Watch the islands

In my view, we will learn a lot about Covid-19 over the next 4 weeks from the data coming from various island countries. Many of these countries are now pretty isolated from the rest of the world, and will provide a useful test for certain key questions.

1. Greenland had 11 cases, and now has zero. It was the first island to exterminate the virus.

2. Faeroe Islands had 185 cases. There are only 11 active cases today, and no new infections since April 6. No deaths, and no one is in serious or critical condition. They will likely eliminate the disease within a few weeks.

3. Iceland has had 1771 cases and gets about 10 new ones each day, with the number steadily declining.  They will probably no longer be getting new cases after a few more weeks, and then in another 6 weeks or so will be virus free.  They’ve had 9 deaths.  Iceland is important because unlike Greenland and Faeroe Islands it’s a statistically significant sample.  Within a month or so we’ll have a good idea as to how many Icelanders will eventually die of the disease (I’d guess about 15), and this will begin to pin down the actual fatality rate.  Testing is extremely comprehensive in Iceland, and hence the data is more accurate than elsewhere.

4.  New Zealand has had 1431 cases and gets about 10 new ones a day, with the number steadily declining.  In other words, very much like Iceland.  As in Iceland, active cases are also falling very fast.  They’ve had 12 deaths, a modestly higher rate than Iceland.  This makes sense given that they’ve tested less comprehensively than Iceland, and thus missed a few more cases.  The NZ government intends to drive the case total to zero, at which time normal life can resume.

You see a similar pattern in other islands.  Taiwan had a spike of new cases today from a ship in their navy, but otherwise has almost stopped community transmission. Hawaii has bent the curve more than other American states.  I’d also like to point to some quasi-islands:

1.  Australia looks a lot like New Zealand and nothing at all like Canada (which it closely resembled during the early weeks of the crisis.)  Its active caseload is falling fast, as is community transmission.  The mortality rate so far is a bit over 1%.  That will rise modestly, but of course they missed some cases.

2.  South Korea’s hard border with the North makes it a quasi-island.  Active cases are falling very fast, with rapidly declining community transmission.  The reported mortality rate is over 2%, but of course they missed some cases.

3.  Hong Kong has only 4 deaths in 1026 cases, and only 8 are in serious or critical condition.  Community transmission has almost stopped and active caseloads are falling fast.  Macao had only 45 cases, no deaths, and community transmission has stopped.

Overall, the various island and quasi-island data suggests a far lower mortality rate than what one sees in Europe, or even the US and China.  The obvious explanation is that these islands are missing fewer cases, as demographics alone can’t explain the 13% reported mortality rates you see in some European countries.  Singapore’s particularly interesting, with only 11 deaths out of 6588 cases.  I attribute that to pretty complete testing and the fact that many cases are recent, so some of the infected will eventually die.

Of course recall the previous post.  Everything here could turn on a dime, as infections surge again in some of these countries.  I did not expect the recent surge in Singapore.  But we’ll know much more in about 4 weeks.

PS.  I’ve generally discounted the possibility of “herd immunity” building up to any significant extent.  Not enough cases where I live in Orange County.  But I’m going to change my view on New York.  They look likely to end up with at least 30,000 deaths.  That suggests that at least 3 million New Yorkers will be infected in this wave of the epidemic.  That’s 15% of New York State’s population.  (And the ratio would be even higher in the NYC metro area.)  Furthermore, the infected would skew heavily toward socially active people that interact a lot with others, including the so-called super-spreaders.  The non-infected 85% will have a lower than average R0.  That’s certainly not complete herd immunity, but given the skew toward the socially active, the 15% figure understates the true amount of herd immunity.  The next wave in NYC (next winter?) will be milder, especially if people continue to be cautious, wear masks, etc.

Maybe it’s wishful thinking on my part, but I believe we can ride out the rest of this pandemic (until a vaccine) without any more government-mandated lockdowns after this one is lifted in May.  But people will still be very cautious, and hence I expect a weak economy for quite some time.

PPS.  This picture of Faeroe Islands looks like an optimal optical illusion.  Unfortunately, I’ll never get there:


Tags:

 
 
 

41 Responses to “Watch the islands”

  1. Gravatar of DeservingPorcupine DeservingPorcupine
    19. April 2020 at 14:31

    Aren’t you retired? Go there!

  2. Gravatar of agrippa postumus agrippa postumus
    19. April 2020 at 15:16

    and your point is islands are equivalent to mass populations? stick to sticky wages, radar. they are your excellent faults.

  3. Gravatar of Benjamin Cole Benjamin Cole
    19. April 2020 at 15:22

    Japan has had a low rate of infection but now it appears to be growing. They have made the curious decision to keep schools open, but to go to lockdowns. Children are famously immune to C-19 but can spread the disease to adults.

    I have seen estimates that as high as 40% of New Yorkers are probably infected or have antibodies to the virus in their system and are thus now immune. If New Yorkers really wish to protect the remaining population from becoming infected with the cold virus, they will have to adopt long-term draconian measures, to put it mildly.

    The US just extended the lockdown on the Canadian-US border, but of course the southern border remains a defacto open border, where perhaps three thousand people a day crossing over. There are reports that the nursing-home population of San Diego is in retreat.

    Regarding policy towards C-19, the important thing is what you want to believe, not what you have to believe.

  4. Gravatar of ssumner ssumner
    19. April 2020 at 15:36

    agrippa, You asked:

    “and your point is islands are equivalent to mass populations?”

    Yep, I’m claiming that the Faeroe Islands are essentially identical to India. After all, I said that in the post, didn’t I?

  5. Gravatar of Ray Lopez Ray Lopez
    19. April 2020 at 15:47

    Ssumner: “I’ve generally discounted the possibility of “herd immunity” building up to any significant extent. … But I’m going to change my view on New York. … at least 3 million New Yorkers will be infected in this wave of the epidemic. That’s 15% of New York State’s population.”

    My god, does Dr. Sumner not know how herd immunity works? Let me give a quick lesson. Herd immunity is given when the infected population exceeds = 1 – 1/R0 Covid-19 has a R0 that varies between 1.4–5.7 (Wikipedia on R0). Hence if R0 = 1.4, herd immunity is achieved at a population infected percentage of 1-1/1.4 = 29% (or greater). If R0=5.7 however, you need a population infection of at least 82%, much higher, which means more people will die (recall 20% of C-19 victims have severe reactions that may lead to death).

    Herd immunity changes according to behavior, so ‘social distancing’ will reduce R0 and hence herd immunity goes lower, however, some studies suggest R0 does not change as much as you might think. I hope that clarifies things.

  6. Gravatar of Brian Donohue Brian Donohue
    19. April 2020 at 15:52

    15% infected + social distancing could go a long way.

  7. Gravatar of JG JG
    19. April 2020 at 15:55

    The antibody study take in Santa Clara County suggests actual cases are underreported by 50x to 85x. In a week or so the same team will release another antibody study covering LA County and 20+ other cities around the nation. Numbers will vary pending ages , population density, and general healthiness of residents and other factors. Still based on study and current reported cases it is reasonable to estimate that between 30M to 60M Americans have already contracted corona. Because people don’t know they have disease, once it starts to move it is virtually impossible to stop sAve for cure or robust testing in place that can detect and isolate very early on. Good news is that fatality rate is on order of that of a bad flu season. With or without lockdown numbers in one year likely to be similar. After spread starts in earnest we can Protect elderly and at risk and avoid hospital overcrowding and some mitigation (ie masks, etc) but otherwise corona will run its course one way or another.

  8. Gravatar of JG JG
    19. April 2020 at 16:02

    A very highly regarded epidemiologist in Sweden noted that one of the reasons Norway has fewer corona fatalities per 1M Than Sweden is because norway’s senior homes have far fewer residents per home than Sweden.

  9. Gravatar of Brian Mccarthy Brian Mccarthy
    19. April 2020 at 16:39

    Small sample, but interesting:

    https://www.businessinsider.com/coronavirus-test-200-chelsea-massachusetts-finds-32-percent-exposed-2020-4

  10. Gravatar of Benjamin Cole Benjamin Cole
    19. April 2020 at 18:22

    JG commenter above: A+

    To paraphrase Joe Louis: We can run, but we can’t hide.

  11. Gravatar of Elizabeth Harris Elizabeth Harris
    19. April 2020 at 19:12

    Covid 19 wiki if anyone is interested

    https://lymediseaseassociation.org/home-page-featured/covid-19-sars-cov-2-helpful-info/

  12. Gravatar of Matthias Goergens Matthias Goergens
    19. April 2020 at 19:28

    Scott,

    Singapore’s recent spike of cases is mostly amongst the young and physically fit foreign construction workers.

    As you suggest, the spike is too recent to say anything about death rates, but the demographic of infected also differs from Singapore as a whole in a benign way.

    Ray, your math is probably too simple. I don’t think we can completely ignore the properties of the social graph.

    Silly example: imagine a society of many recluses and one rock star. The recluses never interact with each other, the rock star interacts with everybody.

    No matter the numbers, if the rock star becomes immune and unable to carry, any disease will die out.

    Scott’s argument is that if you selectively remove the most connected nodes from your social graph, what’s left over is a much sparser graph.

    The explanation of your formula on Wikipedia starts with the caveat:

    “This threshold can be calculated by taking R0, the basic reproduction number, or the average number of new infections caused by each case in an entirely susceptible population that is homogeneous, or well-mixed, meaning each individual can come into contact with every other susceptible individual in the population,[9][29][42] and multiplying it by S, the proportion of the population who are susceptible to infection:”

    With emphasis on “population that is homogeneous, or well-mixed”.

  13. Gravatar of Market Fiscalist Market Fiscalist
    19. April 2020 at 20:21

    To put the Iceland numbers in perspective: California has over 100 times the population of Iceland and has just over 100 times the number of deaths (the death rate per million is probably slightly lower for California)

    Not sure what this tells us but it does suggest to me that all statistics about covid-19 have to be treated with extreme caution in terms of learning anything useful about the disease as I am sure that California has just been lucky in some as yet unknown way to have seemingly avoided being harder hit.

  14. Gravatar of Christian List Christian List
    19. April 2020 at 20:46

    I don’t think that herd immunity is just a whole-or-nothing trait, as Ray seems to imply, but also a gradual process, therefore gradual immunities are still useful.

    Thus, if we assume that there are 5-15% who already had Covid-19, plus a substantial background immunity, as cruise ships and a French aircraft carrier seem to imply, plus the summertime with higher temperatures and maybe more humidity, plus all the measures that lower R0 (like social distancing and masks), then hopefully we are on the right track.

  15. Gravatar of mbka mbka
    19. April 2020 at 21:32

    Scott,

    for all islands, and indeed for all countries – “normal life” includes international trade and travel and therefore, it can’t resume until a vaccine, effective treatment, or sufficient herd immunity are achieved. This even more so for islands, because they depend very heavily on international exchange of goods and labour, not to mention tourism. In which case NYC’s “strategy” or lack thereof, will allow it to open for normal business much sooner than NZ, HK or SG.

    Christian List,

    I kinda share your hope. Remember just a short while back at the beginning of this crisis, I wrote more or less the same thing, minus numerical estimates and plus the hope that subclinical cases would spread background immunity and/or less lethal variants of the virus would do the same thing… and you ridiculed me.

  16. Gravatar of dtoh dtoh
    20. April 2020 at 02:36

    I curious why you left out….

    Singapore
    Japan
    Ireland
    Great Britain, and
    Manhattan

    But I’m glad that you now agree with me that travel restrictions do matter.

  17. Gravatar of ChrisA ChrisA
    20. April 2020 at 03:25

    I still think we are trying too much to rely on natural experiments. They are subject to too much interpretation. Let’s do some proper controlled experiments and get some data we can rely on. We need 1,000 healthy young volunteers to be deliberately infected and then we can monitor them closely as they progress through the illness. For those arguing it is unethical – this is a classic trolley problem. I thought utilitarians had already answered these problems.

  18. Gravatar of W. Peden W. Peden
    20. April 2020 at 06:07

    “This picture of Faeroe Islands looks like an optimal illusion.”

    Ignorance is bliss?

  19. Gravatar of Michael Rulle Michael Rulle
    20. April 2020 at 06:12

    Both Germany and the Santa Clara tests had a 15% of population estimate for having had the virus. Scott, how did you derive your estimated forecast of 15% for NY? Based on current rules? If we really believed 15% is the number (a big if of course—not suggesting you are locked into 15%)—-but if we did—-this fact is the pathway to a social solution—-to be followed up, hopefully, by some incremental benefits from vaccines and therapeutics.

  20. Gravatar of Michael Rulle Michael Rulle
    20. April 2020 at 07:03

    Do total deaths matter? The CDC in April has released statistics showing that on an “all deaths basis” that we are so far at a multi-year low.

    It is reasonable to assume that deaths fluctuate and causes of death fluctuate. It does not matter how precise their stats are—-it only matters if they are representatively accurate.

    So this year, Covid was a surprise player, and others were also surprise players. One possible reason excess deaths are not present might be many died from Covid who would have died anyway from something else.

    Or, since we know how e count Covid deaths—-maybe they caused far less

    But putting aside unknowable explanations, at least as we understand things today—-Does this raise any concerns regarding our shelter in place strategy?

    It has to.

  21. Gravatar of myb6 myb6
    20. April 2020 at 07:56

    @JG and anyone else who can pitch in here,

    Not an epimediologist, and it’s really hard to know which experts to trust, but I’ve seen data ppl I respect argue that the Santa Clara antibody study has a serious problem with handling false positives (esp the uncertainty in the antibody test’s false positive *rate*).

    However, seems like Iceland, with its widespread genetic testing and thus far better accuracy, is a much better data source? As of 4/4 https://www.nejm.org/doi/full/10.1056/NEJMoa2006100
    the random testing was coming back 0.6% positive, as of today (which should be about the right lag) they’ve had 9 deaths, which means an actual death/infection of 0.4% which isn’t far from that Lancet study which got 0.66% studying China https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30257-7/fulltext

    The difference might make sense considering the overload in Hubei plus treatment improvements as practitioners learn?

    This would suggest the Santa Clara antibody study is quite probably exaggerating the infection rate 2X or more.

  22. Gravatar of Monday assorted links – Marginal REVOLUTION Monday assorted links - Marginal REVOLUTION
    20. April 2020 at 08:04

    […] Scott Sumner watch the islands.  This piece seems to imply that in-migration is a major source of heterogeneity.  I’ve […]

  23. Gravatar of ssumner ssumner
    20. April 2020 at 08:08

    Ray, LOL, Do you know how to read?

    JG, That Santa Clara study has been completely discredited. It’s garbage.

    Matthias and Christian, No need to respond to Ray, he’s a troll.

    mbka, I agree, I should have said normal domestic life. Just to being able to go out and do things without worrying about Covid-19 has substantial value.

    dtoh, Again for the 100th time. Travel restrictions matter if you have domestic transmission under control. Travel restrictions make absolutely no difference if you are a place like America or Europe in February, just twiddling your thumbs for 4 to 6 weeks. Thus travel restrictions in the US did not help.

    Why did I leave out those places? I picked places that help us understand mortality rates. The places you cite mostly do not. That’s the main purpose of the post.

    Michael, People need to stop citing the Santa Clara study—it’s garbage.

    Where did I get 15%? It’s in the post. 30,000 times 100 is 3 million, which is 15% of 20 million.

    You asked:

    “Do total deaths matter?”

    Matter for what? Matter for human welfare? Yes. Matter for whether we are overreacting? No. For that, we need death rates if there were no social distancing. We lose 35,000 a year in traffic deaths—I’d expect that to fall sharply this year. Also, social distancing will reduce flu and pneumonia deaths.

    You asked:

    “Does this raise any concerns regarding our shelter in place strategy?”

    No. It may well be a bad strategy (I oppose domestic bans on travel), but not because of the data you cite.

  24. Gravatar of myb6 myb6
    20. April 2020 at 08:12

    Going to reiterate that it’s super hard to sort through all this, but the flip side of [Iceland death rate applied to estimate actual US infections -> we’re nowhere near herd immunity] is that a big part of the concern with covid getting out of hand was running out of ventilators, but now that treatment protocol is changing that’s considerably less of a concern.

    So right now I’m (very tentatively) feeling that ppl who can’t work from home should be allowed to return to work, enforcement needs to leave individual outdoor activity the hell alone, but obviously keep going w low-cost interventions like masks, bans on large gatherings, limiting discretionary interaction, etc.

    Closed schools are pretty costly, but just my personal experience as a parent that they are petri dishes makes me think that’s still worth it.

  25. Gravatar of Monday assorted links | AlltopCash.com Monday assorted links | AlltopCash.com
    20. April 2020 at 09:46

    […] Scott Sumner watch the islands.  This piece seems to imply that in-migration is a major source of heterogeneity.  […]

  26. Gravatar of usvi usvi
    20. April 2020 at 10:19

    If you want to control your theory against correlated variables, check out the US Territories. For example, here’s the USVI’s numbers: https://doh.vi.gov/covid19usvi While the US Virgin Islands haven’t cut off entry to the point that an island country can, the tourism influx has gone to near zero in a months where 2 – 3x the population would have otherwise visited. They seem to be doing relatively well at the point but could also be lack of testing.

  27. Gravatar of j w j w
    20. April 2020 at 10:37

    @ssumner

    Scott – you state that the Santa Clara study is garbage. What are you basing that on? I’ve read the peer review on medium https://medium.com/@balajis/peer-review-of-covid-19-antibody-seroprevalence-in-santa-clara-county-california-1f6382258c25
    The reviewer spends all his time discussing false positives but does not spend anytime on false negatives. It’s like he considers only the possibility of false positives affecting the results and not the possibility of false negatives. Both most be considered when evaluating results. The reviewer presents the table of how to figure the false positive rate but does not seem to actually step through the calculation.

    The study may not be valid, but i’m not sure this reviewer has made the case for why it is not so.

  28. Gravatar of DanC DanC
    20. April 2020 at 13:33

    Watch the Ohio prisons. That should give you a good feel for mortality numbers in 21 days. Some are approaching 90% infection rate

  29. Gravatar of DanC DanC
    20. April 2020 at 13:36

    BTW in Ohio, about 50% of reported cases are prisons, health care workers, and nursing homes. Not exactly super spreaders. More like super proximity.

  30. Gravatar of Tom Brown Tom Brown
    20. April 2020 at 14:10

    Optimal illusion? (your last paragraph)

  31. Gravatar of Monday assorted links | Great News Monday assorted links | Great News
    20. April 2020 at 15:18

    […] Scott Sumner watch the islands.  This piece appears to suggest that in-migration is a serious supply of heterogeneity.  […]

  32. Gravatar of dtoh dtoh
    20. April 2020 at 16:11

    Scott

    “Travel restrictions matter if you have domestic transmission under control. Travel restrictions make absolutely no difference if you are a place like America or Europe in February, just twiddling your thumbs for 4 to 6 weeks. Thus travel restrictions in the US did not help.”

    They also matter when the infection is just getting started. And it would be very hard to characterize the U.S. as twiddling its thumb. Most of the behavioral modification was undertaken by firms and individuals, and the travel restrictions in the U.S. provided a huge head start by delaying the initial spread of the disease. And I guarantee you the hospitals were not twiddling their thumbs for three weeks waiting for the CDC and FEMA to act. And it partially offset the time lost to CDC incompetence.

    I just don’t see how you can seriously or objectively argue that the travel restrictions had no effect in the U.S. especially when travel restrictions provided similar results in other countries where they were implemented.

  33. Gravatar of ssumner ssumner
    20. April 2020 at 16:11

    myb6, I mostly agree, but I don’t believe hospital overcrowding was the main motivation for the shutdown. It was to prevent a million deaths. Just to be clear, I don’t believe there was ever any plausible outcome with a million deaths in the US, but that’s because the private sector would have shut down even if the government had not ordered it.

    jw, Maybe that term was too harsh, but read this:

    https://statmodeling.stat.columbia.edu/2020/04/19/fatal-flaws-in-stanford-study-of-coronavirus-prevalence/

    Tom, Well, it’s both!

  34. Gravatar of ssumner ssumner
    20. April 2020 at 16:21

    dtoh, I’m not sure you understand what was going on here in late January, all through February and early March. Almost nothing was being done, and nothing would be done until the problem got very severe. We still had regular NBA games every day, and Mardi Gras. If your policy is “we’ll sit on our thumbs until there are 2000 known cases”, then travel restrictions just delay the inevitable by a week or so; it doesn’t change the eventual outcome at all. It was the same in Europe, where countries are ending up with similar mortality rates even if they got hit later than Italy. I explained this in an Econlog post.

    In contrast, Taiwan was dramatically ramping up its ability to produce surgical masks, while we did nothing. US Hospitals may have drawn up some “plans”, but they were largely ineffective. The private sector has no incentive to stockpile a billion surgical masks, that’s the government’s job.

  35. Gravatar of dtoh dtoh
    20. April 2020 at 16:55

    Scott,
    I know exactly what was going on. I was in the U.S. for most of February. Everyone I know was being much more careful about social contact and hygiene. I’m also on the Board of a manufacturing company. Management of the company was working 12 or 15 hours a days to implement plans to prevent the spread of the infection. At the same time they were in contact with local hospitals to figure out how to increase the supply of PPE.

    Every other company in the area was doing the exact same thing. Other board members were affiliated with hundreds of companies across the country including a very large nursing home chain and one of the world’s largest suppliers of medical equipment. Every single one of the companies was doing exactly the same thing. And “incentives” was absolutely the last thing on anybody’s mind. Anyone who doesn’t understand this knows nothing about human nature or how businesses operate.

    And it didn’t just delay the curve, it substantially reduced the magnitude of the curve.

    And BTW – there was never was a shortage of PPE (or ventilators) in the U.S. Everyone was worried about it, but it never happened. (If you don’t believe me, find one credible account where a hospital (not a disgruntled attention seeking employee) reported that it had run out of PPE or had to ration it at the expense of safety.

    Tell 3M they were doing nothing to increase mask production and they would laugh you out of the room. Or Prestige Ameritech where they’ve been working 80 hours a weeks since the beginning of February.

  36. Gravatar of Benjamin Cole Benjamin Cole
    20. April 2020 at 17:14

    I don’t believe there was ever any plausible outcome with a million deaths in the US, but that’s because the private sector would have shut down even if the government had not ordered it.–Scott Sumner

    I wonder abut this. In 1957-8, no one blinked an eye when 116,000 Americans died of the flu, on a population of about one-half today’s.

    In present-day Sweden businesses are not shutting down. In the US, meatpacking-plant owners complain about being shut down by government ukase. I think they would not shut down otherwise, along with many other industries.

    I suspect larger restaurants and indoor venues might shut down for lack of business, or alter operations. Movie theaters perhaps, or perhaps they would run films intended for teenagers and young adults.

    All over the world, schools seem to stay open or close depending on government edict. They are not closed in Japan.

    A new study on LA County, from USC–

    “221,000 to 442,000 adults” in LA County have COVID-19 antibodies in their blood. No one knows how many additional children have anti-bodies to COVID-19.

    There are 617 COVID-19 related deaths in the county, usually people with co-morbidities.

    That’s a 0.01% to 0.03% death rate from COVID-19—but even that measures only adults. The death rate would sink if children were added into the mix.

    No doubt the USC study will now be attacked or defended on its methodology, because COVID-19 is no longer about policy options, but defending one’s previous positions or even ideologies.

    Frankly, I think hysteria beat science.

  37. Gravatar of Elizabeth Harris Elizabeth Harris
    20. April 2020 at 18:54

    Canada and Katy Tur are reporting that Kim Jung Un is brain dead following heart surgery. What a weird year.

  38. Gravatar of ssumner ssumner
    20. April 2020 at 19:16

    dtoh, They may have thought they were taking effective steps, but they were not. The fact is that people lived their lives pretty much as normal until the caseload got high in mid-March, and then everything turned on a dime. With a tighter travel ban it simply would have delayed the point at which we got serious for a week or two, and then we’d be in exactly the same boat. The NBA season would have gone on a week longer.

    The same thing occurred in Europe. Countries watched the disaster unfold in Italy and did almost nothing as the Italian death toll mounted. They when it starting getting bad in their own country they began to take steps. As a result, Italy will end up not much worse than average when it comes to the final death toll. Other countries are now catching up.

    I agree that ventilators are not the big problem. But we should have been producing masks and shutting down major events. We didn’t, and the problem spread like wildfire. This country was simply unwilling to act until it was too late.

    BTW, Trump just announced he intends to suspend all immigration to the US. Let’s see if the courts have the guts to stand up to this dictator. Congress sets immigration policy.

  39. Gravatar of Steve Steve
    20. April 2020 at 19:57

    This is a smart post; I like the case studies.

    A bunch of thoughts:

    1- Aus & NZ are heading into winter. I don’t think this matters much (low-density seems more important for coronavirus), but worth watching.

    2- Travel hubs seem more affected than travel destinations. Most Islands are destinations, although Iceland has some hub aspects.

    3- Singapore: what a previous person said, young (and likely thin) workers, heavily from Bangladesh, should have lower infection fatality rate.

    4- I think Japan and Korea are the most interesting cases, since they have very dense cities that are constantly vulnerable to reinfection from travel abroad. If they succeed it’s reason to believe masks work, and also that my herd immunity fear is overblown.

    (virus thoughts next post)

  40. Gravatar of Steve Steve
    20. April 2020 at 20:13

    This virus has some unique properties which is one of the reasons “Nobody knows anything.”

    It’s a respiratory virus that can linger in the air for several minutes to several hours. There are reportedly also very high rates of “viral shedding” early in the infection. There seems to be a multiplier effect, any time people share a small amount of air. I am less concerned about physical distance and outdoor activity, but confined air space is problematic. It’s almost like the virus was intelligently designed to maximize spread…IN A CAVE!

    What are the most cave-like man-made structures? Obviously, subways. But airport walkways, airplanes, dive bars, and gyms also have characteristics a batvirus could love.

    Transportation infrastructure is especially problematic, because it nullifies the potential of test and trace. You’ll never be able to identify contacts, if the virus spreads inside the big underground people movers at Detroit airport, or in the 6 Subway in Manhattan.

    It’s a huge unknown, if transit hubs and crowded metros, will be able to sustain the lack of travel necessary to prevent reinfection from other hot spots.

  41. Gravatar of ssumner ssumner
    21. April 2020 at 09:16

    Steve, Very good points. I do think warm weather helps, but the very similar patterns in New Zealand and Iceland make me think it’s not decisive. If Iceland can control it (remains to be seen), New Zealand will probably also succeed.

Leave a Reply