Is the curve bending?

Left uncontrolled, the coronavirus would probably kill about 0.5% of a country’s population. (Assuming 1% mortality and 50% infected.)
One of my more innumerate commenters doesn’t think that’s very many. But 1.7 million dead Americans is actually kind of a big deal. Most would be over 70, but perhaps 100,000 would be under 70. And since when do 70 year olds not matter?

Of course there won’t be 1.7 million dead Americans from the virus, maybe not even 1% that many. And that’s because we’ll take measures to prevent that outcome. And those measures will likely cause at least a brief recession. (A longer slump will depend on monetary policy choices.)

I look at the data each day, and today’s the first time I’ve seen a tiny sliver of evidence of a slowdown in the virus in Europe. The growth rate finally seems to be slowing in a few countries (especially Scandinavian), from the roughly 33% daily growth rate of recent weeks. Of course East Asia still leads the developed world in controlling the spread of the virus, and the tropical regions still have low totals (albeit rising.) And there have been previous false dawns, so it will take a few more days to ascertain whether the growth rate is actually about to slow.

I’m still confused by the fact that the mortality rate in Italy is 100 times higher than in Scandinavia (7% vs. 0.07%). Commenters always give me explanations when I raise this point, but the explanations are the sort of thing that would explain a 2 or 3-fold difference, not 100-fold. And yet there must be some reason.  Presumably one of my commenters is right, even if the explanation seems implausible.

PS.  I am including Finland as one of 5 Scandinavian countries, although I believe it is actually Nordic, not Scandinavian.

PPS.  I have a new column at The Hill on the Fed and the epidemic.

PPPS.  I was really glad to hear that Walmart and other firms will be opening testing facilities, as our pathetic government is unable to even tie its own shoe laces.  The Pelosi/Trump spending deal seems like a nothingburger, at least in terms of stimulating the economy.  Perhaps it can be justified on egalitarian grounds, as lots of low-income people will suffer during this period.


Tags:

 
 
 

55 Responses to “Is the curve bending?”

  1. Gravatar of Ray Lopez Ray Lopez
    13. March 2020 at 15:38

    Racism by Sumner noted. Actually I want to see if this comment shows up unmoderated since I notice being first to post ends up not posting automatically. But, to be fair, Sumners post is culturally racist, just saying, not that there’s anything wrong with that, when comparing countries.

  2. Gravatar of Mike Hinton Mike Hinton
    13. March 2020 at 15:47

    Speaking of curves… is it possible that the 10-year ended above 1% today because the Fed is actually sounding serious enough with their announcements for QE throughout the curve and the (pretty much) guarantee of a 100 bps cut next week?

  3. Gravatar of Benjamin Cole Benjamin Cole
    13. March 2020 at 16:12

    Some observers contend the mortality rate of coronavirus is actually far less than 1%. Outside of Hubei Province, China mortality rates are much lower than inside Hubei.

    Interestingly, once infection rates get towards 50% to 60%, the number of new infections starts dropping. In other words, each new infected person then infects less than one new person, the opposite of exponential growth.

    The Scandinavian mortality rates for coronavirus are telling. Those are well-run countries with administrative capabilities, and small populations. They probably did a good job on tracking data.

    So… apply 0.07% mortality rate to 170 million infections in the US…about 119,000 deaths.

    Of course, that is absolutely a large number and in that regard stunning. In relation to the total population it is, of course, much less worrisome. It is about two to three times the number of opioid deaths per year.

    Sensibly, the United States doesn’t collapse its economy and financial system to prevent opioid deaths. And yes, to some extent opioid deaths are self-inflicted and therefore incur less sympathy (by some). And no, you cannot become dead from opioids because the person next to you overdoses.

    Still, comparing opioid deaths to potential coronavirus deaths provides some perspective.

    And in signing off let me say – – – Danger, Will Robinson, Danger!

    Ray Lopez:

    https://fred.stlouisfed.org/series/CLVMNACSCAB1GQEL

    That is a chart of Greece GDP. It looks like an outline of Mount Fuji. Money may be neutral, and your car may also be in neutral (as you go downhill).

  4. Gravatar of Benjamin Cole Benjamin Cole
    13. March 2020 at 16:35

    Headline from Fox News—

    “Bernie Sanders warns coronavirus deaths could exceed armed forces casualties in WWII”

  5. Gravatar of Physicist in Paris Physicist in Paris
    13. March 2020 at 17:19

    To solve the puzzle, check the Total Recovered column.

    The important ratio is Total_Death/Total_Recovered. It is similar everywhere.

  6. Gravatar of Benjamin Cole Benjamin Cole
    13. March 2020 at 17:37

    This is an interesting perspective, and I think the right one:

    “Chief scientific adviser wants 40million Britons – 60 percent of the entire population – to catch coronavirus for herd immunity to take effect

    Sir Patrick Vallance said millions fighting off the virus will help in the long run

    He said: ‘Sixty per cent is the sort of figure you need to get herd immunity’

    “The UK’s chief scientific adviser has revealed that around 40 million people could be left to catch the coronavirus as part of a controversial Government plan for the country to develop herd immunity.

    Sir Patrick Vallance said around 60 per cent of the population will need to catch the virus to build up a national tolerance strong enough to stop the virus circulating.

    Chief medical officer, Professor Chris Whitty, did not say how many people could die during the time this takes to work, but the current global death rate – around three per cent – suggests it could be 1.2million people in a worst-case scenario.

    But officials expect it to be considerably lower – Professor Whitty said even one per cent – 400,000 people – would be an overestimate.

    The Government is now aiming to control the speed of the UK’s epidemic and predicts the worst of it could come in around 12 weeks’ time – mid-June – suggesting it could taper off to levels similar to the current situation by September.

    It will attempt to slow down the spread by gradually introducing stricter measures, such as potentially banning large gatherings or closing schools in future, as the outbreak speeds up.

    Sir Patrick admitted trying to stop the virus completely would be more dangerous in the long run.

    He said: ‘All of the evidence from previous epidemics suggests that when you do that and then you release it, it all comes back again… this is to make sure we don’t end up with a sudden peak again in the winter, which is even larger, which causes even more problems.’

    —30—

    https://www.dailymail.co.uk/news/article-8108271/Chief-scientific-adviser-wants-60-percent-entire-population-catch-coronavirus.html

  7. Gravatar of Christian List Christian List
    13. March 2020 at 18:01

    Scott,

    great column at The Hill.

    Benjamin,

    This guy is really bold. His government has failed and now he’s calling his failure an intentional strategy.

    The great thing about the Daily Mail is that its commentators are usually nailing it. This time again. The comments about this buffoon are hilarious. One great comment after another, sorted for best rated of course.

  8. Gravatar of Brian Brian
    13. March 2020 at 18:37

    The discrepancy between Italy and Scandinavian countries is not surprising at all if you consider that Italy reported its first case nearly a month before cases were reported in Scandinavia. For example, a bit of Googling shows Italy confirmed its first case January 30 and Norway February 27.

  9. Gravatar of Benjamin Cole Benjamin Cole
    13. March 2020 at 19:34

    Christian List:

    I think the Brit is just telling it like it is. You cannot lock down an highly infectious virus, and we don’t have vaccines.

    BTW, did you know cold viruses generally mutate to less lethal forms? Yet people who contact the less lethal form are inoculated. The less lethal forms tend to win out over the lethal forms, for obvious reasons.

    This may in part explain the higher reported fatality rate in Wuhan, and then descending rates elsewhere.

    Italy? An outlier for sure. I would like to know if there is some benefit for declaring that someone who died was afflicted with coronavirus. More funding for the facility in question, or a benefit for survivors, etc.

    BTW, the Trump-Pelosi economic “plan” for responding to the coronavirus is weak and pathetic.

  10. Gravatar of Esteban Esteban
    13. March 2020 at 19:56

    Scott,

    You’re probably the most clear-headed monetary economist alive today. Please focus on the Fed’s job of 2% inflation and that control of the money supply gives them infinite ammo to do everything they can.

  11. Gravatar of Todd Kreider Todd Kreider
    13. March 2020 at 20:47

    “Of course East Asia still leads the developed world in controlling the spread of the virus,”

    Look at Germany. There are too many variables to make this claim. Scandavian data is still too sparse to say much even if encouraging, as is German data since the death rate is still 0.2%. Both 50% contracting the virus in the U.S. and a death rate of 1% are both too high, especially the mortality rate. A good guess is 100,000 to 500,000 U.S. deaths over two seasons.

  12. Gravatar of Michael Sandifer Michael Sandifer
    13. March 2020 at 21:26

    Benjamin Cole,

    There may be no herd immunity for this virus, at least on a permanent basis. You might want to watch this interview with an infectious disease expert concerning the COVID-19:

    https://www.youtube.com/watch?v=E3URhJx0NSw

  13. Gravatar of ssumner ssumner
    13. March 2020 at 21:29

    Mike, It’s possible, but with the 10-year yield it’s always hard to be sure.

    Physicist, Interesting. But even if you are right, doesn’t the 100-1 ratio I describe remain something of a puzzle? Why such a large discrepancy?

    Brian, That’s part of it, but the size of the discrepancy is still surprising. Deaths often come quite quickly after someone starts showing symptoms, and most people are not tested until they show symptoms, right?

    Thanks Esteban.

    Todd, I expect even fewer than 100,000 deaths. I was talking about what would happen without social distancing. But since people will try to avoid the virus, the number catching the virus will be only a tiny fraction of the predicted 50%.

  14. Gravatar of Brian Brian
    13. March 2020 at 22:31

    Scott,

    This observational study (https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)30566-3/fulltext) says the average time to death is 18.5 days. So it seems pretty early to be expecting to see a lot of deaths if the first cases were reported just a couple weeks ago.

  15. Gravatar of Benjamin Cole Benjamin Cole
    13. March 2020 at 23:43

    Michael Sandifer—

    Some experts are concerned. That was a spooky show.

    If people do not gain immunity to a highly infectious and highly lethal virus, then I suggest we restructure our economy and life into a more agrarian economy, of isolated communities among which travel is restricted.

    Have you got calluses on your hands, enough to handle a pitchfork and a hoe? Can you dig a well for water?

  16. Gravatar of Ray Lopez Ray Lopez
    13. March 2020 at 23:44

    @Benjamin Cole – my apologies, as it turns out you were right. Not about money neutrality, but about herd immunity with nCoV19. See Marginal Revolution today or this Twitter thread: https://twitter.com/iandonald_psych/status/1238518371651649538

    @ (the fake, pretend, ‘internet doctor’) Christian List – you owe B. Cole an apology, see above. If you’re really a doctor I hate to be your patient, as you’re incompetent and unethical, did you study under Dr. Josef Mengele? Trying to make friends with Cole after you’ve insulted him, to get him to side with you, won’t work, Verzögerung. Me and Cole live in SE Asia and are kindred spirits in that sense. I also see some of your countrymen here. The Germans, like cockroaches, are everywhere.

    As for the Sumner post, indeed, the John Hopkins site, if we trust the China data, shows the new Covid19 cases plateauing, which is good news, though UK scientists (see above) still are predicting 80% of the people in the UK will get Covid19, which might mean the China data is fake. Time will tell.

  17. Gravatar of sty.silver sty.silver
    14. March 2020 at 01:57

    I’m somewhat surprised by your optimism about people’s ability to prevent the spread of the virus (my estimates on how many people will die are much higher than yours). What I’ve observed (in Germany, but will it really be better in the US?) is that people are absurdly bad at following the most basic safety rules (if someone coughs/sneezes, go away; don’t touch your face; wash your hands immediately upon coming home).

    Conversely, if people are this bad what they’re doing is still enough to significantly reduce the spread, I think that implies that anyone who is properly careful has very low risk.

    Or, I suppose, people might get better at it now that the consensus on the severity of the pandemic is shifting. But I would be surprised. Everything I’ve seen in my life thus far indicates that most people are amazingly incapable of changing habits.

  18. Gravatar of sty.silver sty.silver
    14. March 2020 at 02:03

    Also – 1% is a reasonable (probably pessimistic) guess about the mortality rate under *normal* circumstances where treatment is possible. If, however, tens of millions of people get the virus simultaneously, many of them will inevitably receive no professional treatment since the system can’t handle that many cases. This could increase the mortality rate from perhaps 0,6% to 3% (just making up numbers here). So I wouldn’t rule out scenarios where *more* than 1.7 million Americans die.

  19. Gravatar of Mads Lindstrøm Mads Lindstrøm
    14. March 2020 at 02:27

    With respect to death rates, I think Brian is spot on. In Denmark at least, according to official numbers, we only really saw exponential growth five days ago. Therefore, the death rate for Denmark cannot be reasonably compared to Italy yet. Also in Denmark there is still so few infected, that the hospital services can treat everyone who needs.

    See https://videnskab.dk/coronavirus . It is in Danish, but scroll down to the graf titled “Corona-smittede i Danmark”. We stopped testing as much, so the leveling off may be fake or may not be.

  20. Gravatar of Mads Lindstrøm Mads Lindstrøm
    14. March 2020 at 02:33

    For people interested in comparison studies, there is a natural experiment going on with Denmark and Sweden. Two fairly similar countries. The infection rates are roughly similar, but Danish measures to combat the virus are a lot harsher. We (Denmark) closed down the borders (not completely, but for most people), most office workers work from home, non-essential public employers are on paid leave, … Sweden do a lot less. Time will tell if we (Denmark) are being hysterics or the Swedes are taking this much too lightly.

  21. Gravatar of MikeF MikeF
    14. March 2020 at 03:23

    The difference between Scandinavian countries morbidity rates is that the smoking rates are very low and the air quality is very good…not rocket science…the health of a patients lungs is one of the primary determinants if you live or die from this. China, Italy, Korea…high pollution and high smoking rates. I suspect the accumulated lifetime exposure to PM2.5 but could also be related to current levels…pollution in China has plummeted due to the outbreak and now the deaths are leveling off…probably more correlation than causation but worth studying. Of course once the healthcare system gets overloaded and mechanical ventilators are rationed, the fatality rates take another leap up.

  22. Gravatar of Benjamin Cole Benjamin Cole
    14. March 2020 at 03:27

    Ray Lopez:

    Thanks for noting how right I am. I encourage all to read the twitter mentioned by Ray:

    https://twitter.com/iandonald_psych/status/1238518371651649538

    I want my $50k from Tyler Cowen. I suppose he wants a white paper from somebody with credentials. Fast, quick and right does not count.

    The best government program for coronavirus is no government program (other than a payroll tax holiday).

    Where are the libertarians all of the sudden? Now they are all whining government is doing enough? The government should have a program of free adult diapers for libertarians.

    Ray, keep working on “money neutrality.” You’ll come around.

  23. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 04:40

    Italy is odd. Age matters, but not that much. A friend of mine who lives in Spain and whose mother is Italian had been telling me how unreal the panic is there—-now. He suggested one hypothesis for its unusual stats. He claims there is a large mislabeling scheme in the Italian garment industry, i.e., many clothes labeled “Made in Italy” are actually made in China. There is consequently a well above average amount of travel between Milan, and it’s environs, with China.

    Therefore, it is not implausible that Italy may have had similar rates of the disease at about the same time as China, but not being aware of it, thus leading death rates China would have had.

    While I can easily believe the garment scandal, it’s hard to believe there are not other places in the world where Chinese heavily travel back and forth from.

  24. Gravatar of Christian List Christian List
    14. March 2020 at 05:04

    Benjamin,

    you are deadly wrong. And the great twitter thread you gave is by a psychologist. I hope he has not understood the strategy of the British Government correctly, because it would be an absurd strategy.

    Just two examples: It is absurd to assume that they control the exponential growth curve (with their inaction) in such a way that hospital admissions and hospital discharges keep the balance. Secondly, it is absurd to assume that they control the exponential growth curve in such a way that children are infected en masse, but not people at risk. If they really pursue this high-risk strategy, the situation quickly spirals out of control.

    So far we have only heard the opinion of this weird adviser Patrick Vallance. Thank God, I have no other indications that this is really the main strategy of the UK government.

    The British healthcare system is notorious for its inefficiency and chaos, especially the hospitals. If Johnson is really letting the virus run free, then Britain is heading for disaster.

    However, the real strategy of the British government can easily be googled on Sky, BBC, Guardian, or wherever. The strategy has four phases and was announced by Boris Johnson on March 12th. According to Boris Johnson we are in phase 1 or phase 2 right now. Media reports vary here. These phases are called, fasten your seatbelt: Containment, delay, containment, delay, and guess what, containment. This is the standard WHO strategy. So stop your fake news please.

    You cannot lock down an highly infectious virus

    This is just more fake news. 90% of what you write about Corona is fake news. I might have believed this a few weeks ago, but already in January the WHO sent a team of experts to China. This team of experts quickly realized that the virus can be locked down. The Chinese followed the advice and locked it down. Large parts of Asia have followed the advice and locked it down. Scott has listed the countries several times. If the facts change, I change my mind. No big deal. But hey, don’t let stupid facts get into your way.

    The comments and “strategy” from parts of the British Government, by the way, rather remind me of this bit from Yes Minister:

    https://twitter.com/NorthmanTrader/status/1237393224714481667

  25. Gravatar of Benjamin Cole Benjamin Cole
    14. March 2020 at 05:15

    Christian List:

    Well, don’t worry, my comments on this blog have not yet affected the policy of any government. (I am curious where Donald Trump got his excellent idea for a payroll tax holiday—that is exactly my idea. A YouTube I did on the topic got a bunch of hits before Trump announced his plans. But the world is a big place).

    See Tyler Cowen’s treatment of the Brit policy.

    https://marginalrevolution.com/marginalrevolution/2020/03/the-british-approach-to-coronavirus.html

    I dunno, Tyler seems okay with the Brit policy. Evidently, the Brit experts feel the virus will get out anyway. You say the virus has been “locked down” in China, but now that they are going back to work….

  26. Gravatar of Aleksander Aleksander
    14. March 2020 at 05:15

    ssumner:
    “I’m still confused by the fact that the mortality rate in Italy is 100 times higher than in Scandinavia (7% vs. 0.07%).”

    I’m 98% certain that the mortality rate is much lower than 7% in Italy, and MUCH higher than 0.07% in Scandinavia.

    It looks high in Italy, because “confirmed infections” are only a small fraction of actual infections. We’ve known for a long while that Italy has only identified a small minority of infections; there are probably closer to 100,000-200,000 infected in Italy. They are busy with other things than trying to find the remaining 80,000-180,000 infections right now.

    The “confirmed infections” numbers are probably much more accurate in Scandinavia, but their cases are much newer, so people haven’t started dying yet. I’m 90% certain that at least 10 Norwegians, 8 Swedes and 8 Danes are going to die in the course of the next month. I’m willing to bet money on this if anyone will take my bet.

    Finally, the death rate probably IS higher in Italy right now, since they don’t have enough hospital capacity. I don’t have a strong sense of how much higher, but I wouldn’t be surprised if it’s 2-4 times higher in Italy.

  27. Gravatar of Benjamin Cole Benjamin Cole
    14. March 2020 at 05:17

    Of one thing I am fairly sure: Pasta and garlic do not fight the coronavirus.

  28. Gravatar of Christian List Christian List
    14. March 2020 at 05:32

    About the differences in mortality rates: I agree with other commentators.

    The Italians were simply the first to be hit, and very hard. I think that is the main reason. It overloaded their hospital system massively and this then led to the death rates.

    In addition, there may be other reasons: such as the quality of health care in general, the age of the population, the health of the population, etc. But I think that’s less important.

    There may also be differences in the way the cause of death statistics are filled out. There are relevant differences from country to country quite often. There are media reports that say that the Italians, for example, also test post mortem, which is something that is not necessarily done in other countries (like Germany), especially if we don’t suspect corona yet.

    Or the Italians simply evaluate every respiratory disease death in an epidemic area as corona death, regardless of whether it was really corona or not. The death rates are so high, one must almost hope that the differences are due to this.

    In any case, the expert opinion is that the death rates will eventually converge. The question is only in which direction.

    You say the virus has been “locked down” in China, but now that they are going back to work….

    @Benjamin
    Yes, that is the sensitive part of the WHO strategy. But it is not risky. If the virus comes back, you just go back into lockdown.

    What you described is, as I said, hopefully not the British strategy, but let’s assume for a second that it is. So what will Johnson do if this “strategy” goes terribly wrong? Good luck with that.

    The British would quarter Boris Johnson alive. You saw the Daily Mail commentators already, didn’t you. And this is the conservative part of the country, people that usually support Johnson all the way.

    If Trump and Johnson keep this up, they’ll be done for in no time.

  29. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 05:36

    It has struck me as peculiar that “no one” is suggesting randomized testing. One of the many unknowns is whether we know what the “denominator” is—-i.e., how many people have it.

    It is easy to construct the designs of a test, (yes, there can be loads of selection bias for many reasons—-but that is always true to a greater or lesser degree in randomized testing). Even wide standard error results would be useful. Pay people to be tested.

    PS. I once heard James Simons of Math and Trading fame, discuss how absolutely idiotically lucky he was when he first began trading (in the late 70s well before he had models or even much common sense). He allocated money to a commodities trader. Within weeks he was up 7 fold. It then occurred to Simons “how do I decide to get out of a trade”. He had no idea and winged it.

    I think of that today. How do we decide to say we are no longer in an emergency? It is a virtual certainty Covid 19 is now a fellow traveler with earthlings. I still believe we are suffering from excess risk aversion disease. But that could not be more irrelevant at this time. We will need to create a “myth” story to exit this. Bent curves, low death rates, low spread rates—-global, local, and personal actions “worked” and we now have created an “equilibrium” we can all live with. In two months, we need to say this.

    PPS. As “unexplainable” as Italy’s high rates are, more unexplainable are non Hubei China’s low rates. It seems impossible. But if true—-what does that mean? If NOT true, what does THAT mean?

    Bottom line is we know not much——-except historical examples of how communicable diseases have spread in history.

  30. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 06:06

    RE: Britain. I do not know what they are really doing, but assuming they are doing something like letting the virus roam free, I am glad someone is doing this. It is a natural experiment——

  31. Gravatar of Todd Kreider Todd Kreider
    14. March 2020 at 06:11

    Scott, there is no way that “the number catching the virus will be only a tiny fraction of the predicted 50%” although it may be on the low side, around 15% to 20%.

    @Aleksander who wrote:”I’m 98% certain that the mortality rate is much lower than 7% in Italy, and MUCH higher than 0.07% in Scandinavia.”

    With over 8,000 cases South Korea’s mortality rate is 0.9% and with almost 4,000 cases, Germany’s mortality rate has been consistently at 0.2% after its first 1,500 cases. It will be higher than 0.07% but still 0.2% to 0.9% is very likely.

    Time will tell if we (Denmark) are being hysterics or the Swedes are taking this much too lightly.

  32. Gravatar of Todd Kreider Todd Kreider
    14. March 2020 at 06:14

    I included the last line above by mistake.

  33. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 06:25

    Re Unintended consequences:

    A Doctor friend of mine runs a large Department at NY Presbyterian. They believe if NYC closes schools, as much as 30% of the nursing staff will stay home to take care of children. Fortunately rumors of lockdowns and school closings seem not true.

  34. Gravatar of Todd Kreider Todd Kreider
    14. March 2020 at 06:50

    “Fortunately rumors of lockdowns and school closings seem not true.”

    16 states have closed K-12 schools for 2 to 4 weeks. It will be interesting to see if NY follows next week.

  35. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 06:53

    Re: individual opinion versus Public action

    I do not think I am exaggerating. But I have not spoken with one person who does not think this is a massive overreaction. But, including myself, no one has a clear idea of what should be done instead.

    Of course, many experts have semi apocalyptic forecasts. I do believe innumeracy is rampant among specialists. The mechanical extrapolation of exponential growth is high level innumeracy. Merkel’s 70% is pretty out there. Opportunity costs are ignored. More practical ideas involving protecting the elderly are not considered—-maybe because it is difficult.

    We confuse the worry the “scientists” have——it is their job to worry——-with proper public policy.

    Leaders do not believe the people trust them——because they don’t. Just view “approve” numbers for the last 10-20 years. So they go to the extreme to play it safe—-thus intensifying the feeling of unsafeness.

    Future historians will write thousands of books about this. We will see.

    In medical terms this virus is not H1N1. But at this stage of the virus versus H1N1, the projections were high then too—-but we did not have hundreds of count-o-meters to stare at all day.

    Wierd.

  36. Gravatar of Aleksander Aleksander
    14. March 2020 at 07:14

    Todd Kreider:
    “with almost 4,000 cases, Germany’s mortality rate has been consistently at 0.2% after its first 1,500 cases”

    What do you mean by this? Germany’s first fatality happened 5 days ago, and since then 1-2 people have died every day. But we know that these people must have been infected weeks ago, and have little to do with the recent surge in new infections/widespread testing. Even if Germany in fact has identified every single one of their cases (which is unlikely), it’s overwhelmingly likely that more than 0.2% of them will die (since 0.2% already has died, and most of the infected are likely to be relatively early in the process).

    Again, I’ll bet you on this. At least 40 people will die in Germany over the next month. I’ll take almost any odds on this bet; the patterns in every infected country so far is consistent enough for me to be confident in this.

  37. Gravatar of Aleksander Aleksander
    14. March 2020 at 07:23

    Todd Kreider:
    “Time will tell if we (Denmark) are being hysterics”

    It’s ultimately a question of values, but given the current uncertainties I think it’s very hard to be too careful right now. Italy had identified 20 infections when the first person there died, and the number of dead has been rising exponentially since then, despite more and more stringent quarantines. There is every reason to believe that this is because most of the people who die now were already infected then (first death was three weeks ago). Assuming a death rate of 1%, that means there more than 50,000 infected in Italy when the authorities thought there were only 20. (A higher death rate would mean even more infected.)

    The same is unlikely to happen in Denmark today. But unless your government has run extensive tests on a representative sample of the population already, I doubt they can say with any certainty how many are currently infected. Even if they’re not off by a factor of 2500, they might well be off by a factor of 10, or even 100.

  38. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 07:32

    Re: “waiving rules”

    One of the more amazing things I heard at the press conference Friday was how many different times—-about different issues——the Administration announced the “waiving of rules”.

    We waived rules because it is safer. Why do we have so many rules that need to be
    “Waived”? That is an obvious question.

    Press will never discuss it, but they will make sure I get a pop up that an 82 year old woman with emphasima was the “first” coronavirus death in NY. About 450 other people died yesterday in New York. Along with that cheery news, the pop up added the “officials “fear” that hospitals “might” get overcrowded.

    I keep forgetting one industry that loves this is the media.

  39. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 07:36

    PS—-“waiving Rules”

    I forgot to give the answer. We have so many rules to waive because of human nature. It is human nature to expand one’s mission and to grow. This includes regulators. We have so many rules to waive, because regulators expand their mission.

  40. Gravatar of Todd Kreider Todd Kreider
    14. March 2020 at 08:04

    Aleksander: “Germany’s first fatality happened 5 days ago, and since then 1-2 people have died every day. But we know that these people must have been infected weeks ago, and have little to do with the recent surge in new infections/widespread testing.”

    People who died five days ago had to have been infected within the past three weeks or so, although an outlier (within reason) is possible. Germany’s death rate has held steady at 0.2% from 1,500 cases to 4,000 cases, which is telling. Germany can easily have 40 deaths over the next 40 days, but I bet their mortality will remain under 0.5%.

    An OECD country can’t be off by a factor of 10 to say nothing of 100 with respect to how many really have been infected by the coronavirus because an upper bound of the mortality rate is 2% outside of Italy, which has been an outlier the past few weeks. You don’t need extensive tests in your own country to have a good idea because countries like South Korea have tested extensively.

  41. Gravatar of Student Student
    14. March 2020 at 08:19

    Scott,

    What matters is if the ratio (delta new cases day t / delta new cases day t-1) is > 1 or not. If it is > 1, then we are still in the exponential phase of the logistic curve. When it is 1.0, we are at the inflection and we can expect the current number of total cases will double and then top out.

    Suppose the ratio is 1.15, then 20,000 cases will become like 100 million in 61 days. If it’s 1.05, 20,000 cases will become 400,000 cases in 61 days. It’s all about this ratio and how we limit the spread and when. People are still treating this like a snow day. It’s not. It’s time to get drastic now. #doitnow.

  42. Gravatar of Student Student
    14. March 2020 at 08:22

    Check out Alex’s post on MR about MRU and the coronavirus. He posted an excellent relevant 8 min refresher on the logistic curve. Very well put together. My elementary school kid could understand even…

  43. Gravatar of Aleksander Aleksander
    14. March 2020 at 08:30

    Todd Kreider:
    “People who died five days ago had to have been infected within the past three weeks or so, although an outlier (within reason) is possible.”

    More specifically, all of them were almost certainly already infected before the first of them died five days ago, and the official number of “confirmed cases” was 850. So the 4000 number has nothing to do with the people who have died so far. The sudden increase in “cases” either reflects how many people got infected since the first person died, or it reflects more extensive testing (I think both). If you think we can trust the number of confirmed cases, you should at least calculate the death rate from 850, giving a death rate of 0.95% (although I don’t think this is a very good method either).

    “An OECD country can’t be off by a factor of 10”

    So you think there were 200 infected in Italy three weeks ago? That would mean almost everyone who died, died within a couple of days of being infected. We KNOW that’s not what happens. They get infected, stay asymptomatic for a 5-10 days, get sick, go to the hospital, and die within a month. Maybe some of them die really quickly, but we know that the average isn’t anything like 2 days.

    More specifically, China’s rate of deaths/day peaked 22 days after the lockdown in Hubei. That suggests that the average time from infection to death is about 3 weeks.

    “because an upper bound of the mortality rate is 2%”
    What do you mean? Most infections in the world right now are new, so it will take a few weeks for people to start dying from them. I agree that we will have a better overview of the number of infected once all of them are dead, but right now, they’re not. That’s why Italy managed to be off by a factor of AT LEAST 100; there’s almost no way to be sure until people start dying in droves.

  44. Gravatar of anon/portly anon/portly
    14. March 2020 at 08:43

    Physicist: “To solve the puzzle, check the Total Recovered column. … The important ratio is Total_Death/Total_Recovered. It is similar everywhere.”

    SS: “Physicist, Interesting.”

    Physicist said “check,” right? Here’s the top 13 in deaths according to JH, with the ratio in parentheses:

    3075 Hubei 52960 (.06)
    1266 Italy 1439 (.88)
    611 Iran 2959 (.21)
    191 Spain 517 (.37)
    79 France 12 (6.58) [note: “France France” in TR column]
    72 SK 510 (.14)
    37 WA 1 (37)
    22 Henan 1250 (.018)
    21 Japan 118 (.18)
    21 UK 18 (1.17)
    13 Heilongjiang 447 (.03)
    13 Switz 4 (3.25)
    10 Netherlands 0 (undefined)

    A few more:

    8 Germany 46 (.17)
    8 Guangdong 1299 (.006)
    7 Diamond P 325 (.02)
    2 Sweden 1 (2)
    1 Zhejiang 1211 (.001)

    These ratios are, of course, if anything LESS similar to each other than the Deaths/Confirmed Cases ratios. Physicist’s comment was silly, not interesting.

    Maybe the Asian ratios are interesting, but even if they are we should ignore them so as not to be culturally racist.

  45. Gravatar of ssumner ssumner
    14. March 2020 at 09:12

    Brian, Fair point, but compare France and Germany. Their graphs track each other over time really closely, and yet the deaths are vastly different.

    Sty.silver, Maybe I shouldn’t be so optimistic. But given we know that that the East Asian approach seems to work, there’ll be enormous pressure to adopt it here.

    Mads, I suspect the Danes will be proved right—within days.

    Michael, Yes, and even I knew about the clothes mislabeling years ago. It wasn’t exactly a secret.

    Aleksander, That’s plausible.

    Ben, The UK has already abandoned your idea.

    Michael, Some of the non-Hubei deaths were labeled “pneumonia” as people did get that disease toward the end. My dad died of emphysema/flu, take your choice.

    Student, Good point,

    Anon, I should have checked. Oops.

  46. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 09:15

    RE; K-12 school closings

    15 states have done this. Even in a world of safety at all costs, this decision seems like a particularly bad one. WHO said (on 3/2/20) that in China 1.2%.all cases are people of age 10-19 and .9% if under 10. So this translates to about 800 children having the virus. Assuming children are 25% of population this means (using Hubei population, but China cases) the odds are 1/25000 of a under 19 year old contracting the virus. Of the 800 cases let’s assume 25 died, even though the reports are less than 3. So, the odds of dying is 1/800000 (using Hubei population and assuming 20 million out of 80 million are 19 or under).

    In the US, assuming 80,000,000,children, the odds of a child dying in a vehicle fatality is 1/30000 approximately. (2800 die in vehicle crashes)

    Ok—-you get the point. .

  47. Gravatar of Todd Kreider Todd Kreider
    14. March 2020 at 09:16

    Aleksander wrote: “So you think there were 200 infected in Italy three weeks ago? That would mean almost everyone who died, died within a couple of days of being infected.”

    No. I wrote earlier that this does not apply where a country has just started to reported cases and deaths. Germany with 4,000 cases is likely past that point. Symptoms arise in 5 days on average (97% within 11 days) so 5 to 10 days is not how to describe it, and it takes 3 weeks to go from symptoms to death, not 4 weeks.

  48. Gravatar of MJ MJ
    14. March 2020 at 09:49

    Norway’s infections started about two weeks after Italy’s Lombardy cluster, so I suspect we have yet to see the true mortality rate there.

  49. Gravatar of Michael Rulle Michael Rulle
    14. March 2020 at 10:14

    Let me go positive on this, or try at least. Assuming this will result in a lower equilibrium rate than if we did less, or at least not worse, it’s good practice for other times when we Will need to coordinate. And it will be over, I think. And there will be optimism when we believe it is “over”.

    As I write this, though, I really wish we were less afraid

  50. Gravatar of Ray Lopez Ray Lopez
    14. March 2020 at 10:19

    Ssumner: “Ben, The UK has already abandoned your idea.” – what? Source please? Fake news, like Christian List, the fake internet “MD”. In fact, a mere day ago 3/13/20, it is stated here the UK policy remains: https://www.wbur.org/hereandnow/2020/03/13/uk-covid-19-coronavirus (“There are more than 500 cases of COVID-19 in the United Kingdom and Prime Minister Boris Johnson says his government has moved from trying to contain the coronavirus to delaying it. The strategy is controversial and not everyone agrees with it.”)

    Seems then the UK strategy is still in effect? What else are Christian List and Scott Sumner misleading us about? I personally have apologized to Ben Cole, who called the UK strategy well before it became news. Call it luck, but B. Cole, who literally wears a tin foil hat on YouTube when vlogging on economics, called for the UK strategy well before it became news. Give credit where it’s due. Heck, if somebody can prove strong money non-neutrality I’m willing to give up my stance on that issue as well.

    “When the facts change, I change my mind. What do you do Sir?” – attributed to Lord Keynes, speaking to somebody like C. List or S. Sumner.

  51. Gravatar of Peggy Simelon Peggy Simelon
    14. March 2020 at 10:33

    Re Coronavirus death rates.

    Currently loos like 95% of people who contract Coronavirus recover without medical assistance. 2% need a respirator, are unlikely to live if none available. The U.S. has about 3 hospital beds per 1000 people, and about 0.5 respirators per 1000 people.

    To have enough respirators, need number of exposed U.S. residents < 10,000,000

    All the respirators in (northern) Italy are already in use, so the marginal coronavirus patient can’t have one

    I think the US is 8-14 days behind Italy, let’s hope extraordinary measures were taking are enough to bend the curve a lot

  52. Gravatar of Aleksander Aleksander
    14. March 2020 at 17:00

    Todd Kreider:
    “No. I wrote earlier that this does not apply where a country has just started to reported cases and deaths.”

    Fine, so you think there are currently 22,000 cases in Italy? That means 6.8% mortality rate.

    And even if you do, or if it doesn’t count or something; why do you think OECD countries can’t have measured just 1/10 of carriers? We already know deaths are going to increase over the next few weeks, as the people who are currently infected died off; the question is how many. And according to my models, current rates of deaths/day suggest that many first-world countries are off by a factor of 10.

  53. Gravatar of Christian List Christian List
    14. March 2020 at 18:00

    compare France and Germany. Their graphs track each other over time really closely, and yet the deaths are vastly different.

    Interesting indeed.

    Either the French detect too few people that are alive and infected — or the Germans detect too few people that are dead and infected. It’s usually a combination of both.

    This may also explain the Italian death toll. Many more people in Italy would then be infected than the official figures indicate. Probably there were many more infected people in Hubei as well.

    Or maybe the French were just unlucky and the virus has gone through a few retirement homes, which has not happened in Germany yet.

  54. Gravatar of Thaomas Thaomas
    14. March 2020 at 18:34

    We don’t what fiscal policy to “stimulate the economy;” we want it to invest in positive NPV activities and transfer income to temporarily and sometime permanently unlucky folks. Only the Fed can enforce Say’s Law, making sue that what is not spent on goods that can’t be produced will be spent on other goods ~ maintain NGDP on track.

  55. Gravatar of ssumner ssumner
    15. March 2020 at 08:47

    Michael, Schools are not closing to protect children, they are closing to protect the grandparents of children.

Leave a Reply