There are no mask opponents in a foxhole
On March 1st, I asked why masks don’t help average people. In the comment section, I was shot down by people who assured me that masks are not appropriate for the general public.
If the NYT represents the conventional wisdom, then I think it’s fair to say that the tide is turning on this question:
More Americans Should Probably Wear Masks for Protection
Experts have started to question whether masks may offer at least some protection to healthy individuals and essential workers.
“Started to question”? I started to question 4 weeks ago.
I’d also like to get something else off my chest. Please remember that observations are not predictions and also that observations are not endorsements. If I observe a surprisingly low death rate in Japan, or Germany, or China, or Taiwan, or anywhere else, it is not an endorsement of their public policies. How could it be, given that these countries pursued radically different policies (with tight controls in China and more laissez-faire in Japan.)
Nor are observations equivalent to predictions of the future course of the epidemic. South Korea, Iran, and Italy all had the epidemic under control in mid-February. In all three countries the epidemic was out of control by the end of February. I have no crystal ball.
I still think some of my observations (such as the mortality gap between southern and northern Europe) are interesting, and I’ve noticed other people discussing the same data points.
If I talk about the data from a particular country, it does not mean I believe the data is accurate; rather that it’s accurate enough to make my point. I’m sure that both Italy and South Korea have missed a number of cases, but the data is accurate enough to see that South Korea has been doing much better than Italy in recent weeks.
If I harshly criticize the new fiscal stimulus, that does not mean I would have voted against it. While it contains many objectionable components, such as $1200 checks to people who are still employed and corporate bailouts, it also has good features such as improved unemployment insurance and a Treasury backstop for the Fed. I probably would have held my nose and voted for it.
If I say that I’m not convinced that warm weather will solve the problem, it doesn’t mean I don’t believe that warm weather will help moderate the problem.
When a pundit says that a huge number of people would die if no precautions are taken, it doesn’t mean he’s prediction a huge number of people will actually die.
Tags:
28. March 2020 at 11:10
It is perfectly legitimate to say that there aren’t enough masks for everyone and that the first priority should be to insure that the medical people taking care of patients have enough. Hardly anyone would disagree with that.
But that’s not what people did. Instead, they claimed that masks are useless to most people and that you’re a covidiot if you disagree. Even people as dim as Peggy Noonan saw through that one.
28. March 2020 at 11:14
The biggest objections I’ve seen to the use of masks by the general public are:
1. We should be reserving masks for healthcare professionals, while there’s a shortage in that sector.
2. Many in the general public use the wrong types of masks for the wrong purposes, and then may engage in behavior that’s riskier for infection than they realize.
3. Even many in the general public who use the right kind of masks don’t know how to use them properly, meaning putting them on properly, failing to understand their limitations, and failing to use proper hygiene after using the masks.
For example, some fail to realize that, especially if not worn properly, some contagious material can slip out the sides, particularly if they take no proper care when coughing or sneezing, for those who may not realize they have the virus, but have mild symptoms.
Also, failing to wear the masks properly will reduce their effectiveness at protecting the wearer, who may engage in riskier behavior than warranted with a false sense of security.
I fail to see how people in the general public can fail to benefit, however, if they wear the right masks, the right way, and behave in an informed manner, as long as there are no shortages among those who need them most.
28. March 2020 at 11:17
Masks cannot be that important if you look at the death figures in Germany. Close to no one wears a mask here. Conventional wisdom also says that keeping the distance and washing your hands should do the job.
I don’t think anyone has ever said that masks do harm, unless of course you neglect other measures like washing your hands and keeping your distance.
If one still gets too close to strangers these days, without being an essential worker, then you are doing something wrong entirely.
I don’t think the NYT article is well thought through because it recommends something that is sold out everywhere anyways. They should recommend something to people that they can actually implement and that does no harm to healthcare workers.
28. March 2020 at 11:28
@Jeff
The result would have probably been the same: some individuals will always buy up everything.
But you’re right, your way might have been better because it’s more honest and because it might have helped to build up a bit more social pressure against mask wearers. Shoulda, coulda, woulda.
28. March 2020 at 11:51
Scott, it’s even worse
Masks reduce the “infectious dose” (amount) that a person gets – the lower the dose, the small the maximum viral load they must deal with.
For many healthy people, a smaller dose is basically a poor man’s vaccine. In a world where we need MORE OF THOSE.
28. March 2020 at 11:56
NEXT TOPIC you can wrestle with…
WE ALSO KNOW that kids going to school is how we all get infected by the flu. This is the flu but people die.
—–
In the Mass Economy, workers generally have small tight groups we interact with daily.
With 2M tests a day we can put masks on everyone and run 80%+ of the Economy
BUT WE CANT if we send all of the little asymptomatic angels into an 8 hour Petri dish to come home and INFECT ALL THE WORK CIRCLES.
28. March 2020 at 11:58
Guy’s we’re not talking about N95 masks, people wearing just a cloth mask cut down their intake by 70-90%.
28. March 2020 at 12:00
One more Scott I think we agree on:
In order to run the Economy, we have to move to Red / Green Zones
this requires NOBODY CAN LEAVE RED ZONES
Doing this frees up medical personnel and equipment to flood red zones.
28. March 2020 at 12:12
The “fight for the masks” is a good example of why populism (and Trump?) are so successful. The message of the health organisations was not wrong, but it was complicated and muddled. The message of the other side was simple and straight to the point, everyone can understand it. And the simple message always wins in such a case, especially in cases of fear and urgency, even quite regardless of how intelligent the recipients of the message are.
Probably more honesty and more appeals to donations would have been better. In Germany, large companies have donated their masks to hospitals after the situation in the hospitals was explained. People are willing to make great sacrifices in such a crisis if you ask them the right way.
28. March 2020 at 12:53
A mea culpa by Sumner, who walks back on some of his more outrageous misstatements (like warm weather won’t help defeat Covid-19). It’s OK, Sumner at least acknowledges his faults, as I have, so he’s to be commended.
@Nazi and Morgan: 7 out of the 9 posts above are from you two. Now Morgan is more fun to read and logical than Nazi, but can’t you guys focus and collect your thoughts before you post? Actually List, if you stop posting here altogether it would suit me fine. Tell you what: if you promise not to post here anymore, so will I. I bet a lot of readers would be happy with that decision. How ’bout it?
28. March 2020 at 13:03
Here is one way of looking at it.
The total supply of masks Is limited at any given time. Modelling suggests that such a pool would probably not meet the demand from healthcare professionals.
A well designed mask may be shown to reduce the relative risk of infection by – and I am making this up for arguments sake – 40%.
An average person in theor day to day life 4 weeks ago would come across an infected person maybe once every 3 days with an infection probability of 5% per encounter. Therefore the daily absolute risk of infection was extremely small
A health professional 4 weeks ago, simply through selection, would encounter several a day.The daily absolute risk of infection would be a magnitude higher than the average person.
The value of the relative risk reduction to restricting the use of masks to exposed health professional is much higher than making them available to the general public especially given the supply constraints.
As the prevalence of the virus increases, than the absolute risk of transmission increases and therefore the differential between health professional/vernal public reduces. And as supply increases so the. Does the value of broader distribution.
28. March 2020 at 13:05
Here is one way of looking at it.
The total supply of masks Is limited at any given time. Modelling suggests that such a pool would probably not meet the demand from healthcare professionals.
A well designed mask may be shown to reduce the relative risk of infection by – and I am making this up for arguments sake – 40%.
An average person in theor day to day life 4 weeks ago would come across an infected person maybe once every 3 days with an infection probability of 5% per encounter.
A health professional 4 weeks ago, simply through selection, would encounter several a day.The daily absolute risk of infection would be a magnitude higher than the average person.
The value of the relative risk reduction to restricting the use of masks to exposed health professional is much higher than making them available to the general public especially given the supply constraints.
As the prevalence of the virus increases, than the absolute risk of transmission increases and therefore the differential between health professional/vernal public reduces. And as supply increases so the. Does the value of broader distribution.
28. March 2020 at 13:43
Christian, You said:
“Masks cannot be that important if you look at the death figures in Germany.”
If you want to rely on anecdotes, how about the fact that East Asians use masks far more heavily than Westerners, and they are doing far better than even Germany.
A shortage? Even a homemade mask with one piece of tissue paper between two paper towels is 80% to 90% as effective as a regular surgical mask.
You said:
“The message of the health organisations was not wrong,”
That remains to be seen.
Ray, You said:
“A mea culpa by Sumner, who walks back on some of his more outrageous misstatements (like warm weather won’t help defeat Covid-19).”
Tell you what Ray. If you cannot find where I said that, then you must apologize to me in the comment section. If you do not apologize, there will be severe repercussions.
Does that seem fair? I’ll circle back later, until you apologize.
Andrew, You said:
“The total supply of masks Is limited at any given time.”
Taiwan had a huge mask shortage in January and now they have plenty. Has it ever occurred to you that perhaps the solution is to eliminate the shortage, not tell people not to wear masks? We could have done the exact same thing as Taiwan. Why didn’t we?
The rest of your comment has no bearing on this post, which discusses the question of whether masks help average people avoid catching C19, not whether they are better left to professionals.
28. March 2020 at 14:28
Christian List:
“Masks cannot be that important if you look at the death figures in Germany. Close to no one wears a mask here.”
Germany has one of the highest reported corona-death counts in the world. Why are they evidence of mask-wearing being ineffective?
28. March 2020 at 14:39
Similar to warm weather, I suspect the question of the effectiveness of masks is primarily a question of the overall reproduction number. The purpose of masks outside the clinical area is not primarily self-protection. In a crowded area such as a subway station or a supermarket, tens of thousands of possible infection events take place within a very short time. If there are a few people walking around with masks to protect themselves, that practically doesn’t lower the reproductive number at all, even if they wear very good masks. But if there is a social norm that means that practically everyone wears some kind of mask that prevents their cough from spreading or that they can touch their noses, this can contribute substantially to reducing the reproduction number. So masks are not so much about personal risk, but about large numbers. And, of course, the bottom line is that there is nothing that, on average, protects the normal individual, that doesn’t hide in the woods, better than a reproduction number less than 1, because that means that the pathogen becomes very, very rare.
28. March 2020 at 14:42
@Aleksander:
“””
Germany has one of the highest reported corona-death counts in the world. Why are they evidence of mask-wearing being ineffective?
“””
Right question, wrong metric. If asking about the effectiveness of masks one should not look at death counts but at cases/infections.
28. March 2020 at 14:46
Alright Scott, I have no problem admitting when I’m wrong. When one looks at all the facts from all the different countries, it seems you hold the winning hand once more, in the sense that you were more right and I was more wrong.
This seems to be a problem of quite some “professional idiots” like me, when our subject is affected and when we should generalize at the same time. In German we say: One misses the forest because of all the trees.
The health organisations had a contradictory and complex message, spun from many different threads of detailed expertise. But the other side has exposed the weakness of this muddled theory and plunged a dagger right into its heart.
I still think that Taiwan had an advantage here because they have a mask industry that Germany doesn’t have and that the US probably doesn’t have either. Most masks came from Asia before. These countries have expertise and factories that the US und Europe do not have.
Taiwan “only” has to scale, Europe has to create from scratch. A famous textile manufacturer in my region now also makes masks, I have seen them: They are really expensive but still not comparable with Asian quality, and the quantities are also way too low. It will take months to make enough masks, and most of these masks will come from Asia.
Back to the mask theory again: there is also the equally logical explanation why certain countries in Asia are so little affected: SARS. Their institutions were prepared and ours weren’t.
I don’t think that another pandemic will catch Europe and the US on the wrong foot again. IF we survive this one, then from now one we will be prepared. 😉
28. March 2020 at 15:13
@Jens and Alexander
Yes, yes, you are right. If we look at cases per capita Germany is not in the top but it’s certainly up there, not looking too good.
Is this due to testing or do we really have so many more cases than others? One could say, look at Africa, nobody wears a mask there, yet there are hardly any cases. I assume they don’t test enough. So we have to rely on Western countries vs. Southeast Asian countries, and here we really have to admit that you can’t exclude masks as a contributory cause anymore.
It’s not an unplausible story, but as I said, we don’t have the masks to equip our population anyway. So I’m not sure how this is helping us right now. Maybe, as Scott has already suggested, the population can now make their own masks and wear them?
I assume that quite some Western governments are reluctant to admit mistakes and make a U-turn. Pick the stage from the Yes Minister sketch, I assume it’s now stage 3-4.
Paradoxically, Trump doesn’t seem to have this problem. Isn’t most of his policy one big U-turn? He simply claims that there is no U-turn at all and that he always favored plan B.
28. March 2020 at 15:50
@Christian:
The thing about testing is that while it’s true places that test more will show higher incidence of afflicted people, even places that don’t test will have their numbers show up in the deaths.
So for example in the US, most of the problem appears to be in New York and New Jersey, and California has far less of a problem (at least so far). Yes NY is testing a lot more than CA, but there are far fewer deaths in CA as well. So it clearly is less prevalent there.
Africa doesn’t have huge waves of corona deaths (yet) and Europe does, so even with different testing rates we know Africa doesn’t have the same problem.
28. March 2020 at 15:55
Not sure if the data on these material is accurate but…
https://mobile.twitter.com/CMichaelGibson/status/1239718351573843973
28. March 2020 at 16:04
@msgkings:
“””
even places that don’t test will have their numbers show up in the deaths.
“””
Meticulously accurate: only if Covid-19 diseases also increase all-cause mortality. To assign a death to SARS2 you also need a test. And whether this happens everywhere under the same conditions is an interesting question.
28. March 2020 at 16:18
@Student:
“””
Not sure if the data on these material is accurate but…
https://mobile.twitter.com/CMichaelGibson/status/1239718351573843973
“””
Yeah, I guess it is much more important how many people wear something in front of their mouth and nose than what exactly they are wearing. And I suspect that the causality is not to filter tiny particles, but to break the air flow (lower range) when exhaling / coughing and to keep your fingers out of your face.
28. March 2020 at 16:48
Sumner: “Ray, You said: “A mea culpa by Sumner, who walks back on some of his more outrageous misstatements (like warm weather won’t help defeat Covid-19).” Tell you what Ray. If you cannot find where I said that, then you must apologize to me in the comment section. If you do not apologize, there will be severe repercussions.” –
Ok I’ll find it. Seems your memory is slipping in your old age…hold on. OK I found it, like ten seconds later. Link: https://www.themoneyillusion.com/warm-humid-weather-will-take-care-of-the-problem/
Sumner: (quote) “Warm, humid weather will take care of the problem” Maybe, but that’s not obvious to me: [graphs show with Covid-19 taking off in warm weather countries, as it always done in the exponential phase, but as I said in the comments it’s the dry season now in most of these countries save Brazil-RL] (unquote)
But, since I enjoy trolling you, “I apologize if I misread you”. I know when to suck up to the teacher, it got me through school most of the time except those three [two] professors that went out of their way to low grade me (I still graduated though, despite those pricks).
28. March 2020 at 16:54
https://www.researchgate.net/profile/Katy-Anne_Thompson2/publication/258525804_Testing_the_Efficacy_of_Homemade_Masks_Would_They_Protect_in_an_Influenza_Pandemic/links/53fefd8f0cf21edafd154e85/Testing-the-Efficacy-of-Homemade-Masks-Would-They-Protect-in-an-Influenza-Pandemic.pdf?origin=publication_detail
“Filtration Efficacy:
All the materials tested showed some capability to block the microbial aerosol challenges. In general, the filtration efficiency for bacteriophage MS2 was 10% lower than for B atrophaeus (Table 1). The surgical mask had the highest filtration efficiency when challenged with bacteriophage MS2, followed by the vacuum cleaner bag, but the bag’s stiffness and thickness created a high pressure drop across the material, rendering it unsuitable for a face mask. Simi- larly, the tea towel, which is a strong fabric with a thick weave, showed relatively high filtration efficiency with both B atrophaeus and bacteriophage MS2, but a high pressure drop was also measured.”
28. March 2020 at 17:58
Prof. Sumner, a question, if I may. A genuine question, not a counterargument with a question mark. Why are you against giving everyone $1200? Let me explain my befuddlement. Extra money given to working people “for free” is just a contribution to inflation, but we want more inflation. Maybe it’s not the role of Congress to generate inflation, but it doesn’t hurt in current circumstances. And extra unearned money in the hands of well-to-do people is actually a form of “loan”. Some time in the future they will have to return it in form of higher taxes or higher prices or both.
28. March 2020 at 20:08
Don’t fault Ray, he’s just trying to provide some light relief in these tense times. Laugh with him, not at him.
“I still think some of my observations (such as the mortality gap between southern and northern Europe) are interesting, and I’ve noticed other people discussing the same data points.”
Perhaps, but it seems to me that the raw international “count” numbers have been so meaningless, that there’s not much there except the really obvious things, like “wow, Italy.”
With one exception, which is how did China manage to have just the one big cluster? The size of the Wuhan cluster relative to the next largest cluster seems surprising (maybe only to me). How did they manage that? The obvious answer is sheer luck – it just happened that no one infected early on in Wuhan left and started the ball rolling elsewhere, I guess.
I haven’t seen this discussed anywhere, but maybe I’ve just missed it.
This kind of thing is interesting, though:
https://twitter.com/nextstrain/status/1243968958681600000/photo/1
28. March 2020 at 20:18
How can we un-mask the truth about COBID-19?
28. March 2020 at 20:23
On the mask thing, it really is kind of bizarre. I thought following the first NYT piece linked in the OP, this one was worth reading also:
https://www.nytimes.com/2020/03/17/opinion/coronavirus-face-masks.html?action=click&module=RelatedLinks&pgtype=Article
The Scott Alexander piece was of course very good too.
The only explanation that makes sense to me is that there was no reward for the being the expert(s) or bureaucrat(s) who pushed the “mask stockpiling” point. So somehow wearing masks never became an element of our response, and now the horse is out of the barn, and no one wants to push it now because of the “this makes us all look like we don’t know what we’re doing” problem.
Or something like that, some larger bureaucratic failure where you have to look at the big picture. Not unlike the failure of the MP bureaucracy to pivot to a focus on NGDP more quickly….
28. March 2020 at 21:13
Ready to cry uncle yet?
Canadian Drillers Face Nightmare Scenario As Oil Crashes To $5
28. March 2020 at 21:22
Christian, Masks are really easy to make; any country can do so.
Ray, I’ll accept your apology, as I know that you struggle with reading comprehension.
D.O. I favor helping those in need, not throwing money at families making $150,000/year who also are seeing their cost of living fall this year. That’s money better spent on fixing our health care system to deal with the epidemic.
And I don’t believe that fiscal stimulus creates inflation.
anon/portly. You are wrong, many other Chinese provinces had 500 to 1500 cases.
And I already posted on that earlier NYT article.
28. March 2020 at 21:41
Why don’t we pick a group of people in NYC right now and do a RCT with a few different types of masks to answer the question more definitively? Let’s see if surgical masks and N95 masks help protect people from COVID19 infection beyond control. Until then, I remain skeptical a few bloggers can infer causality from the circumstantial evidence. And y’all probably criticized Trump when he tweeted the CQ “cure” when it failed the RCT a few days later…
Regarding “observations”, COVID19 testing frequency, accuracy and protocol differs greatly country to country. Compare Iceland’s testing intensity to France’s. I hope no one is naively using CFR and other relevant metrics as given, because “confirmed” cases do not meet the same criteria across borders.
Not to mention, even fatality numbers may be wrong, due to how comorbidities are accounted and whether post mortem testing is conducted. Once again, NO ONE SHOULD TRUST THE CHINESE NUMBERS.
http://shanghaiist.com/2020/03/27/urns-in-wuhan-far-exceed-death-toll-raising-more-questions-about-chinas-tally/
28. March 2020 at 21:44
Jesus WTF
Scotts made of kinds of mistakes here
1. CHINA DID THIS – the CC = CHINA = FUCKWITS WHO DESERVE TO LOSE GLOBAL FACE
2. We should have shut down travel with China in JANUARY prior to Chinese New Yar – Trump f*cked up by not being Trumpian enough.
3. Same for EU
4. Trump should have seized NYC and declared martial law no later than March 2, when De F*ckwit was telling everyone to go out on the town.
—-
I of course, have been FRONT RUNNING TRUMP while Scott made excuses for his wife’s home country.
At SOME POINT we must blame Countries for their CULTURE and the Chinese have sh*t culture. REALLY SH*T CULTURE. The Chinese peole at this pont ARE RESPONSIBLE for the CCP.
—
Scott, I front run Trump…
ANY Chinese foreign national with CCR family is going to be expelled from US colleges and workplaces.
AND you Scott should get behind this…
The USA should LOVE AND FAVOR ALL NON-CCP Chinese families, and PUNISH GLOBALLY all Chinese families with CCP relatives.
They only have one kid it’s not hard.
29. March 2020 at 00:20
@MORGAN WARSTLER – Hey-Zeus Morgan, you’re such a Texas racist… What do you have against CCP people? They are as hard-headed and ‘patriotic’ as you! I dated in Cali a CCP student (all CHN students in the USA are well-connected CCP party members; no poor farmer folk from CHN in the US) and am proud to say I had s*x with her and it was great. Completely godless, pragmatic, tall and beautiful (yes tall! She was a volleyball player as well as brilliant in her field, she got a write up in a prestigious paper and some patents). Later I found out when I was with her she was seeing a younger man, and, I think she actually got an abortion when she thought I made her pregnant (speculative but TMI for here). She ended up dumping me for an even richer man (my family is worth over $10M). Pragmatic, ruthless, beautiful, not at all traditional (she never taught me any Chinese words and could care less about ancient Chinese traditions, lol, it was pitiful but interesting, some Greeks are this way also) I learned a lot about China from her. She’d make a good honey trap. Probably equivalent to a gold-digging Texas southern belle femme fatale too. Did you know Sumner’s wife is Asian? So unlikely he’ll join your racist campaign.
@Sssumner – thanks for not banning me. I’m actually surprised you haven’t already, to your credit too, since I troll you mercilessly. You should however ban Christian List (clearly a troll with no humor either) and that Ruelle verbose guy, though Ruelle’s status went up when he mentioned he once interviewed for a job with the Fed. Bye.
29. March 2020 at 03:58
A few observations…
1. There are 18 million health care workers (HCWs)in the U.S. That’s 5% of the population. So regardless of occupation, there should currently be 6000 cases of Covid-19 among HCWs and 110 fatalities in the U.S. The actual number is a fraction of this, and given the infection rate in the general population, it would be hard to conclude that these were workplace acquired infections. If one wanted to get flamed here, one might even suggest that working in hospital is one of the safest things you can do during the Covid-19 outbreak.
2. The annual occupational death rate is 20 per million for HCWs. It’s 100 per million for construction workers. 200 per million for truck drivers.
3. There is currently no shortage of PPE for HCWs. There’s a fear that there may be a shortage in the future, but there is no current shortage.
4. There are a lot of factors influencing transmission, they all interact, and I think it’s difficult to draw conclusions about mask efficacy… especially for surgical (non N95) type masks worn by the general public.
29. March 2020 at 05:36
Re: coronavirus stats
I am glad Scott reiterated what we should have already known—-he is making observations—-not claiming to be a wizard. It is what I try to do—as this comment is just that.
I.Ratel wrote a bizarre and interesting essay named “Covid 19 and US Mortality” on the “according to Hoyt” blog. The numbers are so “off” one would just dismiss them—-but these are intelligent people who have Had others check—-and all agree “they make no sense”—yet there they are. Data derived from CDC.
Key Points
1. 2019-2020 flu season has had the highest (by a decent amount) doctor visits for flu like symptoms versus last 6 years.
2. However, the number of pneumonia deaths have declined sharply over last 30+ plus days——at what looks like a pretty high standard error level—-versus last 5years.
3. Here is the kicker——Total mortality ALL CAUSES at the last 30 days has declined by an even larger percent——again—-at what appears to be a pretty high standard error level (meaning, if random, it happens very infrequently)
4) Of course, one’s first guess is it is wrong——but I believe that if wrong, it is the CDC numbers——which means they should have already noticed it and corrected it. Or the writer and his checkers just screwed up.
5)So I constructed a “curve fit” (meaning a hypothesis which fits the data after the fact—-I.e.—not scientific method) to see if I could at least make a logical construct which fits the data (correct logic does not make it true)
6) Hypothesis——The Covid 19 virus is a natural occurring vaccine for certain respiratory diseases. The supporting “evidence” out side the data include the following—a) we have observed a much much higher death rate among the old relative to the young—- it is as if the decline in deaths is a result of the young getting this “vaccine” and dying less this year. Like all vaccines, if you survive the vaccine, your chances of not dying goes way up.
OK—-do I believe it? No. But if the data the CDC is producing is accurate, I would like an explanation. Oddities happen all the time—-including this data itself
29. March 2020 at 06:38
RE: no bends in curves of deaths
Many Corona Virus sites have pretty good data. The ones I like are where the start date of each Country’s curve is when deaths or death rates reach a predetermined level. In theory, we can compare each country on a comparable basis.
One first observation is this really does not look good at all. Ourworldindata.org has a great interactive sight. Some initial points on death growth.
1) It is virtually certain that China’s numbers are almost meaningless. They hit “doubling every 10 days” (which very few countries are at a level that “good”) very fast, faster than anyone—-plus the curve looks like it was drawn to a specification. In addition, the virtual zero numbers outside Hubei make no sense at all——Hubei people did travel to Milan—-but not Shanghai or Beijing—to pick 2? This will come back to haunt the world, one way or the other.
2) Yes, we all know about “catch up counting” and no real randomized testing etc——but 2 or 3 observations.
A) Japan, 29 days into the disease, is still at the doubling every 10-12 days level——It is not bending but it’s linear line appears to be constant For last 15 days at something like doubling every 15 days.
B) South Korea, 36 days in, is not bending—-it has angle that has been pretty straight for 20 days—-it is a better than 10 day doubling—-maybe 12-15 or so—-but it is not bending.
C) U.S. —27 days in— we have basically been doubling every 3 days since day 1—with zero bend and an angle slightly worse than every 3 days. This really looks terrible—-we should have started to bend. Germany looks like us but only 15 days in.
D) Italy—-35 days in—-has had a slight bend since day 15—but it’s current doubling rate (again “tangent” slope of curve) is six or seven days and bending slowly. It started bending from 3 days doubling about 15 days in
Ok—-what does it mean? This thing is way bigger than I thought. China appears insane. They have Practically stopped counting or never started outside Hubei. They really are engaging in the “herd” immunity approach. However, this might be good.
But we will never know for years.US counts accurately and the numbers really suck—-As of now—-it could hardly look worse—-27 days in and not a bend yet in sight—-and no way is N.Y. the only bad place. It may be the worst in US but not by this much.
My view changes every few days—-but this sucks.
Yet, my other comment above—-what the hell is that?
29. March 2020 at 07:00
RE: masks
As of now, assume masks help—how risky is that?
New cases are increasing and are constant on percent basis in NY—-yet we quarantined 14 days ago—-the magic number—-let’s see if we start to bend the N.Y. curve——US cases exhibit first slight bend very slight—-this week
29. March 2020 at 08:58
RE: extra deaths due to Covid 19
I assume Covid-19 increases death rates. But, by how much? Does that matter? I think so. Each year we certainly know death rates by age and disease—-including all inferential stats from that. How much has it added? It is not just a numbers game—-but also a cause and effect analysis. We have all heard that in Italy a Covid-19 death is assumed if the patient had Covid-19—without regard to other diseases the person had. Each case may very well be intricate—-but it’s totality can be measured against years when this virus was not around.
Perhaps the people on the frontlines have no time to worry about such “trivia”—-because they have lives they are accountable for—-and that is the correct approach. But we have a ton of qualified counters who certainly can do this analysis. Yes, we may not know which person really died from this or that cause—-but we can know how much COVID-19 has likely added.
29. March 2020 at 09:08
“anon/portly. You are wrong, many other Chinese provinces had 500 to 1500 cases.”
HUH?
Read what I said more carefully. (I’m not suggesting anyone else is a sloppy reader; no one is a more habitually sloppy reader than I). I know there were cases elsewhere. I wasn’t asking “why weren’t there any cases outside of Wuhan?” I was asking “why were the number of cases/deaths so large in Wuhan relative to the next largest province?”
Here’s today’s death tally at Johns Hopkins:
Hubei 3182
Henan 22
Heilongjiang 13
Beijing 8
Guangdong 8
etc.
This doesn’t suggest to me that there was in any sense a “similar” number of cases in any other province, compared to Hubei. It suggests to me that the number of cases in Hubei, even accounting for a higher case fatality rate, must have been much, much higher.
Now consider than in Hubei there was a first person to be infected, then a second, then a third, and so on.
What if, say, the 14th or the 53rd or the 177th person infected was a traveler headed back to Beijing?
Surely if we ran this as an “experiment” 1000 or 10000 times, there would be some outcomes where at least one other cluster got started elsewhere *before* China figured out what was going in Hubei? And that therefore that cluster lead to a death total in some other part of China “similar” to the Hubei death total?
Am I crazy? Doesn’t it seem like China may have gotten incredibly lucky? Isn’t the (primary) explanation for why some places in the US have more cases now, simply that their first case was earlier?
Maybe there’s no mystery in the Chinese numbers, or nothing interesting about them. But I don’t see how making the point that other areas of China had “many” cases makes the mystery go away.
29. March 2020 at 09:20
Again, just to be sure I’m not overstating anything, this is a putative mystery – I don’t know whether it’s a real mystery or not. I’m kind of saying, “if we ran this experiment 10000 times, where one place in China saw 3182 deaths, what is the percentile value of the outcome where the next highest death total was 22?”
I’m just saying it seems like it could be a P1 or P2 outcome, maybe. Then again, maybe it’s a more typical outcome. Maybe it’s a P90 outcome – maybe China got *unlucky* that some other place got 22 deaths.
29. March 2020 at 10:12
In addition to the relative (not absolute) Chinese death numbers, there are two other “dashboard total” numbers I’ve been following with some interest.
Beijing 8
Singapore 3 (844 confirmed cases)
I remember when the place where I live was tied with Beijing at 8 – our death total has doubled four times (approx.) since then.
And the Singapore numbers – if you can’t trust theirs to be halfway accurate, whose can you trust?
To me, these numbers have suggested that successful containment is perhaps “easier” than the analyses from some quarters have suggested. Everyone likes to say “this isn’t the flu” but maybe if you make everyone wear a mask and test aggressively this thing really does become “kinda flu-ish.”
Neither Singapore not Beijing is super shut-down, right?
Maybe this wrong, maybe something else is key. Maybe it’s some combination of “draconian central government and quiescent citizenry” or something.
Or maybe it’s the “dumb-ass” factor. My guess is, if the entire population of Singapore emigrated to the US, the percentage of them that would become “dumb-asses,” in the American sense, would be very small.
My thinking is that you can get away with being a dumb-ass in Beijing or Singapore, but not the *type(s)* of dumb-ass(es) that may make pandemic control more difficult in the USA.
(This is, needless to say, a very vague and probably worthless hypothesis – but still).
29. March 2020 at 10:13
DF, On that silly “urn” story, check out my new post.
Morgan, Not surprised to discover that you are racist scum. (Pretty bad when someone makes Ray look good by comparison.)
Anon/portly, OK, I see your point. But the cases in Italy were originally concentrated in Lombardy. The difference is that Italy did not put a tight control on escape from the province. China did.
29. March 2020 at 14:21
“The difference is that Italy did not put a tight control on escape from the province. China did.”
Okay, well, I didn’t make my point clearly enough, as always.
Let me first quote myself:
“I’m just saying it seems like it could be a P1 or P2 outcome, maybe. Then again, maybe it’s a more typical outcome. Maybe it’s a P90 outcome – maybe China got *unlucky* that some other place got 22 deaths.”
Now let me fix my quote:
“I’m just saying it seems like it could be a P1 or P2 outcome, maybe. Then again, maybe it’s a more typical outcome. Maybe it’s a P90 outcome – maybe China put a tight control on escape from the province (and/or responded in other ways) in a timely enough fashion that it actually got *unlucky* that some other place got 22 deaths.”
I hope this makes more sense now. When I said “maybe it’s a P90” outcome I really meant it!
Remember, when we look around the world at these statistics and the different countries, there’s clearly a lot of randomness to them. For an obvious example, why did King County WA have more than half of all US deaths for awhile? Obviously (well, as I understand it) because of one person arriving from Hubei on January 19 and getting the ball rolling there early.
Clearly there was a period of time between the initial infection or recognition of a cluster of some sort (Wikipedia: December 27?) in Wuhan and the actions China took in response (Wikipedia: January 19?). It’s not like if all this had happened (say) a year earlier or a year later, we would the exact same outcomes, would we? There’s some randomness?
Maybe if I had all the values of all of the relevant variables and could run an accurate simulation exercise, I would discover that an outcome similar to the one that actually occurred was almost inevitable. But maybe not! I don’t know. You don’t know (I don’t think anyway). Maybe some epedemiologist has looked into this issue; I don’t know.
Maybe I’m looking at this all wrong and the data we’re seeing is telling us that what happened in China is exactly what we should have expected to happen. But I don’t see why we should conclude that. I’m thinking of all this as a “extrapolating information from a sample of size 1” problem but maybe that’s not an insightful way to look at it.
Again, the “mystery” to me (and that’s the wrong word, of course, but I can’t think of a better one) is simply that *possibly* it should surprise us that it circulated to the point it did in Hubei without a more significant escape to another province. That’s all. And *possibly* it shouldn’t surprise us, to be sure!
29. March 2020 at 14:24
“For an obvious example, why did King County WA have more than half of all US deaths for awhile? Obviously (well, as I understand it) because of one person arriving from Hubei on January 19 and getting the ball rolling there early.”
I meant to add that another reason might be that everyone in King Country WA is a complete bastard, but I assume everyone already knows that.
29. March 2020 at 16:43
Something about this theory cannot be entirely correct. The price of the masks has increased by 3000% and more. This is money lying on the street, yet it’s oftentimes not picked up. If it was as easy as you say, someone would pick up the money.
95 percent of surgical masks and 70 percent of respirators used in the US were made overseas,
There is no way to close this gap in such a short time. The articles I read say that all mask companies in the US have started scaling up massively already in January. But it is not enough, the demand is simply too massive.
Let’s assume that the US companies have previously produced 5% of the surgical masks, so they would have to scale by a factor of 20 to meet normal total demand, but this total normal demand has now increased by a factor of 10 to 100 as well. Wow, this is a nightmare assignment.
29. March 2020 at 21:59
Two evolutionary biologists discuss the COVID-19 pandemic. They are firmly in favour of masks.
https://www.youtube.com/watch?v=ym-WGOq96G0&t=165s
It is fun watching two evolutionary biologists be very careful about being precise and accurate. It would never occurred to me that identifying which species virus comes from matters for expectations about transmission mechanisms.
30. March 2020 at 00:16
Jens:
Death counts is a proxy for infection count. A much better proxy, as it happens, than official “confirmed infections” numbers.
30. March 2020 at 00:19
Christian List:
“If we look at cases per capita”
Don’t look at cases per capita. Don’t look at “cases” (confirmed cases) at all. Almost all countries are undercounting by at least an order of magnitude. The case numbers are useful for some things, but not for estimating CFR (except in some particular locations, like the Diamond Princess, South Korea, and Iceland in a month).
30. March 2020 at 00:28
“””
Death counts is a proxy for infection count. A much better proxy, as it happens, than official “confirmed infections” numbers.
“””
That’s correct. But to be able to say something about the effectiveness of masks, it would be even better to know a little more about actual infection rates.
30. March 2020 at 07:32
In this thread, my point is disagreed with. In the other thread, things are somewhat different:
“Not enough time to produce a caseload that would lead to 40,000 deaths. If it were, we’d be seeing 40,000 deaths in many other cities. But we aren’t.”
So in this thread, if we get 2500 (Wikipedia: 2524) deaths in Wuhan, we should *not* expect to see more than 10 or 20 deaths anywhere else, because of how quickly China reacted, and how they reacted.
In the other thread, if China had reacted a little more slowly, allowing things to get more out of control in Wuhan, then the relative effectiveness of the Chinese response completely disappears.
Hmmm, okay maybe if Wuhan got to 40000, we wouldn’t expect to see the same 125 to 1 or 250 to 1 ratios. But would we expect those ratios to be larger or smaller? I don’t think it’s obvious.
If the “Sumner Conjecture” from this thread – which I take to be that the 2^7 or 2^8 Wuhan vs. next city ratio was not likely due to “luck,” it was more something like “inevitable,” based on the Chinese response, then I think it suggests that this ratio would have only gone *up* if things in Wuhan had gotten worse! Maybe 2^9 or 2^10.
If there is something “inevitable” about the distribution of outcomes we see, then maybe it would be the case that if we had seen 40000 deaths in Wuhan, we’d expect to see 80 (2^9) or 40 (2^10) deaths in other cities.
As always, my thinking could be crazy wrong, of course.