Today in is social distancing worth it, lets check in on Sweden.

Submitted by ak47 on May 8th, 2020 at 10:20 AM

I think its important to have an educated discussion about social distancing and its value, I tried to do that yesterday and think it was somewhat successful and I'm going to keep trying. I should note, if you watched plandemic and think it made some good points just don't bother posting, you are stupid and your views are worthless. Now that we've gotten that out of the way lets check in the natural experiment Sweden has provided to see if the partial re-openings many states are moving towards right now have a strong evidence of support.

For comparison purposes throughout I will compare Sweden to Norway, Denmark, and Finland since those countries most closely resemble Sweden in terms of culture, density, average health of citizens, and economy which provides a good baseline of comparison. Trying to make Sweden look bad by comparing it to South Korea doesn't make much sense from a natural experiment, neither does trying to make it look good by comparing it to Italy.

First lets check on the economic impact. Obviously one of the arguments about re-opening is it is necessary to save the economy. Recognizing that we live in a global economy every economy was going to shrink, including Sweden's regardless of what they did. However, with more domestic activity the hope would be that Sweden's economy shrank by less and showed lower unemployment than its neighbors. That is could things being open with social distancing provide enough economic value to those places to really matter. Lets check the numbers.

Sweden: GDP expected to shrink by 7%

Norway: GDP expected to shrink by 5%

Finland:GDP expected to shrink by 6%

Denmark: GDP expected to shrink by 6.5%

So not a great start but I haven't done all the research and maybe Sweden had a uniquely export based economy even relative to its neighbors. However, at the very least Sweden's economy is not doing significantly better than its neighbors as the result of staying partially open.

So lets move on to the public health impact. Since total positives is really a testing driven number, the most accurate way to track the virus's impact is through deaths and deaths per million.

Sweden: 3,175 deaths and 314 deaths per million

Norway: 217 deaths and 40 deaths per million

Finland: 260 deaths and 47 deaths per million

Denmark: 522 deaths and 90 deaths per million

So these are pretty ugly numbers. Sweden has 3x the number of deaths as Norway, Finland, and Denmark combined. It is pretty clear from these numbers that aggressive social distancing has an impact on the total number of deaths in a country. Whether that trade-off in deaths is worth it for what society is giving up can be debated, whether or not it saves lives can't really be.

There is going to be a lot of retrospective looks at things once this is all over and much of the story has yet to be written including the potential of a second wave, but for now the natural experiment that Sweden and its nordic neighbors provide has given us these numbers. In my opinion its pretty clear that in a globalized economy, staying partially open isn't enough to have a meaningful impact on the economy and it does cost lives. In contrast to Sweden, Norway announced today a plan to essentially re-open the entire economy by June 15th, meanwhile Sweden had more deaths yesterday than Germany.

SugarShane

May 8th, 2020 at 10:26 AM ^

Comparing Sweden to only its Scandinavian neighbors is nonsensical to me. It’s just cherry picking to support your narrative. 
 

Your last sentence is another example of cherry picking. You say they had more deaths than Germany Yesterday ? I thought you’re not comparing those populations. What about Spain and Italy?
 

the world is more than the cluster of neighboring countries. When you talk about US states, do you only compare Massachusetts to Vermont and New Hampshire and call data out of Louisiana and Florida meaningless?  
 

All data from all countries needs to considered. All responses by different countries need to considered. Location, climate, everything. Limiting your analysis to a tiny sample size just undermines your conclusion
 

 

SugarShane

May 8th, 2020 at 10:37 AM ^

I have made multiple comprehensive posts on here about Sweden and the data

 

theyre faring much better than Italy, uk, France and Spain while worse than their immediate neighbors. They still have 30% capacity in their ICUs.  They are 3 weeks past their record death and new case numbers, and now on a slight downtrend in new cases and deaths. Still more of a plateau overall but they succeeded in flattening the curve 

rob f

May 8th, 2020 at 10:44 AM ^

Not all countries are "created equal" when it comes to ICU capacity.  The US has relatively low capacity compared to most other developed nations, largely because profitability is a higher priority here than maintaining extra reserve capacity for times like these. 

WindyCityBlue

May 8th, 2020 at 11:26 AM ^

It could be argued that the US is one the best prepared with regards to ICU beds per capita. 

https://www.statista.com/chart/21105/number-of-critical-care-beds-per-100000-inhabitants/

 

We were also one of the best when setting up field hospitals to prepare for overflow.  Many of these were rarely if ever used which highly suggests our ICU capacity has been optimal compared to other nations. 

https://www.npr.org/2020/05/07/851712311/u-s-field-hospitals-stand-down-most-without-treating-any-covid-19-patients

BoFan

May 9th, 2020 at 12:46 AM ^

It turns out the source used for the US data in your Statistica chart is very different than the source used for the European data so much so that the numbers are not comparable.  The study used for the European data in the statistica chart actually stated that there are no standards across European countries where they can accurately count and compare actual beds and that access to data is limited.  Further the European study doesn’t count ICU beds. It counts a combination of ICU beds and other acute care beds.
 

The US study is focused on US regional differences in terms of ICU beds per population. But they are not actually measuring a national average.  That study does two things that indicate Statistica is overstating the US number compared to the European data. The first is that they don’t just take the total ICU beds divided by total population, they only are looking at about 2/3 of the population. Second, it appears that the ratio used by Statistica is a regional median and not a national average. Both of these issues in the US data means that statistica is overstating the ICU beds per population compared to the data from the European study.  All of this says that the data is severely flawed as a comparison.  
 

I only looked into the studies because I was surprised that the US had such a high number of ICU beds compared to Europe. The physicians per hundred thousand population are the opposite. I would’ve been fine if it was a valid study that was specifically for ICU beds in every country. But I am even more surprised that statistica did such a poor job of compiling different sources of data and then stating that the chart represents ICU beds when it’s actually does not.  

At the end of the day, the US may have more ICU beds, but this source from statistica provides no facts to support or refute that statement.

WindyCityBlue

May 9th, 2020 at 7:06 AM ^

So what you’re telling me is that different countries use different methods to measure/analyze things? Shocked I tell ya!

Next thing you’re going to tell me is that different countries use different ways to determine if someone dies from COVID or not.  And that China’s numbers are way off. 

ak47

May 8th, 2020 at 10:54 AM ^

No offense but you clearly have no understanding or training in setting up natural experiments and social science research. If you are trying to measure the impact of something you want to eliminate as many confounding factors as possible, not increase them. You also don't know much about Europe if you think Italy and Sweden are comparable in things like density, cultural norms, average age, etc.

For example, if heat has an impact on the virus comparing Sweden to an equatorial nation would make your findings useless because you don't know how much was driven by policy versus difference in temperature. Its the same reason just comparing Mass to Florida isn't going to be super helpful. Just like comparing NYC with high rates of public transit wouldn't make sense to compare to Houston because everyone there drives. Same thing with the average age of a country, the virus kills people who are older more. How much is the change in the per capita death rates between Italy and Sweden a function of average age of the population versus policy. The less factors you can control for, the less helpful your analysis is. Therefore, a subset of a few countries that most closely match each is the ideal sample.

SugarShane

May 8th, 2020 at 11:01 AM ^

I guess we’ll see.
 

Sweden’s goal was to let this burn faster and end sooner while other countries have this linger for a longer time. 
 

No point in arguing now, the results to date are to be expected two months into the experiment. for all we know, Sweden’s in the fourth quarter while the rest of their neighbors are just getting started. 

LewisBullox

May 8th, 2020 at 12:51 PM ^

It's not apples to oranges in terms of the countries included. As others have said, it makes a lot more sense to compare Sweden to Norway and Denmark than Italy.

The picture does not change if you include other non-Scandinavian countries similar in population and GDP like the Netherlands and Switzerland. Beyond those two, you start running low on apples to apples non-Scandinavian countries.

In terms of strategical approaches, I agree there is nothing definitive that can be said yet.

J.

May 8th, 2020 at 12:30 PM ^

It's not a status report when it's presented without the additional context that would be necessary to understand the data.

Here in Texas, they sometimes use the programmable road signs to say something like "2,631 deaths on Texas roads this year."  I mean, that sounds bad -- but I have no way of knowing if it's an improvement, if things are getting worse, etc.

The old saying about "lies, damn lies, and statistics?" That comes from cases like this -- using statistics without context.

blue in dc

May 8th, 2020 at 12:57 PM ^

I keep hearing that Sweden’s plan was herd immunity, interestingly, key people in Sweden deny that.

“Anders Tegnell, chief epidemiologist at Sweden’s Public Health Agency – the nation's top infectious disease official and architect of Sweden's coronavirus response –denied that "herd immunity" formed the central thrust of Sweden's containment plan, in an interview with USA TODAY.”

https://www.usatoday.com/story/news/world/2020/04/28/coronavirus-covid-19-sweden-anders-tegnell-herd-immunity/3031536001/

 

“Tegnell told The Daily Show's Trevor Noah (scroll down to watch the video) that accepting a higher death toll initially if it meant a better result in the long run was never part of Sweden's strategy when it came to fighting the novel coronavirus. ‘
 

We calculated on more people being sick, but the death toll really came as a surprise to us," he said. "We really thought our elderly homes would be much better at keeping this disease outside of them then they have actually been.

https://www.thelocal.se/20200506/coronavirus-in-sweden-the-death-toll-really-came-as-a-surprise-to-us

 

BroadneckBlue21

May 8th, 2020 at 1:28 PM ^

"There is going to be a lot of retrospective looks at things once this is all over and much of the story has yet to be written including the potential of a second wave, but for now the natural experiment that Sweden and its nordic neighbors provide has given us these numbers."

He acknowledges that his data is not comprehensive with regard to time. He is not ignoring it. He also continues to contextualize his own conclusion as "my opinion." If one is a reader with good comprehension skills, he's done his job in acknowledging his limits. Now, you, you don't like what the current data suggests because it does not align with your view, so you are emphasizing fallacies where there is none.  Incomplete data does not make for fallacious statements when the data is prefaced as such. You can be outraged--but if the numbers hold true over time--then what is your deal? What will you think? Will you be man enough to say, well, damn, "herd immunity" theory didn't work for Sweden in comparison to its neighbors? 

Right now, you are creating the fallacy because you are relying on the data favoring your own conclusion for what it will be in a year. You are the one projecting the OP's stance in a year. You are straw manning the argument to some distant hypothetical couched as your truth. 

bronxblue

May 8th, 2020 at 10:59 AM ^

Actually, most of what you said at the top as being non-sensical is exactly what people who analyze this do.  They try to find a data set that is most representative of the question they are trying to answer.  It's why people are pointing out, for example, that the US "plateau" seems mostly due to NY, MI, and WA (amongst a couple of early-hit states) declining while there are surges in other areas/regions, so assuming the worst is over may be premature.

It's not cherry- picking data as much as trying to do a thorough, reasonable analysis to find the actual answer.

ScooterTooter

May 8th, 2020 at 11:14 AM ^

Of course, the people talking about "surges" and the rest of the country having an increase in cases  always fail to mention the increase in testing + who is being tested. To not mention this would be...cherry picking data correct?

Here's Nate Silver mentioning this:

https://twitter.com/NateSilver538/status/1258404881347510272

Kent County is going out of their way to test the homeless and nursing homes (which is smart and should have been done 45 days ago). So is Kent County experiencing a surge or are they just catching more cases? 

Did you know that we tested more people than ever before yesterday?

https://covidtracking.com/data/us-daily

Pure coincidence that our case count increased over the prior 3-4 days. 

Here is New York's daily case count.

Funnily enough, it increases when they test more!

https://covid19tracker.health.ny.gov/views/NYS-COVID19-Tracker/NYSDOHCOVID-19Tracker-DailyTracker?%3Aembed=yes&%3Atoolbar=no&%3Atabs=n

 

Rabbit21

May 8th, 2020 at 11:39 AM ^

Yep, more tests = more cases and that was always going to happen.  The flip side is more exposure is going to lead to more cases as well.  If you want to use these spikes as a reason to keep up your own version of social distancing, go for it, but for a lot of people "waiting for a vaccine" is simply unsustainable.  Give people good information in order to make informed decisions, its the only way out of this as the dam is already breaking.

bronxblue

May 8th, 2020 at 1:00 PM ^

I'm not arguing about case count; obviously the more you test the more positives and negatives you'll find.  Now, the ratio from this testing (you'd hope) would start going down the more you go on.  The fact it hasn't really started to change despite the increase in testing is a little dispiriting; it means that previous testing was more representative of the population than we had hoped, and I think at the time the assumption was these were the most obvious patients.  The silver lining there, I guess, is that perhaps more people have been exposed than expected and didn't have particularly harsh responses to the disease.  So it's a mixed bag on that front.

But most people don't focus on just raw testing numbers; they look at things like hospitalizations (which do have some limitations due to capacity and social factors) and deaths (which do rely on proper categorization).  And in that case, death rate nationally has been persistently high especially if you factor out the NYC area.  That's a problem because of the long tail of this disease (it seems like the average death occurs 3-4 weeks after exposure).  And we've been social distancing for some time longer than that, which could mean other regions are going to be hit harder as regions like MI and NY (who were exposed earlier) die down.  

  

ScooterTooter

May 8th, 2020 at 2:36 PM ^

Except it has:

Three weeks ago we had ~1.2 million tests and 168k positive tests.

Last seven days we had ~1.7 million tests and 170k positive tests.

So a 14% positive rate has fallen to 10% positive rate in two weeks. 

What's likely is that our peak was far higher than the 30-35k cases we were seeing in April and that we've come quite a bit further down the curve in actual cases. 

The death rate is kept high because we keep finding more and more elderly people dying in nursing homes or at home (likely due to heart attacks, strokes or perhaps even COVID) that are being added to the totals.

I will keep repeating this: The disaster is that we ignored the most obvious dangerous infection point to take a broad approach because "experts" and their megaphone in the media spent too much time telling us that this was dangerous to everyone, that schools were major infection vectors and that we needed to close beaches and parks instead of focusing on the thing that has been obvious from the beginning: This is extremely deadly to the elderly and they should be protected. 

ZooWolverine

May 8th, 2020 at 11:50 AM ^

I think your comparison is undermining your point. I would expect Massachusetts is often compared to other states in New England, because they have a lot in common. I certainly often read articles comparing Michigan to other midwest states, or Michigan's economy to other rust belt states. It's pretty logical to do so. Comparing New York and Montana would give little information about how one handled COVID-19 vs the other.

Second, it wouldn't make sense to compare Sweden to Italy because the question is really whether proactive shelter-in-place is a good idea. Italy only shut down after becoming totally swamped, not proactively. If you really want to include Italy in the analysis, it should get grouped with Sweden for trying to remain open until it couldn't anymore, not compared against it. (In fairness, Italy got hit so hard so early that it probably makes more sense to leave it out of the comparison entirely--maybe I'm remembering wrong, but I don't think most countries had a coherent response plan in action before it was already too late in Italy.)

 

ChuckieWoodson

May 8th, 2020 at 10:27 AM ^

In today's edition of winning over hearts and minds, AK47 tells you you're an idiot for watching a video (and if you think it made good points you're a shithead).  Great way to start a post, champ.

ak47

May 8th, 2020 at 11:23 AM ^

I’m trying to have a reasonable discussion with reasonable people. If you believe anti vaxxer propaganda and clear lies because it’s on YouTube or your friend shared it on Facebook you aren’t a reasonable person worth discussing this with. I’m not trying to have a discussion on the relative value of social distancing with my 5 year old niece either 

ChuckieWoodson

May 8th, 2020 at 12:18 PM ^

In regards to the comment as it pertained to the video specifically, I didn't take it personally at all - there are just other, less insulting ways of saying that - and I've found that if you do (are respectful), there's actually a small chance of making people reexamine what they think is truth.  Insulting them puts up the wall immediately. I realize the OP doesn't care about that as evidenced by his post but just a personal pet peeve of mine.

jasgoblue

May 8th, 2020 at 10:28 AM ^

Can you please share links? With all the FUD and conspiracy theories around this thing, it would help to have links to reputable sources for the stats. I'm genuinely interested in the data.

ThereWillBeNoHugs

May 8th, 2020 at 11:51 AM ^

I read an article in the 5-2-2020 issue of The Economist that is complimentary to the economic point he made about Sweden, although it lacks stats. It's interesting nonetheless. Here is a paragraph from the piece titled Not All Quite There: "Some indication that the spending effects of a lockdown will persist even after it is over comes from Sweden. Research by Niels Johanssen of Copenhagen University and colleagues finds that aggregate-spend-ing patterns in Sweden and Denmark over the past months look similarly reduced,even though Denmark has had a pretty strict lockdown while official Swedish pro-visions have been exceptionally relaxed.This suggests that personal choice, rather than government policy, is the biggest fac-tor behind the drop. And personal choices may be harder to reverse."

 

CaliforniaNobody

May 8th, 2020 at 10:35 AM ^

I literally could not care less what some guys on MGOBlog thinks about this. I am listening to the scientists who say we need to keep distancing. Lots of annoying boomers insisting kids go back to work.

GET OFF YOUR H…

May 8th, 2020 at 11:34 AM ^

Eyes, ears, and a functioning brain is a start.  Think back to early March.  We were basically thinking a couple weeks of lockdown...here we are 2 months later.  They could have basically just put a shrug emoji up for every news conference at the beginning.  Predictions being made based off of models taking staying home and social distancing into account, numbers aren't even close to what the models were predicting.  It probably would have been more productive if at the beginning they just said, frankly we know this is a big time issue and we have no clue what is going to happen.