an earlier version. I’d say 50 times fewer deaths is enormous and almost all Western countries have a pretty narrow range of deaths with Belgium at the high end and Germany at the low end and others quite close like the U.S., Sweden and France.
30. June 2020 at 08:56
@Todd Kreider
When left unchecked by public health measures, the virus ripped through Wuhan like a New York City nursing home. To explain the difference between contagion rates in Wuhan and the rest of East Asia without positing that it was due to differences in public health measures, you’d have to make the claim that the Wuhanese lack a background immunity to SARS-COV2 that everyone else in East Asia has.
30. June 2020 at 09:30
@Carl
I don’t think enough is known about Wuhan or even China as a whole with respect to this virus. I wish China had a mostly open government, but it doesn’t. I’m always open to good information, though.
There is also a difference between public health measures and public health in general. The public health in general is far better in Japan and South Korea than in China (or South East Asia) yet a lot of testing hasn’t given South Korea a better outcome than Japan.
30. June 2020 at 10:00
@Todd Kreider
I agree with your distinction between public health and public health measures. Obesity, diabetes, smoking and immunosenescence, for example, are factors of public health that affect fatality rates. I’m making the claim that lawfulness is a key factor of public health in determining spread within a country. And, I would say that all these public health factors (and a couple of others of course) determine how well public health measures work within a country.
1. July 2020 at 11:03
« Mark, I guess you’d want to look at northern vs. southern Belgium to make that distinction. Anyone have that data? »
i’m taking that you mean flanders is protestant and wallonia is catholic? That’s wrong, there are almost no protestant in belgium overall (around 2% of the population), belgium is catholic throughout.
interesting post though!
1. July 2020 at 15:58
unanimous, I know that. I meant Flanders uses a Northern language and Wallonia uses a Romance language.