The long-term CovidSim predictions from Report 9

A group of us have just had a paper published in The British Medical Journal on the effect of school closures on mortality from the coronavirus disease. The coverage has been rather unfortunate, as it is being interpreted as supporting a herd immunity strategy, which is certainly not what (I think) we were trying to argue. I thought I would write this post in the hope of clarifying what we actually did.

The work was motivated by trying to explain what seemed to be some counter-intuitive results presented by the Imperial College group in mid-March, in a document often referred to as Report 9. Specifically, there are some scenarios presented in this report, where adding an intervention leads to more deaths than a similar scenario without that additional intervention. For example, if you look at Table A1, the model predicts that adding place closures (PC) to case isolation (CI), household quarantine (HQ) and social distancing of those over 70 (SDOL70), would ultimately lead to more deaths than if place closures had not been implemented. A similar effect occurs if you add general social distancing (SD) to a scenario with CI and HQ.

ICU bed demand, against time, for the scenarios presented by the Imperial College group in mid-March in Report 9.

The reason for this counter-intuitive result is illustrated by the Figure on the right, which shows ICU bed demand for the scenarios presented in Report 9. Some of these produce a single wave of infections. However, in some cases, adding a new intervention, substantially impacts the first wave, but means that once the interventions are lifted you can get a second wave, which – if the most vulnerable are not suitably protected – could produce more deaths overall, than the equivalent scenario without this additional intervention.

So, does this mean we should not have added some of the interventions? Well, for starters, these are model projections none of which specifically match what we actually did. Also, as James Annan has pointed out quite forcefully on Twitter, some of the model parameters used in Report 9 were clearly not correct (i.e., the basic reproduction number, R0, was lower than we now know to be the case). We were mostly trying to understand why some of these results presented were counter-intiutive, than make any kind of specific prediction, or update what was presented in Report 9 in mid-March. The result may well be different if we were to redo this using updated parameters [Edit: I should have been clearer here. I mean the results presented in Report 9 might have been different, not that our results would have been different].

Additionally, in all of the scenarios presented in Report 9 where there was a single-wave leading to herd immunity, the ICU bed demand and the total number of deaths far exceed what we actually experienced. If we had followed such a scenario, it would almost certainly have over-whelmed the healthcare system and would have almost certainly been perceived as far too extreme. Hence, I don’t think that our paper specifically supports an argument against the lockdown (although people can, of course, make their own interpretations).

Does this mean that we should now follow some kind of herd immunity strategy? Again, these are model results, so one should bear that in mind when drawing interpretations. We did finish the paper by doing some comparisons with actual data, and the model does do well if you update the parameters (i.e., higher R0 value and the epidemic starting sooner than suggested by Report 9). However, there are lots of things that the model doesn’t include. It doesn’t include the long-term impact on those who get infected and don’t die. It does suggest that limiting deaths would require properly shielding the vulnerable, but it doesn’t tell us if this is actually possible. It doesn’t tell us if we will actually develop immunity. There are many caveats that, I think, should be considered before drawing strong conclusions.

At the end of the day, what we were really trying to do was better understand the results presented in mid-March, which I think we’ve now done. There may well be implications to this, but I do think one should be cautious of drawing strong conclusions from a single study that was more motivated by trying to clarify what’s already been presented than make any specific predictions.

This entry was posted in Research, Scientists, The scientific method and tagged , , , , , . Bookmark the permalink.

205 Responses to The long-term CovidSim predictions from Report 9

  1. Remember to always use your turn signal when changing lanes. 😉

  2. gwws says:

    There is a different effect of shielding the vulnerable which is counterintuitive. This effect is due to the matrix structure of the reproduction number. There are four transmission numbers from fit to fit from fit to vulnerable and from vulnerable to fit and from vulnerable to vulnerable. I can send more details if you are interested

  3. Can you explain why you say this “If we had followed such a scenario, it would almost certainly have over-whelmed the healthcare system and would have almost certainly been perceived as far too extreme.”?

    After all, the healthcare system was never ovewhelmed in Sweden. It was never overwhelmed in any US states that used light touches. Florida locked down hard at first, but has now basically let the virus run its course. Even with a very elderly population, Florida never had any hospital bed shortages, neither ICU nor general. Hospitals were only overwhlemed (and this is using the term very broadly) in a few towns in Italy and in the very poorest borough of NYC where the hospital system is objectively terrible. What’s the reason for continuing to think this?

  4. Eric,
    Fair enough, I should probably have been more careful. What I meant was that if you think the models are reasonable representations of what would have happened under those scenarios, then a scenario in which there is a single wave leading to herd immunity in a few months (but that may reduce the impact by shielding those who are particularly vulnerable) would probably still have been much more severe that what we experienced, both in terms of deaths and ICU demand.

    What I was trying to distinguish between was scenarios where there is likely to be more than one wave, and scenarios where there is essentially a single wave. The latter would (I think) almost always be pretty severe, given that it’s being dealt with in a single wave (I may be wrong, but I don’t think I’ve seen a scenario where there is a single wave that doesn’t produce high ICU bed demand and a lot of deaths – in the UK, at least).

  5. gwws,
    I’m sure there are other potential counter-intuitive results.

  6. Willard says:

  7. gwws says:

    “I’m sure there are other potential counter-intuitive results.”

    I’m interested. Can could send me more details.

  8. gwws,
    Sorry, I was meaning that I wouldn’t be surprised if there are ways of shielding the vulnerable that also lead to somewhat counter-intuitive results. I wasn’t meaning that I have other examples of such counter-intuitive results.

  9. verytallguy says:

    I’m not sure that claiming that in Florida the virus has “run its course” is accurate.

    Unless continuing to kill 100 people every day with no end in sight is what having run its course means.

  10. verytallguy,

    are you suggesting that Florida will face an overwelmed hospital system in the future? If so, care to drop the anonymity and bet? Probably not.

    What you posting is the death by *date reported*. Deaths by date of occurrence peaked between July 25th and August 4th and have been steadily declining since then. Hospitalizations have been been steadily declining for over 6 weeks and are at about 1/6th the level they were at six weeks ago.

  11. verytallguy says:


    I suggest what I wrote, nothing more, nothing less.

  12. unfortunately what you wrote wasn’t accurate.

  13. verytallguy says:

    Neither was what you wrote, so I guess that makes us even

  14. I pointed out what you wrote that was inaccurate. Florida hasn’t had 100 deaths/day since August. I still stand by what I said, and you’ve only suggested it was inaccurate by saying something false.

  15. Willard says:

    FWIW, the claim that Florida “now basically let the virus run its course” is about how the state is dealing with the virus, not a prediction on a future trajectory.

    Perhaps a question for Eric or anyone who followed the lichurchur:

    Increasing COVID-19 caseloads were associated with countries with higher obesity (adjusted rate ratio [RR]=1.06; 95%CI: 1.01–1.11), median population age (RR=1.10; 95%CI: 1.05–1.15) and longer time to border closures from the first reported case (RR=1.04; 95%CI: 1.01–1.08). Increased mortality per million was significantly associated with higher obesity prevalence (RR=1.12; 95%CI: 1.06–1.19) and per capita gross domestic product (GDP) (RR=1.03; 95%CI: 1.00–1.06). Reduced income dispersion reduced mortality (RR=0.88; 95%CI: 0.83–0.93) and the number of critical cases (RR=0.92; 95% CI: 0.87–0.97). Rapid border closures, full lockdowns, and wide-spread testing were not associated with COVID-19 mortality per million people. However, full lockdowns (RR=2.47: 95%CI: 1.08–5.64) and reduced country vulnerability to biological threats (i.e. high scores on the global health security scale for risk environment) (RR=1.55; 95%CI: 1.13–2.12) were significantly associated with increased patient recovery rates.

    Do these conclusions still hold?

  16. JCH says:

    Learn from cows or pigs. Viruses tend to flag in the summer, especially the hottest party of the summer, which in the Northern Hemisphere is mid July to the end of August. The virus weakened. COVID-19 cases started dropping around mid July, especially in USA southern states. Because this virus clearly survived the summer heat blast, it is now invigorating.

    This virus has a lipid layer. Place a pad of butter on the kitchen counter. What happens?

    Drug trials done on summer infected. They’re likely suspect.

    So now the northern USA is starting to burn with new infections. They could be more lethal. Still very warm in Texas and Florida, They will soon follow the northern trend. We’re heading two to three thousand deaths per day.

    Spent my childhood stacking up dead piglets killed by viruses. Do I think there is a pig farm in Sweden that is somehow different? No.

  17. Entropic man says:

    Eric Winsberg

    “Florida hasn’t had 100 deaths/day since August. ”

    That turns out not to be the case.

    Florida had 123 deaths yesterday and 158 today.

  18. Entropic man: read is fundamental, bro. Those are not deaths that happened yesterday and today. They were *reported yesterday and today*. Of the 158 deaths reported today here’s how they were distributed:

    July (+47-1), August (+73), September (+33-2), October(+15-1)

    There has not been a 100 death day since August. Don’t correct things if you don’t know what you are talking about.

  19. Steven Mosher says:

    Err couple points

    1. Florida and sweden have nothing to do with the argument. Zero. ZIP, nada.
    The argument is about the UK and will obviously turn on the ICU capacity there.
    in short, FL and SWE can teach you nothing about agent based modeling in the UK.
    2. For all things Florida please see my buddy. Probably the best analyst of FL data.
    I passed him the line data a while back and he went to town.


    nice paper Dr. R
    I’m not shocked that people mis use it

  20. Steven Mosher says:

    Dr. R
    it might be instructive to explain to people what the mixing matrix captures.

    Short version. We dont know the mixing matrix for FL or SWE and that matters.

  21. Jeffh says:

    Why is this herd immunity nonsense still circulating? By now there is accumulating evidence with Covid-19 that immunity may be transient or else high viral loads are necessary to achieve it. Moreover, a very significant number of those who recover suffer significant physiological legacies including cardio-vascular damage, kidney damage, partial infertility, long-lasting fatigue, respiratory problems etc. And not just the weak or elderly as a very recent study in the Netherlands reported. The ‘let it rip’ crowd rarely factor this in.

    Containment until there is a vaccine is the proper way to go. Here in the Netherlands the second wave is dwarfing the first and even the first one pushed hospitals to the edge. We never had a full containment strategy here anyway, we were ‘lockdown lite’, not appreciably different from Sweden. They most certainly are not out of the woods and are probably not anywhere close to achieving a herd immunity infection threshold. The problem in the Netherlands is that our government is a far right neoliberal administration that constantly promotes the idea of personal freedom and individualism. The characteristics clash with societal altruism. It is hardly surprising to me that the virus is so bad here. In Sweden the prevailing mentality is very different. Look at Norway – they had a lockdown and the second wave there is lower than in Sweden. The key lies in personal responsibility and the type of government that is in charge.

  22. Steven,
    Yes, this is all very dependent on how people interact, the makeup of households, people’s behaviours, so one should be careful about drawing strong conclusions about a particular region based on what is happening somewhere else.

  23. Joshua says:

    Eric –

    How much, do you think that improvements in therapeutics and standard care contribute to lowering the IFR now as compared to early on in the pandemic?

    Do you think there is much likelihood an effective vaccine being distributed in a relatively near-term time frame? If so, how might that possibility inform your thinking about a “herd immunity” strategy?

  24. Joshua says:

    Eric –


    > Hospitals were only overwhlemed (and this is using the term very broadly) in a few towns in Italy and in the very poorest borough of NYC where the hospital system is objectively terrible.

    What about those communities in Texas? Also, what ctiteria/metrics do you use to measure “objectively terrible,” and how many other hospitals in the US (serving what size of population) meet that criteria?

  25. Joshua says:

    Eric –

    If I might – another question.

    I heard a podcast with you back in the day where you discussed the impact of school closures. IIRC (and my memory isn’t as good at Trump’s) you and the other discussants said it appeared that closing schools didn’t affect the course of a pandemic after the first few weeks, and that there aren’t much data where school closures came along with the other SIPs and NPIs that we’ve seen with Covid, and that you found the *no effect* finding to be counterintuitive.

    I’m wondering if after these months have passed you have moved in your views on this topic, and if so, how. Of course, Covid might be a special case, given the age-dependency of outcomes.

  26. Joshua,

    Lots of questions. I’ll do my best: The US has 2.6 hospital beds per 1000 people. That’s already pretty low if you compare us to other OECD countries. (Germany has 4.3) Manhattan has 6.5 beds. Queens has 1.5 beds per 1000 people. That’s abysmal. Also, google around for the various investigative reports on NYC’s care homes. You’ll find stories of people lying in the same soiled diapers for days, catheters being picked up off the dirty floor and used, feces smeared on the wall, etc. I’ll stand by the claim that that’s objectively horrible. I don’t think Texas hospitals were ever severely taxed. There were breathless stories in June about how they were *going* to be overtaxed, but I don’t know of any cases of people being denied care.

    Why was the death rate so high early on? We don’t know, exactly. Therapeutics have improved somewhat, but I have a hard time attributing the huge difference to that, since the new therapeutics are not that terrific. I *suspect* that the early death rates were caused *in part* by fear of hospitals being overrun. If you look, you can find stories from Belgium, Spain, New York city, etc, of old people being taken out of hospitals and sent back to care homes *in anticipation* of hospital overloads that never materialized. People can have it at bashing this claim and I’m not going to argue about it because its just my intuition. But you asked, so I answered.

  27. As for schools, I think the evidence continues to grow that schools are not a major source of spread.

  28. I don’t have any relevant knowledge for predicting when a vaccine will arrive better than anyone else. But I do recall many people promising us one “by october”. And I think the two candidates that looked most attractive early one have seen setbacks. (the oxford one and the moderna one.) We also need to remember that vaccines for respiratory infections are historically not that effective.

  29. Eric,
    My impression is that despite the places where it seems the impact wasn’t as severe as might be expected, most public health experts, and epidemiologists, still regard it as something that requires interventions, potentially quite strong ones (at least). What’s you general view of that? Are they over-reacting, or are these places where the impact wasn’t all that severe more an exception than the rule?

  30. My take is that way back in March, when we had no idea what we were looking at, some targeted interventions were clearly justified. At this point, I think we have painted ourselves into a corner and nobody knows what the end game is. I no longer understand what the *long-term* purpose of these interventions are. the costs of some of these interventions, meanwhile, are staggering. The achievement gap we are going to see for children from underprivileged backgrounds is going to be horrific. Suicides and drug addictions are already skyrocketing. And the costs born by developing nations of the economic losses we are creating are going to be very bad.

    I also think expert opinion on this is much more divided that the press has been letting on. Lots of very smart and well informed people have been alerting us to the fact that these tradeoffs are not good ones.

  31. As with climate science, this paper is being discussed through the disparate lenses of commenters formed by pre-existing political conditions. Call the comorbidities.

    As with climate science, report writing could have been improved by anticipating the visceral reactions of some (many?) readers.

  32. Joshua says:

    Eric –


    > Queens has 1.5 beds per 1000 people. That’s abysmal.

    While number of beds is obviously an important metric for evaluating whether hospitals might get overrun, it isn’t nearly the only relevant metric. And if we’re evaluating whether there was likely to be other hospitals that were as overrun as those were in Queens, then you have to evaluate what other regions have similar ratios of beds to population. I see no particular reason to think, a priori and without investigation, that the hospitals in Queens were significantly different than public hospitals in any other densely populated, low income neighborhoods throughout the country – which would have implications for the care of 10’s of millions of people in dealing with the pandemic.

    > Also, google around for the various investigative reports on NYC’s care homes. You’ll find stories of people lying in the same soiled diapers for days, catheters being picked up off the dirty floor and used, feces smeared on the wall, etc. I’ll stand by the claim that that’s objectively horrible.

    As bad as the conditions might be in NYC’s care homes, I see no particular reason to think that the conditions there are unusual for care homes in a similar community. And I’m not getting how you connect that to the larger issue at hand?’ How does the level of care in care homes connect to the question of hospitals being overwhelmed.

    > I don’t think Texas hospitals were ever severely taxed. There were breathless stories in June about how they were *going* to be overtaxed, but I don’t know of any cases of people being denied care.

    I don’t doubt that there was hype – but again I question the basis on which you have made some authoritative statements. I think that in South Texas, there is a lot of evidence of what you could reasonably consider to be “overwhelmed” hospitals – with some measure of impact on the quality of care.

    > Why was the death rate so high early on? We don’t know, exactly. Therapeutics have improved somewhat, but I have a hard time attributing the huge difference to that, since the new therapeutics are not that terrific. I *suspect* that the early death rates were caused *in part* by fear of hospitals being overrun. If you look, you can find stories from Belgium, Spain, New York city, etc, of old people being taken out of hospitals and sent back to care homes *in anticipation* of hospital overloads that never materialized. People can have it at bashing this claim and I’m not going to argue about it because its just my intuition. But you asked, so I answered.

    OK. that seems better to me. I assumed that as a modeler who deals with these issues you were speaking from more than intuition. My intuitions are rather different from yours and I think I can point to less intuitive information that suggests that your intuition has led you astray. That said, sure, “in part” fear of hospitals being overrun could certainly have been a contributing factor. I think the important question is to try to evaluate which factors are most explanatory. Not asking you to argue about it

    > I don’t have any relevant knowledge for predicting when a vaccine will arrive better than anyone else. But I do recall many people promising us one “by october”. And I think the two candidates that looked most attractive early one have seen setbacks. (the oxford one and the moderna one.) We also need to remember that vaccines for respiratory infections are historically not that effective.

    I wasn’t asking for a prediction, or even speculation about a time frame. The reason why I’m asking the question is because it seems to me that the timetable for a potential vaccine has huge implications for the advisability of a “herd immunity” approach vs. a SIP/NPI approach. IOW, if a vaccine is developed and distributed then it would would reduce the # of infections needed to reach a “herd immunity threshold.” Even without one, a slower spread allows time for the development of and improvement in therapeutics and standard of care (or if you will, interventions to improve nursing home standards) for those who are infected, which can greatly improve outcomes per # of person infected. But with a vaccine, gambling on a faster spread in hopes of there being less economic damage (which is an unproved theory) could mean many unnecessary deaths and much unnecessary disease. As an armchair epidemiologist, I was hoping for your input on that issue as someone who models these phenomena.

  33. “Do you think there is much likelihood an effective vaccine being distributed in a relatively near-term time frame?”

    “I wasn’t asking for a prediction, or even speculation about a time frame. ”

    Sorry, my mistake. I thought we were having a good faith conversation.

  34. Joshua says:

    Eric –
    > the costs of some of these interventions, meanwhile, are staggering.

    This is another unproven assertion. There were all kinds of changes that took place before interventions were put into place and it is very hard to pin down the counterfactual of what would have happened and they not been put into place. For example:

    > The achievement gap we are going to see for children from underprivileged backgrounds is going to be horrific. Suicides and drug addictions are already skyrocketing. And the costs born by developing nations of the economic losses we are creating are going to be very bad.

    This certainly seems true. The impact of the pandemic are wide and deep. However, the differential calculation as to which impacts are, to what degree, the results of the interventions as opposed to a raging pandemic in and of itself is a very complicated task. And extending that into a counterfacual assertion about what would have been better had interventions not been put into place only increases the complications exponentially.

    > I also think expert opinion on this is much more divided that the press has been letting on. Lots of very smart and well informed people have been alerting us to the fact that these tradeoffs are not good ones.

    If you can link to some evidence that really deals with these issues comprehensively, I’d appreciate it. I have seen a lot of really smart and well-informed people who have provided some very poor quality analyses on these issues – in particular analyses that draw conclusions about the relative benefit of a “herd immunity” approach without engaging with obvious counterarguments.

  35. Joshua says:

    Related to school closing:


    FoHM’s decision to keep schools open despite surging cases may also have added to the spread. A report from the agency itself, released in July, compared Sweden with Finland, which closed its schools between March and May, and concluded that “closing of schools had no measurable effect on the number of cases of COVID-19 in children.” But few Swedish children were tested in that period, even if they had COVID-19 symptoms. And the lack of contact tracing means there are no data about whether cases spread in schools or not. When new FoHM guidelines allowed symptomatic children to be tested in June, cases in children shot up—from fewer than 20 per week in late May to more than 100 in the second week of June. (FoHM reversed course in July and returned to recommending that children under 16 not be tested.

    Indirect data suggest children in Sweden were infected far more often than their Finnish counterparts. The FoHM report says 14 Swedish kids were admitted to intensive care with COVID-19, versus one in Finland, which has roughly half as many schoolchildren. In Sweden, at least 70 children have been diagnosed with multisystem inflammatory syndrome, a rare complication of COVID-19, versus fewer than five in Finland.


  36. Joshua says:

    eric –

    > “Do you think there is much likelihood an effective vaccine being distributed in a relatively near-term time frame?”

    >> Sorry, my mistake. I thought we were having a good faith conversation.

    My bad. My point was to get the sense of a yes or no, and then if yes, what you think about the impact of a vaccine on a relatively short-term timeframe. My interest wasn’t in particular related to your views about what the timeframe might look like.

  37. Joshua says:

    eric –

    IOW, my focus was more on the next sentence:

    > If so, how might that possibility inform your thinking about a “herd immunity” strategy?

  38. Willard says:

    > my focus was more on the next sentence

    If every response R from Eric attracts R x N questions, I doubt we can talk about focus.

  39. Joshua says:

    Willard –

    > If every response R from Eric attracts R x N questions

    *every*. If I understand you correctly, I asked Eric some questions and my following 5 responses contained a total of zero questions.

  40. Joshua says:

    Oops. Missed this one:

    > And I’m not getting how you connect that to the larger issue at hand?’

  41. Joshua says:

    Eric –

    Regarding your impression that I’m engaging in bad faith. I would ask you to consider that I have referenced your pandemic modeling here:

    and at Andrew Gelman’s blog:

    And even defended your related expertise here at Andrew’s:


    My questions for you were good faith attempts to ascertain your perspective – as someone who has spent time modeling this and other related issues.

  42. Willard says:

    > If I understand you correctly, I asked

    What you asked is independent from my own understanding of NP problems, Joshua. My “every” was emphatic. There’s no need for you to quote *every single* word from Eric’s response to dis-aggregate points in a way that brings the flow of the exchange to a standstill. All you need is to ask more than one question for every response you get.

    There’s a fine line between sandbagging and asking for clarification.

  43. Joshua says:

    Willard –

    I don’t understand what “NP problems” are.

    > There’s a fine line between sandbagging and asking for clarification.

    Perhaps so, although I don’t know really how you’re applying “sandbagging” here. But it is clear that Eric determined that I was operating in bad faith and it seems likely that asking him focused questions might have set that up.

    Nonetheless, I was asking for information and clarification. I assumed that as someone with experience modeling this stuff, he had an evidence basis for the assertions he was making and I wanted to know what it was – precisely because some of his assertions ran counter to my opinions and some of the evidence I had seen.

    Although I’d cetainly prefer thst he not conclude that I was engaging in bad faith, as it turns out, it appears he was largely going on intuition – and as such in stating that he provided the clarification I was asking for and my questions proved useful in the end. At least for me, if not for you.

  44. “There’s a fine line between sandbagging and asking for clarification.”

    There is also a very broad line for those who are here to argue for the sole purpose of being argumentative. :/

  45. Jeffh says:

    There is nothing illusory about hospitals being ‘overrun’ in the Netherlands. Right now the equivalence of every ward in 2-3 typical Dutch hospitals are filled with Covid-19 patients, 1239 in total and 230 in ICU. And the worst is yet to come as the number of infections here is rising exponentially. As of today on a rolling one-week average the rate of infections is the highest in the world. Within two weeks hospitals here will be bursting at the seams. Thanks to Germany, where containment is still policy, they have agreed to take our overload into some of their hospitals. The death rate lags 2-4 weeks behind infection rate so despite being better prepared a lot of people are going to die.

    Certainly it is important to factor in the social costs of intervention versus non-intervention. But as I said before, and this has been completely ignored in discussions here, the infection fatality rate is not the only nasty effect of Covid-19. A study at the University of Amsterdam last week evaluated the health of 1001 people (median age 48) who were infected by the virus but recovered. Over 90% were still experiencing debilitating systems several months after infection and 48% were experiencing multiple symptoms. These include severe breathing difficulties, damage to the heart, kidneys and other organs, lethargy and fatigue, muscle spasms, repeated headaches and more. Almost 90% of the individuals had no pre-existing medical conditions. A colleague knows a doctor who is a lung expert in a local hospital. He told her that since March he has had to treat a huge number of Covid-19 patients who have almost forgotten how the breathe properly. The virus has somehow desynchronized the automatic breathing response of some people. What this all suggests is that this is a dangerous, novel pathogen and that until we know a lot more simply ‘riding it out’ is not a sensible option.

    What are the societal costs of these coronavirus legacies on survivors? Should they not be factored in?

  46. verytallguy says:

    “I don’t have any relevant knowledge for predicting when a vaccine will arrive better than anyone else. But I do recall many people promising us one “by october”. And I think the two candidates that looked most attractive early one have seen setbacks. (the oxford one and the moderna one.) ”

    Anyone promising a vaccine by October was either ignorant, or deliberately misleading.

    But the Oxford vaccine, at least, has a good chance of efficacy data being available by the end of the month. Indeed, I would stake a small wager on it. I would not, however, bet money on what the results are, though I would say it’s probably more likely than not that they are good enough to justify a rollout.

    The moderna one is more of a long shot, the concept never having been taken to commercialisation before.

  47. verytallguy says:

    Which is a lead in to the question begged by the discourse above: “What should our strategy be?”

    In March, with so much unknown and a doubling time of 3 days, there was no viable strategy other than hard suppression.

    Now, we see disparate approaches, often within countries, we’ve learned a lot about the virus (though disappointingly little about effective treatments), and, as above, have a reasonable but not certain prospect of a vaccine within months.

    Generally, it seems the countries with the least cases have also had the least economic impact, and a combination of stringent social distancing and test, track, trace have achieved that.

    Countries with a more laissez faire attitude continue to kill large numbers – Brazil and US are the poster children.

    So what now? Seems to me a no- brainer really. Suppress the virus, beef up test, track trace and survive the winter. Develop a plan B for coping with a long surge and if a vaccine isn’t here by spring, gradually open up society, and our morgues with it.

  48. Willard says:


    NP is a concept from complexity theory. The idea I’m trying to convey is that exchanges are constrained by time and space. If we want to move them forward, issues that compel us to move sideways need be left unresolved. And each step of an exchange introduces side issues.

    Your question to Eric about the likelihood “of an effective vaccine being distributed in a relatively near-term” can indeed be interpreted as asking for his 8-ball call. If you then reject it as irrelevant, say because it’s not what you ask, then diminishing returns are to be expected. Eric’s comments stand on their own, he fulfilled all his commitments. His job is done.

    Blog comment sections are not parlors. If there are questions that deserve to be raised, you can raise them without addressing them to anyone in particular. Even better, you can offer what you think is the appropriate response right from the start. A show of hands can then occur without having rounds of raises and calls. You say what you think is the case and why, otters do the same, nothing gets settled, everyone return home happy.

    Sandbagging can be felt when the question being asked is followed by a series of counters that could have been laid out beforehand. That the person from whom you expect work feels sandbagged ought to be enough as a cue. The concept is quite clear enough for what I’m trying to express.

  49. “Anyone promising a vaccine by October was either ignorant, or deliberately misleading.”

    People have such short memories.

    EVERYBODY was promising a vaccine by this fall

    “The Company has concluded the first agreements for at least 400 million doses and has secured total manufacturing capacity for one billion doses so far and will begin first deliveries in September 2020. AstraZeneca aims to conclude further agreements supported by several parallel supply chains, which will expand capacity further over the next months to ensure the delivery of a globally accessible vaccine. ”

    “Those are the results from Oxford University’s Jenner Institute the world has been waiting for regarding a potential COVID-19 vaccine.

    The most optimistic timeline is that the vaccine could be available on an emergency basis for high-risk groups as early as October.”

    those are the big three. who am i forgetting?

  50. JCH says:

    If you want to study schools, head for the Dakotas now. Watch it in chilly weather in real time. They are very reluctant to close anything there. Some of South Dakota’s true believers in Dr. Scott Atlas are already quaking.

    In the Northern Hemisphere, only one country that experienced a winter outbreak, and that is China. South Korea got a winter graze. China’s mitigation was extreme. They still had a significant outbreak in Hubei, but pretty much clamped it everywhere else.

    The Northern Hemisphere benefited from the temperature effect starting in April. Our mitigation efforts were nowhere close to as good as the western countries think.

    60,558 new cases in the USA today. If something major is not done soon, it will soon be 70,000 a day, then 80,000, and so on.

    Deaths will be high unless there is a therapeutic intervention. They have been using Regeneron’s REGN-COV2 on 2,000 hospitalized patients in the UK since September 12th. It’s hard to believe a huge ongoing success story in the UK trial would not have leaked to news networks by now.

  51. Dave_Geologist says:

    Why is this herd immunity nonsense still circulating?

    Same reason AGW-is-a-hoax is still circulating Jeffh. People want to believe in the Tooth Fairy.

    And some who don’t believe nevertheless want the public to believe in Tooth Fairies because they sell Tooth Fairies, or because they have a broader political or economic agenda which is advanced by public belief in Tooth Fairies.

    And some who believed in Tooth Fairies, but in their heart of hearts now know they’re not real, still have a lot invested in that belief. The more bad decisions and adverse consequences flowed from their prior belief, the harder it is to admit to the fiction because it also means admitting they were wrong, perhaps with real-world consequences they could and should have avoided.

  52. Dave_Geologist says:

    But I do recall many people promising us one “by october”.

    I take it you’re referring to Donald Trump?

    Certainly not the WHO (maybe late next year), nor any of the leading vaccine candidates (maybe finish Phase III trials this year, well into next year before hundreds of millions of doses are available), not the FDA, not CDC, not ECDC, not Ferguson (otherwise Report 9 would not have had a second wave this winter), nor any mainstream scientific or medical bodies.

    Who else did you have in mind eric, other than The Donald? Don’t be shy, name names.

  53. Dave_Geologist says:

    Ah. Believe it or not I was going to add for the avoidance of doubt that in the context of the thread (whether a vaccine renders second and later waves moot because we’ll be able to prevent deaths, not just defer them), “arrive” doesn’t mean the first shipment of vials, nor does it mean the completion of Phase III trials, and certainly not the start of Phase III trials

    AstraZeneca today received support of more than $1bn from the US Biomedical Advanced Research and Development Authority (BARDA) for the development, production and delivery of the vaccine, starting in the fall. The development programme includes a Phase III clinical trial with 30,000 participants and a paediatric trial.

    AFAIK the lead contenders are about at the start of Phase III trials, exactly what the press release said on the tin.

    “Arrive” in any meaningful context means at least a hundred million doses available in the US and a billion worldwide.

    The Oxford team were only delayed a couple of weeks by their adverse reaction. This is well worth a listen: Sarah Gilbert on developing a vaccine for Covid-19. Particularly for the novel approach of building a backbone for plug-and-play individual vaccines, with the express aim of enabling rapid development to handle new pathogens.

  54. Jeffh says:

    Excellent posts, Dave. Just today an Institute of Brain Research here in the Netherlands will begin studying the effects of Covid-19. Apparently, a significant number of people here who recovered from Covid-19 infection in spring are suffering severe brain-related trauma and damage as a result of the virus. This can be added to harmful effects of viral infection on the lungs, cardio-vascular system, the kidneys, on male fertility and other side-effects. When comparing the costs of intervention versus non-intevention, these really need to be factored in.

  55. Anyone who believed in “by october” most likely bought into “by october” a long time ago. So, for example, I never bought into “by october” at any time by anyone making such a PROMISE, not Small Hands, not the drug companies, not the FDA, not the WHO, nobody-nowhere-nohow!

    People who believe in Imaginationland need to move to Russia and partake of their national healthcare program …

  56. Dave_Geologist says:

    Do these conclusions still hold?

    ISTM no, Willard. There’s been a lot of water under the bridge since their data cutoff in May, and they themselves acknowledge the limitation that many countries were not past their peak so the ultimate First Wave death toll was unknown.

    See this: Have deaths from COVID-19 in Europe plateaued due to herd immunity?, and specially the graphs in the supplementary material. The later into the death toll you leave lockdown (i.e. the more you let the virus spread), the more deaths you get in the first wave, within and between countries. Seems like a slam-dunk to me. Absent magic, vastly different infection fatality rates even within nations, or vastly different pre-existing immunity* between regions and between adjacent nations, it woz lockdown wot dunnit.

    * Immunity in this context being T-cells or whatever, what some have called “dark matter”, because antibody-defined prevalence gives an IFR in line with independent studies (I presume the Spain outlier is where ICUs were overwhelmed and patients were triaged).

  57. Dave_Geologist says:

    This analysis Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe (also data only to May but by then lockdowns were being lifted and we know that most countries had a trough during the summer so they’ve captured most of the first wave deaths) basically says that it was lockdown, and everything else was tinkering around the edges. This is an unwelcome message prone to provoke denial but is a lesson the UK should perhaps have learned from the lack of success of the initial “Happy Birthday” phase, and the failure of the current gradually tightening measures to get R back below 1. The circuit breaker in central Scotland, coming soon I suspect to northern England, perhaps indicates some recognition that basically you have to stop households meeting other households, unpleasant as it is and disruptive to the economy as it is. As with climate change, there are no easy or comfortable options.

    I take a similar message visually from this comparison of Sweden to other Nordic countries: Four months into the COVID-19 pandemic, Sweden’s prized herd immunity is nowhere in sight. Fig. 1 shows that Sweden ended up in much the same places as its neighbours on their Response Stringency Index, albeit more by encouragement and people making their own decisions than by enforcement. But count back 3-4 weeks from Sweden’s peak deaths and see the difference in stringency at the time those dead Swedes were infected. Hit it hard then relax a bit vs. start with baby steps and add more if it’s not working. I know which I’d choose (I note though that some epidemiologists have advised against a single hard lockdown because it makes it hard to know which aspect is working and which can more safely be relaxed).

  58. Joshua says:

    Dave –

    > Immunity in this context being T-cells or whatever,…

    In case you don’t know…people who study this say that although it may well mean a probability of less severe infection, T-cell reactivity aka “T-cell immunity” likely doesn’t mean immunity from infection.

  59. Dave_Geologist says:

    ATTP, now you’re back “oan the Covid”, as Janey Godley would say 😉 , this has been bugging me for some time and I wonder if there has been any discussion of it by you or other modellers.

    I’m aware of the “why did they use 2.x for R when it was obviously 3.y by then” discussions, and remember my own puzzlement when I was seeing two doublings per week in deaths but government sources were quoting a week to ten days as the doubling time. Since then we’ve had the genotyping that shows most of the first wave infections were brought in from Europe, probably after the half-term holidays, and had not grown locally from January importations. Genomic epidemiology of SARS-CoV-2 spread in Scotland highlights the role of European travel in COVID-19 emergence and Preliminary analysis of SARS-CoV-2 importation & establishment of UK transmission lineages.

    That will presumably have distorted estimates of R during that period, because cases that you’d infer had grown very rapidly from the established pool of infection had actually appeared out of nowhere and seeded their own infection chains. 2000 independent importations, each probably affecting at least one household and some affecting an entire coach party or whatever, must have represented tens of thousands of individuals, perhaps many tens of thousands. Once that gets smoothed out I can imagine it would look like transient increase of R from 2.x to 3.y. Had we not locked down that could have been tested against observations because once that surge of imported infections had worked through we’d go back down to 2.x.

    Needless to say, the arrival of a second wave just after the summer holidays, with a month or so lag between the second wave in popular holiday destinations, has me thinking history is repeating itself. Since it coincides by definition with schools and universities returning after the holidays, that adds another confounding layer when policy-makers try to decide what let the cat out of the bag and how to herd the subsequent litter of kittens back in.

    I’ve also wondered in the context of the Ferguson point that in retrospect we locked down too late, could an alternative intervention of quarantining arrivals have had the same impact as an earlier lockdown? Of course in the policy world chances are that if quarantining holidaymakers had bought us a few weeks more time before reaching mid-March levels, the consequence would have been that we’d have delayed lockdown until the beginning of April and would have ended up in the same place as we are now.

  60. Dave_Geologist says:

    I think that’s a much-missed key point Joshua.

    Immunity as in “I don’t get sick and can’t infect anyone” vs. “I don’t get sick but can still infect Granny” have hugely different implications both for any idea of herd immunity and for vaccination strategies. Take the two extremes: measles where we vaccinate the young and healthy and protect the vulnerable who can’t be vaccinated by eliminating circulation of the virus; vs. flu where we vaccinate the vulnerable and accept that there will be widespread virus in the community.

  61. Dave,
    I don’t know that I have any really good answers to your questions. We were mainly trying to understand the model, and the results presented in Report 9, and so used the same parameters (much to James Annan’s annoyance). In the second part of the paper we did then consider a scenario that we thought better matched the lockdown and found that we did indeed need to use a larger R0 value (3 – 3.5) and an earlier start (compared to what was suggested in Report 9) to match the data.

    It is certainly the case that in the latter analysis, the suggestion was that if we relaxed too much (or went back to pre-lockdown behaviour) the we could end up with a second wave that might be similar to what the first wave would have been like had we done nothing (see the figure). Of course, this is a model, so one should treat this with caution. Also, we are very unlikely to simply relax back to pre-lockdown behaviour, so the second wave will probably not be as severe and may well grow more slowly. But it does seem that we need to think carefully about how to deal with this (and I do take Eric’s point that there are potentially severe implications to another hard lockdown).

  62. Willard says:

    > I take it you’re referring to Donald

    I doubt Donald is EVERYBODY, Dave.

  63. Jeffh says:

    ATTP,, congrats on your paper. Food for thought, certainly. Still, seeing it go viral on Twitter amongst the herd immunity ‘let’s ride it out’ crowd is, to me at least, disturbing. For the umpteenth time, we are not only talking about death rates here but nasty medical legacies. My guess is that few if any of the ‘ride it out, develop herd immunity and move on’ advocates have been infected by Covid-19. It’s easy for these people sitting back while feeling well tweeting away about the benefits of riding out a single, quick, mass infection to develop that illusory herd immunity until they themselves are infected, go through terrible symptoms, and then do not fully recover. I just wonder if they would be singing the same tune then, as they struggle to breathe, experience cardiac problems, have difficulty concentrating and even have trouble even getting out of bed in the morning. These are all very common Covid-19 legacies, well documented by now, and not restricted to the elderly and vulnerable. Then, to add insult to injury, they discover that immunity is transient and the second round is potentially worse than the first.

    These points have all been raised countless times by epidemiologists and virologists studying Covid-19 and are ritually ignored by the herd immunity advocates. We have to ask ourselves why.

  64. Jeff,
    Yes, very unfortunate that it’s being promoted in that way. If anything, it partly illustrated that the model being used by the Imperical College team does reasonably well if you update the parameters. So, maybe ironic that some who have doubted the models are using a modelling paper to justify their preferred strategy. Of course, we see the same in the climate context; people who continually criticise models until one produces results that they suddenly like.

  65. Joshua says:

    Willard –

    > If you then reject it as irrelevant, say because it’s not what you ask, then diminishing returns are to be expected.

    I think this was important. I did, in the next sentence explain why I was asking but Eric didn’t make that transition with me. So that inflection point is instructive. I suspect that Eric brought something to that table but I also accept full responsibility there even if I claim unawareness rather than nefarious intent.

    > Eric’s comments stand on their own, he fulfilled all his commitments. His job is done.

    What are our commitments, here? What are our jobs? That is an interesting frame. I have commitments to myself but within this community it’s hard for me to get a concrete picture. It’s an interesting frame for thinking of how to engage, and I’ll examine that some more but I think the commitments and jobs within the community would ultimately need to be more explicit and gain “written consent.”

    > Blog comment sections are not parlors. If there are questions that deserve to be raised, you can raise them without addressing them to anyone in particular. Even better, you can offer what you think is the appropriate response right from the start.

    Yah. That would be a good practice. But it also feels a bit unsatisfying. If we de-identify the dialog it could open many doors. But I feel like it’s a kind of acknowledging defeat. I resist. But maybe the advantages of a more personal engagement require first the establishment of trust and a good way to do that is to open the door by way of a kind of deidentification.

    I thought about this above in my comment to Dave above. I considered that it could have come across as a correction.

    I tried a socio-pragmatic chunk but those are very culturally specific and lean heavily on tone and body language which are lost online.

    Fortunately he didn’t take it that way but that is far from the norm in blog comment exchanges. People are cranky a lot.

    Perhaps I need to think of more structural alternatives to textual socio-pragmatics. But I’d have to be willing to be less cranky first.

    > Sandbagging can be felt when the question being asked is followed by a series of counters that could have been laid out beforehand.

    My goal was to know Eric’s thoughts about the implications of a potential vaccine to a “herd immunity” strategy. He has expertise, I do not. It feels vulnerable in that context to put out my thoughts. And the exercise is potentially irrelevant and unnecessary. But sure, I could expose my neck as a gesture of good faith.

    > That the person from whom you expect work feels sandbagged ought to be enough as a cue. The concept is quite clear enough for what I’m trying to express.

    Yes. Again, we all bring something to the table and I was completely surprised. But there’s no particular reason that I’d have to be surprised again although I’m certain I will be.

  66. Joshua says:

    Jeffh –

    > For the umpteenth time, we are not only talking about death rates here but nasty medical legacies

    I share your frustration. Other things some folks often ignore:

    Extrapolating across very dissimilar contexts is complicated. Sweden is a socialist society with an extensive social safety net paid for by very high taxes, has had equivalent economic damage to other Nordic countries (and far more than many other countries with far lower per capita deaths and disease), has a rather uniquely low average number of people per household, has a uniquely supportive environment for people to not go into work, has a high average baseline health, etc.

    This disparate impact on certain communities.

    The huge sacrifices made by many people, in particular the disproportionate number of essential workers in minority communities, and of course healthcare workers.

    The impact of a vaccine on a “herd immunity” strategy.

    The impact of improvement in therapeutics and standard of care to a “herd immunity” strategy.

    The preliminary nature of the evidence we have so far

    The uncertainties related to understanding “herd immunity” in the ongoing context.

    The difficulty of decision-making about risk in the face of uncertainty.

  67. Joshua says:

    Anders –

    > Of course, we see the same in the climate context; people who continually criticise models until one produces results that they suddenly like.

    I’ve noted that interesting juxtaposition over at climate etc., along with the importance of addressing uncertainty in modeling, but no one there seems very interested in that topic – well, at least with me anyway.

  68. Dave_Geologist says:

    Thanks ATTP. Wasn’t expecting you to have done it 🙂 . It does seem the sort of thing you can investigate with an agent-based model. Even if there’s not an explicit option to seed a bunch of new infections you should be able to stop the run, edit them in, then restart it. Heh heh, reminds me of RESTART files in ECLIPSE or VIP reservoir simulation software. Originally intended for saving a part-job when runs took days or weeks and you saved the previous day’s work in case of a crash. But exploited for coupled geomechanical simulation where you took the snapshot, ran it through a geomechanical model which updated the parameters (porosity loss, whatever) then smuggled the changed file back into the run so the reservoir simulator picked up where it left off none the wiser. IIRC when you put in pre-planned checkpoints it was called a RESTART CARD, even though it was a file entry made in a GUI editor. Not prizes for guessing how far back the practice goes 🙂 .

    I note that the two genomic papers are still only preprints, but they have heavy-duty funders so I presume are top people you’d expect to do a good job. Perhaps epidemiologists need to talk more to genomics people? Tracking lineages is one way to know to what extent imported vs. home-grown infections contribute, and given our lax quarantine or self-isolation practices compared to some countries you’d have thought it was something policymakers would be looking at and asking for advice about. Or could there be some denial there: serious quarantine put into the “too-difficult” box? I’m pretty sure the Scottish authorities are aware of the Glasgow-Edinburgh paper because the results came up in discussion of the Nike outbreak (traced and squashed, despite much panic-stricken press afterwards). Actually Fig. 5 is a real lesson in international spread. Came in from Ecuador, spread to a handful in Scotland before dying out here, but we seem mainly to have exported it to The Netherlands and one found its way to New Zealand of all places.

  69. Willard says:

    > This disparate impact on certain communities.

    This is probably the most striking parallel with ClimateBall. Everybody knows there is no silver bullet, yet we all seem to reason as if there needs to be one. We live in a diverse world. Our policies come with trade-offs that vary according to geography, demography, political culture, etc. They are also based on far from perfect knowledge. Instead of asking ourselves what would be the optimal solution to our problem, I think we should ask: what are the requirements for our institutions to act responsibly? Something like epistemic duty needs to be delineated.

    In a paper with two other philosophers, Eric argues that governments failed (and are still failing) their epistemic duty. I disagree with their argument, but I think they raise the right question.

    If we keep arguing irresponsibly, we won’t get responsible governance.

  70. Yes, I read Eric’s paper where they argue that governments have failed. I guess my simplistic argument was that this was happening so fast that governments were making decisions quickly and may not have really had the time to truly check if they were satisfying their epsitemic duty. It did make me think, though, and I can see a concern that we don’t want this to become something that happens too often, because then you would start to worry that governments were using these kind of events as a mechanism for imposing conditions that violate civil liberties.

  71. I think all three of us agreed that the quick responses were allowable. But the ongoing situation, even when we were writing the paper in, I think, late April, was no longer acceptable.

  72. Below is the relevant passage:

    Even so, it might be appropriate, at the beginning of a potential catastrophe, for policymakers to adopt a very cautious stance. In doing so, it might be excusable to accept, provisionally, the extremely cautious predictions of epidemiologists, despite the problems in their data and models. It might be fine to act first and ask questions later. It should be stressed that even this concession is questionable—after all, governments must have strong and solid evidence, rather than poor evidence, that a potential disaster of a certain size is occurring in order to justify their behavior. Historically, “we must avert disaster” has been the main excuse for government overreach. But even so, as Nozick (1974) rightly observes, the potential to avert “catastrophic moral horror” through speedy action can license many responses that would normally go beyond the pale.

    Regardless, this kind of justification will not do beyond the very short term. Even in the direst emergencies when immediate action is required, we expect policymakers to supply the needed justification shortly thereafter, to rely upon established standards of evidence, to rely on high quality evidence, and to show their work in which they balance various social and ethical values against each other.

    For all the reasons outlined above, it will not do, in more than the very short run, for policy makers to declare, as Governor Newsom of California has done, that they are simply “following the science” in responding to a crisis like the COVID-19 pandemic. In the interest of transparency, they should make it clear that they are adopting precautionary reasoning and inform their constituents what the plan is to quickly move to a more substantive cost-benefit analysis—and explain what values are to undergird that analysis. But states are under more substantive obligations as well. They should begin collecting the data needed to properly assess their strategies and determine whether continued restrictions of citizens’ basic liberties are justified. The longer they neglect to take measures like these, the more their impositions look incompatible with the foundational commitments of liberalism.

  73. Eric,
    Yes, I largely agree, but I may be more relaxed as to what I would regard as short-term. I have worried about the “followed the science” narrative, because I think it does confuse the role that scientific information plays in decision making and potentially the scientists up to be those who will be blamed. I think policy makers should know better and that scientists should be clearer that their role is to inform, not force policy-makers to make specific decisions. I can see scenario might say “what should we do?” and scientists might provide answer. However, the responsibility still lies with the policy makers.

  74. Joshua says:

    So I haven’t read the paper yet – but in response to the excerpt:

    > Regardless, this kind of justification will not do beyond the very short term. Even in the direst emergencies when immediate action is required, we expect policymakers to supply the needed justification shortly thereafter, to rely upon established standards of evidence, to rely on high quality evidence, and to show their work in which they balance various social and ethical values against each other.

    I wonder what the criteria would be that would be used to mark the transition from one stage to the other. I don’t have a suggestion but finding one seems extremely problematic to me. Certainly, I think this problem parallels a problem in climate change – what needs to be weighed in is the scale of possibly low-probability but extremely high damage function outcomes. Of course, high damage function outcomes runs in both directions w/r/t SIPs and NPIs, just as they might with climate change. At root, how does one de-polarize and de-politicize and de-bias the an analysis?

    > The longer they neglect to take measures like these, the more their impositions look incompatible with the foundational commitments of liberalism.

    I wonder about the discussion of “liberalism.” If point of fact, at least in this country, polling shows pretty widespread support for SIPs and NPI, and in fact pretty widespread support for interventions that were more stringent than what we’ve seen. Of course, there’s the problem of politicization, of regional differences, and of disparities of media influence.

    But an analysis of illiberalism, it seems to me, should take into account that there is a structural imbalance (at least in the US), whereas a pretty marked minority of people (being white and living in a rural community, as two among many ways of differentiating) have disproportionate power (by way of the presidency, representation in the Senate, and representation in the House [irrespective of who is currently in the majority there]) in affecting policy outcomes.

    Our structural inequalities, and in particular the disproportionate influence that wealth applies to our policy-formation process, it seems to me, that means that people must be careful to define illiberalism in context if they are going to discuss “foundational commitments to liberalism.”

  75. Joshua says:

    This is a good one…

    My irony meter, which already extends way past 11, just went off the scale…I read this (from Matt by way of Judith):

    > Now, with cases low and the Swedish economy in much better health than other countries,

    after reading this:

    > A fourth mistake is to gather data that are compatible with your guess but to ignore data that contest it.

    Truly remarkable. Definitely time to recalibrate.

  76. Jeffh says:

    Maybe we all ought to go back to late January when all of this was beginning to unfold. I wrote a paper for an English economics journal in Poland with a Dutch economist based there in April and did a TED podcast shortly thereafter in which I addressed this. There is a lot of discussion here about we should or shouldn’t have done once the cat was metaphorically out of the bag, but little in the way of addressing the equally big question as to why governments did virtually nothing when the WHO formally raised the alarm, thereby enabling the pathogen to spread across the world, infecting thousands and then millions. Where did the precautionary principle go then? Well, I think that the inability of governments to respond before Covid-19 reached critical mass – or more accurately, governmental impotence – is tightly correlated with the neoliberal capitalist doctrine that has the world in its clutches. The system is front-loaded to maximize profit and investor’s returns, and nothing whatsoever is put aside in the event of a calamity. The system is so tightly wired that it can be easily derailed if something gets in the way.

    When it became clear that the coronavirus posed an increasing threat, governments faced an unenviable dilemma. The guns were pointed at them from the private sector, and especially from the most powerful transnational corporations. Certainly by early February national health agencies and epidemiologists across the world were informing their heads of state that the virus represented a potentially serious threat, yet little happened or was done to stop it. Essentially, the way I see it is that governments had their hands tied by powerful, vested interests that have enormous influence on policy. We all know what those vested interests are. Governments had the chance to proactively respond before the virus was seriously spreading, and they could have done this in multiple ways. Instead they sat on their hands, closed their eyes, rocked back and forth and hope the virus would ‘fizzle out’ in a similar way that SARS-Covid 1 had done 17 years earlier. Essentially, governments were in a no-win situation: they were ‘damned if you do, damned if you don’t’. If they had implemented forceful travel bans and light lockdowns before the virus spread and took hold, we would probably not be in the situation that we are now. But if they had done that, a large portion of the corporate sector would have gone berserk and demanded proof that the virus posed a real threat and that interventions were necessary. How could that be proven? It is impossible.

    So they waited, and waited until the virus became a pandemic and have been playing catch-up ever since. This mutant form of capitalism holds the world hostage, allowing governments only to respond reactively and not proactively. Indeed, I see Covid-19 as an ‘appetizer’ for the far greater threat posed to mankind by climate change. From my perspective, the prognosis is not good. If we are unable to respond proactively to an imminent threat like Covid-19, how on Earth are we going to deal with one, albeit far greater, that plays out in slow motion by comparison?

  77. Joshua says:

    Jeffh –

    > Essentially, the way I see it is that governments had their hands tied by powerful, vested interests that have enormous influence on policy. We all know what those vested interests are.

    Of course, governments did have a vested interest in not issuing SIPs and engaging other NPIs that would have a clear signal of harm, even if the causality of the harm and the differential harm in the face of obvious counterfactual scenarios were complicated questions. Although the influence of powerful vested interests is indeed relevant, certainly there were more, very powerful motivators that were also in play. It’s interesting to note that many of the most powerful vested interests have actually benefitted from SIPs and NPIs.

  78. Joshua says:

    Jeffh –

    On the other hand, going back again to the excerpt:

    > Regardless, this kind of justification will not do beyond the very short term. Even in the direst emergencies when immediate action is required, we expect policymakers to supply the needed justification shortly thereafter, to rely upon established standards of evidence, to rely on high quality evidence, and to show their work in which they balance various social and ethical values against each other.

    Consider the failure to get a stimulus through our Congress and president.

    There is absolutely no doubt that it is desperately needed by a huge number of people, and that needs is ongoing and getting bigger as time goes on. And yet we not only have a failure to get a stimulus package through, we have a block (of disproportionately wealthy people) with disproportionate power who are dedicated to opposing it, and can effectively block it.

    IMO, it is extremely difficult to know the differential outcome of what would have happened with vs without SIPs and NPIs being issued. There is a very real possibility, IMO, that the economic damages would have been worse in their absence. Although for sure there is no guarantee of that, I have yet to see a compelling argument made otherwise.

    We might similarly argue that a large economic stimulus package could have diminishing or negative returns long-term. I tend to think not, but I certainly can’t prove a case.

    But unlike with SIPs and NPIs to address Covid, where there might reasonably be some doubt related to the uncertainties facing policy-makers early on in the pandemic when they were deciding to implement those interventions, and where there definitely are uncertainties facing them when deciding whether to lift those interventions, there is, I would say, just about zero question as to whether a stimulus would have an IMMEDIATE impact on countless people.

    IMO, that suggests that focusing on getting immediate stimulus in place should be a higher priority than focusing on the inherently complicated task of after-the-fact evaluating counterfactuals about extremely complex matters, particularly when we lack good longitudinal data, control for confounding variables, and the ability to implement variable controls. Not to say that such analyses aren’t of value. I certainly think that they are. But I think it is highly unfortunate that as a society we seem more focused on fighting people who disagree with us about the viability of SIPs and NPIs than we are on uniting to address the immediate needs that clearly are at hand.

  79. Joshua says:

    Perhaps of interest regarding how to decide when to lift SIPs and NPIs.


    States that were quicker to end shelter-in-place rules and to reopen in the spring have paid an economic price. Our Back-to-Normal indices for Arizona, Florida, South Carolina and Texas indicate that their economies have effectively gone nowhere since mid-May.



    How we created the index
    Our Back-to-Normal Index combines 37 indicators, including traditional government statistics and metrics from a host of private firms to capture economic trends nationally and across states in real time. The government statistics cover retail sales, industrial production, durable goods orders and housing starts, to name a few. Private contributors to the index include Zillow for home listings, OpenTable for restaurant bookings, Homebase for its measures of hours worked at small businesses, the Mortgage Bankers Association for data on applications for mortgage loans, the Association of American Railroads for rail traffic, and Google, whose cellphone-based mobility data is a window into how actively people are shopping, going to work and venturing out to play. (Read the full methodology here.)


  80. verytallguy says:

    “EVERYBODY was promising a vaccine by this fall”

    Note that one of the promises was for supply, and AZ are indeed already manufacturing, and the other two were not promises.

  81. David B Benson says:

    Possibly Sub-Saharan Africans are more immune:

    There’s a recent paper in Nature blaming Neanderthal genes for adverse effects of the novel coronavirus infection.

  82. Willard says:

    Vintage 2020-04-17:

    COVID-19: Vaccine may be ready by fall and other reasons for hope.

    Around the same time:

    The best-case scenario is that by the autumn of 2020 we could have an efficacy result from the phase III trial to show that the vaccine protects against the virus, alongside the ability to manufacture large amounts of the vaccine, but these best-case timeframes are highly ambitious and subject to change.

    The newsie minces less words, as it’s called Coronavirus vaccine developed in the UK could be ready by fall, if it works, but it’s still formulated as a truism.

    Around the same time:

    Pfizer says coronavirus vaccine could be ready for emergency use by fall

    A bit earlier:

    A coronavirus vaccine could be ready by September, according to a British scientist who is developing one

    The conditions under which scientific promises are satisfied might deserve due diligence.

  83. So, what is the latest news on actual vaccine ETA’s? You know, an actual vaccine or vaccines where something in excess of tens of millions of people have already been vaccinated. I could really care less about overly hopeful and optimistic timelines from six months ago, as I am much more interested in overly hopeful and optimistic timelines today! 😉

    And if you say Russia already has one, then I have some prime beachfront property in Louisiana to sell you …
    Highlight Russia and their Daily Cases and their Daily Deaths. Small Hands, Bojo and Putin already got their’s, so that is three down and ~7,799,999,997 to go (give or take a few tens of million people) …

  84. Willard says:

    No idea, Everett. But perhaps this could help:

    I’m not sure how all this matters for the price of tea. Was that information fed in the modulz, did it constrain the horizon for policy making? I suppose it was just hopeful.

    All in all, here are my main two takeaways from COVID-19. Long-term policies that we assign to a population need to be as simple as a nursery rhyme or else they backfire. Social nets matter.

    I hope we’ll improve on that last one.

  85. of course, if you go look at that same chart in the wayback machine from just june, you can see that many were in fact predicting full rollouts by october

  86. So, hmm, err, who actually plots a zero on the log axis of a plot anyways? It should be 100 or 10^2. Oh and why plot different x-axis scales, as one hides the second wave.

    While I’m at it, why even plot anything on a log axis, I gave that one up months ago. Only plot dailies in linear-liner space, no cumulative at all, as one is an integral of the of other and then taking the log of cumulative is doubling down on your, quite frankly, lukewarm results.

    And the obsession with just the UK, I have hinted at that in the past, but I have just about have had it and I needed to just say so out loud and directly.

    So here is my spin on all the atrocious modelling efforts going on everywhere. If one were to do the entire WORLD, how many modelling coefficients would be necessary? I’d SWAG tens of thousands at the 100 km^2 level, all needing calibration using actual data. In other words, and IMHO, all you have done is a grand scale curve fitting exercise. Again, and IMHO, no calibration data, for this, or any pandemic, is an order of magnitude analysis. which I think is where all this started from, an ~order of magnitude analysis. So, you have a model that inherently needs a lot of calibration data and updating as underlying factors change (e. g. lessons learned).

    Small Hands still sticks to the very early ~2.4 million deaths estimate to this very day. The only way that estimate makes any sense is in a Groundhog Day sort of way, each day where everyone totally ignores/forgets/disappears the previous day deaths, A true, do absolutely nothing, it will just go away, scenario. A scenario that Small Hands also adheres to, ‘herd mentality’ indeed. :/

  87. Joshua says:

    Funny that a mis-directed question produces an irrelevant answer that becomes a focus of attention.

  88. Jeffh says:

    I vehemently disagree that EVERYONE was promising to have a vaccine ready by autumn. This is the kind of cherry-picking exercise that climate science deniers are adept at. I vividly recall Fauci and many other epidemiologists way back in March saying that it would take a considerable time to generate, test and manufacture a vaccine and that the earliest one would be widely available was in the second quarter of 2021. Some were even more pesimistic and said it would be later than that. The internet is full of quotes along those lines.

  89. Dave_Geologist says:

    On vaccine timing: due diligence requires consideration of the words “could” and “the most optimistic scenario for a working vaccine” (working ≠ rolled out to half the population) and “just about possible if everything goes perfectly”. And that’s certainly not a promise, let alone a promise that it would have been rolled out at a level which allows release from NPIs.

    Not something to bet the farm on, let alone millions of lives.

    “Working” is relevant for decision-making in the sense that knowing you can successfully vaccinate the population over the next six months would and should inform decisions about the nature and timing of NPIs, and of business support (life-support for six months vs. it might be years before that sector re-opens). Although I honestly can’t see that the promise of full vaccination after another six months of economic hardship would sway the minds of those opposed to NPIs one iota.

  90. Dave_Geologist says:

    Err, eric, it’s hard to do due diligence if you don’t know the difference between “first supply” and “full rollout”. Or for a programme covering hundreds of millions in the USA and billions globally, the start and end of rollout. I would hope the signal to unlock would not just be that the President or PM has had his dose, but it will take three or six months to reach herd immunity levels among the population.

    And even using the un-diligent approach 50/50 by the end of the year is a long way short of EVERYONE.

  91. Dave_Geologist says:

    Moving away from vaccines, perhaps the most striking thing about Fig. 5 of Genomic epidemiology of SARS-CoV-2 spread in Scotland highlights the role of European travel in COVID-19 emergence is what happened when it got to New Zealand, vs. what happened when it got to The Netherlands.

    Hmm. Wonder which country was quarantining arrivals?

  92. Jeffh says:

    Certainly not us in the Netherlands. Our testing capacity early in the pandemic was risible, and even now we are lagging. The government stopped testing people arriving at Schipol Airport many weeks ago, and then started recommending people quarantine rather than enforcing it. Unsurprisingly in a country where people are reluctant to listen to authority the virus is currently spreading exponentially and the hospitals are rapidly filling up. A recent survey revealed that 9 in 10 people here showing symptoms of a respiratory infection that may or may not be Covid-19 still regularly leave their homes and enter the public arena. One in six said they would not comply with some covid restrictions and 60% of those said they would ignore all restrictions. The government has still not mandated the use of masks indoors because the head of the health agency believes they are not very effective, despite volumes of evidence to the contrary. Because of that, as expected, most people here do not wear them at all. By the end of the week we expect 2000 or more beds to be occupied by Covid-19 patients. It is bleak here right now.

  93. Dave_Geologist says:

    I did wonder whether the splurge of Netherlands cases was a reflection of the time delay in international notification of contacts or of poor isolation and contact-tracing in-country. Stay safe.

    The UK had an effective, pre-existing contract-tracing infrastructure at the time of the Nike outbreak intended for things like measles or STDs. The latter in particular meaning that the tracers were sensitive to local communities as it was done foot-leather style. Hence the England and Wales strands being snuffed out. Finland, Australia, Portugal and Slovakia too, but they had very low numbers so tracing should have been easier. Plus Australia quarantined IIRC. Interesting that the Belgium and USA ones were stopped though.

    UK contract-tracing was abandoned when it was overwhelmed in March and testing focused on triaging patients between Covid and non-Covid wards. The failing privatised one you may have heard about as the UK system is for England (and Wales?) only but is run by the UK Government. It sounds today as if they’re going to at least partly listen to reason and give more funding and responsibility to local Public Health teams.

  94. gwws says:

    At least the newspapers in Holland are clear:

    We can no longer eradicate this virus – not even with a vaccine, says Van der Hoek. “Not only is it already too widely distributed around the world, but the virus now also exists in animal reservoirs. … Inevitably, SARS-CoV-2 is going to become the fifth common cold virus. … That could take fifty years or happen within two years. There is no road back. ”

    NRC 10-11 October 2020.

  95. Joshua says:


    Is it your impression that in he Netherlands, people think the threat of serious illness is small? Fatigue? Have views on the seriousness of the illness become politicized? Is there block advocating for a “herd immunity strategy” as there is in the US?

  96. Willard says:

    > due diligence requires consideration of the words “could”

    Conceding the point might be more economical than having to appeal to a Clintonian conception of language, Dave. While there are studies on promises, e.g. scientific promises don’t look commissive, sticking to matters that relate to AT’s post looks more promising to me.

    For instance, earlier you said ISTY that the conclusions about obesity, age, and wealth did not hold. Yet the study you linked only looked at per-capita immunity. Scratching my own itch, I found that the CDC updated its evidence basis last week, and obesity is still a risk factor:

    As for age, there is a recent study that concludes:

    Our study revealed that if the mortality rate or the fraction of symptomatic infections among all COVID-19 cases does not depend on age, then unrealistically different age-dependencies of susceptibilities against COVID-19 infections between Italy, Japan, and Spain are required to explain the similar age distribution of mortality but different basic reproduction numbers (R0). Variation of susceptibility by age itself cannot explain the robust age distribution in mortality by COVID-19 infections in those three countries, however it does suggest that the age-dependencies of (i) the mortality rate and (ii) the fraction of symptomatic infections among all COVID-19 cases determine the age distribution of mortality by COVID-19.

    I have yet to find a study of wealth as a risk factor. But I found something staggering:

    The COVID-19 pandemic is estimated to push an additional 88 million to 115 million people into extreme poverty this year, with the total rising to as many as 150 million by 2021, depending on the severity of the economic contraction. Extreme poverty, defined as living on less than $1.90 a day, is likely to affect between 9.1% and 9.4% of the world’s population in 2020, according to the biennial Poverty and Shared Prosperity Report. This would represent a regression to the rate of 9.2% in 2017. Had the pandemic not convulsed the globe, the poverty rate was expected to drop to 7.9% in 2020.

  97. verytallguy says:

    “I vehemently disagree that EVERYONE was promising to have a vaccine ready by autumn”

    It is notable that the evidence presented here which supposedly supports such a claim, actually clearly refutes it, almost all timelines discussed being heavily caveated.

    Nevertheless, let us hope results are soon, and positive. It is not unlikely.

  98. Willard says:

    > actually clearly refutes it

    Scientists don’t make the kind of promise that refutation presumes. One might as well argue that the very idea that scientists promise predictions or forecasts is absurd. Alternatively, one can easily see that announcing promising results with timelines that scoop the competition as some kind of promise.

    On the one hand, we have an interpretation that makes Eric’s claim absurd. On the other, we have an interpretation that makes Eric’s claim a fair description of the timeline. The first turns into a parsomatic affair where only my own ball moves forward. The second returns us to AT’s post.

    I understand that Eric started the gotcha games, but at least he did not rely on semantics to do so.

  99. Joshua says:

    Vague references to “lockdowns” without specifics, without related context, are less than useless. As are empty appeals to authority, as in…

  100. WingNutDaily, I think I’ll wait for an official WHO announcement from
    UPCOMING: Press Conference, Monday 12.10.2020 5:30 p.m. 6:30 p.m. (Geneva Time – CEST)
    T-18.5 hours and counting …

  101. Joshua says:

    In the absence of fleshed-out alternatives, armwaves to “no lockdowns” make complete sense as a theoretical construct. The real world makes “no lockdowns” a bit more problematic.

    When there is no realistic infrastructurie finding for robust testing, tracing, and isolating…

    When there is no realistic infrastructure for “protecting the vulnerable”…

    When powerful political lobbies block funding to make schools safer, to compensate older teachers, to enable comprehensive surveillance to allow for appropriately targeted opening…

    When hospitals in poor communities are understaffed and under-resourced…

    When the most vulnerable communities are those least able to work from home…

    When PPE is in widespread shortage…

    When necessary healthcare workers are overworked and underpaid…

    When key demographics lack access to healthcare and/or are worried about their immigration status I’d they seek healthcare out…

    When the communication in the science of public health policies is explicitly managed for political expediency…

    Then exploiting “the poors” to argue against SIPs and NPIs ia no more helpful than pretending that SIPs and NPIs aren’t mixed bags. The parallels here with arguments about climate change are striking. One thing for sure is that “the poors” are an easy public facing rationalization for whichever biases one seeks to confirm.

    There is no realistic way (imo) to disaggregate the economic damage of SIPs and NPIs from the economic damage of the pandemic itself. At least I haven’t seen any compelling arguments from those who make the claim of doing so.

    One of the most stark ironies here is many of those who argue as spokespeople for “the poors” see no problem with condescending to ignore what polls say that poor people (like other people) say that they believe would be the best policies. SIPs and NPIs have broad support. Of course the “poors” spokespeople consider themselves more informed than “the poors” themselves, and this situated to make better argument on their behalf than they can themselves.

    But it would be nice to at least see some fleshed-out arguments more than just handwaves. What are the real world alternatives to SIPs and NPIs? Where are the nuanced policy proposals such as those Nabarro is asking for?

    If anyone has seen any, I’d appreciate a link.

  102. Joshua says:

    But instead of discussing that, let’s bicker about something really important, such as who is right about who predicted vaccines would be available when.

  103. Joshua,

    SIP = Stay In Place = ?
    NPI = NonPharmaceutical Interventions =? (had to Google that one)

  104. Joshua says:

    Shelter In Place (interventions, policies,mandates) .

    I don’t like the political gamesmanship of “lockdowns.” Just another case of ceding framing, ala Lakoff.

  105. I don’t like the political gamesmanship of “lockdowns.”

    Well neither do I, but we do have so-called militias (terrorists), freedom fighters (terrorists), long guns (terrorists), Proud Boys (terrorists), Boogaloo Boys (terrorists), that tray table of an ass Small Hands (figuratively speaking, that ass is big enough to set a tray table on and eat his shit, Moscow Mitch has eaten there many a time). …

  106. Willard says:

    The World Health Organization is harder to dismiss than petitioners saying in chorus:

    But then, compare:

    The only time we believe a lockdown is justified is to buy you time to reorganise, regroup, rebalance your resources, protect your health workers who are exhausted, but by and large, we’d rather not do it.

    and contrast:

    Building public health capacity is not the same as implementing lockdowns. It is about building up the capability of public health services to interrupt transmission in localities. Experience around the world shows that this capability is key to a successful COVID responses. Lockdowns just freeze the virus in pace they do not lead to elimination

    The two quotes are by the same author.

    So we’re back at “test-trace-isolate-protect services everywhere, with clearly justified performance metrics,” not very different from what Very Tall said.

    All in all, it’s hard to have a consistent messaging when those who recognize that message consistency is important (see David’s post) fall for journalistic let-you-and him-fight games.

    “4SD Sàrl is located in GENÈVE, Switzerland and is part of the Consulting Services Industry. 4SD Sàrl has 4 employees across all of its locations. There are companies in the 4SD Sàrl corporate family.”

    Four employees?

    If David wants to “respond to misleading information sensitively” or some such …
    “What will it mean for all of us? We will all learn to do the right things at the right time in the right way … because that is what we choose to do.

    We will also respond to misleading information sensitively (because we think we understand why some people believe in conspiracies) and firmly (because some of what is being shared is without evidence to support it).”

    Hiding behind 4SD (or whatever) is not the way to go about doing so IMHO.

  108. David B Benson says:

    I really don’t know what all of you are discussing, but it must be important.

  109. Finally a discussion of dispersion WRT growth

  110. Jeffh says:

    Nabarro was in no way supporting the Barrington Declaration, which was prepared by several academics whose predictions about immunity levels and the course of the virus were miles off the mark. Indeed, the Barrington Declaration is something of an embarrassment. Trust the corporate media to try and link the WHO with it.

  111. Dave_Geologist says:

    Willard, you know from previous experience that I find word games tedious and soon stop playing. And when I stop playing it’s not a concession of defeat: it’s just that the only word game I like is Scrabble and I’m bored now.


    Anyone who read any vaccine statements outside the funny papers and concluded we’d have enough to vaccinate to herd immunity by the fall Didn’t Do The Research or was guilty of a wildly optimistic interpretation. The links you quote, word games or no word games, don’t change that fact.

    Anyone who thinks first availability of vials or first rollout to the general population meant we could unlock from that day without mess mortality, and not months later when either all vulnerable people or 60-70% of the population had been vaccinated, was being wildly optimistic or doesn’t understand viruses and vaccines.

    You or I can safely engage in word games or unresearched musings, but people who for whatever reason are trusted public figures should consider due responsibility as well as due diligence.

    Anyone who Did The Research and knows it will take months at least to get from FDA approval to completion of a rollout that enables safe removal of NPIs should be agreeing with me, and not claiming that they’d been conned or promised something that was not delivered.

  112. Dave_Geologist says:

    BTW if the obesity, age, and wealth comment was aimed at me you must be confusing me with some other poster.

    Of course obesity is a risk factor. There appears still to be some debate as to whether once you’ve removed all the co-morbidities it remains an independent factor, but show me an obese person without co-morbidities. Even if true, that’s a difference that makes no difference. It makes sense biologically that there would be a residual, because the weight and bulk puts more strain on the heart, lungs and circulation, all organs which the virus damages. So you’d be starting with a handicap even without co-morbidities.

    Age has been a known risk factor since January, and I’ve never said otherwise.

    Wealth (poverty) ditto since early in the year, although again there are linked factors like crowded housing, low-paid public-facing jobs with no or poor sick pay, lack of resource to buy sanitisers and PPE, no car so use crowded public transport, health-care inequities, all the co-morbidities that go with poverty and low self-esteem, physical and psychological (the latter impacting compliance vs. fatalism), etc. etc.

  113. Ben McMillan says:

    Mass availability of a vaccine (which seems fairly likely) within the next six months would be a triumph, and pretty much unprecedented speed.

    It would also vindicate the idea of suppression/elimination (playing for time) over the ‘let everyone get it’ approach. I think the main lesson from ‘Report 9’ was really that all those options look bad, rather than the distinctions between a bunch of bad options being interesting.

  114. Dave_Geologist says:


    To steal from vtg: It is notable that the evidence presented here which supposedly supports such a claim, actually clearly refutes it.

    And eric, you must know that apart from a well-deserved reputation for a certain political and economic stance which bears on their objectivity on matters lockdown, the reputation of all Murdoch rags on this blog has been pushed well into negative territory by their decades of science denial over climate change. That link is not an argument from authority. It’s a dent in your credibility.

  115. verytallguy says:

    I didn’t know, but it turns out the Great Barrington Declaration is an output of the American Institute for Economic Research.

    Typically of such “institutes” or “thinktanks”, its title tells you nothing of its aims; it’s a right wing libertarian organisation which promotes small government.

    Now, who could ever conceive of such an organisation presenting fringe science and using right wing media to promote it as mainstream ?

    What an unprecedented state of affairs!

  116. Jeffh says:

    One final thought from me. Whenever I read pundits allegedly tearing their hearts out saying how much pain they feel about Covid-19 restrictions hitting the poor and the vulnerable, especially in the south, be wary. Many of these pundits and politicians support the dominant political and economic neoliberal system that it is driving poverty, social injustice and environmental destruction around the world. When President Donald Trump referred to El Salvador, Haiti and African nations as ‘shitholes’ last year, he of course completely ignored the pivotal role played by the U.S. and its proxies in driving and maintaining deep poverty and inequality in the south over many decades. The irony and hypocrisy are hard to miss. Even with interventions implemented to limit the spread of Covid-19, there is absolutely no reason whatsoever that the wealthy nations cannot step up and help those suffering disproportionately. The problem is that the will is not there. The corporate elites are too busy plundering resources and capital from the underdeveloped nations to pay anything other than lip service to the pressing problems there. Jason Hickel, Patrick Bond, Tom Athanasiou and other economists have written extensively about this.

  117. Dave_Geologist says:

    Back to the Netherlands Jeffh, It occurred to me when I recalled that there were so many identical variants that there must have been lots infected at one time, probably in a super-spreader event. A quick Google found this as the top hit:

    Dated May but the event was on 8th March, about the right timing. So that could be an element of bad luck: consider the Korean religious sect and nightclub outbreaks in an otherwise compliant country. Of course that doesn’t detract from your assessment of the country more generally, and indeed even if there was broader compliance then, there may be less now as fatigue sets in.

  118. Willard says:

    Dave, none of what you said made any dent on Eric’s point, and if you quote “do these conclusions still hold,” try to read these conclusions before opining on them.

  119. Jeffh says:

    Joshua, Dave, many people here, as elsewhere, downplay Covid-19 by referring to it as ‘simply a slightly worse flu’. The Dutch, as I said, are stubbornly independent and rules are something many here do not like. The English site had a good article about the prevailing mentality here yesterday and why it is not effective during a pandemic. There is also a movement, ‘Viruswaarheid’ or ‘Virustruth’ based in Amsterdam that is gaining traction here, especially in the media. It is vehemently opposed to any even modest measures to tackle the virus. They are closet herd immunity followers. Bear in mind again that there were never any curfews here and the first lockdown was nowhere close to measures that were implemented in Italy, Spain and France. I would say that we fell more on the side of the Swedish strategy, which, by the way, is still heavily criticized by many epidemiologists there.

    Every week there are multiple stories published here of Covid-19 restrictions being broken. Large church gatherings, secret parties with hundreds of people attending and increasingly large protests. Today the country, with just over 17 million inhabitants, recorded almost 7,000 new infections. We expect around 5,000 hospital beds to be occupied by Covid-19 patients in two weeks and, at this rate, twice that in mid-November, by which point the health service will be on the verge of collapse. The virus is running amok and tomorrow the government will issue more restrictions, but still probably not mandate the use of masks. When I told this to German friends and colleagues they were literally gobsmacked. The Swiss and Norwegians wonder what on Earth the Dutch government is thinking. They believe that the government here responds not to the rising number of infections, but only to rising numbers of patients in hospitals and ICU wards. Of course this lags reported new infections by two or three weeks, because that is how long it often takes for the viral symptoms to become serious enough for hospitalization.

    Unrestricted freedom comes at a price, and we are clearly seeing that here. Right now the prognosis is dire.

  120. Willard says:

    > Nabarro was in no way supporting the Barrington Declaration.

    From news dot com dot au’s newsie:

    When asked about the petition, Dr Nabarro had only good things to say. “Really important point by Professor Gupta,” he said.

    Contrary to the article’s spin, that point is not very different than what the WHO holds since the beginnings, e.g.:

    We must all learn to control and manage this virus using the tools we have now, and to make the adjustments to our daily lives that are needed to keep ourselves and each other safe.

    So-called lockdowns enabled many countries to suppress transmission and take the pressure off their health systems.

    But lockdowns are not a long-term solution for any country.

    We do not need to choose between lives and livelihoods, or between health and the economy. That’s a false choice.—21-august-2020

    While News Corp Australia has its spin to sell, people say what they say. It won’t be the first nor the last doctor to have problems dealing with Freedom Fighters megaphones.

    Being able to trace back the list of positions the WHO held over the time along its empirical support should be easier.

  121. August 21 is “the beginning?”

    This is the WHO in June

    “”A targeted approach can avoid an “endless” cycle of lockdowns. “These measures are the best way to suppress and stop transmission, so that when restrictions are lifted, the coronavirus doesn’t resurge,” said the Director General. “The last thing any country needs is to open schools and businesses, *****only to be forced to close them again**** because of a resurgence.””

    IOW: resurgence forces closures of schools and business. nice try.

  122. IOW: Not following best practices (e. g. measures) leads to resurgences. Nice try. :/

    It would appear that good intentions count for nothing if few follow best practices. Best practices meaning lessons learned from around the world, in total.

    Please understand that what one ought to do is not the same thing as what one actually does.

  123. Dave_Geologist says:

    Willard, the conclusions concluded that lockdowns and border controls neither reduced nor increased mortality, but that lockdowns were associated with increased recovery rate. My bad if increased recovery rates meant less mortality as opposed to (say) shorter times in hospital, less severe post-Covid symptoms, whatever. I misread it in that case because they seemed to be saying something different from mortality, which they mentioned earlier. If I’d been the editor I’d have urged them to use mortality all through the paragraph to avoid confusion.

    If they meant what I think they mean, that lockdowns did not reduce mortality (OK another possible source of confusion is do they mean infection fatality rate not deaths per head, but that would be silly because you don’t do that by lockdown, you do that by shielding the vulnerable only), my examples from numerous countries of correlations between mortality reduction and lockdown refutes the conclusions* (and please, no philosophical debate about refutation** – I won’t be drawn into word games but am clarifying possible confusion).

    * That doesn’t necessarily mean the conclusions for that particular dataset are wrong. (Although I would note that absence of evidence is not evidence of absence. Just because a correlation doesn’t show, that doesn’t mean it’s not there. It may be lost in the noise or confounded by a counter-variable and emerge with a larger or longer-duration sample.) Just that they don’t generalise. On one of our favourite topics, that’s why this sort of science develops by consilience, not by replication and Auditing. Each Petri Dish is unique, and the same Petri Dish can behave differently now than six months ago. Consensus emerges from lots of Petri Dishes being consilient.

    ** No debate in the sense that I won’t be drawn into it. I would of course not advocate censorship.

  124. Willard says:

    Fair enough, Dave. I think the following is uncontroversial:

    [L]ow levels of national preparedness in early detection and reporting, limited health care capacity, and population characteristics such as advanced age, obesity and higher unemployment rates were key factors associated with increased viral spread and overall mortality.

    Let’s also assume, as the authors suggest in their discussion, that full lockdowns (whatever that means for the moment) increase recovery rates. Same with having closed down the border early. By contrast, increasing tests were not correlated with less deaths or critical cases.

    (I disregard their discussion about smoking because it looks too noisy, and because it does not seem to affect where I’m going with this.)

    This creates a dilemma that may explain our current pickle. We can’t go back in time and close borders, and we can’t go full lockdown forever. If we want to “focus” our protection, we need to test, trace and track. The evidence how or if that works isn’t quite clear from that paper, and partial lockdowns had less effect. Yet there’s little else we can do.

    However, the idea that institutions ought to fulfill their epistemic duties will also apply if we choose to focus protection. After all, it’s up to them to lead “better protecting those who are at highest risk.” Alluding to costs and benefits might work for a letter, its performative effect don’t last long enough to reassure the essential workers I know, including me. Worse, it may carry more risks than we presume:

    Your [Sam’s] intuition on the efficacy of an airport profiling system is wrong. The psychology of security is complex, and there is a great deal of of research about how our brains systematically get security decisions wrong. This is an example of that. Profiling at airports gives us less security at greater cost.

    We have more time to geek up AGW than Covid-19. The longer restriction rules are in force, the simpler and fairer they need to be. And I don’t know how to pull this one off.

    This sucks to no end.

  125. Willard says:

    > This is the WHO in June

    And here it is in June May:

    Countries put these stringent measures in place, sometimes called lockdowns, in response to intense transmission.

    Many have used the time to ramp up their ability to test, trace, isolate and care for patients, which is the best way to track the virus, slow the spread and take pressure off the health systems.

    The good news is that there has been a great deal of success in slowing the virus and ultimately saving lives.

    However, such strong measures have come at a cost and we recognize the serious socio-economic impact of the lockdowns, which have had a detrimental effect on many people’s lives.

    Therefore, to protect lives and livelihoods, a slow, steady, lifting of lockdowns is key to both stimulating economies, while also keeping a vigilant eye on the virus so that control measures can be quickly implemented if an upswing in cases is identified.—11-may-2020

    Better luck next time.

  126. Jeffh says:

    Gupta predicted in March that up to 50% of the British population may already have Covid-19 antibodies.


    She said in an Unherd interview in May or June that a second wave was unikely.


    A large number of highly qualified epidemiologists have been highly critical of the Barrington Declaration. Nabarro may or may think there is merit in it. Many qualifed experts disagree. I return to my much earlier question: given the prevalence of medical legacies of Covid-19 infection as well as the knowledge that immunity may be short-lived, why is the highly discredited concept of herd immunity with this pathogen still on the table?

  127. Willard says:

    > Many qualifed experts disagree.

    Citation needed.

  128. Jeffh says:

    Come on Willard, you can do better than that. The internet and media are filled with rejoinders. Start with the Sciencemediacenter site – some heavyweights weigh in there. They also do in the NY Times, Mother Jones and elsewhere. Indeed, the staunchest defenders of the Declaration are mostly far right souces like the National Review and National Post. The document includes signatures by ‘Dr. Joe Bananas’ and other parody names. It is the anti-lockdown version of the OISM climate denial petition of 1998.

    I am not personally advocating lockdowns, but the declaration leaves a heck of a lot of fog. How to open up virtually everything while protecting the elderly and vulnerable? The declaration was scant on details. And there is the matter of herd immunity I referred to – the idea that we simply ‘ride it out’ is untenable unless we are willing to accept a huge death toll. The Dutch healthcare system is heading for the brink. Gupta has taken this route out of sheet, stubborn pride. Her previous predictions bit the dust, but this is her personal Alamo it would appear.

  129. Jeffh says:

    The Guardian article by Ian Sample also interviews a number of critics. By now the declaration has been roundly criticized for its lack of detail by many experts. It has found its niche among those on the political right.

  130. Willard says:

    > The internet and media are filled

    You’ve already been proven wrong once this morning, JeffH, and right now you’re using the Q Anon trick.

    Your claims. You back them up.

  131. Dave,

    If they meant what I think they mean, that lockdowns did not reduce mortality

    This was, unfortunately, one of the conclusions from our paper. If all we do is mitigate to try to avoid the waves from growing too large, then it’s very hard to avoid eventually ending up with a large number of deaths. In Report 9, the total number of deaths exceeded >200k deaths in all mitigation scenarios. Of course, there are obvious caveats. A very effective test, trace, isolate programme may well avoid this and there may be updated age distributions for deaths that could also change this result. Also, there is a timescale issue. The overall impact of a single wave with >200k deaths would be much more severe than a similar number of total deaths spread over a much larger time interval (I’m not suggesting we aim for this, just that it may be difficult to avoid). Of course, we may also develop improved clinical treatments and may also get a vaccine, so there are other advantages to spreading things out.

  132. Willard says:

    Just to give an example of how burdening others to find one’s own citations is unfun.

    Searching “The Guardian Ian Sample” leads me there:

    Just for October there are 20 articles. None of them are obviously about the petition. So I need to scan them one by one. And that’s for a source that I can read, I have no access to the NYT.

    Add to that “Gupta predicted in March that up to 50% of the British population may already have Covid-19 antibodies” and “She said in an Unherd interview in May or June that a second wave was unikely.”

    So far my reading list includes:


    All this to double check a mere bandwagon effect.

  133. I should add that the Report 9 analysis also presented some suppression scenarios, which involved case isolation and household quarantine being on all of the time, but having other interventions like social distancing and place closures turning on when ICU demand crosses a threshold, and turning off when they dropped below about 25% of that threshold. This kind of strategy can keep total deaths below 100k (over a 2 year period) but do seem to typically require that these extra interventions are on a lot of the time.

  134. Joshua says:

    Willard –

    > Yet there’s little else we can do.

    I’m not sure what you mean by this. Surely there’s lots we can do. So I’m thinking that your comment there must be more circumscribed than how it reads to me?

  135. Joshua says:

    Anders –

    > Of course, we may also develop improved clinical treatments and may also get a vaccine, so there are other advantages to spreading things out.

    Fatality rates have already dropped very significantly. The implications of that is one thing that the “herd immunity” crowd consistently ignores (I have yet to see one “herd immunity” advocate address the implications of improved treatment and/or the impact of vaccines, and I notice that Eric declined the invitation to address that issue above ).

    X number of infections four months ago resulted in far more deaths than X number of infections now. Of course we don’t know that fatality rates may not trend higher going into the winter – perhaps more people being indoors will lead to more serous infections (via higher viral load) as well as more infections overall, but what are the odds that effect would really be strong enough to fully negate the benefits of better therapeutics and standards of care? I would think pretty low.

    And yes, if a vaccine is developed and distributed on a fairly short time frame, SIPs and NPIs could mean that the # of total infections before a “herd immunity threshold” is reached would never equalize the number of infections reached with a faster infection rate in their absence.

  136. Willard says:


    I think we have little choice than to follow this recipe:

    We continue to urge countries to focus on four essential priorities.

    First, prevent amplifying events.

    Second, protect the vulnerable.

    Third, educate, empower and enable communities to protect themselves and others, using every tool at their disposal.

    And fourth, get the basics right: find, isolate, test and care for cases, and trace and quarantine their contacts.

    This is what works.—25-september-2020

    The dilemma I’m referring to is that we can’t remain at the first step forever. Life would come to a standstill. Yet that’s the step we understand best. Even then there will be deaths. So we need to get to the fourth step, but that’s when we risk more. Nobody from the front line (whose opinions I value more than anyone from any armchair including myself) wants to exacerbate a second wave.

    I think the crucial steps are the second and the third ones. We need the cooperation of most everyone. Yet the more we wait, the more everyone gets impatient and act dumb. The tension is palpable, including in this thread.

    I’m not prescribing anything more specific than that. I think this is a sound plan. In fact I think it’s the only sound plan.

  137. Joshua says:

    Willard –

    In broad stokes I agree.

    But I see so much handwaving about protect the vulnerable that I have become jaded. Coming from the “herd immunity” crowd it reminds me of “let’s get nuklur” from “skeptics.” It’s all fun and games until someone loses an eye (i.e., has to pony up to pay for protecting the vulnerable).

    I watched a video from Gupta et al. at the Great Barrington publicity campaign. They make much sense in theory. But the reality gets complicated and it rubs me the wrong way that there is so much overlap between those advocating a policy of “protect the vulnerable” and those who have for decades attacked the sorts of social safety net policies that would make it possible to protect the vulnerable. What am I to think when soshlist Sweden becomes the darling of rightwing libertarian extremists? That doesn’t make “protect the vulnerable” wrong, of course, but I can’t just overlook the cynical exploitation for the sake of beating an ideological drum of identity warfare.

    So to the extent that SIPs and NPIs come at the expense of action, at the expense of actually protecting the vulnerable, then they seem like a legitimate target to me. But in absense of real world proposals and funding to protect the vulnerable it looks like “skeptics” dressed up in Covid clothing to me. We can walk and chew gum at the same time.

  138. Joshua says:

    Jeffh –

    Thanks for your 4:11 pm comment.

  139. Joshua says:

    I don’t like speaking to “an audience” rather than an interlocutor (as to me doing do suggests bad faith), but since Eric isn’t responding to me (after mistakenly thinking I was enhancing in bad faith)…

    Eric said:

    > Deaths by date of occurrence peaked between July 25th and August 4th and have been steadily declining since then.

    That is inaccurate. Deaths dropped from early July to early August and have remained steady since then.

    That pattern is very much in line with the rate of infection, with a lag for symptoms to worsen and a lag for reporting. The trend line in infections, with the same lag, suggests another six week at the same level of deaths.

  140. Joshua says:


    > “Just look at what’s happened to the tourism industry, for example in the Caribbean or in the Pacific, because people aren’t taking holidays. ”

    Yeah. ‘Cause the tourism industry in the Caribbean would have been peachy keen during a raging pandemic without “lockdowns.”

    Consider in the US: With SIPs we have stimulus, and loans, and extended unemployment checks. Imagine what might have happened without SIPs. Would people have gotten benefits? Would people have been fired if they had comorbidities and didn’t want to go into work in a crowded and poorly ventilated office? Working grandarent primary caregivers who didn’t want to send their grandchildren to school for fear they’d become infectious but couldn’t stay home to take care of them or they’d get fired.

  141. Joshua says:

    Sorry – I meant dropped from early August to early September (Seven day average = 101 on Sept 05) and remained steady since then (=98 on Oct 12).

  142. Willard says:

    Here’s a mesmerizing page:

    If we click on the somewhat hidden link at the bottom of the John Hopkins’ page (see JHU CSSE), we get to the GitHub repository:

    The dashboard is rather neat:

    There is another version for mobile.

  143. Ben McMillan says:

    Once you try to avoid hospitals being overwhelmed, a year or two of interventions on a lot of the time is pretty much what you get anyway (unless you go for elimination). I think Report 9 did a very good job of explaining this which isn’t coming across clearly in the comments here.

    The authors of Report 9 thought that the suppression scenarios would be less awful than the mitigation scenarios (the direct mortality predictions are not the point), and this is a big part of the reason there was a lockdown in the UK.

    I think to go beyond ‘academic’ issues, you have to grapple with the reality that almost everyone has ended up in a suppression scenario.

    (as a side note, on multi-year timescales, ‘everyone gets it once and then it is over’ is very dubious)

  144. Ben,

    Once you try to avoid hospitals being overwhelmed, a year or two of interventions on a lot of the time is pretty much what you get anyway (unless you go for elimination). I think Report 9 did a very good job of explaining this which isn’t coming across clearly in the comments here.

    Yes, I tried to highlight that a few comments ago. This, is, I think an important. Based on the models, you can try to mitigate, which slows but doesn’t stop the spread,m or you try to suppress, which trys to brings numbers to a low level and keeps it there. The former could avoid over-whelming healthcare systems but would probably lead to lots of deaths on a relatively short timescale, while the latter would seem to require – as you say – interventions being in place a lot of the time.

  145. Jeffh says:

    Willard, as a scientist, I take your QAnon comment as a smear. Since when have I intimated that a petition set up by a right wing think tank is a conspiracy? You place a lot of faith in the Great Barrington Declaration; I don’t, and neither do a lot of statured experts. Get over it.

  146. Jeffh says:

    Critics of the GBC include:

    Dr. Michael Head, University of Southampton;
    Dr. Rupert Beale, Francis Crick Ibstitute;
    Dr. Simon Clarke, University of Reading;
    Dr. Simon Stevens, NHS Trust;
    Dr. James Naismith, University of Oxford;
    Dr. Jeremy Rossman, University of Kent;
    Dr. Stephan Griffin, Leeds University;
    Dr. Gregg Gonsalves, Yale University.

    Gonsalves wrote a scathing critique in The Nation. I am on my IPad and cannot link to these but I will later. But again, being accused of peddling QAnon-type responses is poor, Willard. You write as if there is no qualified opposition to the tenets laid out by Gupta et al. There is plenty.

  147. Dave_Geologist says:

    Of course I find that uncontroversial Willard. It’s a major reason for the difference between the US and UK, between Sweden and its Nordic neighbours, between the Netherlands and Germany, between Australia and New Zealand, China, Taiwan and the rest etc. The measures the low-mortality countries took suck too. Some of them, like China’s, would be unacceptable in the West. Even the ones with quarantine and closed borders instituted local lockdowns. And their small outbreaks died out or are dying out.

    With the caveat (or addition) of “other things being equal”. One of the other things that is not equal is highlighted in Fig. 1b of the supplementary material I posted above. The timing of lockdown during exponential growth of the epidemic. Deaths before lockdown should be a good proxy for prevalence a few weeks earlier, and assuming exponential growth, for the existing infections baked in at lockdown which lockdown can’t prevent because it doesn’t un-infect people. As Deputy CMO Van-Tam warned UK viewers yesterday (you know bad news is coming when he’s on, because he’s the bad cop to CMO Whitty’s good cop). Leave it a week late when deaths are doubling each week and you’re committed to double the deaths. Doubling twice a week, four times the deaths. If lockdown doesn’t check virus spread, why the correlation? It’s not a perfect correlation for the reasons you mention, because each Petri Dish is different, and at least in the UK because for a week or two before lockdown Apple and Google Maps data showed we were beginning to self-isolate off our own bat. OTOH because each point represents one country, many of those co-factors were the same before and after lockdown, except in countries where ICUs were overwhelmed.

    But attributing all the differences to those co-factors is rather like attributing all the forest fires in western North America and eastern Australia to poor brush clearance, poor suburban zoning, careless campers and arsonists while denying the impact of a longer, drier fire season and the underlying cause of that.

  148. Dave_Geologist says:

    Ah, thanks ATTP. A third misunderstanding then 😦 . Obviously, if there is no step change in the infection fatality rate or a working, widely deployed vaccine, then sooner or later, however hard we try, the entire population will be infected and as many people will die as would have died in a single massive first wave. With the caveat that the IFR will rocket if ICUs are overwhelmed. There was a kinda tacit acceptance of that in Report 9, with triggers for subsequent lockdowns based on approach to an ICU threshold which meant that by the time the pool of infectees had worked its way through, ICUs would be overwhelmed.

    But those are big “ifs” and those future deaths haven’t happened yet and may be avoidable (not this coming wave obviously, it’s too soon for a vaccine). Nevertheless, we’ve seen improvements getting on for half in the ICU fatality rate, so people will live in December who’d have died in April. Trump might have died in April. (Can the modelled IFR change between waves? – that would seem a obvious knob to twiddle on the “buying time” scenario.) You then get into a discussion about the value of deaths deferred but that is standard NICE QALYs stuff (quality adjusted life years – which would of course also have to include Long Covid in the quality measure and collateral health and economic damage on the other pan of the scales).

    To take an extreme example, even if Covid-19 becomes an annual disease like seasonal flu, Norway would take ten years to reach Sweden’s per capita death toll if they both rinse and repeat spring 2020. And based on seroprevalence, it will take Sweden the best part of decade to get to herd immunity, even if mild infection confers long-term immunity. Do we really think there will be no medical advances in a decade? Or that giving our elderly relatives two, five or ten years extra life has no value? Obviously it also has a cost if we do that by lockdown.

    As it happens I’m with Nabarro in the 4SD piece and the need to look for a middle path, and with his previous comment, also taken out of context by the likes of the Murdoch press, that Sweden was a possible model for handling Covid long-term, once we’d overcome the first wave and had the mitigation measures we should have put in place this summer but didn’t. Coronavirus: WHO head calls herd immunity approach ‘immoral’. No, they’re not back-tracking and saying we should have gone for herd immunity in the first place, nor are they advocating the Barrington/Gupta herd immunity let-it-rip approach (BTW what are Gupta’s qualifications? – she was on BBC News yesterday with a virologist and talking what even I could see was nonsense: that measles was suppressed by infections bringing herd immunity, and that we don’t get sick from common colds because of herd immunity; the virologist pointed out that most common colds are re-infections with the same virus and that vaccines bring measles herd immunity, which is why there was a surge in cases when vaccination levels fell in the wake of the Wakefield scandal).

  149. gwws says:

    Shielding the vulnerable in an epidemic: a numerical approach at addresses some of the issues raised by the Great Barrington Declaration.

  150. Dave_Geologist says:

    The GBD crowd remind me of a number of bosses I was unfortunate enough to have.

    The oil price has fallen and we need maintain profitability. The plan is to cut costs by 10% while maintaining production.

    That’s not a plan it’s an objective. How will we do it?

    Cut costs by 10% while maintaining production.

    Yes, but what’s the plan?

    Cut costs by 10% while maintaining production.

    Rinse and repeat.

  151. Dave_Geologist says:

    In the BBC interview the virologist mentioned a recent Nature Medicine paper which tracked re-infections with the common-cold coronavirus in a population whose serum was sampled every three to six months and tested for antibodies. Re-infection was judged by reported symptoms (within 28 days) and a sharp increase in the same antibody, usually after months or years of decline. Some as short as six months with no decline.

    Reinfections occurred most frequently at 12 months after infection, indicating that protective immunity is only short-lived.

    From which I would speculate that immunity probably lasts less than 12 months, but the chance of re-infection is much higher going into the next cold/flu season than it is in the summer, when there is much less virus about.

    Seasonal coronavirus protective immunity is short-lasting

  152. Dave_Geologist says:

    On a more positive note, a study which cultivated virus from samples collected from individuals’ upper respiratory tracts and correlated it with PCR values and onset of symptoms supported the current guidance of infectiousness being for up to ten days after onset of symptoms (for about 95% of cases). There is evidence from elsewhere of longer infectiousness among seriously ill, hospitalised patients, but I imagine they’re routinely treated as if infectious until cured.

    Duration of infectiousness and correlation with RT-PCR cycle threshold values in cases of COVID-19, England, January to May 2020

    I presume this and similar studies are what Trump’s doctor was relying on when he said the tests showed Trump was not infectious: they established a semi-quantitative threshold below which a positive PCR test was returned from non-infectious individuals due to residual viral RNA.

  153. Willard says:


    If you want to be taken as a scientist, act like one. Scientists cite their sources. Scientists attack arguments using arguments. Scientists don’t cherrypick. Scientists try to remain calm and objective. Scientists don’t smear other scientists. Et cetera.

  154. Willard says:

    Gregg raises an important point:

    There are other differences that would make the Swedish experiment an exercise in folly in America: Swedes have greater faith in their government and were willing to comply with voluntary restrictions; because of their national health system, they would not lose their homes or their livelihoods if they got sick; many Swedes live alone (not in multigenerational households as is common here); Swedes also have lower rates of the comorbidities I described above that put millions and millions of Americans at high risk for complications of Covid-19.

    The argument can be turned on its head, however: Americans don’t comply with voluntary restrictions, and so we should not expect they will; because of a lack of social net, Americans will lose their homes if they stop to work; Americans are at a higher risk of complications, so we should expect their numbers to be worse ceteris paribus.

    Gregg also proposes “nuanced, targeted approaches using data on our local epidemics” and cites an editorial in BMJ that underlines a very important argument:

    Evaluating the implementation of more nuanced strategies will require us to look deeper for evidence as to how and why specific activities helped reduce epidemic spread, for whom, when, and under what conditions. The media has commonly portrayed that the primary factor differentiating “successful” and “unsuccessful” responses is the speed and intensity with which broad-scale policy change has been enacted. We hail leaders, for example, whose responses have been swift and strong while decrying those whose responses have been delayed or less stringent. But this reaction may selectively overlook specific cities, regions, and even countries where there have been disconnects between the level of stringency and incidence rates of COVID-19

    The authors are led to this conclusion because they emphasize:

    Translating fundamental insights from epidemiological models into effective public health practice implies transparency about the conditions on which insights are projected into estimates with the most common condition of COVID-19 models being of a homogenous population with relatively homogenous social networks. However, there are consequences to a top-down mandate of nonspecific responses onto a heterogeneous population. Nonspecific primary prevention approaches are known to increase disparities by further marginalizing those already at highest risk of severe outcomes—including people living in congregate settings and people of disadvantaged communities who may also have poorer existing access to care and higher rates of comorbidities that increase risk of poor COVID-19-related outcomes. As an example, the closure of businesses has had a dramatic economic effect on populations already generally at the margins, with tens of millions of people filing for unemployment in the United States alone and countless more seeing their “gig economy” incomes fall dramatically. There have also been significant disruptions to the health system with currently unmeasured, but potentially substantial, increases in morbidity and mortality associated with diversion of resources away from prevention and treatment of cardiovascular disease, mental illness, acute illnesses, reproductive health, cancer, and other infectious diseases—to name a few. Similarly, disruptions to the educational system at all levels may have downstream effects on the health and well-being of individuals, populations, and the economy .

    Even if we accept and implement the WHO’s recommendations, they’re not enough. We need different strokes for different folks. We need to stop falling for journalistic baits. We also need to fulfill our epistemic duty.

  155. Willard says:

    In other news, here are three billion worth of social nets:

    Figures revealed today appear to show the huge sum is unaccounted for in the torrent of contracts handed out to private firms since April.

    Last month the Department of Health said £11bn of contracts had been agreed between April 1 and September 7 – most of them related to coronavirus.

    Yet analysis of publicly-available contracts data could find less than £8bn of contracts awarded by the government.

    The analysis by Tussell, a data provider on government, found £7.7bn of Covid contracts and £74m of others.

  156. an_older_code says:

    there does seem to be an implied assumption by some commentators that we have binary outcomes – 100% Death (well death tends to be 100%) and 100% Survival

    aka the long term health effects of Covid19 are discounted – especially when looking at the long term economics

    I think the reality is more nuanced – covid19 seems to be quite debilitating to some survivors

  157. Joshua says:

    Willard –

    > The argument can be turned on its head, however: Americans don’t comply with voluntary restrictions, and so we should not expect they will; because of a lack of social net, Americans will lose their homes if they stop to work; Americans are at a higher risk of complications, so we should expect their numbers to be worse ceteris paribus.

    As I see it, Americans’ (in comparison to Swedes) likely resistence to voluntarily complying with restrictions, their lack of faith in public health officials, and their relative lack of compatibility with “collectivist” approaches, along with how those factors align with a less extensive social safety net, are all reasons why comparing cause and effect between policies and outcomes in different those counties is fraught. From the same policies we’d have much worse outcomes than other countries such ass Sweden.

    Of course, given the primacy of outcomes in Sweden to the argument presented in the Great Barrington platform, that is only one category of the fraughtness in the US adopting policies like Sweden’s – in line with one size doesn’t fit all thinking (which a thesis crucial to Gregg’s article, imo).

    Another key category is related to the demographic and structural dissimilarities, such as in Sweden a better baseline health (and fewer cormorbidities on average), less economic economic inequality (given the signal of SES in Covid infections and morbidity/mortality rates), more single-person households, less multi-generational household, less grandparent primary-caregiving of grandchildren, greater ability for a higher % of people to work from home, better access to healthcare, less concern about the costs of healthcare, etc.

    And of course, I go back again to the joke that when Swedes heard they’d have to stay six feet apart, they said “That close?)

    And then the question that sits aside those considerations is what do we make of a lack of a clear signal in how much adopting a policy of less restrictive SIPs and NPIs has advantaged Sweden over other countries with more restrictive policies.

    Yes, more restrictive SIPs and NPIs are associated with negative outcomes. So are less restrictive SIPs and NPIs. Yes, the ground is wet. Does that mean it’s raining?

  158. Willard says:


    Herd immunity proponents who invoke Sweden failed to pay any kind of due diligence:

    Earlier this month, Tegnell told the Financial Times that national lockdowns were like “using a hammer to kill a fly.”

    But he added that contrary to popular belief, Sweden’s goal was never to allow the virus to run rampant until the majority of the population was exposed. Instead, Tegnell said, leaders relied on residents to exercise personal responsibility.

    Their numbers increased for other reasons than deep stratergery.

    Moreoever, the WHO’s latest press conference shot the idea down (h/t Dave):

    At the same time we must remember that this is an uneven pandemic. Countries have responded differently and countries have been affected differently. Almost 70% of all cases reported globally last week were from ten countries and almost half of all cases were from just three countries.

    For every country that’s experiencing an increase there are many others that have successfully prevented or controlled widespread transmission with proven measures. Those measures continue to be our best defence against COVID-19. There has been some discussion recently about the concept of reaching so-called herd immunity by letting the virus spread.

    Herd immunity is a concept used for vaccination in which a population can be protected from a certain virus if a threshold of vaccination is reached. For example herd immunity against measles requires about 95% of the population to be vaccinated. The remaining 5% will be protected by the fact that measles will not spread among those who are vaccinated. For polio the threshold is about 80%.

    In other words herd immunity is achieved by protecting people from a virus, not by exposing them to it. Never in the history of public health has herd immunity been used as a strategy for responding to an outbreak let alone a pandemic. It’s scientifically and ethically problematic.

    First we don’t know enough about immunity to COVID-19. Most people who are infected with the virus that causes COVID-19 develop an immune response within the first few weeks but we don’t know how strong or lasting that immune response is or how it differs for different people. We have some clues but we don’t have the complete picture.

    There have also been some examples of people infected with COVID-19 being infected for a second time. Second, the vast majority of people in most countries remain susceptible to this virus. Seroprevalence surveys suggest that in most countries less than 10% of the population have been infected with the COVID-19 virus.

    Letting the virus circulate unchecked therefore means allowing unnecessary infections, suffering and death. Although older people and those with underlying conditions are most at risk of severe disease and death they’re not the only ones at risk. People of all ages have died.

    Third, we’re only beginning to understand the long-term health impacts among people with COVID-19. I have met with patient groups suffering with what is now being described as long COVID to understand their suffering and needs so we can advance research and rehabilitation.

    Allowing a dangerous virus that we don’t fully understand to run free is simply unethical. It’s not an option but we do have many options. There are many things that countries can do and are doing to control transmission and save lives. It’s not a choice between letting the virus run free and shutting down our societies. This virus transmits mainly between closed contacts and causes outbreaks that can be controlled by implementing targeted measures.

    Prevent amplifying events, protect the vulnerable, empower, educate and engage communities and persist with the same tools that we have been advocating since day one; find, isolate, test and care for cases and trace and quarantine their contacts.—12-october-2020

    I’m not sure why anyone would argue for something that is unethical according to the World Health Organization, but here we are.

  159. Willard says:


  160. David B Benson says:

    Man, 25, has 2nd infection worse than first:

  161. Pingback: Honest brokering | …and Then There's Physics

  162. Jeffh says:

    Willard, imho you were using the same strategy I often see from climate science deniers. They will push a meme and when contradicted demand their opponents provide proof. When it is argued that the peer-reviewed literature is full of studies proving them wrong, they demand to have them presented on a plate in front of them. You merely did the same thing. It is easy to find croticism of the Great Barrington Declaration online but you made it seem as if it is my job to produce it, otherwise it does not exist. Is that it? As is clear, some pretty qualified people have views ranging from mildly critical of the GBD to damning. They are not necessarily exceptions.

    Covid deniers are emulating climate science deniers by pushing petitions. The GBD is an example. The point is that the declaration is more pushing a neoliberal political agenda than it is about science. It was publushed by a right wing think tank. The two major proponents of the GBD appeared on a Fox News interview and even met with senior members of Trump’s pandemic task force recently. They know where the bread of the GBD is buttered.

    You have your opionions on this issue and I have mine. There seems to be little middle ground which is a shame, i am not a proponent of draconian lockdowns but I also vehemently oppose the ‘no restraints’ strategy that Trump and the GBD appear to be advocating. Arguing, as some of them do that Covid-19 is generally no worse than the flu is wrong and dangerous. Here in Holland our hospitals are fast approaching their breaking point. Today, finally, the governmemt intervened. Perhaps the paper discussed on this thread is correct that lockdowns do not save lives. But then again, Britain is a poor example. They locked down much, much too late. Norway has has a very low mortality rate and locked down early. Ditto China, Taiwan, South Korea etc.which intervened. Sweden will look bad compared with Norway whatever happens from here on in.

  163. Jeff,
    I should clarify that the paper isn’t so much saying that lockdowns don’t save lives. It’s more that if all you do is something akin to lockdowns, you’ll eventually probably have a similar total number of deaths. We can, hopefully, develop better strategies.

  164. Willard says:


    My position aligns with the World Health Organization’s. Either yours does too, or it does not. If it does not, and assuming the WHO plays the IPCC’s role regarding Covid, you would be the contrarian here.

    I doubt we really disagree, in fact I think everyone here is mostly in violent agreement. As I see it, the difference between Gregg’s “nuanced, targeted approaches using data on our local epidemics,” Barrington’s “focused protection,” and the WHO’s recommendations is mostly cosmetic.

    I’m asking for citations because I care about evidence more than opinion. I think that readers may profit more from a comment thread with citations than with opinions only. Having citations also helps dispel violent agreements.

  165. Craig Thomas says:

    Our lockdown included “schools are closed”, but you need to be careful what you read into this – my partner works in a school that serves a very disadvantaged area with high rates of unemployment, drug use, crime, missing fathers and neglected children and they have a very strong program for working with problem families and monitoring children at risk. Although the school was “closed”, the teachers ensured that the children who require monitoring continued to attend. I would assume many schools did the same.
    This would be hard to model if you don’t know it’s happening.
    Our end result was for our local population an infection rate of 0.02% and a fatality rate of 0.0006%

  166. Craig,
    Indeed, I think our local primary school was still open for children of critical workers. You’re right that it’s not simple to model this properly, so these models results should be treated with some caution.

  167. verytallguyv says:


    I don’t think the Great Barrington declaration is at all in line with the WHO.

    Particularly, but not exclusively advocating deliberately increasing exposure risks for low risk groups:

    “The most compassionate approach that balances the risks and benefits of reaching herd immunity, is to allow those who are at minimal risk of death to live their lives normally to build up immunity to the virus through natural infection, while better protecting those who are at highest risk. We call this Focused Protection. ”

    And also the whole concept of herd immunity through exposure is regarded a doubtful.

  168. Joshua says:

    Willard –

    > As I see it, the difference between Gregg’s “nuanced, targeted approaches using data on our local epidemics,” Barrington’s “focused protection,” […] is mostly cosmetic.

    I don’t think that either party would agree with you. Characterizing their views in the manner you have srikes me as being condescending.

    The GBers effectively minimize the difficulty of “protecting the vulnerable,” and advocate everyone else getting infected quickly.

    My guess is that Gregg would say that a key component of “protecting the vulnerable” is *minimizing* infections (within reason) among everyone else…

    Of course, there’s a lot of wiggle room in the phrase “within reason” but that’s about as non-cosmetic a difference as you can get.

    And the GBers also spend time minimizing the virulence of the virus, Bhattacharya in particular.

  169. Willard says:

    Joshua, Very Tall,

    “Focused protection” is just the new luckwarmism. A think tank could fool thousands of people, including at least one member of the WHO itself, which Newscorp used to fool Eric and many, many outlets. To me that indicates a lack of substantial difference and the usual exploit of the narcissism of small differences.

    Compare and contrast:

    [GB] to allow those who are at minimal risk of death to live their lives normally to build up immunity to the virus through natural infection, while better protecting those who are at highest risk.

    [WHO] a slow, steady, lifting of lockdowns is key to both stimulating economies, while also keeping a vigilant eye on the virus so that control measures can be quickly implemented if an upswing in cases is identified.

    I challenge you to design a set of policies that would satisfy one and not the other. Unless and until we see that, whether the protagonists disagree or not is immaterial.

    The real doctrinal conflict will happen when the Barrington letter will be used to go against the WHO’s recommendations. There are many ways this could happen. Take how Matt King Coal used the luckwarm brand to stretch the limits of justified disingenuousness beyond any plausible interpretation of Mosh orthodoxy.

  170. Joshua says:

    Willard –

    I left out the WHO in my comment. I think it’s a tall order to characterize “the” WHO with respect to their policy recommendations, and I certainly haven’t put in the work to be able to do so. I was looking at whether the difference between the GBers and Gregg is merely cosmetic.

    As for the WHO and the GBers, I’ll say that if we equate them beyond anything cosmetic we run smack dab into retweets like this one:

    I’ve been following Bhattacharya a bit, mostly w/r/t the Santa Clara study and that group’s follow on work. There is much that’s relevant in his advocacy that goes beyond the narrowly focused point of agreement that we would all like “the vulnerable” to stay uninfected.

    I think this is quite parallel to the argument that “skeptics” make, that they are in agreement with IPC w/r/t whether humans contribute to global warming.

  171. Joshua says:

    Sorry, what got lost in my post was that Judith retweeted Robert’s tweet.

  172. Willard says:

    > I was looking at whether the difference between the GBers and Gregg is merely cosmetic.

    I doubt you did, as the evidence has already been laid out.

  173. Joshua says:

    This is interesting:

    Sweden and the U.S. are unique in their failure to reduce coronavirus mortality rates as the pandemic progressed


    Which references this:


    On September 19, 2020, the US reported a total of 198 589 COVID-19 deaths (60.3/100 000), higher than countries with low and moderate COVID-19 mortality but comparable with high-mortality countries (Table 1). For instance, Australia (low mortality) had 3.3 deaths per 100 000 and Canada (moderate mortality) had 24.6 per 100 000. Conversely, Italy had 59.1 COVID-19 deaths per 100 000; Belgium had 86.8 per 100 000. If the US death rates were comparable to Australia, the US would have had 187 661 fewer COVID-19 deaths (94% of reported deaths), and if comparable with Canada, 117 622 fewer deaths (59%).

    While the US had a lower COVID-19 mortality rate than high-mortality countries during the early spring, after May 10, all 6 high-mortality countries had fewer deaths per 100 000 than the US. For instance, between May 10 and September 19, 2020, Italy’s death rate was 9.1/100 000 while the US’s rate was 36.9/100 000. If the US had comparable death rates with most high-mortality countries beginning May 10, it would have had 44 210 to 104 177 fewer deaths (22%-52%) (Table 1). If the US had comparable death rates beginning June 7, it would have had 28% to 43% fewer reported deaths (as a percentage overall).

    In the 14 countries with all-cause mortality data, the patterns found for COVID-19–specific deaths were similar for excess all-cause mortality (Table 2). In countries with moderate COVID-19 mortality, excess all-cause mortality remained negligible throughout the pandemic. In countries with high COVID-19 mortality, excess all-cause mortality reached as high as 102.1/100 000 in Spain, while in the US it was 71.6/100 000. However, since May 10 and June 7, excess all-cause mortality was higher in the US than in all high-mortality countries (Table 2).


  174. Joshua says:

    Anders –

    There’s some interesting discussion in the Time article re closing schools and infection rates in Sweden.

  175. izen says:

    There is a factor that I think vitiates most of the modelling whether it is a simple statistical pattern matching exercise or a complex agent based system that attempts to match ther way infection spreads by mirroring individual behaviours.
    It is the unstated assumption that ALL response options are available for the government to impose and that the general population will acede or comply with any that are indicated by the best evidence.

    We know what works, from the Nations that succeeded in keeping the number of excess deaths to a few thousand and a fraction of the total average number.
    Fast lockdown to limit further spread as soon as cases are detected, the aggressive and comprehensive testing, tracking, treating and isolating once you have developed a system capable of tracking/tracing and treating/isolating the number of cases and spread rate that is present.
    If a government tries to impose a BAU, ignore the deaths/morbidity and just carry on policy excess deaths becvome a significant percentage higher than the usual average (Brazil, US) and it is likely that sooner or later local governance ort community groups will decide to impose a ‘lockdown’ or curfew and social distancing measures independently.

    I think it is inconceivable that China would or could have followed a ‘let it spread’ policy. Their overt ideological commitment to the well being of the society (however token) makes a direct coercive intervention by the State inevitable. The general population would have been surprised and disappointed if the State did NOT respond in that way. The relationship of citizen to State in China and the welfare system means the individual was much more prepared to accept that level of State intervention, and much better protected from the economic and health impacts of that policy, further increasing acceptance.

    In the same way I can not envision the US applying the China solution, or even the New Zealand variant. The willingness and ability of the National government to apply such Draconian (from a Western POV) measures, and the population to comply is MUCH lower. While the lockdown of regional and international travel has the same economic impact, the value placed on maintaining certain businesses is much higer in some Nations than others. This applies to other economic areas as well.
    A brief look at Flightradar or planefinder reveals there are still large daily movements of people around the globe. Most without pre-flight testing or strong post arrival monitoring. The Australian outbreak demonstrates how risky that may be.

    Governments vary in just what policies they can legitimately try and impose on their population, and what aspects of the Nation they govern are accorded the highest values. Populations vary in how much government coercion they will accept, and what logistical protection there is for their basic welfare if the ‘normal’ operation of society is disrupted. China did not have to make supermarket shelf-stackers ‘essential’ workers, it could weld the doors shut of a apartment building and have food delivered.
    The US has trouble getting reliable mask use.
    Sweden tried to avoid the lockdown and go straight to a test and trace containment policy hoping that the willing compliance of the population and a good welfare safety net would be sufficient. The story that when the Swedish government announced that social distancing required a 2m gap between people, some Swedes (jokingly?) complained this was too close and well inside their personal space in normal circumstances, indicates the social factors within the general population that can alter the impact of the policy a government follows.

    In the UK the first ‘lockdown’ was porous, factories making biscuits and snacks had to gear up for increased demand, along with toilet roll makers. Having eased the lockdown and reopened schools, colleges and work a second wave of infection was inevitable without a competent test and trace system in place. But the lack of effective welfare support beyond some limited furlough and loan schemes meant the general population could not sustain a strong lockdown any more than the government was willing to impose it in the face of strong vested interests that wanted the economic impact of the travel and service industries shutting reversed.

    In the event there is no effective virus, or it mutates and returns semi-annually like seasonal flu, herd immunity will be the only option or outcome. If you can restrict the rate of excess deaths to less than 10% of the monthly average that may be acceptable. But if the spread is very rapid and the death rate high, even though it will predominately affect the old, retired and those with morbidities that probably exclude them from the workforce, around 50% of the population have experienced the death of a parent or grandparent, they will act independently of any government policy. A high death rate is also unsustainable on the way to herd immunity from a logistical consideration. Our funeral services can cope with an average of about 2000 deaths a day, double that and you have to lease refrigeration capacity and have year long waiting lists for service and burial/cremation. I would suspect that a majority of individuals will impose a lockdown as a matter of self-preservation.

    It can also be worth remembering that ‘low risk’ is relative. COVID may much less frequently cause death of persistent damage in the under 30s. But the annual ‘Flu’ also has much lower morbidity in younger age groups. COVID is STILL around an order of magnitude more serious than ‘Flu’ for the young. I think the herd immunity argument is less appealing if framed as exposing the young to something ten times as dangerous as ‘Flu’ or Measles in the interests of protecting the economy and the fast acquisition of a communal protection.

  176. Joshua,
    I should stress that our paper was not claiming that closing schools would have no effect, it was trying to understand why doing so (in the model) led to more deaths compared to a similar scenario without school closures. The answer is essentially that closing schools is very effective so that when the interventions are lifted you can end up with a large susceptible population. If the infection starts spreading again and you’re no longer shielding the vulnerable, you can end up with a large second wave that could result in more deaths than the similar scenario without school closures.

    I think I did put something like this into the paper, but I can’t remember if it survived. I shall have to check.

  177. Joshua says:

    Anders –

    Thanks. I was wondering about how closing schools would lead to more deaths as opposed to no difference. I figured maybe it was because kids would be hanging out more at home where most transmission takes place.

  178. White House informally endorses letting pandemic spread unchecked


    Our so-called plan to do almost nothing is working and has been working since March. Don’t worry, be happy and we’ll get to 2.4 million lost soles ASAP. In fact, we are looking for volunteers to sacrifice themselves for the cause, the older the so-called volunteers the better.

  179. Grim new analyses spotlight just how hard US is failing in pandemic


    The US is actually succeeding in killing off the most people on a per capita basis as long as they continue on with their non-plan plan.

  180. soles should be souls above, kind of felling stupid right about now

  181. Willard says:

    Simple three steps:

    Mask-up, help bars, test and trace.

    Easier said than done, and the gap between saying and doing increases each day.

  182. Willard says:

    From Our Man in Victoria:

  183. Joshua says:

    On Florida –

    Given what I posted above I should note that in Florida, deaths are down quite a bit since then. Also, that Steven’s friend’s forecast lines up with reality fairly well.

  184. Chebyshev says:

    Is herd immunity a “strategy” or an inevitable result of mechanics of disease propagation?

    I am confused.

  185. jacksmith4tx says:

    All herds have a base level of immunity but after we discovered germ theory science leveraged that knowledge and technology to accelerate the process. Science also says if you immediately crush the first appearance of a novel contagious disease it will be technically extinct.

  186. Dave,

    Yes, I saw that (I get an acknowledgement 🙂 ). One of my issues with that is that they don’t seem to have really interograted the parameters themselves. For example, I think one of the parameters is already associated with a distribution. If that distribution represents the range for that parameter, then varying that parameter doesn’t necessarily make sense. They also mix parameters associated with the infection, with parameters associated with the interventions. Varying the former could represent some uncertainty with how the infection might behave, but the latter will depend on what we choose to do, so the range due to the latter more represents how our choices could influence the outcome. So, I’m not entirely sure what’s telling us that we didn’t already know (the outcome is going to depend a lot on what we choose to implement).

  187. Dave_Geologist says:

    Thanks ATTP. I’d only just picked it up from a news briefing so hadn’t seen your name in lights 🙂

    I guess it depends on what you’re trying to do, as with all models. “What is the range of outcomes we’d expect” (varying response as well as epidemic characteristics), vs. “What if we do nothing, what if we do this, what if we do that”. If you’re an NHS planner who can’t second guess what government and the public will do, the first might be appropriate. If you’re a minister or adviser deciding what to do, you need discrete scenarios for the response.

    I do think it was unfortunate that critical models (not just at IC as they were not the only modellers) were being run laboriously on PCs, and that government decisions were being based on systems of differential equations calibrated to a few model runs. Hardware should not have been a limitation – a, dare I say it, superforecaster should have said “Hey, doesn’t the Met Office have supercomputers that could speed this up a thousandfold?”.

    I agree that you can often tell what the outcome of many MonteCarlo runs will look like. It’s in line with what I said in the coding thread: an SME will spot garbage-in, garbage-out because it looks wrong, whereas an expert coder with no subject-matter expertise is vulnerable to GIGO errors or to making assumptions and shortcuts that look fine to a non-expert but which an expert would immediately recognise as silly. Or to wasting time pursuing something that an expert would know is not an issue and can safely be ignored.

    I have had counter-intuitive results, and they’re often a delight, but the first thing I do before believing them is check for mistakes. I remember being taught tricks at school to mentally calculate the rough answer to something, on the basis that if you make a mistake in long arithmetic or in hand-cranked or electronic calculator input, it’s more likely to be a long way out than a little bit out.

  188. Dave,
    Yes, I agree that you can often tell when a result is obviously wrong and that you should always check a counter-intuitive result. Every now and again when I’m working with undergraduates, they’ll spend ages trying to solve some computational issue that I end up fixing in a few minutes. They’re often impressed, but I then point out that I’ve been doing this for a few decades and sometimes it just becomes quite straightforward to work out where the problem must be, or what it probably is.

  189. Bob Loblaw says:

    You mean experience counts, Ken? You should do a blog post on that. You could call it

  190. Dave_Geologist says:

    Reminds me of the time I was helping an intern with her project when her supervisor was in holiday. She was trying to demonstrate that a structure was thin-skinned not thick-skinned as had been assumed. She’d come up with a nice thin-skinned fit and I asked if she’d tried to make a thick-skinned model fit. She said yes, but she couldn’t. I knew it should be possible because the existing interpretation was based on a mechanistic boundary-element model (a simplified finite-element model where the element-boundary properties vary, but the material properties of the element interiors are fixed), not just a geometric one. Harder and more expensive to do, but truer to life.

    You would normally use a tri-shear model for thick-skinned geometric restoration, which as the name implies has three independent knobs to play with. The advantage and disadvantage is that you can fit almost anything. I sat at her desk and had a match within an hour. Once you’ve done something often enough you have an intuition for what works. As well as non-uniqueness, the disadvantage of three independent knobs, which interact in non-obvious ways, is that if you’re inexperienced you flounder around in circles and never get anywhere.

    My colleague later made a thin-skinned BEM fit so the real answer was that we didn’t know which was right. Which was a pity, because they had very different implications for deeper drilling targets we couldn’t image on seismic 😦 There’s a rule going back to the origin of section-balancing, which is what she was doing, that you can prove an unbalanced section is false, but you can never prove a balanced one is true. It’s still valuable because sometimes it’s enough to prove that one hypothesis is false in order to answer the question of the day .

  191. Carl Bergrstrom is talking about Goodhart’s Law and game theory wrt pandemic modeling.

  192. Dave_Geologist says:

    Nothing new under the sun Paul.

    I recall a quote from a public-health official during the 1918-19 pandemic to the effect that he’d be blamed afterwards whatever the outcome. If lots of people died, he’d be blamed for not saving them. If few people died, he’d be blamed for destroying the economy needlessly.

  193. Willard says:

    There’s XKCD 793, and then there’s XKCD 435:

  194. Joshua says:

    Given what I posted above about what I posted above, I should note that the mortality numbers in Florida are up again quite significantly, and given the trend with infections, are likely to go up a good bit more and then maybe plateau. Although, that may not tell us much of anything regarding the effects of interventions.

    That said, this comment from Eric has not aged well:

    > Hospitalizations have been been steadily declining for over 6 weeks and are at about 1/6th the level they were at six weeks ago.

    If he were paying attention suppose he’d show up to accuse me of bad faith because I criticize his argument.

  195. Pingback: 2020: A year in review | …and Then There's Physics

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 )

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.