Some of the biggest mistakes with the pandemic response have been hinted at but they haven't been discussed quite as much or as clearly as they should be.
Big Mistake #1 was that infectious disease experts failed to discover that (most likely) all, fast-spreading, respiratory pandemics are driven by airborne transmission via aerosols, despite having had plenty of time and resources at their disposal to find out about this crucial, elementary, and not-hard-to-determine fact.
Big Mistake #2 was that most of the experts failed to recommended switching to respirators ASAP, even after it was becoming fairly obvious that cloth and surgical masks were ineffective at stopping the variants.
Big Mistake #3 is that respirator use still isn't widely recommended in countries that don't have widespread access to vaccines, even though respirators could be rapidly deployed and be just as (or more) effective than vaccines.
Big Mistake #4 is not holding these experts accountable for their mistakes.
I think that any Covid retrospective should begin with asking why we ignored everything written 10+ years prior to this pandemic. There's seemingly no recognition on that failure by every public official, health or otherwise.
2) Your second point is we could have emulated Taiwan. Outside of islands (NZ and Hawaii did similar), with limited access points and better ability to isolate/test entry and exits, this is likely an impossible ask and I dare say not a serious suggestion. To this point, the document above specifically says, paraphrasing, "screening passengers or closing hubs are not recommended".
Do generally appreciate the thoughts, particularly on better mitigation strategies (keep outside open!) and think globally.
I read the article yesterday (Friday, Mar 12, 2022) and thought it was very good. We (the country) need to look at the grand picture to see whether we can avoid the pitfalls the next time. Thanks.
Polite critique of your NYT piece and overall thesis:
Zeynep: “What if the world had reacted as quickly and aggressively in January 2020 as Taiwan did?”
1) The premise that the US (or any other country) could have had better results had they only followed the Taiwan playback ignores that Taiwan is an island roughly the size of Maryland which is far easier to close down. That's the difference of outcome, as the other variables (tests, contact tracing, masks) were all tried in 100 other countries and failed in 100 other countries. To celebrate Taiwan's success would be like telling the rest of the US they should have been more like Hawaii.
2) Contact tracing done early in the pandemic was built around a faulty understanding of how the virus behaves. The old model relied on 6 foot rule, didn’t know it was airborne, didn’t know it could go through HVAC systems and infect patrons across the rooms of restaurants [1], so it's impossible for it to have actually worked by intention. Additionally the PCR test not being able to detect early infection meant that contacts screened were almost certainly cleared before they were truly confirmed virus free. Considering that contact tracing failed for every other non-island country should give pause on its efficacy.
3) Claims the mass testing halted Covid ignore that all of these island nations actually did very little testing. By April 2020 the US (and pretty much all of Europe) had overtaken South Korea and Taiwan in tests/population. [2] Japan didn't even bother with testing or contact tracing (so we praised their "3 C's instead") [3] and gave generous optimism to their cloth mask usage despite the masks not helping their last bad flu season [4]. Meanwhile testing leaders like Denmark wound up having among the highest case rates in the world.
4) Despite being able to put off the inevitability of Covid, all of these island nations eventually realized “zero covid” was impossible as Hong Kong, New Zealand, and South Korea have recently joined Australia in eventually having the virus run rampant through their populations. I’m not clear how we can cite masks, contact tracing, testing, or Japan’s “3 C’s” as effective ways to halt Covid in 2020 when they aren’t working in 2022.
5) Why single out Taiwan or South Korea as an example of “doing it right” when Sweden had a better outcome? The once pariah of Europe enters year 3 of Covid having the second lowest excess deaths in all of Europe, only beaten by Norway (5% more deaths than usual for Sweden 2020-2021 compared to 2.5% for Norway). Sweden is also lower than South Korea (+8.2%) and Israel (+12.3%) and their approach is at least easy to replicate (and carries the least collateral damage) [5]
Zeynep: “Mass testing could have detected people who were infectious before they even knew they were sick and sometimes those who never had symptoms at all.”
MD: Unfortunately we don’t have tests that work like this. The rapid tests are notoriously bad catching infection before you show symptoms. The PCR tests are notoriously good at finding infections you cleared months earlier, giving not “false positive” but more like “shouldn’t have been positives”. Again, the fact that every non-island failed to halt Covid no matter how many tests they rolled out indicates this isn’t a reliable solution.
Zeynep: “South Korea beat back that potentially catastrophic outbreak, and continued to greatly limit its cases. They had fewer than 1,000 deaths in all of 2020. In the United States, that would translate to fewer than 7,000 deaths from Covid in 2020. Instead, estimates place the number of deaths at more than 375,000.”
MD: South Korea actually had roughly 20,000 excess deaths in 2020, so it’s possible their low testing resulted in that undercount. Additionally the demographics of their elderly population are much different than the US (their elderly are much less obese, hypertensive) which is a key difference in potential outcome.
Zeynep: “When vaccines were developed, rich countries hoarded them”
MD: That is true, but what are some examples of countries which had worse outcomes in 2021 than wealthy highly vaccinated countries? The lowest vaccinated countries all appear to be in Africa which has largely evaded Covid.
Zeynep: “Widespread earlier vaccination could have helped limit the possibility for these variants emerging”
MD: What evidence can you provide that widespread vaccination could have prevented variants from emerging especially considering Delta appeared late 2020 before we rolled out vaccines (indicating even “warp speed” isn’t fast enough)? For over 80 years we have been unable to solve the problem of flu vaccines evolving beyond our vaccines – why would we believe now we finally solved that problem despite every highly vaccinated nation having surges beyond anything previous as 2021 ended? That Omicron was selected because of our high vaccination rates seems like a more plausible hypothesis than one where 100% vaccination rates could somehow prevent variants (especially given the issue of animal reservoirs).
[5] See mortality.org, compare average deaths in 2020-2021 to historical 4 year average, note that many countries lag reporting 2021 data (Canada, Australia, etc)
That’s based on projections, but now we have actuals. The source of the OWID charts you show is mortality.org, and if you run the data, you’ll see that Sweden is below SK, below Israel, equal to Finland, and above Norway as I noted.
I'm not a statistician, so I'm not going to bother trying to interpret the raw data. If you think OWID is doing something wrong, you might want to talk to them about it.
I figured it out, in September 2021 OWID switched from using a 5 year historical average to a linear regressed forecast to make predictions for expected 2020 deaths.
For a country like Sweden, it makes a huge difference, as they had their lowest number of deaths in history in 2019 (85,936). Going back 10, 20 years, they had always had 89K deaths +/- 1,000. The blip of 2019 dropping so much skews the prediction for 2020 lower than it should be.
Using the linear forecast put expected deaths in Sweden at 85,850 in 2020 (and 2021) instead of back to 89K like you would expect.
Column J shows using forecast to predict 2020 deaths, column K shows using 5 year average (I personally use 4 year average).
Using their new method, Sweden would have ~2500 excess deaths in 2021 when it should be zero (you can follow row 22 and see if you think if it makes sense to predict 2021 having 89K deaths or 86K deaths).
But maybe that 2019 dip is okay, even though it may mess up the 89K trend. After all, trends can end.
I've also noticed that there's a significant decrease in deaths for 2019-2021 compared to 2018 (I haven't looked at other years) in the 65-74 age group but a significant increase in deaths in the 75+ age groups in 2020.
On the other hand, maybe all of this can be countered by demographic arguments but you'd have to apply them to the other countries you're talking about and make sure to adjust for factors such as age, health, population size, etcetera.
Before you know it, you're overwhelmed by the complexity. That's why I prefer to keep it simple and stick with excess death stats (adjusted by pop size if necessary) and maybe broken down by age group.
The 2019 dip is fine, for such small populations having variation like that is normal, just have to make sure if using trendline to predict anything you keep eye on dips as they can overly impact trajectory.
Example, Wisconsin usually has 51K deaths a year, but in 2015 they had a low of 46,100 for whatever reason. Using linear regression forecast would then assume 2016, 2017 should see 46,000 deaths as well. In reality they bounced right back to 51,184 (2016) followed by 52,399 (2017). If we were scrutinizing excess deaths we might have been alarmed and proclaimed 12,000 more people died in Wisconsin over 2 years than predicted using that methodology.
Both approaches of course have limitations, no denying that.
Agree on simplicity, while I keep pop density, obesity rates, and age groups in mind, I typically just stick to high level views unless something is really standing out.
You said: "I've also noticed that there's a significant decrease in deaths for 2019-2021 compared to 2018 (I haven't looked at other years) in the 65-74 age group but a significant increase in deaths in the 75+ age groups in 2020."
Was that an observation specifically on Sweden or on multiple countries as a whole? I hadn't noticed it on the latter, though I was aware that Sweden had no excess deaths in the <65 category throughout the pandemic. I hadn't run for the other countries though as many are still lagging in their 2021 data (Sweden has a good report here that is updated faster than mortality.org:)
So good to hear from you, hope you are doing well!
I thought that piece was just excellent, even though I have some skepticism about comparing other countries' successful strategies to ours. I have a friend who is back in the states (Minn-SP) after being 'stuck' in Japan for two years, so spent her pandemic there. She was amazed at the efficiency of their contact tracing, and good for them. But how far back would we have had to go for that to work? We know contact tracing weakens as a tool once you are at the point of community spread, so that means before presence in NYC. Probably before even the Diamond Princess.
So here's my hypothesis: maybe the countries that suffered most to SARS Co-V2 suffered because, as developed, open and mobile societies, they were particularly hospitable to the virus. Restaurants, offices, conventions, travel-hubs: these are a few of SARS's favorite things.
And no, I don't consider Taiwan, S. Korea or Japan to be 'open' countries, certainly not in the way Italy, the US, the UK, or even Brazil are.
Best of all in your article was the indisputable notion that this could have been shut down in Wuhan had they acted correctly. I'm not sure what to do about that. I AM sure what NOT to do, and that's blame random Asian-looking people.
I'm still frustrated by how badly the CDC and FDA did on making testing available early. It was clear very early on from South Korean data that if you had one death that you couldn't trace to an infection path, there were ~100 people wandering around infected, symptomatic or not. The FDA and CDC were hung up on approving the tests as medical diagnostics, rather than as public health tools, and didn't really seem to do the math on being able to accept lower reliability tests for screening and brief isolation as a tool to prevent spread. On top of that, the CDC just completely failed on ramping up test availability, which further led to rationing of the tests that we did have. And then we didn't learn - we essentially failed the next pandemic already when Omicron rolled around just before the holidays in 2021 and the US couldn't supply enough tests for people to test and isolate themselves, so people just went ahead and met together and spread it.
We also could and should have been sequencing at much higher rates to detect variants. The US probably has more RT-PCR and more sequencing capability per capita than the rest of the world, but it's all tied to academic and corporate research labs and there's generally no mechanism to bring it to bear for public health purposes in an emergency. The state of Washington did finally start to do that on its own, and California seemed like it might but didn't as far as I can tell.
When a new pandemic plan gets written, it needs to have a few features:
- Demonstrable ability to ramp up production and distribution of testing capabilities at the scale and speed of South Korea.
- Same as above but for basic PPE that are used for universal precautions. There's a lot of logistics involved to make sure it all gets cycled out of stocks and into use rather than stored and rotting.
- Prepared plans for how to use university research lab capabilities at least for population surveillance, and how medical systems will get samples to them and data back to public health systems quickly.
- Identified production capacity for vaccines that the government can redirect/reconfigure, probably by maintaining production of vaccines for other diseases at levels below capacity limits. This might even be done by using the capacity for vaccines for "orphan" diseases that the private sector has decided aren't economically valuable enough to produce vaccines for.
I am struck by how futile these “this is how we should prepare for what happens next” pieces are. It is almost all theoretical, not connected to action, and time and again not leading to action. Maybe we should write about things that could lead to action?
A phrase comes to mind: " . . . an exercise in futility . . . " I don't mind exercising in futility as long as I can keep in mind that whatever muscle I am exercising will eventually become stronger until the exercise is no longer futile but successful.
I wish to express my thanks to you for two years of good information provided in high profile forums such as the New York Times and the Atlantic which has had a real chance of reaching decision makers in a timely manner.
Unfortunately, you can bring the horse to water but you can't make it drink. The political situation in the US has made not only for some bad responses but some that were willfully bad with for example Fox pushing what was essentially anti-countermeasure propaganda while their own staff were masked and vaccinated, and plenty similar behavior from de Santis, Abbott and others.
But there were a lot of well-meaning but mistaken responses, such as dismissing the role of airborne transmission while cracking down on outdoor activities, and for those willing to listen and to .make course corrections, the information you and others like you provided was invaluable and I'm sure in some cases lifesaving.
March of Folly is a great book and so relevant now. The long list of blunders societies arrived at, despite clear and mounting evidence is sobering. Now Covid too.
But one doubt. Even countries that seemed to be smart is beginning to feel rattled, like New Zealand and Taiwan. Are the leaders who messed up royally earlier, like Trump, Modi, Bolsanaro et al, going to use that to paint over their crime?
I'm confused. Is this an announcement of an intention to write a counterfactual history? Notice that the history has been written but then how to access it?
If the history is yet to be written I hope it will deal with really specific counterfactual? Why were CDC/FDA hostile to massive asymptomatic testing? Why were HCT not used to speed up vaccine development (and why were no updated boosters made available for Delta and Omicron)? Why did DCD not provide information and a methodology for how information should be gathered and used by policy makers for making changing nuanced; data, time, and place sensitive cost effective policies about NPI? Why was the messaging about vaccines almost exclusively about safety (almost 'doth protest too much") for the vaccinated person and not on the benefits to the person ad the people that the vaccinated person would not infect?
My brother, as a manager of industrial processes that could go wrong or be dangerous, advocated finding the "Root Cause." To me this is separate from "Why." The Why in much of what we did was the individuals in place and the byzantine expectations we have of underfunded Federal and other level processes. We also use Why to ask why the pickup truck hit the golf team a few days ago, or why one person shot another. Often we'll never know why a one-off thing happens. But if I broaden the focus from Why to Root Cause, for me it can take away the blame and make it possible to study how to make it better. Of course on social questions we first need to acknowledge that the Root Cause is what we want government to do and whether we are ready to pay for it. My brother's oil company had already answered those questions.
In regard to these failures, I'd like to have seen journalists ask the decisionmakers what their model was. What data into what model led to the hostility to early massive screening tests? That in turn might lead to deeper questions about why that model and not some other model was used, why that data and not some other data. etc.
By "DFA" do you mean the FDA? As Trump said, if you test more then you have more cases. That would mean you might have to do something like lockdowns or quarantines that slow down the the full-time pursuit of profit.
Yes FDA. Let's stipulate that Trump was bad news, but most of the errors DFA and CDC made were own goals.
Trump did not tell FDA not to approve quick and dirty screening tests (which enable less drastic lockdowns/quarantines and the "full time pursuit of profit, as if there is anything wrong with that). He did not hype the safety issue of vaccines over their effectiveness at protecting the unvaccinated. He was not responsible for the misconception of NPI's as self protection. He did not veto human challenge trials for vaccine development. He was not responsible for closing parks and closing venues according to activity (in particular schools) rather than according to conditions contributing to spread.
I sometimes think that there is a Karma that produces one vaccine resister for every outdoor mask wearer.
I agree Trump was bad news and that he didn't order any of those things. But, he and his advisers did appoint many of the public health officials on the front lines and the attitude of the boss can easily permeate the ranks of those who report to him, directly or indirectly.
The whole approach to the virus has been a clusterfuck since the beginning and no one in power has done much to change that.
I disagree that there is nothing wrong with the full-time pursuit of profit when people are dying because of it. Besides, the massive increases in profits we have seen in so many industries is evidence that they could have given up a few weeks of profit to protect people where transmission risk was/is high.
Be careful in analyzing the conditions contributing to spread. It took over a year for the WHO and CDC to accept the fact that the main avenue of transmission was aerosols. Governments and institutions have spent untold amounts on hygiene theater and buying plexiglass dividers that just make the situation worse by inhibiting ventilation.
In addition, asymptomatic people and those who have been vaccinated are still able to infect others. The vaccines were never meant to be sterilizing or reduce transmission, regardless of the narrative we have been fed ceaselessly. They only reduce the risk of hospitalization and death, which is fine as far as it goes, but it's not what the government has been pushing.
"It took over a year for the WHO and CDC to accept the fact that the main avenue of transmission was aerosols." And they were not appointed by Trump. :)
With no reliable therapy and the IFR of SARS-CoV-2, human challenge trials wouldn't have been ethical. If paxlovid or other therapies turn out to be highly effective, it might be possible to do challenge trials for future new strains.
Thank you. I enjoyed several Tuchman books growing up, but later learned they might be not the best sources. Calling this one out-of-date is kind. Thank you so much for pointing me to a better source. I was looking for a readable but more "up-to-date " book.
Bill, thank you so much for reminding me of Sleepwalkers.
What an amazing writer Mr. Clark is. Exact, concise, energetic, engaging. the book would probably have been 1400 pages instead of 700 if written by anyone else, and almost certainly less readable.
In chapter one he gets to something I was trying to explain in an earlier response, the difference between asking "Why" vs. "How." If we think of the Why (in the context of what living beings do), we get a one-off answer. If we ask How we can get at personal or social forces that we can perhaps set up so that we do better at some time in the future.
Anyway, I am fully engaged in the book and I thank you very much.
Thanks, Bill. I do know of this book and was not quite sure I was up to 700 pages. HOwever it is quite reasonable on Kindle so I will tackle it. Also couldn't find the Mombauer contribution. I clicked on your link and it just took me to a very large number of reviews.
Yes, I had the same problem with finding her reviews in the Amazon reviews. The Amazon reviews are not useful in that aspect. If you access to scholarly journals then
there is
Guilt or Responsibility? The Hundred-Year Debate on the Origins of World War I
Mombauer, Annika.
Central European History; Cambridge Vol. 48, Iss. 4, (Dec 2015): 541-564. DOI:10.1017/S0008938915001144
Thank you. I would love to have a transcript of Nancy Messonier's radio broadcast. That was the most effective and informative commuication got from anywhere at that time.
Some of the biggest mistakes with the pandemic response have been hinted at but they haven't been discussed quite as much or as clearly as they should be.
Big Mistake #1 was that infectious disease experts failed to discover that (most likely) all, fast-spreading, respiratory pandemics are driven by airborne transmission via aerosols, despite having had plenty of time and resources at their disposal to find out about this crucial, elementary, and not-hard-to-determine fact.
Big Mistake #2 was that most of the experts failed to recommended switching to respirators ASAP, even after it was becoming fairly obvious that cloth and surgical masks were ineffective at stopping the variants.
Big Mistake #3 is that respirator use still isn't widely recommended in countries that don't have widespread access to vaccines, even though respirators could be rapidly deployed and be just as (or more) effective than vaccines.
Big Mistake #4 is not holding these experts accountable for their mistakes.
Hi Zeynep,
1) I did not see any reference to this, or other similar if exist: https://www.aier.org/wp-content/uploads/2020/05/10.1.1.552.1109.pdf
I think that any Covid retrospective should begin with asking why we ignored everything written 10+ years prior to this pandemic. There's seemingly no recognition on that failure by every public official, health or otherwise.
2) Your second point is we could have emulated Taiwan. Outside of islands (NZ and Hawaii did similar), with limited access points and better ability to isolate/test entry and exits, this is likely an impossible ask and I dare say not a serious suggestion. To this point, the document above specifically says, paraphrasing, "screening passengers or closing hubs are not recommended".
Do generally appreciate the thoughts, particularly on better mitigation strategies (keep outside open!) and think globally.
Thanks,
Otto
I read the article yesterday (Friday, Mar 12, 2022) and thought it was very good. We (the country) need to look at the grand picture to see whether we can avoid the pitfalls the next time. Thanks.
Oops. March 11.
Polite critique of your NYT piece and overall thesis:
Zeynep: “What if the world had reacted as quickly and aggressively in January 2020 as Taiwan did?”
1) The premise that the US (or any other country) could have had better results had they only followed the Taiwan playback ignores that Taiwan is an island roughly the size of Maryland which is far easier to close down. That's the difference of outcome, as the other variables (tests, contact tracing, masks) were all tried in 100 other countries and failed in 100 other countries. To celebrate Taiwan's success would be like telling the rest of the US they should have been more like Hawaii.
2) Contact tracing done early in the pandemic was built around a faulty understanding of how the virus behaves. The old model relied on 6 foot rule, didn’t know it was airborne, didn’t know it could go through HVAC systems and infect patrons across the rooms of restaurants [1], so it's impossible for it to have actually worked by intention. Additionally the PCR test not being able to detect early infection meant that contacts screened were almost certainly cleared before they were truly confirmed virus free. Considering that contact tracing failed for every other non-island country should give pause on its efficacy.
3) Claims the mass testing halted Covid ignore that all of these island nations actually did very little testing. By April 2020 the US (and pretty much all of Europe) had overtaken South Korea and Taiwan in tests/population. [2] Japan didn't even bother with testing or contact tracing (so we praised their "3 C's instead") [3] and gave generous optimism to their cloth mask usage despite the masks not helping their last bad flu season [4]. Meanwhile testing leaders like Denmark wound up having among the highest case rates in the world.
4) Despite being able to put off the inevitability of Covid, all of these island nations eventually realized “zero covid” was impossible as Hong Kong, New Zealand, and South Korea have recently joined Australia in eventually having the virus run rampant through their populations. I’m not clear how we can cite masks, contact tracing, testing, or Japan’s “3 C’s” as effective ways to halt Covid in 2020 when they aren’t working in 2022.
5) Why single out Taiwan or South Korea as an example of “doing it right” when Sweden had a better outcome? The once pariah of Europe enters year 3 of Covid having the second lowest excess deaths in all of Europe, only beaten by Norway (5% more deaths than usual for Sweden 2020-2021 compared to 2.5% for Norway). Sweden is also lower than South Korea (+8.2%) and Israel (+12.3%) and their approach is at least easy to replicate (and carries the least collateral damage) [5]
Zeynep: “Mass testing could have detected people who were infectious before they even knew they were sick and sometimes those who never had symptoms at all.”
MD: Unfortunately we don’t have tests that work like this. The rapid tests are notoriously bad catching infection before you show symptoms. The PCR tests are notoriously good at finding infections you cleared months earlier, giving not “false positive” but more like “shouldn’t have been positives”. Again, the fact that every non-island failed to halt Covid no matter how many tests they rolled out indicates this isn’t a reliable solution.
Zeynep: “South Korea beat back that potentially catastrophic outbreak, and continued to greatly limit its cases. They had fewer than 1,000 deaths in all of 2020. In the United States, that would translate to fewer than 7,000 deaths from Covid in 2020. Instead, estimates place the number of deaths at more than 375,000.”
MD: South Korea actually had roughly 20,000 excess deaths in 2020, so it’s possible their low testing resulted in that undercount. Additionally the demographics of their elderly population are much different than the US (their elderly are much less obese, hypertensive) which is a key difference in potential outcome.
Zeynep: “When vaccines were developed, rich countries hoarded them”
MD: That is true, but what are some examples of countries which had worse outcomes in 2021 than wealthy highly vaccinated countries? The lowest vaccinated countries all appear to be in Africa which has largely evaded Covid.
Zeynep: “Widespread earlier vaccination could have helped limit the possibility for these variants emerging”
MD: What evidence can you provide that widespread vaccination could have prevented variants from emerging especially considering Delta appeared late 2020 before we rolled out vaccines (indicating even “warp speed” isn’t fast enough)? For over 80 years we have been unable to solve the problem of flu vaccines evolving beyond our vaccines – why would we believe now we finally solved that problem despite every highly vaccinated nation having surges beyond anything previous as 2021 ended? That Omicron was selected because of our high vaccination rates seems like a more plausible hypothesis than one where 100% vaccination rates could somehow prevent variants (especially given the issue of animal reservoirs).
[1] https://www.washingtonpost.com/food/2020/12/11/korean-restaurant-coronavirus-airflow-study/
[2] Consider comparing tests per 1,000 people for South Korea, United States, and Taiwan: https://ourworldindata.org/coronavirus/country/south-korea#how-many-tests-are-performed-each-day
https://ourworldindata.org/coronavirus/country/united-states#how-many-tests-are-performed-each-day
https://ourworldindata.org/coronavirus/country/taiwan#how-many-tests-are-performed-each-day
[3] https://www.vanityfair.com/news/2020/05/masks-covid-19-infections-would-plummet-new-study-says
[4] https://www.upi.com/Top_News/World-News/2019/02/01/Millions-in-Japan-affected-as-flu-outbreak-grips-country/9191549043797/
[5] See mortality.org, compare average deaths in 2020-2021 to historical 4 year average, note that many countries lag reporting 2021 data (Canada, Australia, etc)
Except for Israel, Finland, Norway, and South Korea have significantly less excess deaths than Sweden and Taiwan and Japan have none.
https://ourworldindata.org/explorers/coronavirus-data-explorer?zoomToSelection=true&time=2020-03-01..latest&facet=none&pickerSort=asc&pickerMetric=location&Metric=Excess+mortality+%28count%29&Interval=Cumulative&Relative+to+Population=true&Color+by+test+positivity=false&country=SWE~NOR~FIN~KOR~JPN~TWN~ISR
That’s based on projections, but now we have actuals. The source of the OWID charts you show is mortality.org, and if you run the data, you’ll see that Sweden is below SK, below Israel, equal to Finland, and above Norway as I noted.
As far as I can tell, OWID is using the most recent HMD data (https://www.mortality.org/Public/STMF/Outputs/stmf.xlsx). So, what OWID is showing should be correct.
I'm not a statistician, so I'm not going to bother trying to interpret the raw data. If you think OWID is doing something wrong, you might want to talk to them about it.
I figured it out, in September 2021 OWID switched from using a 5 year historical average to a linear regressed forecast to make predictions for expected 2020 deaths.
I put this together so you can see the difference in outcome: https://imgur.com/a/OjgmUEm
For a country like Sweden, it makes a huge difference, as they had their lowest number of deaths in history in 2019 (85,936). Going back 10, 20 years, they had always had 89K deaths +/- 1,000. The blip of 2019 dropping so much skews the prediction for 2020 lower than it should be.
Using the linear forecast put expected deaths in Sweden at 85,850 in 2020 (and 2021) instead of back to 89K like you would expect.
Column J shows using forecast to predict 2020 deaths, column K shows using 5 year average (I personally use 4 year average).
Using their new method, Sweden would have ~2500 excess deaths in 2021 when it should be zero (you can follow row 22 and see if you think if it makes sense to predict 2021 having 89K deaths or 86K deaths).
But maybe that 2019 dip is okay, even though it may mess up the 89K trend. After all, trends can end.
I've also noticed that there's a significant decrease in deaths for 2019-2021 compared to 2018 (I haven't looked at other years) in the 65-74 age group but a significant increase in deaths in the 75+ age groups in 2020.
On the other hand, maybe all of this can be countered by demographic arguments but you'd have to apply them to the other countries you're talking about and make sure to adjust for factors such as age, health, population size, etcetera.
Before you know it, you're overwhelmed by the complexity. That's why I prefer to keep it simple and stick with excess death stats (adjusted by pop size if necessary) and maybe broken down by age group.
The 2019 dip is fine, for such small populations having variation like that is normal, just have to make sure if using trendline to predict anything you keep eye on dips as they can overly impact trajectory.
Example, Wisconsin usually has 51K deaths a year, but in 2015 they had a low of 46,100 for whatever reason. Using linear regression forecast would then assume 2016, 2017 should see 46,000 deaths as well. In reality they bounced right back to 51,184 (2016) followed by 52,399 (2017). If we were scrutinizing excess deaths we might have been alarmed and proclaimed 12,000 more people died in Wisconsin over 2 years than predicted using that methodology.
Both approaches of course have limitations, no denying that.
Agree on simplicity, while I keep pop density, obesity rates, and age groups in mind, I typically just stick to high level views unless something is really standing out.
You said: "I've also noticed that there's a significant decrease in deaths for 2019-2021 compared to 2018 (I haven't looked at other years) in the 65-74 age group but a significant increase in deaths in the 75+ age groups in 2020."
Was that an observation specifically on Sweden or on multiple countries as a whole? I hadn't noticed it on the latter, though I was aware that Sweden had no excess deaths in the <65 category throughout the pandemic. I hadn't run for the other countries though as many are still lagging in their 2021 data (Sweden has a good report here that is updated faster than mortality.org:)
https://www.scb.se/en/finding-statistics/statistics-by-subject-area/population/population-composition/population-statistics/pong/tables-and-graphs/preliminary-statistics-on-deaths/
Note that link is a direct XLS download if you click on it.
So good to hear from you, hope you are doing well!
I thought that piece was just excellent, even though I have some skepticism about comparing other countries' successful strategies to ours. I have a friend who is back in the states (Minn-SP) after being 'stuck' in Japan for two years, so spent her pandemic there. She was amazed at the efficiency of their contact tracing, and good for them. But how far back would we have had to go for that to work? We know contact tracing weakens as a tool once you are at the point of community spread, so that means before presence in NYC. Probably before even the Diamond Princess.
So here's my hypothesis: maybe the countries that suffered most to SARS Co-V2 suffered because, as developed, open and mobile societies, they were particularly hospitable to the virus. Restaurants, offices, conventions, travel-hubs: these are a few of SARS's favorite things.
And no, I don't consider Taiwan, S. Korea or Japan to be 'open' countries, certainly not in the way Italy, the US, the UK, or even Brazil are.
Best of all in your article was the indisputable notion that this could have been shut down in Wuhan had they acted correctly. I'm not sure what to do about that. I AM sure what NOT to do, and that's blame random Asian-looking people.
I'm still frustrated by how badly the CDC and FDA did on making testing available early. It was clear very early on from South Korean data that if you had one death that you couldn't trace to an infection path, there were ~100 people wandering around infected, symptomatic or not. The FDA and CDC were hung up on approving the tests as medical diagnostics, rather than as public health tools, and didn't really seem to do the math on being able to accept lower reliability tests for screening and brief isolation as a tool to prevent spread. On top of that, the CDC just completely failed on ramping up test availability, which further led to rationing of the tests that we did have. And then we didn't learn - we essentially failed the next pandemic already when Omicron rolled around just before the holidays in 2021 and the US couldn't supply enough tests for people to test and isolate themselves, so people just went ahead and met together and spread it.
We also could and should have been sequencing at much higher rates to detect variants. The US probably has more RT-PCR and more sequencing capability per capita than the rest of the world, but it's all tied to academic and corporate research labs and there's generally no mechanism to bring it to bear for public health purposes in an emergency. The state of Washington did finally start to do that on its own, and California seemed like it might but didn't as far as I can tell.
When a new pandemic plan gets written, it needs to have a few features:
- Demonstrable ability to ramp up production and distribution of testing capabilities at the scale and speed of South Korea.
- Same as above but for basic PPE that are used for universal precautions. There's a lot of logistics involved to make sure it all gets cycled out of stocks and into use rather than stored and rotting.
- Prepared plans for how to use university research lab capabilities at least for population surveillance, and how medical systems will get samples to them and data back to public health systems quickly.
- Identified production capacity for vaccines that the government can redirect/reconfigure, probably by maintaining production of vaccines for other diseases at levels below capacity limits. This might even be done by using the capacity for vaccines for "orphan" diseases that the private sector has decided aren't economically valuable enough to produce vaccines for.
I am struck by how futile these “this is how we should prepare for what happens next” pieces are. It is almost all theoretical, not connected to action, and time and again not leading to action. Maybe we should write about things that could lead to action?
First you have to define what went wrong, and what could have been done about it, and when it needed to be done. High-value stuff!
Next come the action steps.
A phrase comes to mind: " . . . an exercise in futility . . . " I don't mind exercising in futility as long as I can keep in mind that whatever muscle I am exercising will eventually become stronger until the exercise is no longer futile but successful.
I wish to express my thanks to you for two years of good information provided in high profile forums such as the New York Times and the Atlantic which has had a real chance of reaching decision makers in a timely manner.
Unfortunately, you can bring the horse to water but you can't make it drink. The political situation in the US has made not only for some bad responses but some that were willfully bad with for example Fox pushing what was essentially anti-countermeasure propaganda while their own staff were masked and vaccinated, and plenty similar behavior from de Santis, Abbott and others.
But there were a lot of well-meaning but mistaken responses, such as dismissing the role of airborne transmission while cracking down on outdoor activities, and for those willing to listen and to .make course corrections, the information you and others like you provided was invaluable and I'm sure in some cases lifesaving.
March of Folly is a great book and so relevant now. The long list of blunders societies arrived at, despite clear and mounting evidence is sobering. Now Covid too.
But one doubt. Even countries that seemed to be smart is beginning to feel rattled, like New Zealand and Taiwan. Are the leaders who messed up royally earlier, like Trump, Modi, Bolsanaro et al, going to use that to paint over their crime?
I'm confused. Is this an announcement of an intention to write a counterfactual history? Notice that the history has been written but then how to access it?
If the history is yet to be written I hope it will deal with really specific counterfactual? Why were CDC/FDA hostile to massive asymptomatic testing? Why were HCT not used to speed up vaccine development (and why were no updated boosters made available for Delta and Omicron)? Why did DCD not provide information and a methodology for how information should be gathered and used by policy makers for making changing nuanced; data, time, and place sensitive cost effective policies about NPI? Why was the messaging about vaccines almost exclusively about safety (almost 'doth protest too much") for the vaccinated person and not on the benefits to the person ad the people that the vaccinated person would not infect?
My brother, as a manager of industrial processes that could go wrong or be dangerous, advocated finding the "Root Cause." To me this is separate from "Why." The Why in much of what we did was the individuals in place and the byzantine expectations we have of underfunded Federal and other level processes. We also use Why to ask why the pickup truck hit the golf team a few days ago, or why one person shot another. Often we'll never know why a one-off thing happens. But if I broaden the focus from Why to Root Cause, for me it can take away the blame and make it possible to study how to make it better. Of course on social questions we first need to acknowledge that the Root Cause is what we want government to do and whether we are ready to pay for it. My brother's oil company had already answered those questions.
In regard to these failures, I'd like to have seen journalists ask the decisionmakers what their model was. What data into what model led to the hostility to early massive screening tests? That in turn might lead to deeper questions about why that model and not some other model was used, why that data and not some other data. etc.
There are a lot of good answers to your questions, depending on who, what, and where; incompetence and indifference being a big part of those answers.
However, IMNSHO, the overarching reason for the way this all went down, is best encapsulated in the two rules of Neoliberalism:
1. Because markets
2. Go die.
(h/t Lambert Strether, https://www.nakedcapitalism.com/)
??? How does "becasue markets" explain DFA hostility to mass asymptomatic texting? non-use of HCT?
By "DFA" do you mean the FDA? As Trump said, if you test more then you have more cases. That would mean you might have to do something like lockdowns or quarantines that slow down the the full-time pursuit of profit.
Yes FDA. Let's stipulate that Trump was bad news, but most of the errors DFA and CDC made were own goals.
Trump did not tell FDA not to approve quick and dirty screening tests (which enable less drastic lockdowns/quarantines and the "full time pursuit of profit, as if there is anything wrong with that). He did not hype the safety issue of vaccines over their effectiveness at protecting the unvaccinated. He was not responsible for the misconception of NPI's as self protection. He did not veto human challenge trials for vaccine development. He was not responsible for closing parks and closing venues according to activity (in particular schools) rather than according to conditions contributing to spread.
I sometimes think that there is a Karma that produces one vaccine resister for every outdoor mask wearer.
I agree Trump was bad news and that he didn't order any of those things. But, he and his advisers did appoint many of the public health officials on the front lines and the attitude of the boss can easily permeate the ranks of those who report to him, directly or indirectly.
The whole approach to the virus has been a clusterfuck since the beginning and no one in power has done much to change that.
I disagree that there is nothing wrong with the full-time pursuit of profit when people are dying because of it. Besides, the massive increases in profits we have seen in so many industries is evidence that they could have given up a few weeks of profit to protect people where transmission risk was/is high.
Be careful in analyzing the conditions contributing to spread. It took over a year for the WHO and CDC to accept the fact that the main avenue of transmission was aerosols. Governments and institutions have spent untold amounts on hygiene theater and buying plexiglass dividers that just make the situation worse by inhibiting ventilation.
In addition, asymptomatic people and those who have been vaccinated are still able to infect others. The vaccines were never meant to be sterilizing or reduce transmission, regardless of the narrative we have been fed ceaselessly. They only reduce the risk of hospitalization and death, which is fine as far as it goes, but it's not what the government has been pushing.
Karma is a bitch.
"It took over a year for the WHO and CDC to accept the fact that the main avenue of transmission was aerosols." And they were not appointed by Trump. :)
HCT? do you mean the blood test? Seems to me 'fast vaccine development' was one thing we got right.
Human challenge trials. They could have speeded testing.
With no reliable therapy and the IFR of SARS-CoV-2, human challenge trials wouldn't have been ethical. If paxlovid or other therapies turn out to be highly effective, it might be possible to do challenge trials for future new strains.
How ethical is it to allow arbitrary rules of "ethics" to delay literally life-saving vaccines?
I think this is referring to this article: https://www.nytimes.com/2022/03/11/opinion/covid-health-pandemic.html
Yes! The link should be there in the article. (I thought I had it in, sorry! On the road today).
Wecome back!!
Dear Zeynep,
I wonder whether you are aware of this doctor's report, interesting:
https://amidwesterndoctor.substack.com/p/adverse-reactions-to-covid-vaccines
Nationally, we need to emphasize access to HEALTH CARE (not health insurance).
Barbara Tuchmann's book is out of date. See https://www.open.ac.uk/research/impact/history/who-started-first-world-war
Thank you. I enjoyed several Tuchman books growing up, but later learned they might be not the best sources. Calling this one out-of-date is kind. Thank you so much for pointing me to a better source. I was looking for a readable but more "up-to-date " book.
You might also like this Judy,https://www.amazon.com.au/Sleepwalkers-How-Europe-Went-1914/product-reviews/0061146668/ref=cm_cr_dp_d_show_all_btm?ie=UTF8&reviewerType=all_reviews
where Annika Mombauer contributed to the discussion.
Bill, thank you so much for reminding me of Sleepwalkers.
What an amazing writer Mr. Clark is. Exact, concise, energetic, engaging. the book would probably have been 1400 pages instead of 700 if written by anyone else, and almost certainly less readable.
In chapter one he gets to something I was trying to explain in an earlier response, the difference between asking "Why" vs. "How." If we think of the Why (in the context of what living beings do), we get a one-off answer. If we ask How we can get at personal or social forces that we can perhaps set up so that we do better at some time in the future.
Anyway, I am fully engaged in the book and I thank you very much.
Thanks, Bill. I do know of this book and was not quite sure I was up to 700 pages. HOwever it is quite reasonable on Kindle so I will tackle it. Also couldn't find the Mombauer contribution. I clicked on your link and it just took me to a very large number of reviews.
Yes, I had the same problem with finding her reviews in the Amazon reviews. The Amazon reviews are not useful in that aspect. If you access to scholarly journals then
there is
Guilt or Responsibility? The Hundred-Year Debate on the Origins of World War I
Mombauer, Annika.
Central European History; Cambridge Vol. 48, Iss. 4, (Dec 2015): 541-564. DOI:10.1017/S0008938915001144
Hi Zeynep.
Thanks for your article and the Free article idea: tracing the first three-four months of the United States response, but not just Trump.
I have been developing a "Citizen Commission" to explore just that.
The goal is to learn from the pandemic through focus on what can citizens do.
More details to come with likely public launch in May.
Since the Media shapes public opinion I am starting with the initial CDC briefing January 17, 2020
https://www.citizencommission.org/january-17th-2020-press-briefing
The Press, Media, Journalism and Social Media https://www.citizencommission.org/media
I am highlighting the major journalists and plan to include your excellent work.
Here is a link to the time line
https://www.citizencommission.org/covid-19-time-line
Thank you. I would love to have a transcript of Nancy Messonier's radio broadcast. That was the most effective and informative commuication got from anywhere at that time.