The awful arithmetic of herd immunity

The ABC has an article quoting University of Melbourne epidemiologist Tony Blakely as saying (approvingly) that the object of the current “flattening the curve strategy is to smooth the path to herd immunity. Key quotes

You don’t go in too hard because you actually want the infection rate to pick up a bit and then hold,” he said.

“What they’re not saying is [that] ‘flatten the curve’ likely means [that] by the time this is over, 60 per cent of us will have been infected, to develop herd immunity,” he said.

The arithmetic here is pretty horrifying. 60 per cent of the population is 15 million so with a 1 per cent fatality rate, that would be 150 000 deaths. The number surviving but with long-term lung damage could easily be over 1 million.

It gets even worse. If herd immunity is supposed to be achieved over 12 months (by which time we are hoping for a vaccine) that would imply 40 000 new cases every single day. If even 10 per cent required hospitalization for several weeks, they would fill every bed in the country.

As for intensive care, we have a total of just over 2000 beds. Even with a hospitalization rate of 1 per cent, they’d be full in five days. And given that 1 per cent is the estimated fatality rate with treatment, that implies triage on a massive scale, which in turn would greatly increase the death rate.

This is simple arithmetic. If Blakely’s explanation of the government’s strategy is correct, it should be spelt out.

58 thoughts on “The awful arithmetic of herd immunity

  1. You can, of course, try to flatten the curve of infections without necessarily pursuing “herd immunity”. Isolating people so that you can seek to identify every case of coronavirus in Australia and then addressing those will flatten the curve by limiting cross infections but also lead to a long term controllable level of the virus where large numbers have not been infected. At least this should be the preferred objective even if ultimately it fails. The risk of such failure increases the longer we delay decisive action – I have been advocating strong controls for weeks but the perspective of policy-makers (and health authorities) has always been, “let’s wait and see if this thing becomes more serious”. There are no option value advantages to waiting – the economics/health tradeoff seems to me illusory – minimizing deaths in a myopic way seems a good policy both from the viewpoint of health, of costs and in terms of minimizing long-term economic disruption. The costs of control become much higher with each day of delay.

  2. Personally I cancelled my Cub Scouts group, took my kids pout of school and stopped visiting family on March 13th.

    It would be nice if the public was given some information about what proportion of the cases so far are requiring hospitalisation, and what proportion need ventilators. (Perhaps that information does exist and I just haven’t seen it)

    I’m just hoping that there are many mild/asymptomatic cases in Australia which are not being detected.

  3. More arithmatic from David Walsh of gambling and Mona fame.

    Comments with nuance.

    “it should be spelt out” to all staff at woolies coles etc to see and understand this as they may take a dim view at current total lack of ppequip.

    “On the sixteenth day, the first (linear) series will reach 1600. The second (exponential) series will be 1541. Still good. Now let’s go another sixteen days. The first series is now 3200. The second series: 28,485. And another sixteen days? 4800 versus 526,645. And another sixteen? 6400 versus 6,761,701. If that latter number were people, and a disease spread at that rate, then starting with 100 people, there would be more than six million infected in sixty-four days. Well, it is a disease, and it’s called COVID-19, and that’s a bit slower than it’s growing in Australia (and there are already more than 100 victims). The 1.2 that it is multiplied by each day, that means that each person infected with COVID-19 is infecting about 2.4 other people (not 1.2 because the incubation period is more than one day). If we can get that 2.4 down to less than one (and thus the 1.2 down to less than one), the exponentially expanding sick population will fall, and it could decline as fast as it rose if the 1.2 becomes 0.83. That’s why the Chinese COVID-19 population plummeted when people stopped hanging around with each other. But so far, that stuff isn’t working in Australia. People are still interacting, because people are social, and people don’t understand exponential growth.

    Indulge some speculation. If the chance of catching COVID-19 from a carrier during a short contact is 0.05 (one in twenty), then the 2.4 infections that each such carrier causes means they are having forty-eight interactions while they are sick (on average). To get that to below 0.83, they need to reduce the number of interactions they have to sixteen or less. It’s really insidious because an uninfected individual is really unlikely to catch COVID-19 going about their normal day. In Tasmania, for example, there are seventeen people that are known to be infected. Because of the incubation period and untested positives, there are probably really around fifty carriers (presuming the seventeen have been removed from the population). I can guess from the 2.4 infections per carrier that over an eight- or nine-day period infected people interact with an average of forty-eight people (or more times with less people). Presumably uninfected people interact with more people (because they’re less sick), so say sixty-four people (twenty-five per cent more). If there are only fifty people who can infect you (today, but exponential growth means that number will skyrocket), then your chance of being infected over the next week and a bit is (50/500000*64) = 0.0064, or less than one per cent. So why does the exponential growth occur?

    Here’s another (apparently unrelated) problem. How many people are required before there is a fifty per cent chance of two of them having the same birthday? It’s not 183 (half of 365). It’s actually twenty-three. 183 is the answer if the question was how many do you need before someone is fifty per cent likely to share yourbirthday? But all of those present can share a birthday with the others, so each one of the twenty-three can share a birthday with the other twenty-two, and each one of those with the remaining twenty-one, and so on. Clearly, you need 366 people to guarantee that two people have the same birthday (in a non-leap year), but with seventy people the chances of two sharing the same birthday is 99.9 per cent. Your chance of being infected when you wander the streets might be small (over the incubation period) because there aren’t many COVID-19 positive people. But with lots of people wandering the streets, all potentially interacting with everyone else, interactions happen with COVID-19 positive people, and the 0.05s add up quickly. It’s just like birthdays: more interactions means unlikely outcomes become likely, and then almost certain. That’s called a combinatorial explosion, and COVID-19 uses it to blow us all up.

  4. Hats off to all the economists who blog on this site, including of course J.Q. They all understand the epidemiological basics and what the public health and economic consequences of this pandemic will be. I put this down to the fact that they are mathematically literate and in particular they understand what exponential growth means. A great number of the general public and most of our politicians clearly do not understand what exponential growth means when a quantity is doubling in days.

    J.Q.’s mention of the option value concept is also very useful. “Nearly all the time, this (option value) reasoning favors aggressive early action.” Of course, this is especially so when a problem (like a pandemic) has a doubling time measured in days. A stitch in time will not save nine. It will save nine thousand… or 90,000 or (900,000 or 9 million or 90 million lives when talking globally). Even in Australia It would have saved net not $9 million but more like $900 million or $9 billion.

    I say “would have” because it is all too late for that now. We are in for an event in a range from France’s and Germany’s experience to Italy’s experience. This is because our PM Scott Morrison and his government failed to understand the parameters and potential scale of the event we faced. This is the same PM and government which has denied global warming, spurned bush-fire season warnings and now ignored the clear warnings of the COVID-19 pandemic crisis that the events in the rest of the world from China to Italy were giving us.

    This government is not just inept. It is lethally dangerous to all Australians. People have to act themselves and commit to a personal lock-down regimen as strict as possible. We are in for an enormous amount of personal, financial and social pain and many deaths now. There is no best course of action possible. There is only a “least worst” course. The least worse course is to go into complete lock-down immediately except for absolutely essential services. The curve must be bent back to below the medical and hospital capacity. It is almost certain we will badly overshoot this level and then will have to fight to bring it back.Our total deaths will be assuredly in the 1,000s, very probably in the 10,000s and possibly above 100,000.

    When this crisis is over we must not forget. We must never elect a neoliberal government again. We must change our political economy completely from a system which has failed us so comprehensively. We allowed out hospitals to decay, we removed safety nets for people, we allowed the part-time and gig economy with no security, we failed to make long term and short term preparations for any of the crises hitting us. Then we failed to act in time with compete lock-downs when we had ample warning of the coming of this pandemic crisis.

    The Morrison government has a string of first ball ducks to its name and nothing else.

  5. And I wonder what Tony Blakely would say to this person;

    “The blog I didn’t want to write
    March 21, 2020
    “I am currently infected with Covid-19 in Brisbane, Australia.

    “I would tell them that their parents, any friends who have a chronic disease, anyone friend recovering from an auto immune disease will likely die when the hospitals are full in western countries unless you stay the hell home.”

    I have been asked to write a public post to share what is going on, primarily after writing the above statement.  This comes from my research, being on the inside of it all, and what my opinion is regarding this situation.

  6. The perspective of the game of mates has been to protect their wealth, their commercial interests, their objectives. No creative destruction of capital, no viral destruction, no decimation of the population is to be allowed to interfere with said objectives.

    By this we are doomed, some many of us – “Morituri te salutant”.

  7. There are quite good grounds to question the idea of “herd immunity” acquired naturally as well. Coronaviruses play nasty games with our immune system as there’s evidence that “immunity” is both partial and dose-response based – if so we’re going to *need* a vaccine that works other than by simulating an infection.

  8. Our first chance was that China identify the virus epidemic early and inform the world early so the epidemic could be contained and the virus eradicated in one country. China did none of those things. China deserves serious criticism for these (in)actions. China (or their absolutist party government anyway) is seriously culpable. There is even a case that once they realized their problems, they aided and abetted the virus’s spread to the world or at least did nothing to stop it. Our governments were equally culpable in keeping international travel and movements open for far too long. There is plenty of blame to go around.

    Once COVID-19 became a near ubiquitous global pandemic, the option of keeping a a single nation entirely isolated from it and uninfected was taken off the table of possibilities. The sole exception would be if that nation was prepared to remain completely removed from the world economy and isolated from all international people movements indefinitely. This would impractical and finally impossible to achieve in a the “global village”.

    This does not mean you deliberately infect your nation to get herd immunity. This would be the height of stupidity and completely ethically wrong. In addition, it is entirely unnecessary. COVID-19 is so highly infectious and so stealthy, with a long infectious asymptomatic period, it was always going to get into every country once it had gotten into so many that 1/4 to 1/2 of the world population were already at risk.

    Given those facts, the only option is to go hard early on containment and keep the spread throttled way down to a rate below the level at which the medical system would be overwhelmed. This spreads the pain along the time axis but stops the peak going over the medical system capacity. This is what the experts have been telling us for weeks. Again, the pathogen is so infectious we need not worry that our best efforts will throttle the epidemic too much (if there is a “too much” anyway) in one country. We need worry only that even with our best efforts we might not keep the peak below the medical capacity ceiling. (Indeed, that hope is already passed for Australia.)

    Herd immunity is now a given IFF (if and only if) natural immunity and/or a vaccine are proven possible and effective. The possibility of natural recovery and thus likely immunity seems eminently likely as shown by empirical events thus far. One possible concern is fading immunity and reinfection. Another concern is possible bi-phasic infection by the pathogen. This means you can catch it a second time and more and re-infections are worse, which is a very frightening thought. The idea that COVID-19 could be bi-phasic is basically conspiracy theory and shock-jock talk at this stage, But not enough is know yet to totally rule it out. Re-infection (of a non-bi-phasic nature) does seem maybe possible though. This is probably a science reporter writing, not a scientist

    If COVID-19 proves as bad as it possibly can be then only evolutionary natural selection will select for a human remnant with the ability to resist it. This most dire of outcomes seems highly unlikely at this stage but it cannot be ruled out with 100% certainty.

    Now back to our regular programming… click …Days of our Lives… Tommy realizes this new pathogen could wipe out humanity.. click … The Zombie Apocalypse strikes in a way never expected… click… Humanity’s remnant flees the ruined earth in multi-billionaire Mylon Dusk’s “Screw This I’m Outta Here” … and crash on Mars…

  9. Tom Davies: Agree on asymptomatic/ mild cases. There seems to be growing evidence that the proportion of these is large. 7% of the population of Veneto province in Italy has tested positive, two-thirds with no or mild symptoms. The estimated death rates are iffy because of this. Seems to be 3-4% of those with serious symptoms, much harder data. Herd immunity and estimates of huge death tolls from do-nothing are very uncertain. ICU overload is not.

  10. I think it is very encouraging that the number of new cases has been pretty stable in the last few days. That means linear growth rather than exponential. (Some figures at Can we hold this? Let’s hope so. Should we make the extra push to get it down further? If we get enough testing resources, maybe that will feasible.

  11. I saw that on the news last night and I “felt” it was misleading and deceptive, but I don’t know enough about epidemiology. Now that you have put those numbers to me, it seems absolutely ludicrous. Isn’t it sloppy/irresponsible journalism to present one side of the story. I have worked in medical regulation and I can tell you that there a doctors, many who have great credentials, who hold some outlandish and divergent views.

    I feel there’s a lot of sociopathy going around at the moment and Trump is the most glaring example of that. It appears that mass murder sanctioned by the government is preferable to saving lives. I do feel for people who’ve lost their jobs. It really is terrible for them. But the unfortunate reality is that we have 3 ICU beds per 100,000 people (slightly more than Italy) and 2,300 ICU ventilators to use among a population of 26 million.

    I can’t help but think this is a consequence of underfunding our public health system for a generation. An ICU ventilator cost $50,000 (that’s a top range model), which is about the same price as a Toyota Camry. Why don’t we have a warehouse in Dandenong and/or Cambellfield with 50,000 of them and a workforce paid to audit them and ensure they’re in working order? I did a back of the envelope caculation and for the cost of a F-35, you can purchase 3,000 ICU ventilators. That’s more than our current stock and pretty good value when you compare it to a plane that’s not even fit for purpose. We spend billions on a defence force to fight enemies that may or may not exist and, even if we do, the inflated kit we do have us probably useless. What about defence against the real threats, the micro-organisms like COVID-19.

    Do you think that we could have taken less drastic lockdown measures if we built sufficient surge mechanisms into our public health system? If so, isn’t that analogous to property insurance? Nobody whose insured house burns down questions paying premiums all of those years.

  12. “As for intensive care, we have a total of just over 2000 beds. Even with a hospitalization rate of 1 per cent, they’d be full in five days.”

    Since the average ICU stay is 10 days, a constant flow of 40,000 cases per day with 1% needing intensive care would require 4,000 beds, an upgrade in capacity which would certainly be achievable (although we are still ignoring the needs of non-COVID-19 patients). But the actual outcome would be much worse than this because the cases are not spread evenly throughout a year. Blakely considers a number of “flattening the curve” scenarios here:

    The scenario with the lowest peak (which he regards as too constrained) still has cases peaking at about 100,000 per day. The top of the curve is rather flat, so over a 10-day period this amounts to about 1,000,000 cases. With 1% requiring intensive care, the peak demand is 10,000 ICU patients.

    At a less optimistic assumption of 2.5% of cases requiring intensive care (5% of identified cases, and allowing for 50% of cases as undetected), there would be 25,000 patients chasing 2,200 to 6,600 ICU beds (the federal government has raised the possibility of perhaps tripling ICU capacity).

    There is a reason that Boris Johnson ran screaming from the herd immunity strategy, and the premiers, who are ultimately responsible for the hospital systems which will be so grossly overwhelmed, are revolting. And there is a reason that the dominant expert opinion is now to use strong social distancing measures to suppress transmission of the virus in the short run and drastically improve testing and quarantine enforcement so they can substitute for social distancing over the medium term.

  13. There seems to be a conviction amongst many immunologists that everybody eventually getting this is the best we can do.

    I have not read a convincing argument as to why the disease can not be stamped out entirely, at least within Australia, with a massive testing and isolation effort. If there is a compelling reason why it would not work I wish the damn epidemiologists would spell it out for the rest of us because, as you say, the only route to herd immunity without a vaccine involves unprecedented suffering in our lifetimes. Just about any expenditure on testing and tracing is worth making to avoid it.

  14. The estimate of 1% of cases needing intensive care might be too optimistic unless it is of all known and unknown cases. Aggregate numbers of cases to date on Worldometers shows that of known cases with an outcome 16% were deaths.

    Known cases with a known diagnostic outcome are not figures affected by “exponential lag effects” at all. These cases are done and dusted. I assume that all deaths were ICU cases except for those that died before admission to ICU or were triaged to go and die at home. All these outcomes have occurred in Italy so this point is real. But let us assume deaths before ICU and “triaged to die” are very low or even nil. To bring 16% or rather let us say conservatively 48% (assuming two ICU cases survive for every death) down to 1% we have to divide by 48. Conversely, we have to assume that the number of unknown cases with outcome are 48 times higher than the number of known cases with outcome at a given point in time.

    Is it credible that this factor is so large? We might as well use a round number of 50. Is it credible that known cases are only 2% of unknown cases. This would imply 148,566 x 50 total cases, known and unknown. We might as well say 150,000 x 50 to get a round number which equals 7,500,000 unknown cases with outcomes (got well unaided and in many cases not even thinking they had COVID-19). This is just possible though would epidemiologists notice and note “large numbers of what appear to be ordinary colds and mild influenzas seem to be accompanying this COVID-19 epidemic and self-resolving. Either this is the case or there are many mild COVID-19 cases.”

    This is just possible I suppose. If it is true it would be some good news. Herd immunity might be progressing further than generally thought. I don’t know though. These seem to be too many unknowns in all this.

  15. > why the disease can not be stamped out entirely

    A bit of wild speculation from a highly trained professional not-an-epidemiwhatsit 🙂

    The easy way, possibly the only way, to do that is a vaccine. We might get one, or we might not (the flu vaccine, for example, is maybe 60% efficient, but to wipe a disease out you need much more than that).

    Stopping the spread seems to require pretty drastic action. China, Japan, Korea all took dramatic action, and as The Shovel pointed out, so did North Korea. Getting Australians to accept that level of restriction, and getting the Australian military to enforce it, would be hard. Keeping it in place for more than a couple of months seems tricky even for China, but the experts don’t seem to have ruled out a second wave yet. Australia might avoid that just by having a smaller population (too small for a wave to survive in) and vigorously policing our closed borders.

    I suspect we’d end up with nested enclaves for significant numbers of people. Each being an inner core of vulnerable people with one or more groups of isolated carers surrounding it, then a quarantine barrier around them. Isolated groups inside so that when one carer gets sick that group can be removed/isolated without killing the dependents. Then a new set of carers can go in to replace them, with a suitable waiting period before they contact the dependents. Sounds expensive, but I can’t see how else you can preserve the dependents when the disease is active in the community. You’ll likely have to bribe or draft carers because we don’t have enough of them now.

    Oh, and that assumes kids can’t give the disease to adults. Kids who can’t socialise have problems, but if one kid in a school gets sick and can spread that to adults… what now? An isolated enclave big enough to hold a small secondary school would be pretty big, and having to have extra teachers and parents on hand in case one of them got sick… tricky.

  16. That’s a fascinating article, Luke. I’m surprised no one had already gamed this out. Surely it would make sense for a bunch of epidemiologists, health economists and public health people to get together and run scenarios. Like modern armies doing war games (critical as you might be about their politics, ethics and actions, modern armies are superb at logistics).

  17. “Coronavirus is too radical. America needs a more moderate virus that we can respond to incrementally.” – Anon.

    “One of many perils lies in normalizing the “batshit crazy” presently underway…” – “COVID-19 and Circuits of Capital” by Rob Wallace, Alex Liebman, Luis Fernando Chaves and Rodrick Wallace.

    “… commodity agriculture serves as both propulsion for and nexus through which pathogens of diverse origins migrate from the most remote reservoirs to the most international of population centers.42 It is here, and along the way, where novel pathogens infiltrate agriculture’s gated communities. The lengthier the associated supply chains and the greater the extent of adjunct deforestation, the more diverse (and exotic) the zoonotic pathogens that enter the food chain. Among recent emergent and reemergent farm and foodborne pathogens, originating from across the anthropogenic domain, are African swine fever, Campylobacter, Cryptosporidium, Cyclospora, Ebola Reston, E. coli O157:H7, foot-and-mouth disease, hepatitis E, Listeria, Nipah virus, Q fever, Salmonella, Vibrio, Yersinia, and a variety of novel influenza variants, including H1N1 (2009), H1N2v, H3N2v, H5N1, H5N2, H5Nx, H6N1, H7N1, H7N3, H7N7, H7N9, and H9N2.43

    However unintended, the entirety of the production line is organized around practices that accelerate the evolution of pathogen virulence and subsequent transmission.44 Growing genetic monocultures—food animals and plants with nearly identical genomes—removes immune firebreaks that in more diverse populations slow down transmission.45 Pathogens now can just quickly evolve around the commonplace host immune genotypes. Meanwhile, crowded conditions depress immune response.46 Larger farm animal population sizes and densities of factory farms facilitate greater transmission and recurrent infection.47 High throughput, a part of any industrial production, provides a continually renewed supply of susceptibles at barn, farm, and regional levels, removing the cap on the evolution of pathogen deadliness.48 Housing a lot of animals together rewards those strains that can burn through them best. Decreasing the age of slaughter—to six weeks in chickens—is likely to select for pathogens able to survive more robust immune systems.49 Lengthening the geographic extent of live animal trade and export has increased the diversity of genomic segments that their associated pathogens exchange, increasing the rate at which disease agents explore their evolutionary possibilities.50

    While pathogen evolution rockets forward in all these ways, there is, however, little to no intervention, even at the industry’s own demand, save what is required to rescue any one quarter’s fiscal margins from the sudden emergency of an outbreak.51 The trend tends toward fewer government inspections of farms and processing plants, legislation against government surveillance and activist exposé, and legislation against even reporting on the specifics of deadly outbreaks in media outlets. Despite recent court victories against pesticide and hog pollution, the private command of production remains entirely focused on profit. The damages caused by the outbreaks that result are externalized to livestock, crops, wildlife, workers, local and national governments, public health systems, and alternate agrosystems abroad as a matter of national priority. In the United States, the CDC reports foodborne outbreaks are expanding in the numbers of states impacted and people infected.52

    That is, capital’s alienation is parsing out in pathogens’ favor. While the public interest is filtered out at the farm and food factory gate, pathogens bleed past the biosecurity that industry is willing to pay for and back out to the public. Everyday production represents a lucrative moral hazard eating through our shared health commons.” – “COVID-19 and Circuits of Capital” by Rob Wallace, Alex Liebman, Luis Fernando Chaves and Rodrick Wallace.

  18. IMO, JQ’s analysis is correct, as usual. In Australia, dying in hospital is OK; dying in the streets is unseemly. So flatten the growth curve to try to limit the rate/number of people in intensive care and then dead to the number of available hospital beds. In the interim (until most of us have been infected and hopefully become immune to repeat infections of the same virus), spend money on looking for a cure and/or vaccine.

  19. You know Ikonoclast, your “plenty pf blame to go round, but let’s accept it is China first and foremost” seems to me to miss the importance of foreknowledge. Perhaps you dismiss this as irrelevant. An interesting proposition.

  20. witters,

    Never mind foreknowledge, we had forewarning. We would have had more forewarning if China had been more open earlier. China also might have even contained this virus if they had not suppressed the early-warners in their own system.

    We had forewarning for sure. As early as mid to late January the internet was full of videos and commentary about the Wuhan crisis. Some were authenticated and most could not be dismissed as hoaxes: videos of hospital corridors full of sick people, screaming people and staff breaking down from fear, stress and fatigue- videos of people collapsing, doing face plants in the street and staying there inert.

    My son (a stock market investor and former software engineer) called it in mid-January. He said there was going to be a world pandemic and stock market volatility and crashes. He began operating on that assumption, shorting the market, day trading and so on. I held off calling it. I remembered that SARS and MERS had been contained. One did not really know the infectiousness of the new pathogen at the stage. But by the end of January I was convinced a global pandemic was in the offing. At that point it was possible to make the call to isolate Australia from the world and quarantine all returning Australians. It should have been done. No magical foreknowledge required.

    The trouble is that the bias to business as usual and the disbelief in serious black swans blinds people to patently visible black swans when they are already spreading their wings.

  21. It’s peculiar that our death to cases ratio is so small (13 / 3700).

    So far, in any case.

    There’s several possible explanations. One is that people who are less likely to die have been over represented in our cases. Another is that, compared to other countries, our hospitals are well resourced and the quality of our doctors and nurses is high.

  22. It’s early in the growth curve. Not enough cases have run long enough to cause deaths. Our death count is 13 at this time on Worldometers and our recovered count is 170. Thus, of known cases progressed to conclusion so far, our death rate is 7.6%. That is not good at all but admittedly it is a small sample so far.

  23. @Smith9

    It’s not 13/3700 it is 13/n where n= original infection. If it’s a 2 week infection then the number 2 weeks ago was ~249 or 5.2%.

    But this is fairly meaningless if you don’t do mass screening to pick up those who are asymptomatic.

    The range of symptoms ie from nothing much to total organ failure and death could be down to viral load ie, if the infected person only receives a very low dose it’s possible that their immune system would be sufficient to overwhelm the infection. Conversely a high dose would be beyond the immune system.

    It’s all conjecture.

  24. Smith9, your comment is based on faulty reasoning as you have not taken into account the lag time between diagnosis and death.

  25. “ you have not taken into account the lag time between diagnosis and death.”

    I am aware of this. But on a comparison with other countries at the same stage we still look better.

    Might all change for the worse, of course.

  26. Iko, on your reasoning, your calculation is a bit off.

    You did 13/170 = 7.6%. You should have done 13/(170+13) = 7.1%.

  27. KT2, fwiw:

    Walsh: “183 is the answer if the question was how many do you need before someone is fifty per cent likely to share your birthday?”

    var n = 1;
    while (Math.pow(364 / 365, n) > .5) {

    n = 253

  28. As Nick says ,Tony Blakely’s full argument is at his blog , and it is a much more interesting argument than the bastardised version reported by the abc. Tony is a very good epidemiologist who has done excellent work on the negative impact of inequality on health, mostly in NZ. His argument uses utilitarian logic to compare the flatten the curve approach to the go early go hard and eradicate approach (which NZ is now following). He argues that the flatten the curve approach could lead to 60% of the population being infected which would lead to an unacceptable number of deaths. He then introduces a modified flatten the curve scenario where only those of us over 60 go into home quarantine, (but quite probably for an extended period) and compares that to the go early go hard and eradicate approach where almost all of the population are in home quarantine, but hopefully for not too long a period.
    Its an interesting set of calculations.
    But Tony’s blog was written 5 days ago, so is now out of date. I think that an alternative scenario has emerged for Australia. The corona virus case curve has flattened in the last 4 days, presumably because of border closures, and the introduction of social distancing starting two weeks ago. So with continuing of strict social distancing and border closure and lots more testing, it may be feasible to get to eradication without the go early go hard approach that NZ has adopted. The data over the next few days will indicate whether such an approach is feasible.

  29. @Harry Clarke:

    “There are no option value advantages to waiting – the economics/health tradeoff seems to me illusory – minimizing deaths in a myopic way seems a good policy both from the viewpoint of health, of costs and in terms of minimizing long-term economic disruption.”

    This argument is supported by new research into the impacts of the 1918 flu pandemic in the United States:

    “Comparing cities by the speed and aggressiveness of NPIs [non-pharmaceutical interventions], we find that early and forceful NPIs do not worsen the economic downturn. On the contrary, cities that intervened earlier and more aggressively experience a relative increase in manufacturing employment, manufacturing output, and bank assets in 1919, after the end of the pandemic.

    “The effects are economically sizable. Reacting 10 days earlier to the arrival of the pandemic in a given city increases manufacturing employment by around 5% in the post period. Likewise, implementing NPIs for an additional 50 days increases manufacturing employment by 6.5% after the pandemic.

    “…Anecdotal evidence suggests that our results have parallels in the COVID-19 outbreak. Countries that implemented early NPIs such as Taiwan and Singapore have not only limited infection growth. They also appear to have mitigated the worst economic disruption caused by the pandemic. Well-calibrated early and forceful NPIs should therefore not be seen as having major economic costs in a pandemic.”

  30. I have myself been an advocate of an isolation policy from very early on in this pandemic. John Goss above calls it the “go early, go hard and eradicate approach” and notes N.Z. is following it. The world has become one big experimental laboratory and we may see which approach works best. The paper quoted by Luke Elford supports this approach by examining the results of the 1919 flu pandemic.

    However, there is one fly in the ointment. The world is far more connected now, particularly by air travel, than it was in 1919. Also, the automobile is much faster and has much greater range and endurance than the horse. This means local, regional and national suppression or eradication are far more difficult. Modern modes of travel could rapidly re-introduce the virus back into eradication areas when isolation controls are lifted. The fact that COVID-19 is a pandemic, uncontrolled in many countries at once, makes an eradication policy in one country also an extended isolation policy. As soon as isolation suppression is lifted (e.g. New Zealand lets in flights from the rest of the world again), the epidemic presumably starts up again in that country without herd immunity.

    I am not sure of all the implications of this. I am still trying to think them through. It might mean that letting the virus generate herd immunity at a “flattened curve” rate, so long as the curve is kept flattened significantly below the ICU bed capacity, is the best policy. In practice, the combination of the high infectiousness of the virus and the high connectedness of the world mean that our very best efforts at contact tracing and distancing / isolation seem at best to be able to achieve only a flattened curve anyway. Eradication, or rather “eradication in one country”, before herd immunity might not be a possibility unless that country intends isolation indefinitely or at least until a vaccine, which might never eventuate. Autarky anyone?

    I am actually an advocate of less connectedness in the world, though not an advocate of complete isolation and autarky. It is possible for the world to be too connected and to have too much people movement. I think this is the case with the world now. The substantial removal of almost all effective boundaries and borders by open, rapid, mass travel renders the world system a kind of social, cultural, and epidemiologically fragile mono-culture, susceptible (in this and other cases) to the rapid sweep of a pathogen right around the world. Petit bourgeois tourism itself is a kind of plague and it carries “plague”, especially in the form of new zoonotic diseases, around the world. This is not an argument for a haute bourgeoisie to exist and to have special rights over everyone else.

    Zoonotic diseases are a rapidly burgeoning danger we face due to our industrialized food system, our global over-connectedness and our continuous encroachment into the biosphere’s dwindling wild areas. Again, this is another form of the limits to growth reality this finite biosphere imposes on us. We need to heed all these warning signs of wild and human ecology just as much as we need to heed rising CO2 levels and dwindling glaciers. We must stop growing population, built infrastructure and cleared land areas indefinitely. We must stop expanding travel and the many indulgences of consumerism indefinitely. Humanity needs a new ecological and social ethic. One better than mere endless self-indulgence and consumerism. We need to stop the sixth mass extinction or very soon it will be our turn for extinction.

  31. akarog,

    Agreed, scientific analysis is dependent on data. However, scientific hypotheses for testing are dependent on deductive, inductive and speculative thinking involving both pure reason and imagination. The role and importance of speculative thinking is very significant provided it remains disciplined in the service of science. It may be capricious for the purposes of fiction and fun.

  32. Ikonoclast
    If you go for an eradicate strategy, in whichever form you do it, then you can’t fully open up until the vaccine is available (which is 12 to 18 months away). But once you have the numbers under control you can allow limited entry (and exit) by requiring an antibody blood test (or perhaps 2 tests seeing it is less reliable than the virus test), before you let them loose in the community (with monitoring of their health). The antibody blood test gives results in 15 minutes, so it is I think a gamechanger.

  33. 60 cases in Canberra by Saturday 28 March.
    Given what we know about the doubling time and the potential to pass it on before showing symptoms, that could easily represent a thousand people walking around town who caught it yesterday and don’t yet know they’ve got it.

  34. > analysis is dependant on data and quality data is difficult to obtain.

    I read a wonderful article the other day from a highly trained medical academic saying that we risked making bad decisions from lack of data, and implying that we should BAU until we had better data. Evidently just being a professor of medicine apparently isn’t enough to connect someone to observable reality.

    A bunch of links just appeared on another blog suggesting that “most over 60” is true, but not as dramatically true as some people are hoping:


    Canada’s chief medical officer says almost 1/3 of Covid-19 hospitial admissions are under 40

    This story from 9 days ago for the US says 38% of hospitalizations are 20-54 years old

    This undated articles say 50% of France’s ICU patients are under 60

  35. @Luke Elford

    There is current evidence on the economic case for going hard. Taiwan is a non-country in the eyes of the WHO. But it has extraordinary achievements in controlling CV-19: As of last week, it had only 250 infections and 2 deaths despite being highly exposed to mainland China. Taiwan acted early and thoroughly on the basis of good data. It Insisted on masks for everyone. Taiwan learned a lot from SARS.

    We can learn from Taiwan. Arrivals from overseas were subject to 14-day quarantine and taken to their place of quarantine by a special fleet of taxis. Their adherence to quarantine was checked by their mobile phone every day. The fine for not adhering was $33,000.

    Schools remained open and, in many respects, life continued as usual.

  36. Taiwan’s Vice President is an epidemiologist. In a crisis, it’s good to have smart people who know what they are doing, not political hacks, at the highest level of government.

  37. The most common former professions of federal parliamentarians are lawyer and business person. These are clearly the worst people to have in charge. For lawyers “truth” is determined by rhetoric. For business people “truth ” is determined by money. Many in these professions do not fully understand empirical or scientific truth or if they do they pay it no heed, ranking it lower than self-interested power, wealth and gain.

  38. I’ve put an argument that “Command economies work best” under the “Crisis and the case for Socialism” topic. Of course my statement comes with caveats. I think I make a good case if I do say so myself.

  39. Nick says: at 3:44 pm
    KT2, fwiw:

    I was astounded DWalsh was corrected. That is why I wrote “nuanced comments”, someone as you have, corrected him.

    Trivia. Mentioned in post. the biggest gambler on the planet is in Tas. One of David’s original gambling partners…

    “Ranogajec reportedly accounts for 6–8% of Australian bookmaker TabCorp’s $10 billion annual revenue.[9] His betting on Betfair is believed to account for one third of the company’s Australian operations.[10]”

    I wonder what he has placed a bet on recently?

  40. The tail of the herd is still a risk.

    Why do dozens of diseases wax and wane with the seasons—and will COVID-19?

    “So no one knows whether SARS-CoV-2 will change its behavior come spring. “I would caution over-interpreting that hypothesis,” Nancy Messonnier, the point person for COVID-19 at the U.S. Centers for Disease Control and Prevention, said at a press conference on 12 February. If the seasons do affect SARS-CoV-2, it also could defy that pattern in this first year and keep spreading, because humanity has not had a chance to build immunity to it.

    Even for well-known seasonal diseases, it’s not clear why they wax and wane during the calendar year. “It’s an absolute swine of a field,”

  41. Harry Clarke
    Did Taiwan shut down all their shops (except food and drug stores) as Daniel Andrews wants to do? No. Controlling the borders, lots of testing, masks and good hygiene, good contact tracing, good primary health care services, strict social distancing but not extreme social distancing, are the lessons from Taiwan.

  42. John, I think Taiwan acted so promptly (early January) that almost all the possible transmission was caught by focusing on inbound travelers. They were obsessive about this. If Morrison is right and most of the sources of infection are “still” via returned travelers then the stricter measures we have just adopted should help. But I am skeptical about this. We left it too late – Morrison this morning presented aggregate data on levels of infection that can be attributed to inbound travel but a lot of this occurred in the past – what we really want to know are the incremental sources of infection now. My assumption is that there are potentially significant levels of within-community transmission so that Daniel Andrews may have it right. But, yes, I don’t know. I’d go really hard on imported cases – as we now seem to be doing – and pursue the strong shutdown policies of Andrews just in case within-community transmission is now taking over from imported transmission as the major source of infection. We will know in a week nor so what is the right answer here but that will be too late. Important to act now even though we missed the boat on imports.

  43. On Friday night a bunch of travellers flew in to Sydney from a cruise ship and transferred to domestic flights without quarantine. They were not stopped from travel and have relocated. They were doctors.

  44. Yesterday, Scott Morrison said Australia’s testing rates were the “best in the world”. That is a LIE. Australia’s efforts as at 28/03/20 were 13th in the world per million people. Italy’s efforts were two and one half times better than our efforts to that date. So, another BIG FAT LIE from our Prime Mendaciousness.

    As John Ralston Saul said, “Politicians lie every day whether they need to or not, just to keep in practice.”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s