A new study of face masks

We use the synthetic control method to analyze the effect of face masks on the spread of Covid-19 in Germany. Our identification approach exploits regional variation in the point in time when face masks became compulsory. Depending on the region we analyse, we find that face masks reduced the cumulative number of registered Covid-19 cases between 2.3% and 13% over a period of 10 days after they became compulsory. Assessing the credibility of the various estimates, we conclude that face masks reduce the daily growth rate of reported infections by around 40%.

That is from a new paper by Timo Mitze, Reinhold Kosfeld, Johannes Rode, and Klaus Wälde.


Oh goodie. Remember when WHO and the CDC said masks were worse than no mask?

No, I don't.

Why are you wasting everyone's time with something that didn't happen when you could criticize what they actually did do? Or didn't do, as the case may be.

As for the WHO, it did happen. I give a link below. Perhaps the CDC was more circumspect. But in substance you are wrong.

'"...In fact, there's some evidence to suggest the opposite in the misuse of wearing a mask properly or fitting it properly," Dr. Mike Ryan, executive director of the WHO health emergencies program, said at a media briefing in Geneva, Switzerland, on Monday.'

Wow, that's a lot of blood on his hands. Or perhaps phlegm in this case. I didn't think they'd be have someone stupid enough to say that. Saying there is some evidenced that an improperly worn mask is worse than useless is a long way from masks are worse than useless, but given the evidence the good evidence that masks are effective against droplet spread diseases that was a horrible thing to say.

Australia's no mask requirements killed perhaps 30 people and was infuenced by WHO stupidity, but it was WHO stupidity before Mike Ryan's statement.

There is a large difference between someone wearing a mask in a clinical setting seeing multiple patients in a shift, and someone going shopping once a week. It is easy to forget in March that the major concern, and not only on the part of WHO, was in ensuring the proper supply and use of PPE by those who could in no way avoid being in contact with potential infection - hospital staff in particular.

It turns out that those countries where PPE was notably lacking for medical personnel (Italy, Spain, UK, US, the four representing more than half of all current covid19 deaths) - had much higher fatality rates than those countries where PPE was not lacking - South Korea or Germany.

And it is not necessarily stupidity - a professiojnal tends to look at the world through a professional perspective, where an ill fitting mask, for example, means spreading a disease to multiple people, in a context where mask wearing and social distancing is impossible.

But I do wonder when the WHO will start recommending mandatory mouth disinfection before visiting the dentist. Which is precisely what happened this morning by a dentist visit, where the first step was being timed (40 seconds) using a disinfectant. Seems a simple and obvious step to reduce, in a better than nothing sort of way, spreading a virus while having an open mouth worked on.

The other issue is behavioral compensation. From what I can tell, regular cloth masks (which is what most people have) are better than nothing but still not very effective. You still want people to maintain distance and not go out too much even if they have cloth masks. Masking could backfire if people think the masks are a panacea and stop social distancing.

Absolutely. But how many people think they can cough in someone's face wearing a cloth mask?

Masks are a tool, not a silver bullet. And this research shows they seem to work effectively enough (without being fine grained enough to make a distinction about mask type), as the paper makes the reasonable assumption that the German population was practicing roughly the same social distancing, regardless of when mask wearing became mandatory in a region.

What’s never acknowledged, in the weasel defense of ‘PPE shortage therefore we should tell public masks benefit them’ is the front line effects. A reduction in daily infection rate at the order of magnitude noted in this study would surely have some sizable impacts downstream re PPE need.

There are two different figures cited as reductions - the growth rate, and the number of new infections.

The cut in the growth rate is a bit more questionable, though not in Jena's specific case. A fact the authors point out,

The reduction in total infections over ten days is 2.3-13%, which is useful, but not astounding. In the U.S., with roughly 20,000 new infections, a reduction of 400 to 26000 new cases would be worthwhile, but sizable impact is certainly open to discussion.

Sure. Very likely. But how was the WHO and CDC supposed to know this BEFORE this research was done? There was limited research on masks' effectiveness in the real world. Some of the better research pointed out potential problems (like people engaging in risky behavior while wearing a mask, thus eliminating it's effectiveness). Based on the data available at the time, many people reasonably concluded that wearing masks may not be very effective. Science is hard, it turns out.

The problem was that instead of telling the truth, which is we don't really have much, good or otherwise, evidence either way whether it helps prevent infections either individually or the population as a whole, but we need to ensure that PPEs are available to medical professionals, the public health authorities asserted as if it was established fact that they were not useful for the general public. We now know that even if it is indeed the case that it is of limited help for preventing someone from getting infected, wide spread use lowers infection for the whole population by lowering the chances that an infected individual will transmit the disease to those he/she interacts with. Once you lose credibility, it is hard to regain it. It is in hindsight, but there is no question that both the WHO and CDC handled the crisis very poorly in the Jan/Feb and unfortunately, that was the whole ballgame ........ Just think about the travel restrictions. Again, it is in hindsight, but the minute China imposed the quarrantine on Wuhan, every single country in the world should have closed itself off to China and required returning citizens to be quarrantined for a minimum of 14 to 21 days. That would have been hugely disruptive, but by not doing so, the world has incurred probably 1000x the damage in both lives and economic activity.

You really don't need *data* or *science* to conclude masks can reasonably mitigate infection of a *respiratory* disease.

It's like the so-called experts can't treat the population like adults assuming they can't handle the nuance of 'wearing a mask can help cut down the transmission risk for the wearer and those the wearer come in contact with though note that one should not engage in behaviors that go against social distancing protocols and please do not hoard as it is vital our medical professionals have access to adequate PPE.'

This is backwards analysis. The PPE shortage (which was certainly vastly overstated) occurred because of the large number of cases, not the other way around. Do you think when covid19 cases were just beginning that a minor increase in admissions somehow wiped out PPE supplies?

Why would South Korea need massive amounts of PPE when they never had any significant number of cases?

Texas and Florida have done about as well as Germany has. Did they have extra supplies of PPE?

+1. There is probably some effect of stockpiling in certain regions created by a false panic about explosion though.

It seems like a combination of sudden surge in demand in some regions, which were not particularly poorly prepared, plus localized hoarding due to these false ideas of "overwhelmed healthcare" preventing moving resources about.

" had much higher fatality rates than those countries where PPE was not lacking - South Korea or Germany."

That causation is completely backwards. South Korea and Germany didn't have many cases, therefore they didn't run out of PPE.

Germany currently has the world's 9th largest number of cases, according to worldometers. India, Peru, Brazil, and Russia were not in the top ten 10 back in mid-April to mid-May.

Germany has always been an outlier, because its raw numbers of cases was comparable to France, the UK, Italy, and Spain over a couple of months. A number of caveats apply, but there is no way to classify 186,525 German cases, most of which were before May, as not many.

South Korea? Certainly.

Yes, that's a fair point.

WHO and CDC told us to eat babies too. Remember?

No one would follow that advice, it's clearly absurd and it's not their area of expertise. However, mask usage is there area of expertise. So, there, a lie is a pretty grave matter.

What they "said" was that mask don't protect the wearer.

The amazing thing is the number of people who moved seamlessly from "they told us not to wear masks" to "I will not wear a mask .. freeeedom."

To the issues creating a partisan divide in state legislatures across the U.S., add this one: masks.

Many Democratic lawmakers are wearing them amid the coronavirus outbreak while many Republicans refuse.


Heck, maybe some of you nimrods like to complain about the CDC not warning you *while* refusing one yourself.

Cognitive dissonance, and this time not for any kind of win.

Oops, should have closed the block quote after "many Republicans refuse."

You write this at the same time all the worst hit states are Dem. Cognitive dissonance indeed.

As for the face masks, we've received every possible instruction on their effectiveness, so you can pick whatever advice you want and be supported by the CDC and the WHO. I think the most recent is for masks for the caregivers of the sick, but WHO knows.

"You write this at the same time all the worst hit states are Dem"

I think that's the kind of comment you should have considered two, three, times before posting.

For extra credit though, how does it relate to this:


Thanks for posting this. People are treating Democrat/Republican states as if it's some exogenous variable, when the "dense, urban cities" variable is sitting right there. A variable that seems incredibly relevant to the transmission of a viral disease.

Here in Colorado, the area centered around Denver is hardest hit, while out here in the suburbs there's ~4 new cases per day in the entire county.

Blue/Red is relevant though. Red-state governors almost universally took more rational choices. Blue state governors first claimed it was a racist Trump hoax, then panicked HARD.

Trump wanted to genocide the blue states because they didn't vote for him. He did it on purpose in collusion with the Chinese. It was part of his genocidal plan.

It was a really shitty genocidal plan, seeing as how it's "genocide" with a virus that is going to leave something like 99.9% of the population alive. Maybe 99.8%, or even 99.7%, in a particularly hard hit area.

He can't even do genocide right

-10, lying, links that aren’t even relevant to your claim, and the entire comment is just culture war trolling and attacking your outgroup

More Trumpian comments from anonymous at 5am on a Wednesday.

“The amazing thing is the number of people who moved seamlessly from "they told us not to wear masks" to "I will not wear a mask .. freeeedom."”

How many? Who are these people? Oh, it’s out of your ass. How surprising. Any numbers on these “many people”?

No. Of course not.

Click the link, surely it has data on the people who switched from “CDC said not to wear masks and now I won’t wear one because freedom”

Nope, the article isn’t even about that. It’s about state legislatures and mask requirements legislation.

Any numbers on the difference in mask wearing in legislatures?


Keep trolling

lol, do you feel especially panicked this week?

It's funny, because your state of belief is that you don't care who's president. Another stated belief is that you don't like Trump.

So you would think rationally it would be no skin off your nose if Trump somehow stepped out and Pence somehow stepped in.

And yet here you are every day, to throw up clouds of smoke and chaff, to distract the fact that we do in fact have a dotard for president.

Still and sadly you are not unique. At least 40% of the country is not rational on this. They recognize that we have a dotard in charge, but they can't get off the stick.

Which of course is why I feel a bit cantankerous.

By the way, it is truly classic that you answer a link, with an external story, with "lying."

I'll laugh about that one for a while.

In case you're not salty enough,


The WHO recommended against masks. They also stated they might exacerbate the danger. https://www.google.com/amp/s/amp.cnn.com/cnn/2020/03/30/world/coronavirus-who-masks-recommendation-trnd/index.html

yeah they wanted the sheep not to buy up all the masks so governments could buy them all up cheap for their little government employee pets. the proles were meant to die off because their lives are worthless in the governments eyes.

The WHO said there is no good evidence that masks are effective for previous respiratory disease outbreaks. They have a long track of stupidity on this topic. This would be a serious problem if people listened to them.

You are moving the goalposts. A commenter said the WHO recommended against masks. You rudely and dismissively controverted him. You were wrong.

It is classic Crikey. He's not an arguer in good faith at all.

"Remember when WHO and the CDC said masks were worse than no mask?"

No, I still don't remember that. I do recall them not recommending masks. Then you kindly provided a link which contained this quote from Dr. Mike Ryan of the WHO -- "...In fact, there's some evidence to suggest the opposite in the misuse of wearing a mask properly or fitting it properly." As I stated, that's was a terrible thing to say, but it's still not saying masks are worse than no masks. This is an important distinction in my country as criminal charges can hinge upon it.

As for being rude, that's okay with me as long as I'm not being impolite.

It didn't take much critical reading to see what they were really saying even in those early reports:

"There also is the issue that we have a massive global shortage," Ryan said about masks and other medical supplies. "Right now the people most at risk from this virus are frontline health workers who are exposed to the virus every second of every day. The thought of them not having masks is horrific."

So they're saying right up front that masks work for healthcare workers.

And they say up front that they're worried about other people using them improperly.

It is not hard to put two and two together.

The guidance from both the CDC and the WHO was not to wear or purchase masks.. It was specifically not “save them for health workers”

This resulted in thousands of deaths

It’s not hard to put two and two together

Not correct. Here is an example of a press release issued by WHO in early March: https://www.who.int/indonesia/news/detail/06-03-2020-media-statement-the-role-and-need-of-masks-during-covid-19-outbreak

The article even gives advice on the proper use of masks for those who have access to them and want to wear them. The end of the release says, "The sudden surge of mask purchase, unfortunately, has happened globally. Prices of medical and N95 masks are being inflated and availability of stock is depleting. A shortage of personal protective gear, including masks, can leave the frontline responders of COVID-19 dangerously vulnerable and may worsen the spread of disease as is happening in some countries. Indonesia must learn from other countries and not let such a situation occur.

“If you are healthy, please spare the masks for those who are exhibiting flu like symptoms or for the healthcare workers and caregivers looking after them,” Dr Paranietharan said."

I think that agrees with Skeptical's point more than it disagrees.

If the WHO says: "If you are healthy, please spare the masks ..."

That sounds like they are saying not to purchase masks, assuming you are indeed healthy.

It is different than saying"don't wear masks, they don't work", which is the caricature presented.

Though it is bad advise anyway---wearing masks helps prevent transmission and so reduces people going for medical care.

"It is different than saying"don't wear masks, they don't work", which is the caricature presented."

They did say that:

U.S. Surgeon General: Seriously people- STOP BUYING MASKS!

They are NOT effective in preventing general public from catching #Coronavirus, but if healthcare providers can’t get them to care for sick patients, it puts them and our communities at risk!


"CDC does not currently recommend the use of face masks for the general public. This virus is not spreading in the community."


That CDC quote is from February 12 and looks correct to me. If you complaint is that they did not use moral suasion to encourage the private sector to step up mask production, maybe you are right. Certainly, the whole apparatus of government did not perform well when it came to having stockpiles of PPEs and plans in place to ramp up production. Taiwan used its army to help local factories max out production and implemented rationing and, even then, was still unable to get a steady supply of masks out to the general public by early March, as this article makes clear: https://www.straitstimes.com/asia/east-asia/coronavirus-how-taiwan-is-cranking-up-mask-production-to-meet-shortfall

Well I disregarded the advice immediately because it was clearly illogical.

My complaint was that they deliberately mislead people by telling them masks aren't effective in stopping the spread of Covid19, so save them for the hospital workers.

If the the message had been save them for the hospital workers because they have a greater need, that would have been fine. But they were probably wrong (and probably lied) about them being ineffective for the general public.

There's really no reason they couldn't have recommended that the general public wear cloth masks and reserve the more effective masks for health care workers. But they wouldn't even go that far till April.

Skeptical says, "The guidance from both the CDC and the WHO was not to wear or purchase masks.. It was specifically not 'save them for health workers'"

WHO says, "'If you are healthy, please spare the masks for those who are exhibiting flu like symptoms or for the healthcare workers and caregivers looking after them.'"

Moreover, the claim that WHO cost thousands of lives is unsupported. One would have to show that free availability of masks to the general public in March would not have caused shortages among frontliners or that those shortages would not have cost lives.

WHO explained the reasoning behind its message very clearly and the denunciations here smack of Monday morning quarterbacking.

I quote the text directly that they worry about masks for healthcare workers, and you somehow find a space to deny that they are worrying about mask for healthcare workers!

Do you just post nonsense for the fun of it?

"Masks were imposed by local authorities and were not the outcome of some process in which the population was involved."

Many people were making and using their own cloth masks in the weeks before they became mandatory, including the fact that the mask wearing requirement had a more than week long lead in time. The mandatory aspect covered things like scarves, ski masks, and face shields, as all were part of the mask definition. We are not talking, in later April in Germany, about surgical grade masks. As noted in the paper - " We cannot identify differential effects since mask regulations in German regions do not require a certain type." Nonetheless, it appears that even using impromptu masks may be very effective, allowing higher quality masks to be more effeciently directed to where they would have a greater impact than someone just going shopping with a FFP2 mask (In contrast to the US or UK, German medical personnel never faced an actual lack of PPE as they worked). One could speculate that this is a (partial) explanation of what happened in Germany in terms of its fatality rate, with many people making and using their own masks instead of diverting higher quality equipment from various health care facilities for personal use.

Looking forward to reading such detailed information about the U.S.. In light of how it has the world's leading health care system, it should be child's play (especially with an Emergent Ventures grant) to produce such a paper, as the variation in American states and their mask wearing rules is so much larger than in Germany.

It is a bit unfortunate that they so exclusively focus on Jena, compared to Rottweil, which was much more affected by corona virus. They at least acknowledge the point here -'This result needs to be interpreted with some care as the regional variation of Covid-19 cases in Germany is very heterogeneous the longer we go back in time.' On the other hand, the Jena example is quite clear cut, as is the fact that Saxony was also a week earlier than other federal states in making mask wearing mandatory.

Maybe the people involved in this work deserve an Emergent Ventures retrospective grant, seeing as it represents the sort of data based results that would lead to the American economy being able to recover faster by making mask wearing in public indoor spaces mandatory, immediately. Reducing the last month's American average of roughly 20,000 new cases a day by between 2.3% and 13% in ten days would be a major accomplishment at this point.

On the topic of mask compliance, I'll point again to the study using facebook data I mentioned in a previous assorted links comment - https://www.medrxiv.org/content/10.1101/2020.05.09.20096388v1.full.pdf

(It seems to be the only international Covid19 health behaviors study I'm aware of; if there are others by big polling companies, we could compare).

Italians seem to have masked up very quick, moving from 40%-80% participants between w11 and w16. Spain less so, but moved up from 40%-60%.

The Netherlands, one of the Northern Euro societies that listened to what the WHO said, rather than do as the Southern Europeans and observe what East Asians actually did, is around 5%. UK and Germany between these extremes, but more towards the Dutch end.

The US also seems to have masked up very rapidly - contrary to some expectations; anti-mask people were probably more sound and fury than real in the first place, and in the second probably, even when real probably more people who objected to being forced to comply, rather than people who wouldn't willingly choose to comply.

If mass use of cloth masks is pretty effective, the high shifts to mask compliance in Spain and Italy may have been responsible for some of reduction of R0 and slowing of increase that is often attributed to their "early lockdowns".

Reading the paper may be useful. "Second, the timing of the introduction of face masks in Jena is -by and large- not affected by
other overlapping public health measures related to the Covid-19 spread. Since March 22 the German economy had been in a general “lock down” coordinated among all federal states." Austria's lockdown, much like its decline in new cases, started a week earlier than Germany's (Austria's public mask wearing requirement is not as easily compared). General mandatory mask wearing came more than month later than the lockdown in Germany, and the slowdown in new cases was already apparent. Mask wearing was part of the reopening process in Austria and Germany. The Austrian and German declines in new cases was apparent by mid-April, while mask wearing only became mandatory in later April, as part of the preparation to re-open without causing an increase in new infections.

We continue to learn about how to best handle this virus, and the role of PPE seems to be hard to overestimate in terms of stopping its spread. Lockdowns tended to occur first in countries where PPE was seriously lacking, such as Spain or Italy. And retrospectively, where the spread of the virus was significantly accelerated through medical personnel not having access to adequate PPE.

I certainly believe reading that paper would be useful.

The gist of your comment seems to suggest you believe you have found some text which contradicts my comment though, but for the life of me I cannot see how.

I'm not certain how much spread occured within hospitals from lack of PPE in any country. Probably very little of the total. This can be estimated at some point from looking at data on contact with hospitals and whether persons who likely subsequently contracted CV and got it after visiting a hospital, with little PPE. I expect the effect to be pretty small tbh, but there may be some.

The lockdowns in Germany and Austria started a month before mandatory mask wearing, and provide a contrast to "If mass use of cloth masks is pretty effective, the high shifts to mask compliance in Spain and Italy may have been responsible for some of reduction of R0 and slowing of increase that is often attributed to their "early lockdowns".

Germany and Austria provide a direct contrast to Italy and Spain, particularly as both former countries only implemented mandatory mask wearing as the decline in new cases became apparent after their respective lockdowns. Since mask wearing was less in Germany than Italy or Spain, it is clear that the lockdown played a larger role in the German new case decline than German mask wearing.

Different countries have different factors, but mask wearing in Germany and Austria was a step taken as part of their reopening, after their lockdowns successfully slowed the growth in new cases.

Possibly, mandatory mask wearing would be more effective than a lockdown, but that is speculation that this paper does not support, since the lockdown and the resulting decline in new cases happened before mask wearing became mandatory in Germany and Austria.

Sure, but I didn't say mask wearing did have a role in Germany. Only that it is a factor that might weaken confidence that the "lockdown" was as important in Italy and Spain as some have put about.

"Lockdown" may also not have had a very strong effect in Germany (although it probably didn't do nothing).

I guess I am thinking here that "lockdown" seems more effective in Spain and Italy than in my country, UK. But that may actually be due to masks, not that lockdown was more or less effective (however effective that is or isn't).

>> If mass use of cloth masks is pretty effective,

It is not. This has been studied to death in the decades prior. The conclusion is "Cloth masks are better than nothing, but not by much"

And thus, they are nothing more than a modern billboard or fashion statement or cloak behind which rioters hide.

Surgical mask? Yes, it help a great deal. N95 mask? Yes, it is nearly 100% effective.

Cloth mask? No, not at all.

I believe the cloth masks are rated as 50% effective. Vs 95% for surgical masks.

And early on people were saying get a cloth mask that uses an insert, so 'purely a cloth mask' research is not necessarily relevant.

When surgical masks were in short supply, you could argue cloth was better than nothing. But surgical masks are no longer in short supply. Amazon has 50 for $33 with overnight delivery.

In the 2013 paper below, the conclusion is that cloth masks don't do anything really, but are better than nothing.

So, why is anyone wearing a cloth mask anymore when the science is clear AND now that surgical masks are readily available?


From 2013, "Testing the Efficacy of Homemade Masks: Would
They Protect in an Influenza Pandemic?" by Davies et al.

"2 tests comparing the proportion of particles greater
than 4.7mm in diameter and particles less than 4.7mm in
diameter found that the homemade mask did not significantly
reduce the number of particles emitted (P 5 .106). In contrast,
the surgical mask did have a significant effect (P , .001).

Surgical masks are much less comfortable to wear, and I'm just walking down hallways and keeping the 6' distance. Granted, I probably should have worn my N95 on the last plane flight.

Feel free to wear a surgical mask yourself. Whatever floats your boat.

@JWatts. "rated as 50% effective" means what, exactly? If a condom was on average able to stop 50% of sperm from entering the uterus, would that be 50% effective or nearly useless? The supplier's marketing department would have one answer, the mother with 5 children another. It is expected (but remains unclear) that dose of inoculum is one principle determinate of disease severity. It is not expected that the relationship will be linear. A 50% reduction in dose will give a proportional increase/decrease in disease severity only in certain areas of the dose-probable severity distribution function. If the relationship is logistic, then there would be 3 'zones' which would approximate linearity, very low doses, doses (mid-way) between the two inflection points, and very high doses. But it is obvious that the difference in (real) severity between a dose of 0 virus particle and 1 virus particles can not be linear. (i.e. viruses are discrete). It is also obvious that there will be no difference in disease outcome if test subject has its body weight of virus particles administered or twice or half that (i.e. dead is dead).

I believe that's in stopping virus sized particles at close proximity However, please feel free to look up some of the various studies your self and post the results.

But keep in mind,
Surgical masks 80% effective
N95 masks 95% effective

"This has been studied to death in the decades prior. The conclusion is 'Cloth masks are better than nothing, but not by much.' "

I wouldn't say that's conclusive, because we obviously haven't had decades to study if they matter with this new virus.

It's still very possible they don't matter much for COVID-19, but there are reasons that they might matter more than for influenza: droplets vs. aerosol and differences in minimum infectious dose.

The studies I read were looking at using a cloth mask in a healthcare setting, where the study was simulating a massive exposure at a small distance. I haven't seen one where they compared a mild exposure at 6 feet away.

'The US also seems to have masked up very rapidly' Remember those pictures from the NYC subway? Or the recent Lake of the Ozarks pictures?

In a big country with lots of geographical variation, you can find photos of many, many things. I can only report on this paper's results from data collection. (Which results I would guess you want to chisel away at in some form as part of your weird pattern of posts that drip with bitterness directed towards the US and UK, even when this has very little relation to the thread topic).

Note, even that survey data only shows about 60% use of measure in, and probably variable by location, and probably some degree of self selection into the survey by folks more concerned about Covid19. So there is plenty of room to find photos of people not wearing masks.

Week 16, when mask compliance got high in the US according to this data, is also late April, and I don't know when that NYC subway photo was taken. I am guessing not late April.

Your linked paper covers this time span - "In this paper, we present
results for the period from March 13-April 19, 2020" Looking at what happened in the U.S., and the fact that the U.S. represents the largest single outbreak of this pandemic till now, there is no way to write 'The US also seems to have masked up very rapidly' in light of the events in the last two months. Certainly not if one is attempting to argue that masking plays a larger role in slowing the virus spread than lockdowns.

The German paper makes fairly plain that masking slows the new case growth rate significantly - 40% to 60%. Looking at the American numbers between April 19th and now, either covid19 is much, much more infectious than currently believed with a large number of Americans wearing masks compared to the German experience, or Americans did not actually mask up much.

A claim about how fast cases increased in the US (not very rapidly) and then you trying to somehow infer mask usage from that (how, given many other factors?) seems distinct from actual, direct data from their participants on mask usage.

I really don't get why you're so compulsively long-winded on here to such little benefit in adding information.

You could have said "I believe that papers data on mask use increase in the US is inconsistent with the increase in infection rate and the effectiveness of masks. Therefore their sample is unrepresentative". I don't agree with it but such a statement would be a clear and not pointlessly nasty and petty way to express your argument. -1 for brevity and style, prior.

The troll lives purely for the purpose of detailing discussions here. Don't waste time arguing content and tone. He'll post something contradictory soon enough, and of course his goal is to irritate.


Coronavirus hospitalizations rise sharply in several states following Memorial Day
Confounding factors in my natural experiment. Heat may very well kill the virus but who could tell with suddenly crowded beaches, summer parades and crowded swimming pools. Good thing I didn't take Tyler's money.

One interesting thing is that the states where the virus seems to be growing quickly seem to primarily be hot southern states, where the summers get so hot that people spend more time indoors.

Air conditioning reduces humidity, significantly.

And don't forget this is an artifact of increased testing. Look at TX deaths, bottom chart. Decidedly downward. https://www.worldometers.info/coronavirus/usa/texas/

"...where the virus seems to be growing quickly "

They are growing quickly from a relatively small base. Texas, for example, is still doing much better than California. California had nearly 3x new cases yesterday than Florida, which is the worst southern state. And Florida has a high ratio of elderly retirees.

CA: +3K, TX: +1.7K FL: +1.1K

After a long and steady career in immunology, six weeks of immunology for dummies, let me be blunt.

You do not want a complete, permanent covid vaccine in your bodies, You want a partial vaccine, good for only a few years, and all it does is crack the shell. Then the docs will have a flu like protocol follow on.

Cracking a shell then trying something else? There will be many shells to crack, lets crack them first, see what's inside second.

The docs will learn to treat the acute cytokine problem.


Sacramento coronavirus cases are spiking — and family home gatherings are a key cause
Another confounder in my natural experiment. In hot days, people gather indoors also. I need a community where people are used to hotter that hades and go about business as normal, there are not groups like that. Except, maybe Bakersfield.

Try the Middle East, Central America, or South Asia.

Masks (other than N-95 masks) don't protect the wearer, they protect others from the wearer. Hence, they greatly reduce the spread of the virus by someone infected with the virus. Many studies have confirmed this. Indeed, countries where wearing a mask is almost routine have experienced far less spread of the coronavirus. N-95 masks do protect the wearer, if the masks are properly worn. The problem with N-95 masks (besides the shortage of them) is that few know how to properly wear them. The combination of the lockdown and mask wearing prevented about 60 million coronavirus infections in the U.S. alone. https://www.nature.com/articles/s41586-020-2404-8_reference.pdf Unfortunately, politicians are declaring victory in the war against the coronavirus and reopening economies, causing new coronavirus cases to skyrocket. The disconnect between what people know and what people believe is a chasm when it comes to the coronavirus. Everything today is political, even pandemics.

"Masks (other than N-95 masks) don't protect the wearer, they protect others from the wearer. Hence, they greatly reduce the spread of the virus by someone infected with the virus. Many studies have confirmed this."

Can you link to one of the numerous studies? There is a 2015 study that compares using a surgical mask with a cloth mask when used by health workers:

"Results: The rates of all infection outcomes were highest in the cloth mask arm, with the rate of ILI statistically significantly higher in the cloth mask arm (relative risk (RR)=13.00, 95% CI 1.69 to 100.07) compared with the medical mask arm. Cloth masks also had significantly higher rates of ILI compared with the control arm. An analysis by mask use showed ILI (RR=6.64, 95% CI 1.45 to 28.65) and laboratory-confirmed virus (RR=1.72, 95% CI 1.01 to 2.94) were significantly higher in the cloth masks group compared with the medical masks group. Penetration of cloth masks by particles was almost 97% and medical masks 44%."


Yes, the nuttiest among are those that wear cloth masks AND THEN hound those that wear no mask. They might as well be wearing foil on their heads. Both provide the same level of protection.

Actually, it was a bajillion or even a gazillion infections. Numbers like those are meaningless. And since there are actual American states where no lockdown occurred and where one can safely assume mask wearing was less common, one needs to come up with an explanation of how the lack of a lockdown led to 20% of the non-lockdown state citizens being infected, that representing the rate (20% or so of the total American population) of the avoided infections in other states.

This is not an argument against basic measures taken against a new contagious disease, but honestly, such numbers are as imaginary as Epstein saying that American fatalities would only total 5000 (that is, after he corrected his initial error of 500 dead).

"Lockdown" is an unhelpful and imprecise term. All 50 states either closed schools or else ordered certain certain businesses (especially restaurants, bars, and entertainment venues) be closed. The non-"lockdown" states you are referring to are probably those few states that did not also issue shelter-in-place orders to everyone except essential workers.

It could also be that shelter in place orders in very densely populated areas are functionally equivalent to everyone going to a concert or football game.

It seems to have been true in NYC. 84% of the hospitalizations came from people who were at home or in nursing homes


It is certainly the case that the East Asian cities most comparable to New York City -- Tokyo, Taipei, Seoul and Hong Kong -- dodged the worst of this thing, and didn't even have strict lockdown rules.

On the other hand, Metro Manila in the Philippines had a lockdown even more strictly enforced than those in the U.S. and wound up with fewer than 1,000 deaths. Manila is full of slums and dense living conditions in which social distancing is impossible. Tyler is right that we still don't have good answers for all of the anomalies out there.

I think it's because they're not as different as we think. Lockdown vs no lockdown. In the no lockdown scenario, people still take precautions , using masks and some social distancing while out.
In the lockdown scenario, nursing homes and older people in homes are still at risk as we saw in NYC.

Rayward, COVID-19 is spread almost entirely by droplets. As cloth masks reduce droplets breathed in -- albeit imperfectly -- they can reduce infection risk for the wearer. But the main benefit does seem to be keeping people's droplets to themselves.

Almost all masks are this way--much more effective at holding in than keeping out.

And that's the point. This is why when even a regular flu hits in Asia, many people wear masks. It is impolite not to--it indicates you don't care about getting others sick.

Will people leave the city because of the pandemic? I've read many articles about out-migration (to the suburbs) in places like NYC. My nephew is a restaurant broker in a very popular low country small city. The first half of the pandemic he spent hand holding clients who were in a panic about their closed restaurants. The second half he spent working with new clients from NYC, Boston, and DC who were considering moving their restaurants to the low country.

Here's another possible change: the model for real estate development. Tabarrok has covered the model for high density development in California that would allow high density in areas located near the transit line. The measure came close but didn't pass. A favorite model everywhere has been the mixed use development, combing retail, offices, and residential. Is that model dead? https://www.nytimes.com/2020/06/09/business/coronavirus-mixed-use-developments.html

Anyway, will the pandemic change everything or will it soon be forgotten?

>will the pandemic change everything or will it soon be forgotten?

Stay tuned til November 3 to find out the answer!!

It is too late to leave. NYC metro area is probably the safest place in the country now since antibody test show that 20% of the people have probably already had the virus and the rest are mostly staying home. The Daily number of new cases is down over 95% in in jersey city and in single digits.

I am still surprised at all the discussion of the relative merits of various methods of reducing transmission, but very little about already existing medicines that people can take at the onset of symptoms. Publicity is given to ridiculous methods such as snorting lavatory cleaner, but little mention of already existing ideas such as zinc acetate lozenges, lactoferrin, vitamins C and D3. No doubt there are many more. I know about lengthy procedures such as double blind trials and disasters in the past about less regulated medicines, but isn't all the death caused by delay also a disaster?

I have a client that is a citrus grower. The coronavirus has been a godsend for them (the industry has been depressed), as the consumption of orange juice (and the price of fruit) spiked during the pandemic. One assumes that it is because of the vitamin C.

Linus Pauling's fantasy lives on.

" The combination of the lockdown and mask wearing prevented about 60 million coronavirus infections in the U.S. alone."

That Nature study was accepted on March 22nd and approved on May 26th. It had nothing to do with wearing masks. I'm skeptical of a study that was led by a Ph.D. in urban planning, along with someone with a bachelor's degree in symbolic systems and several with "Resource Energy" policy degrees. I'm skeptical.

"....reopening economies, causing new coronavirus cases to skyrocket."

U.S. coronavirus cases increase a day:

Apr 28 to May 4.......2.6%
May 5 to May 11......1.9%
May 12 to May 18....1.6%
May 19 to May 25....1.4%
May 26 to Jun 2........1.2%
Jun 3 to Jun 9............1.1%

(Wikipedia U.S. coronavirus pandemic page)

After more than a month, shouldn't the number of new cases be declining? That is, the number of new cases on March 28, a full month before your series starts, is the same as the number of new cases on June 9. Even more depressing, though the lockdowns may not be sustainable, it is all too plain that the spread of the virus is far too sustainable.

Possibly, making mask wearing mandatory would be helpful, though it is hard to imagine the Trump Administration taking such a science based approach.

> though it is hard to imagine the Trump Administration taking such a science based approach.

It's not the job of the federal government to mandate mask policy. That is up to each state governor.

But if they were going to mandate mask policy, the policy should be surgical mask or N95, because the mask your sister sewed with a witty saying on front isn't doing anything.

And yet, our "science-based" governors accepted those masks. Masks that do absolutely nothing are OK with them.

What we have here is a very sad Hillary voter.

Sorry boy -- the Rona ain't gonna save your team's sorry ass.

We aren't the ones who need saving this time. Orange man go bye bye.

You're *very* confident about the non efficacy of cloth masks. It seems to me that at worst the science on this point is extremely conflicted, and most studies of extremely low statistical power.

I'm curious what evidence you're seeing that is not apparent to pretty much anyone else and I look forward to your reply.

You'd think that after months of this coverage, someone would realize that about 20k new cases on March 28 - with much less testing, meaning a much higher positivity rate - means something very different than the same number of new cases today.

Yet you uncritically quoted CDC's "best estimate" of IFR, even though CDC has yet to explain to the public who it arrived at this estimate. Why are Nature papers to be looked upon with suspicion but proclamations posted on CDC's website to be accepted without question?

And given the death toll in the U.S. and your belief in a low IFR, the actual number of people already infected in the U.S. must be in the tens of millions, since there are already over 110,000 deaths.

Both Nature and Science have been political in the past 20 years as is obvious with some articles on climate change, including a major one that was debunked within days and eventually retracted last year. I don't think a paper on the spread of coronavirus based on some economic growth model analyzed by an urban planner and ten policy people, most in "resource energy," is likely to be remotely accurate.

The CDC's 0.4% number for CFR is consistent with a German study (0.36%) and what Obsterholm now says (0.5%) after weeks of saying 1.0%. The IFR of 0.26% assumes 35% don't know they had the virus, although I've seen higher estimates for that. Well, know more within a month.

People should be a little more self-aware.

Touching bases that people who talk about climate change are wrong, labels you pretty strongly as someone who cannot grasp objective science.

My undergrad degrees are in physics and mathematics, but you are not the first to explain to me how I can't grasp science because I point out some of the B.S. in climate science. For example, the Nature article that made a splash in the NY Times only to be debunked and then finally retracted a year later.

Actually that makes the criticism sweeter.

I'm afraid I've known lots of guys who think they know more than their whole field.

They didn't.

No, it makes your criticism irrational, not that I'm surprised.

On the hope that you can actually grow in this moment, remember the Rodney Brooks blog post which Tyler posted earlier:


The reason gatekeeping works is that most "revolutionary" ideas are false.

You should try a Khan Academy science course. Dip your toes into analytical thinking.

(And I hope to God that you aren't operating at the level where you think "a" retracted paper disproves a body of science. In fact, retracted papers are completely part of the process of consensus science.)

It takes a long, long time for the scientific process to work. And in between there are countless fits and spurts of breakthrough and error.

There are signposts along the way. Regarding climate science, remember 20 years ago we were looking at climate sensitivities that range from 5-10 degrees C for each doubling of CO2. They have fallen far below that, meaning the early science was deeply pessimistic.

Next, consider that if 99% of models have dramatically over-estimated temperature rise. The models correctly responded to backcasting (historic) events. But utterly failed to pick up new events.

In a perfect world where modelers had not bias, you'd expect 50% of the models to overshoot actual and 50% to undershoot actual. But 99% of models overshot actual, and most did by a lot.

This isn't science.

I'm afraid you guys have been making the same mistake for the last 20 years.

You don't consistently oppose new science.

You just reserve the right to reject new science which conflicts with your political beliefs.

Look at how many of you followed every study on chloroquine looking for a political win.

Whereas the correct position, which I actually took, was that it was too soon to tell, and that the president's recommendation was bad whether or not it proved correct.

Because the question is not whether he guessed right, the question was whether he should have been guessing with the nation's health.

But maybe we can put UV of our butts. That still might work.

Only anonymous argues /for a retracted paper. Actually, both of you can be right. While the whole of science suggests there is warming (just not nearly what the climate scientist are hoping for), Nature and Science publish a lot of politicized indefensible garbage. Add Lancet in with that.

I guess I'll patiently explain that one.

I did not argue for the retracted paper to be true. I argued for the retraction process being part of science.

Indeed if papers are retracted that means things are working, right?

To reiterate for redundancy, it would be bad if no papers were ever retracted, and every paper was judged equally valid forever.

It may mean science as a whole is working wellish, but quick retraction is a negative signal on quality of peer review at any particular journals. Todd's point seems more aimed at the politicization of particular journals, and how this affects politicized issues. While you seem to be viewing things through a pro-science / anti-science frame and labeling him on this imaginary axis a bit.

It is better of course that false results are retracted than simply stand forever, but worse for science that they got through peer review initially.

You seem to be claiming that peer review only should/does(?) happen prior to a paper being published (in a peer-reviewed journal). Editorial peer review is only part of the peer review process. Sorry dude, you're way off base here.

No, of course I wouldn't think that.

I'll give you the retraction being part of the process. But Nature and Science get caught publishing crap all the time, sometime they retract but damage is done, no one see the retraction. Similar to the NYT. They knowing publish crap and put the retraction, if any, buried deep.

One paper, or any set of papers are not consensus. And consensus isn't science.

"Nature and Science publish a lot of politicized indefensible garbage. Add Lancet in with that."

By the way, you guys always end up rejecting consensus science and then veering into conspiracy theory.

We should have started a countdown clock.


Touching bases that people who talk about climate change are wrong, labels you pretty strongly as someone who cannot grasp objective science.

Saying Nature had to retract a dubious paper that was debunked is now an attack on science?

Actually that makes the criticism sweeter.

I'm afraid I've known lots of guys who think they know more than their whole field.

He’s now attacking someone for mentioning a debunked paper, he’s starting to realize he swam into the deep idiocy end.

On the hope that you can actually grow in this moment

Full on condescending personal attacks for no reason

And I hope to God that you aren't operating at the level where you think "a" retracted paper disproves a body of science. In fact, retracted papers are completely part of the process of consensus science

Totally irrelevant to the conversation topic, which of course is the credibility of journals that publish junk papers that are then retracted. How good is the peer review if junk gets through easily?

By the way, you guys always end up rejecting consensus science and then veering into conspiracy theory.

Does Todd reject science? What the hell are you even talking about ?

"Does Todd reject science? What the hell are you even talking about ?"

Well Todd rejected the rejected science, so something or other....

> By the way, you guys always end up rejecting consensus science and then veering into conspiracy theory.

Yes, the very same journals that observe no difference between the sexes. Please, don't lecture me on science.

See, that's another perfect example!

You couldn't help but name one paper you don't like to cast aspersion on science itself.


I think he's casting aspersions on Science (magazine), not science. The two aren't the same and often in conflict with each other.

You are proving my point about just how dubious the CDC's unsourced "best estimate" is, though.

The German study you refer to measured infection fatality rate: they took random blood samples and found an IFR of 0.36% (with, of course, a confidence interval around it). Taking this number and multiplying it by the percentage of people who are symptomatic is double-dipping: the 0.36% estimate is already derived from a population that included asymptomatic people.

A study in Indiana found 0.58% and a study in Spain found 1.15%. "Take the lowest estimate out there and then fudge it downward just for good measure" is not a sound estimation strategy.

Numbers, of deaths or anything else, must be looked at in context.


Lambau Stadium capacity: 80,000

U.S. deaths from influenza in 2019: 80,000

Lambau Stadium capacity: 80,000
U.S. deaths from covid19 in five months in 2020: 114,185

June 9 US deaths due to Covid-19 virus: 114,185
Seating capacity of Beaver Stadium, University Park, PA: 106,572

Even North Korea's 114,00 seat Rungrado 1st of May Stadium is too small as of today.

Though with luck, the U.S. covid19 fatality number will never reach the 250,000 person capacity of Czechia's Strahov Stadium.

So Arizona is fine?

I really don't know why you continue to insist that national numbers should drive what must be local policy.

A general article on regional surges here:


How does WaPo write an entire article about trends in numbers and fail to include a single graph?

They also talk about case numbers without ever mentioning if testing positivity rate is increasing or not.

They benchmark Texas' increases in hospitalizations to Memorial Day (May 25). It also happens that day was Texas' lowest number of hospitalizations since April 21. One could just as easily phrase the hospitalization number as up 20% since May 1 (over 1 month ago) rather than up 36% in a couple weeks. Or, to repeat myself from above, WaPo could include a graph so that the reader could get a better overall sense of the trend (which is indeed up over time, though not so dramatically as implied by their chosen baseline).

Also, after mainly using numbers to talk about hospitalizations in Texas, they shift to stating that "28.14 percent of available hospital beds [are] open". How does that compare to 2,056 hospitalized COVID-19 patients? Should I assume that those are the patients filling most of the state's hospital beds?

In fact, Texas has 13,645 available hospital beds. I can only think of one reason that WaPo excludes that number in favor of percent availability. That is, of course, because actually reporting the number of available hospital beds provides context that lowers the sense of alarm that the reporters are trying to create.

If you're going to post links, please pick a newspaper that understands how to report on numbers far better than WaPo.

Link to the data for Texas, BTW - https://txdshs.maps.arcgis.com/apps/opsdashboard/index.html#/0d8bdf9be927459d9cb11b9eaef6101f

I've never said anything of the sort, Ms. Newman. Had rayward said cases are skyrocketing in Arizona, I would have looked at that state.

Arizona coronavirus cases increase per day on average:

Apr 19 to Apr 25...........4.2%
Apr 26 to May 4............4.1%
May 5 to May 11...........3.6%
May 12 to May 18.........3.3%
May 19 to May 25.........2.3%
May 26 to Jun 2.............3.7%
Jun 3 to Jun 9.................4.2%

Looks like a runaway crisis to me. It isn't clear how much of the increase is due to more testing over the past three weeks.

Seriously, you didn't "look at" Arizona until just now?

So wise one, what is happening in Arizona? We have different results in different places who are doing very similar things. In BC we have social distancing, school a day a week, no large gatherings, but mostly normal otherwise and a couple new cases per day. More strictly restricted places have more of a problem.

Are there migrations of people from affected areas? Specific workplace outbreaks? A hospital centered outbreak?

Surely you can enlighten us.

I gave you a link, son.

But really if you didn't hear it ahead of this you should change your news habits.

Do you live in a solo?

You gave a link to a particularly poor quality WaPo article - even by their standards - as I spell at in some detail in a comment above.

I am feeling cantankerous today, but I will also give you some practical advice. Use this link when you wake up in the morning:


It is not human curated, and so not biased in that way. It is run by an algorithm. It is the most trending stories in the US.

If you use that link before MR, it is easier to put the essays in context.

Dr. Michael Osterholm covers the issue of masks in some detail here:


His analysis seems pretty well-reasoned, more supported by the "science" of masks and apolitical than the johnny-come-lately economists who authored this "study". I'm not sure Tyler is doing the public any great service by providing publicity to this.

The German paper seems to be well founded on empirical events and data in a country that did a decent job (till now) of handling the pandemic. Further, the only politics one brings to such a paper are the ones already within the reader, as a personal opinion. The data is clear, the methods of analysis laid out, and the conclusions drawn are looking at actual events and data, with more than adequate caveats applied.

However, the German paper most definitely would not answer this question - "Should everyone be wearing some kind of face covering in public right now?" The German and Austrian laws only apply to wearing masks inside businesses and public transportation. Absolutely no one is required to wear a mask while walking around or bicycling. People generally put on their mask just before entering a store, and take them off as soon as they leave, and very few people wear masks in the parking lot. Or, to tie this a bit more to recent MR posts, not wearing a mask at an outdoor event in Germany would be completely legal, assuming that proper social distancing is followed.

Your link almost reads as if from another world compared to Germany. Cloth masks were always seen as a necessary stopgap, worn to slow the spread of the virus from infected people as the country began reopening. They are not worn to protect the wearer, as has been repeatedly explained. When the first rules mandating their use were announced, easily 50% of the masks worn were homemade cloth ones (or scarves, ski masks, etc.), and the German paper still shows a measurable decline in that initial time frame regardless (the paper notes there is no way to gauge what types of masks were being worn). One reason such stopgap measures were allowed were eminently practical - higher quality masks were still in somewhat short supply, but the benefits in slowing viral spread were seen as worthwhile.

The German paper seems to answer this, 'How many cases of COVID-19 will occur when people using cloth masks and not understanding the limitations of their effectiveness participate in activities with others where virus transmission does occur?' The answer is around negative 2.3 to 13 percent based on actual data, though not exclusively due to cloth masks and recogninzing that by mid-April, no one was likely to be engaging in a variety of activities where the virus would be easily transmitted.

This may have been true in the past - They universally disagreed with the publication of this recommendation based on the lack of information supporting that cloth masks actually reduced the risk of virus transmission to or from someone wearing a cloth mask. - but this research provides a clear example of mask wearing reducing transmission by using data from all of Germany, a country where adequate testing certainly allows a reasonably accurate picture of the disease and its spread to be developed.

Possibly, this paper will, partially, handle this objection - " I urge you go online to the CDC website yourself and you’ll not find one piece of information supporting that cloth masks are effective in reducing respiratory virus transmission." The alternative explanation is that surgical masks are even more effective than anyone knew compared to cloth, and that the U.S. should place an extremely high priority on acquiring hundreds of millions of surgical masks per week, and have all Americans wear them, to have an even higher reduction in virus spread than 2.3-13% due to the negative effect of transmission through cloth masks. Whether or not the CDC has failed in this area, the German health authorities did not, based on actual data showing declines, reflecting the varied timing of mandatory mask wearing rules.

Full disclosure - I stopped reading after "Like everyone else, I'm always looking and hoping for the magic bullet." That is not how the pandemic response was viewed in Germany, starting from the first identified case on January 27.


US Trial Finds Drug Cuts Coronavirus Mortality 40%
For the acute allergics. Dunno any more, the study was 240 patients with the allergic reaction.

Let's hope it works. Unfortunately, when there are dozens of drugs being investigated randomness can result in one looking much better than it is. Hopefully it is effective and, as a bonus, people get wormed.

+1, Yes Crikey nailed it. this could be a great find or it could be 20 studies with 95% confidence intervals

More on Ivermectin


The good news is that this drug (if the study is correct) helps already seriously ill patients, whereas hydroxychloroquine appeared ineffective unless administered early in the process.

This study is flawed. Once masks become compulsory, people spend less time around people because it seems scary.

The German and Austrian mask regulations were part of their reopening plan. That is, more people were able to go shopping for items other than food, got haircuts, or went to the library. Masks becoming compulsory indoors in businesses is what allowed more people to spend more time around each other in previously closed businesses.

Indoors or out? Studies are showing outdoor transmission is rare, it's maddening nobody differentiates.


In Germany, the mask requirement is only for indoors.

Tyler, take your mask off and tell us why Paul Krugman and liberals must be in charge of our profession's speech safety


Justin Wolfers really has upped his game of douche baggery.

Is it churlish to ask if this "study" falls under the 70% that's not replicable? How about waiting for review to determine if this is anything but pure garbage before we react? Is all this crazy?

Yeah, it's crazy. This is because potential downside = had to wear a mask. Potential upside = Grandma doesn't die. So the conclusion is wear a mask.

This is retrospective, being based on past events and data. This would be like asking whether the analysis of a flood following a dam bursting is replicable.

The retracted Lancet HCQ paper was retrospective. It is reasonable to question the paper for a lot of reasons. It appears to be an Economics paper. About a public health subject. It isn't clear (to me, but I'm not skilled in the art) that the paper was subjected to any peer review, but if so then how good the review for a "discussion paper" would be. Anyway, TC posted because it was of interest, had something interesting to say, and I'm sure doesn't necessarily accept the findings as correct. Call it a 'piece of evidence'. But anyway, we here are the peanut gallery, and our main function is to toss peanut shells down into the arena where the work is actually being done.

The retracted Lancet article had faked data. The German paper explains its methodology straightforwardly - it uses public Robert Koch Institute data for infections, and publicly known dates for when masks were mandated to be worn. Since the dates were different in different localities and in different federal states, it is possible to make comparisons using those dates and the RKI infection data.

Which is what the paper does.

Identification problem: government announcing mandatory mask wearing shocks citizens into a higher level of non-mask preventative behavior (for instance, "gee, govt is taking this very seriously, perhaps i ought to stay inside more.")

Comments for this post are closed