The baseline excess deaths is 0 - excess deaths is how many more people died than in an average year. More people died in 2020 and 2021, and when they died it was marked as due to covid. What could possible be a non-Covid mechanism for this?
Most of the more popular theories are difficult to reconcile with this graph. I've been challenging people to come up with an explanation for over 3 years, and I haven't seen any rationalizations that I found persuasive.
You seem to be reading a lot into this post that isn't there.
The point of this post is that there is strong evidence that ~1 million people died in the US of Covid. It is useful to answer questions like this independent of surrounding context.
I think there is a subsection of the COVID truthers who say "it was just the flu" but really mean "there are trade offs between life and what makes life worth living, and we fell on the wrong side of the trade off" but I can't really respect the dishonesty. They are knowingly stringing along a lot of people who definitely don't feel the same way.
I don't think I defended the "general pro lockdown pov". I wrote:
> "Usually I’m the one arguing that we have to do cost-benefit analysis, that it’s impractical and incoherent to value every life at infinity billion dollars. And indeed, most lockdown-type measures look marginal on a purely economic analysis, and utterly fail one that includes hedonic costs. Rejecting some safety measures even though they saved lives was probably the right call. Still, I didn’t want to win this hard. People are saying things like “COVID taught us that scientists will always exaggerate how bad things will be.” I think if we’d known at the beginning of COVID that it would kill 1.2 million Americans, people would have thought that whatever warnings they were getting, or panicky responses were being proposed, were - if anything - understated."
I don't know how better to get across that the kind of lockdowns we did were probably an overreaction, but also it's insane to dismiss something that killed a million people as an overreaction that failed to pan out.
> Who cares if “1 mil is a big number” when divorced from population size?
I think it's useful to put it in the context of other problems. For example, are mass shootings a big problem? They kill about 100 people per year. So COVID killed 10,000x that. Is immigrant crime a big problem? A quick Fermi estimate suggests immigrants commit about 100,000 violent crimes per year. So COVID killed 10x more people than are affected by immigrant violent crime. Is cancer a big problem? About 600,000 people die of cancer per year, so COVID killed as many people as cancer during the two years it was most active.
Divorced from everything else in the comment thread here, I just wanted to say thanks for walking back on something in a public comment. It takes strength and I appreciate that you took the time to do that.
I generally wish we could praise and normalise this kind of thing more ('this kind of thing' being things like changing one's mind, admitting missing something, etc), so, in the interest of being the change I want to see in the world, I wrote this comment. :)
1.2 million deaths is about 0.36% of the US population of 330 million in 2020.
So, my new approach is to figure out how many years of lifespan covid cost Americans. Assume the average American is 38.7 years old and would live another 40 years and that there were 330 million people in the US in 2020. So that would imply the population at the beginning of the covid epidemic had 13,200,000,000 more years to live.
The highest estimate I’ve seen for how many Quality-Adjusted Life Years (QALYs discount years spent in comas, dementia, or horrifying pain, although not by all that much) the average covid fatality cost is 12.4 years. For 1.2 million fatalities, that would be 14.9 million years of life lost.
So, that would be 0.12% percent of U.S. lifespans lost due to covid.
Of course, the next question is: What would the toll have been if we didn't do X, Y, or Z?
But does "not doing X," such as not going to crowded apres-ski discos (big super-spreader sites in the late winter of 2020), means governmental decisions to shut apres-ski discos? Or does it mean people privately deciding to not go to discos?
And how big of a real-life difference is there between crowded discos shutting down by government order and discos shutting down because so few people are showing up its not worth paying the staff?
What it appears is that many people's preferred counterfactual is: What if nobody had noticed and/or nobody had cared about covid and thus kept behaving exactly as they had in 2019?
My guess is the death toll in 2020 would have been considerably higher.
But it's hard to realistically imagine a world where nobody notices or cares about a global pandemic.
Your guess that the death toll would have been higher is not based on reality, you are just echoing people who think they are smarter than everyone else. There was no chance that nobody would notice and no chance people would act in 2020 as they did in 2019. To be intellectually honest, you have to focus on things like this ---- People did notice that Cuomo's policies - brutal, penal isolation of old people as if they were criminals assigned to life imprisonment in solitary confinement, but close - within the same structure - to people who were dying from an infectious disease - killed many people who would have lived to see another day under a more compassionate government (and remember that Cuomo's government used threats of violence to impose its will). People also noticed that other people were being arrested for being alone with their family on a beach. This was an unimaginable disgrace to the people who ordered such arrests. I know you are a smart guy, Steve, but sometimes you look at the big picture too much, because that is what your pals look at (see, e.g., poor Murray and his attempts to count artistic value by the number of times names show up in mediocre art history textbooks). Yes there are stupid people who say there were no excess deaths and you can laugh at them all day but you are no better than them if you just focus on false and falsifying Leninist rhetoric about 'in the long run we were probably right'. It might be intellectually fun for you and others like you to brag about how much you understand numbers, and the big picture, but this was not a numbers game, and not a big picture game, from the point of view of the people who made absurdly bad decisions ---- and, because every problem in history is a problem that is based on particular details, and because we will never know how the counterfactuals would have turned out in detail, it is important to treat with respect those who disagree with you about the bad decisions made by people who were in the government and who backed up their bad decision with threats of violence, such as arrests, prison sentences, and unprecedentedly in the United States, threats of denial of basic decent medical support to those who dared to disagree with clowns like Cuomo and others like him.
I remember during the worst of it in late 2020/early 2021, the hospitals here in Utah were getting a bunch of Covid patients from Idaho and such because their hospitals had been completely swamped. Now I imagine what that would have been like if it had happened nation-wide over a week or two once the Delta strain got going (no more super-spreader events required).
I bet you'd have at least double the 1.2 million in excess deaths. Many of them wouldn't be dead from Covid itself (although there would be more of them with no free ICUs anywhere nor staff to surge locally from elsewhere in the US), but still dead.
Fwiw, I don't think the first article of the series was pro lockdown either. The point of *that* article seemed to be that 1.2 million American deaths is a lot of deaths and that it is an important and under-discussed fact. Sure, I suppose one could say that emphasizing the number of deaths is an argument a pro-lockdown person could use to support their case, but that isn't what Scott is doing.
In fact, he said "And indeed, most lockdown-type measures look marginal on a purely economic analysis, and utterly fail one that includes hedonic costs. Rejecting some safety measures even though they saved lives was probably the right call."
If you want to read something focused entirely on the lab leak hypothesis, I would recommend Scott's excellent review of the Rootclaim lab leak debate.
While I am no covid denier, I think you should really take into account all the people who died of other causes (heart attacks, cancer), because they could not or were afraid to access life saving care, including routine screening.
This. I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases.
Well, there was a strain on healthcare providers more generally. And that strain was caused by COVID. How many of the deaths caused by that were COVID deaths we don't know for sure but it was probably a lot and could well have been a million in the US.
If it's not knowable, even to a degree of uncertainty of approximation, then a meaningful cost/benefit analysis can't be made and the attempt to do so is misguided.
But it is knowable. Not only is it knowable, but the knowledge is easily available to just about anyone who's motivated to get at it. Here is a nifty chart I created tracking the top 8 major causes of death from Jan 2020 through May 2023 (first link below). (I've had problems with Google docs before, let me know if you have trouble accessing it.) If you need a basic tutorial on how death certificates work in the US, I've put together a very high-level description below the links.
If you open my chart, you'll definitely see that deaths due to diseases of the heart, malignant neoplasms (cancers), and cerebrovascular disease (strokes and such) would rise during peak COVID. IIRCC, COVID could only be listed as a contributing factor if the patient had been *diagnosed* with it within the 30 days leading up to death. Either these were due to COVID as a contributing factor outside the 30-day window, asymptomatic or perisymptomatic COVID as a contributing factor, or they were deaths because people couldn't get treatment.
HHS tracks ICU usage, and there were definitely people with non-COVID life-threatening illnesses who couldn't access an ICU during peak COVID waves. The CDC estimated that about 100K people died during the first 2 years of the pandemic because of insufficient ICU beds. We can sort of see this in their excess data. Second link...
And here's how death certificates work in the USA...
1. Drs and coroners are required to issue death certificates. There's a standardized way of filling out a death certificates. On the certificate they list the immediate reason of death and the contributing reason of death. If someone dies of pneumonia while in the hospital being treated for COVID, pneumonia will be the immediate cause of death and COVID will be the contributing cause of death. While some jurisdictions allow a "cause of death unknown" option, that's rarely used.
2. There are 3,244 counties in the US. Each county has a health department. Hospitals, MDs, and coroners, are required to file their death certificates with their county health departments. County health departments are required to upload the data from the death certificates to the CDC. So, probably 99.99% of the deaths in the US have a death certificate with an immediate and contributing causes on it.
3. The CDC tracks this data. The CDC puts all this data online via their Wonder app, which anyone can access to download the data and/or create nifty charts from the data.
It's... hopeless, isn't it? No matter how clear, no matter how easy-to-access, no matter how often trotted out - the data are just ignored. After all, why know, when one can pontificate endlessly, make up clever hypotheses, and "just ask questions".
>So, probably 99.99% of the deaths in the US have a death certificate with an immediate and contributing causes on it.
That sounds high. E.g. about 1% of the population goes missing at some point in their lives and are never found. Some may have left on their own steam, but I suspect that a large fraction are killed and not found (since that would be the same order of magnitude as homicides) - and these don't get death certificates.
We have a population of about 340 million people. You're saying 3.4 million will go missing and never be located during the course of their lives? Well, provisionally, if we accept that, and we take the average lifespan of a US person to be 77 years, that gives us about 44,000 people who will go missing every year — given the current US population. That's still a lot of people! That would be about 3,700/month, so I guess I could accept as a reasonable number. But...
It appears that the 1% number originates from the DOJ's NAMUS database, but it's currently offline. So I can't really evaluate its validity.
I see numerous media quotes surrounding that factoid, though. This one was interesting...
> An NPR interview with Todd Matthews, director of communications for the National Missing and Unidentified Persons System in 2013 delved into the nature of missing persons reports and found the same thing.
> “You know, the missing touches everybody, I think. In 2012, we had 661,000 cases of missing persons; and that’s just from that one year. Very quickly, 659,000 of those were canceled. So that means those persons either come back; in some cases, located as deceased persons, maybe never an unidentified person; or just a total misunderstanding. So at the end of 2012, of those 661,000 minus the canceled, we had 2,079 cases that remained at the end of the year as unresolved.”
> As of December 31, 2021, the National Crime Information Center’s (NCIC) database contained 93,718 active missing person records.
> Children under the age of 18 account for 30,400 (32 percent) of the national recorded 93,718 and 39,114 (42 percent) recorded were children under 21 years of age.
According to the DoJ website these missing person records are never purged. They are retained indefinitely until the individual is located or the record is canceled by the agency that entered it. Some of those number date back to when the database was created. And if Todd Matthews is right, then we're only losing between two and three thousand people each year who never turn up. That's a lot less than 44,000/year.
Approx 2.85 million US citizens died last year. And say we had 3,000 unresolved missing peeps last year. That gives 0.1%. OK, so 99.9% instead of 99.99% of people end up with a death certificate at some point. ;-)
Thanks for keeping me honest! But color me doubtful about that 1% figure.
The comment to which I was replying was, "I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases."
The comment to which I was replying was, "I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases."
For example, this hypothesis suggests that the excess deaths should track COVID hysteria (and lockdowns etc). And it's unlikely that COVID hysteria exactly tracked COVID cases.
But the graph of excess deaths seem to track the graph of COVID cases fairly well. That's some evidence against the hypothesis. (Even if it's not enough to completely rule it out.)
The comment to which I was replying was, "I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases."
My sister died of leukemia last month. It becomes very apparent to those of us who have walked the cancer journey of the extremely long supply chains that are required to support cancer care. Example: In the fall of 2024, one of her chemo cycles was pushed a week, because the hospital had clamped down on “non-essential” fluid usage because of an extreme shortage due to a hurricane in Florida (we are located in Denver) destroying the primary distribution center for IV fluids, and fluids are required to flush the chemo to prevent organ damage (the irony here is apparent to me). Many such examples for just the supply chain, let alone the supplies and available hospital staff being diverted because of hospital strain.
I would rephrase that, since as stated, one who wants a high number for death toll could argue that those SHOULD be counted as "covid deaths." I'd instead attribute those deaths to "pandemic" measures, coördinated by global "health agencies."
Absent the NPIs, the number of such deaths (and associated morbidity) could have been worse. You can't meaningfully assess the casualty absent a fuller context. If someone strangles from a seatbelt in a car accident, how do you attribute the death?
For a more apt metaphor, if people were getting strangled while driving normally due to some new kind of seatbelt mandated due to OTHER people getting into car accidents, I would say attributing those deaths to car accidents is bullshit.
I don't see how thar would be an apt metaphor. Absent NPIs, there's a non-zero possibility that hospitals would have been overwhelmed and people would have been afraid to make visits to the doctor. Think of Bergamo, Italy, early in in the pandemic. The pandemic wasn't a function of "other people."
Let’s take out the “should” and just try to count these things in the specific groups, and worry about the “shoulds” later, when we produce several different interpretations of the same data. Do one where you attribute every single excess cardiac fatality to Covid, and one where you attribute every single excess cardiac fatality to NPIs.
We live in these complex interlocking systems and we're actually not bad at dealing with this complexity... when we're not insisting on reality being black and white for our own psychological comfort.
In fact, to get better at modeling complex systems (like a pandemic rolling through healthcare systems and societies across geographies and time), we need to accept uncertainty and some amount of approximation without freaking out and saying it's hopeless to understand anything.
Science has ever proceeded in this way. Imperfectly and unevenly. But that has never been justification to ditch science and the collection and analysis of data and just start speculating.
Who's the "we" you're talking about? "We" as in the denizens of ACX seem to have problems with complexity and "we" keep coming back to simplistic answers that satisfy our prejudices. "We" as in the scientific community (which does not include me), were actually quite good at modeling complex systems — especially when it came to healthcare outcomes of various NPI regimes and various different healthcare systems. Sure there's a lot of arguments about how to interpret the data, but the data is there.
And amen to this...
> Science has ever proceeded in this way. Imperfectly and unevenly. But that has never been justification to ditch science and the collection and analysis of data and just start speculating.
I know one person who died from covid, but he was in a nursing home. A lack of visitors in the nursing home is, in my experience, very bad for health and mortality. Not only does one miss the benefits of social contact, but the nursing home personnel get the idea that no one cares about you, then you die. This was true for all nursing home patients during covid.
I also know one person, my neighbor, who died of cancer during covid. She was taking chemotherapy at home under the care of her elderly and overburdened husband, and almost certainly died sooner than she would have with proper hospital care.
"The total 'excess' deaths roughly matches the reported Covid deaths. California reported around 50,000 Covid deaths in 2021 and saw a bit under total 55,000 deaths above those expected by the model. "
I think the entire argument is because of overly simplistic models. There's rarely one cause of death. It's usually a this + that + the other thing kind of happening, and if any one of them were missing, death wouldn't have happened (at that particular time).
So. COVID put a lot of stress on the system. Thus people died who wouldn't otherwise have died (at that time). I believe that it was extremely rare that it was the STRONGLY predominant cause of death (say over 70% of the reason). But I suspect that it was frequently the reason that (say) 40% of those who died, died at the particular time that they did. (The particular figures are just numbers pulled out of the air. Don't pay much attention to them. They are just there to paint the shape of this argument.)
Note that being shot in the heart is not sufficient to be the single cause of death. Andrew Jackson was shot in the heart in a duel before being elected president. (Well, it *was* with a black powder dueling pistol...small bullet, low power. But he lived through it.) Causation is almost alway complex. Sharpening things down to a single cause almost always cuts away crucial factors. (That's why designing good experiments is so hard.)
> I think the entire argument is because of overly simplistic models. There's rarely one cause of death. It's usually a this + that + the other thing kind of happening, and if any one of them were missing, death wouldn't have happened (at that particular time).
See my very high-level overview of death certificates in the US of A (above). Also, the models are *not* overly simplistic. While I have issues with standard SIR models, the signals in the data that were gathered as the pandemic progressed was pretty clear. And better yet, we can compare to the outcomes of different countries with different NPI regimes, because many other countries put their data online, too.
You (well, I) *CAN'T* pull them from the published data, because they only list a single cause of death. The documents always want to list a single cause, and that's explicitly the thing that I'm denying. (Admittedly, there's often a major contributor, but I'm asserting that a single cause is the exception rather than the most common event, or even *a* common event.)
Sorry for the delayed response. I poked around CDC Wonder last night, for the year 2020, I found that 378,048 deaths had an ICD code of U07.1 (which is code for COVID). Of these, about 94.5% included at least one additional ICD-10 code indicating other conditions, while another 5.5% listed only U07.1 without any other contributing conditions.
I'll admit it took me well over an hour to dig that data out of wonder. Their user interface is a little clunky. Does that answer your concerns? Because I admit I'm all Wondered out.
So, yes, just about no one *died* of COVID without some other ICD code on their death certificate. And, yes, that's to be expected. Especially since the frequency of comorbidities rise with age in the population. Yes, a perfectly healthy person is less likely to die from a pathogen than a person with comorbidities. But perfectly healthy people are less common than one would expect. According to a 2010 study a hair less than fifty percent of the US population has *no* comorbidities. And one quarter of the US population has multiple comorbidity conditions (MCC).
> In 2010, 49.1% of civilian, noninstitutionalized US adults had no chronic conditions, and 24.8% had only 1 chronic condition. One-quarter of US adults had MCC (Table 2); 21.1% had 2 to 3 MCC and 4.9% had 4 or more MCC. For both sexes, older adults had a higher prevalence of MCC than younger adults. Among adults aged 18 to 44 years, men were less likely to have 2 to 3 MCC compared with women (P = .003). For adults aged 18 to 44 years and 45 to 64 years, men were less likely than women to have 4 or more MCC. In contrast, among adults 65 years or older, men were more likely than women to have 4 or more MCC.
Thank you. That's about what I expected, but I also expect that the original data (that the CDC data was compiled from) oversimplified the circumstances of death. If for no other reason, then it would take a HUGE amount of effort to dig those things out. (How do you detect arterial blockage in a corpse? I don't really think that's possible. Blood tends to coagulate after death. [I'm no expert, so I could be wrong here, but that's what I think.])
I also think that "no chronic condition" is a VAST overestimate. Try "no recorded chronic condition". Most of my life I've noticed chronic conditions that weren't serious enough to ask a doctor about. The same seems to have been true (my evaluation, I didn't ask) of everyone I've known well enough to have an opinion about on the subject. (So perhaps 8 to 10 people, but a 100% positive. [OTOH, it's easier to notice something like that if someone complains about it, so it's a highly biased sample.])
That's the most plausible counterargument I've seen, but it's worth noting that even shaving off 10 percent of 1.2 million leaves us with over a million.
Mortality due to drug overdoses is also tracked by the CDC. Deaths due to drug overdoses during the pandemic (including Fentanyl) happened at the rate of between 6000 and 6500 per month. Drug overdose deaths climbed until 2022, and started dropping off a bit beginning 2023. Last time I checked they were still falling.
Undocumented immigrants, tourists, etc., all get death certificates. So their deaths were also tracked by the CDC. But they're lumped in with everyone else. Additionally, death certificates do not include a field indicating whether the deceased was a U.S. citizen or of another nationality. Only "place of birth."
But there are studies that showed the undocumented immigrants died at twice the rate of people legally here. Given that roughly 0.35 percent of the US population died during the pandemic, and given that there were likely 10.5-11.5 million undocumented immigrants in the US during the COVID years, that suggests that approximately 77,000 of those 1.2 million deaths were undocumented immigrants. Roughly 6%.
Not likely. Most undocumented immigrants could be younger, healthier, working population and less of them would be old people in care homes. Even if they were dying double rate compared to people legally in the US, that would not be a big number.
Don't forget about exponential risk increase by age.
<Post exchange analysis of this thread. Substack sends comments to my email, so I have his pre-edited response. NoPie originally wrote: "I cannot believe that you still forget about exponential risk increase by age." NoPie was aggressively rude, I responded with his "I cannot believe with my own "I cannot believe.">
I cannot believe you didn't check your beliefs against the data before expressing your opinion.
I assumed the studies I checked would be normalizing their data, because all good demographers should age-adjust their datasets. The Minnesota study (third link below), shows that your expectations would be true for all ICD categories *except infectious diseases.* I found that to be interesting.
> Theoretically, there may be expectation of lower COVID-19 mortality among immigrants than US-born residents. Literature documents a healthy immigrant effect—where immigrants tend to be healthier than US-born residents (Kennedy et al., 2015).... In addition, in pre-pandemic times, better health outcomes among immigrants were consistent across all ICD-10 disease categories except infectious diseases and external causes (Aldridge et al., 2018).
Here are links to the studies. If you've got anything other than your supposition to counter with, I'd be interested in the some links to support your claim.
This is the study that I based my claim that undocumented immigrants died at 2x the rate of native-born US residents...
Murder quite famously went up, but probably not enough to change excess death statistics.
Traffic deaths, on the other hand, probably did change enough. They dropped during the initial lockdown period then spiked, and haven't come back down all that much (while murder did drop back down, I think by late 2022).
From 2016 to present, every year has had between 36,355 and 43,230 traffic fatalities in the United States. The lowest year is 2019 and the highest year is 2021. 7,000 deaths is not enough to make a noticeable difference in the total excess mortality number.
Overdoses had been soaring, especially since fentanyl arrived c. 2015. Overdose deaths during the covid years were very high, although it's not clear what they would have reached without covid.
Murders and traffic fatalities shot up after May 25, 2020, especially among African Americans, but that was probably mostly due to the temporary triumph of BLM over law enforcement.
Note that these three causes of death hit younger people on average than covid, so they take away much larger expected lifespans per death. On the other hand, their total numbers are quite a lot smaller than covid deaths.
OD deaths were at approx 6K/month at the beginning of the pandemic (first months of 2020, and they rose to 6.5K/month by the end of 2022. They began to fall off in 2023.
The murder rate spiked in 2020 at 6.42 per 100,000 people, up from 4.99 in 2019. Approx 21,500 people. Meanwhile in the same year <checks the NIH website real quick> COVID-19 accounted 697,000 deaths.
Since most murders occur between people who know each other, we can assume lockdowns had a significant role in the rise in the murder rate. I haven't gone over to the FBI website to check that assumption, though.
So roughly 4800 extra murders happened in 2020.
And a quick Google shows that traffic fatalities went up from ~36,000 in 2019 to ~38,800 in 2020. About 2,800 more traffic fatalities during lockdowns.
These numbers are chump change compared to 697,000 deaths from COVID.
CORRECTION! There were ~375,000 COVID deaths in 2020. The NIH website gave me the numbers for a 20-month period from March 2020 through October 2021. It seemed too high. I should have read the report more carefully. But ODs, homicides, and traffic fatalities are still chump change, against total COVID deaths in 2020, but twice as valuable chump change.
Domestic violence murders weren't up much during covid, surprisingly enough. Instead, there was a huge surge in mass shootings at black social events after George Floyd's death unleashed the racial reckoning. Cops retreated to the donut shop, and so people felt more confident carrying their illegal handguns to parties.
I don't see that in the DoJ data. But I didn't dig very deeply. If you've got some DoJ data or data from other sources, I'd be interested in seeing it. There are hella lot of reports on domestic violence up on the DoJ website, though.
ChatGPT sez:
2019 Data:
Total female victims: 1,948
Killed by someone they knew: 1,773 (91%)
Killed by male strangers: 175 (9%)
2020 Data:
Total female victims: 2,059
Killed by someone they knew: 1,604 (89%)
Killed by male strangers: 197 (9.6%)
I didn't check into male-on-male murders, and what percentage where the victim knew the perp. But yes, you're right, it looks like 4K extra murders in 2020 were not due to domestic violence. But still, we're talking drop in the bucket to total COVID deaths.
Also, you forgot to mention suicides. At one point, suicides from the lockdowns were being blamed as a significant contributor to excess deaths in 2020. But that idea didn't pan out. The suicide rate actually dropped in 2020 from 14.5/100,000 in 2019 down to 13.5/100,000 in 2020.
Sure, but comorbidity was an issue. Americans have never been this unhealthy. And 0.3% hardly matches the nightmare scenario we were lead to believe. Moreover, it's probably why you use the word millions instead of 0.3%.
To be fair, 0.3% was exactly in line with many early estimates I saw. Most of the panic was because the broad CI included >1% estimates. It was always odd to me that people continued to implicitly rely on the broad early CI estimates long after better information came along.
Overall, I agree that not only are there a lot of risk factors in line ahead of COVID, but that improvements in those other health factors are probably a better focus for preventing deaths in the next pandemic.
I used million because that's the title of Scott's post. And more nightmarish scenarios were avoided in part because of the protective measures out in place. I agree some of those measures were overprotective in hindsight - but that was hard to know at first, and the main point is that this pandemic was very real and deadly and worth taking some precautions to minimize.
Yes, if you portray COVID deaths as 0.3% of the population, it all looks rather innocuous compared to <gasp> 1.2 million deaths. But if we frame it another way, during the peak of the worst wave, we were losing the equivalent of eleven 747s full of passengers each day. We shut down US air traffic, brought the country to highest alert, and invoked our NATO treaty to get fighters from European members to help patrol our skies, all because four jets were downed and 3,000 people died. If you look at COVID deaths in units of jumbo jets, the pandemic was pretty frigging significant. But most people just shrugged their collective shoulders and complained about their freedumbs being curtailed.
As for comorbidities, it's important not to look at the US as a singular island of health patterns.
For instance, France, because of its higher rates of smoking (roughly 2.5x US), has a higher average rate of COPD, asthma, and lung cancer, related diseases than the US. Asthma, at least, was one of those comorbidities that increased the risk of developing pneumonia from COVID.
And in Germany, two-thirds of men (67%) and half of women (53%) are overweight. Hard to get an apples-to-apples comparison of obesity, though, because the EU defines it slightly differently from the US. But roughly 25% of their population is seriously overweight. Not US rates, but very high.
The most interesting comparison I took from the pandemic is that Finland has a very high overall comorbidity rate, but during the pandemic, it did better than Denmark, which has a population with lower comorbidities. Obviously, comorbidities are an important factor, but they were far from the only factor determining which populations were able to pull through the pandemic with fewer deaths.
Why not consider that pandemic restrictions may have reduced mortality from other causes. This also seems plausible. Less miles driven. Less trips taken. Less exposure to other illnesses. It’s not just a one way street.
You could have figured this out with the article's central argument: excess deaths over normal, non-pandemic conditions. If you had enough lives saved by the lack of activity that it counter-acted a substantive proportion of COVID deaths, the excess deaths count would be materially different from the COVID deaths count.
Not quite. Since the article is using the excess death count to establish the veracity of the Covid death count, this would still need to be explored. But yes, assuming we know those two variables, no need to.
Among old people in retirement homes, access to family and social networks is, in effect, life saving care, and access to family and social networks was brutally denied to millions of such people for a very long time (in terms of their expected future lifespan, two years could have easily been more than half their expected future lifespan). This effect could easily explain hundreds of thousands of the excess deaths.
Frankly, because no one fights for your life as much as your family. The other people are just paid to be there. If you take the most charitable view, they’ll do their best but they have a million other concerns. If your elderly family are in a period of prolonged medical vulnerability, you absolutely need to have a member of your family be their advocate in conversation with the system, because otherwise the system is often laggardly and obtuse.
I've heard that for end-of-life care distant relatives who didn't visit often and feel guilty are the most insistent on prolonging past the point doctors think is wise. But since these people are actually near the end of their lives (not like most people who died of COVID prior to being infected), there isn't actually that much time that the medical system can give the patient.
Not only that, but contact with family is literally what has many old folks going. Cut that off, and they have nothing to look forward to any longer and shuffle off the moral coil. Anecdotal observation but I have seen it enough to trust there is some validity to the claim.
But even more immediately, yes, lack of care. At the start of the pandemic in Canada some nursing home patients were literally abandoned by staff that fled. Found several days later covered in their own feces etc. Given their pre-existing conditions, such abandonment would cause a sharp decline in their overall health leading to much higher death rates.
Not all nursing homes experienced such drama but I'm sure care was compromised in a very high percentage, resulting in many additional deaths.
I'm sure a lot of this kind of second order effects happened, but again it's kind of ridiculous to suggest that it explains more of the surge in mortality than the respiratory virus that looks like the kind of respiratory virus that kills a lot of people. To me it mostly seems like people are unable to conceive a middle ground between the sniffles and the Black Death; since it wasn't the former, it must be the latter.
A close relative of mine died in a nursing home because there was not enough staff to ensure he was getting his meds (his condition was making it really hard for him to swallow). Previously, this was taken care of by his wife, but, when COVID started, she was denied entrance.
Sorry, that was a joke. When I posted that, I thought it would so patently stupid that no one would actually say things like that unironically, but then I read other comments here, and no longer believe that.
It's amazing how much the "will to live" impacts whether you can recover from illnesses later in life. I wouldn't have believed it until I witnessed this phenomenon first hand.
My grandmother struggled with the sudden isolation during the pandemic. Her husband had recently died and she couldn't take care of herself on her own. The isolation in care facilities caused clear health deterioration for her. We made the decision that the increased COVID risk of taking her into our home to care for her was better than her dying alone. Within days of coming to live with us she became physically better. Her ability to walk and her general gait improved dramatically, her incontinence disappeared, she was happier, she reported less weakness and a greater ability to perform basic daily tasks. Her memory improved.
Measurable, tangible health improvements.
When circumstances prevented us from keeping her at home, she went back to a care facility. Her health rapidly deteriorated, she was quickly moved to a wheelchair, and she died soon after. She told me, "I don't want to live anymore," both when she was in the facility before coming to our home and afterward.
A friend of mine is a mortician. In December a few years back she mentioned she was going into their post-holiday 'busy season'. This was before my experience with my grandmother, so I was shocked that people didn't die at regular intervals.
She said, "No, a lot of people hang on until after the holidays, and then they just lose the will to live." Remember that when you're old you tend to have a lot of old friends, too. If all your friends die, but you don't see other old friends at the funeral because they discontinued funerals, that's a highly significant social factor that's being withdrawn.
I think Scott implied in this post that the 1.2M number must either be fully explained by indirect non-infectious factors, or we must accept the 1.2M number as a result of infectious spread. I doubt he would maintain that stance if pressed, but the defense he gives above strongly implies it. I think it's possible that a non-trivial number of deaths were the result of COVID policies that weren't necessary, and not the result of the virus itself. Those deaths may track with reporting because when we saw deaths go up we pushed harder on the harmful social policies. If the policy-related death number is, say, 200k that's a pretty big number that deserves its own reckoning.
Is that really the most important? I guess other things pointed to as big causes of mortality (smoking, diabetes, hypertension, obesity) are likely to have started before you're elderly and there's less to be done about it now (although I suppose nowadays semaglutides actually can treat the chronically obese).
My understanding is that men's lives shorten when their wives die, but for widows it's the other way around (sometimes phrased as men sucking the life out of women).
When you say, "hundreds of thousands", are you talking 200k or 800k? The former might be in the noise, but in the latter, you wouldn't see the clear correlative pattern in the data the Scott cites.
If your causal model is "intervention -> lack of access to families -> old folks die 'cause sad a la Padme", then you still need a model for the cause of the interventions. This was pretty clearly COVID cases and deaths, especially in the early years. But now COVID deaths are causing such a ruckus that people do a lot of NPIs, which brings us back around to "COVID is actually doing most of the people killing".
This would predict more deaths in retirement homes, while plain old viral disease would predict fewer deaths in retirement homes, so this hypothesis is testable
Yes, this is one of several testable hypotheses. For the record, 1.2 million deaths (and hundreds of millions of survivors) is a lot, but a good researcher can review the details of about 100,000 deaths - after spending a few days determining how to randomize the events leading to the deaths in question - in about a month of research. 'Covid deaths in the western world' is not an impossibly obscure historical event, and I trust that eventually medical researchers will understand what went wrong.
The way test the size of the two effects is by looking at when the deaths in old age homes happened: did they short follow the five main waves of covid or did they happen randomly throughout periods when they were denied visitation.
I can see either argument, so somebody who feels strongly in favor of one side or the other should test.
That is not a good test, because of course those who lost the will to fight Covid were not going to die from that lack of will when the waves of Covid were at minima, rather than maxima.
Steve, because, as you know (I know you know who Bruce Charlton is, and you probably have read his discussion of how flu was considered, fifty years ago, to be the 'friend' of the elderly), loneliness on its own did not kill, but loneliness combined with an attitude of "I don't care anymore" is what killed when old people whose endorphins were near zero got flu symptoms that they did not anymore have a great desire to struggle against (this is simple understanding of how humans work and I am surprised you said "right" to an obviously falsely posed question --- why did loneliness kill in rhythm with reported covid deaths).
I mean , you are too smart to need me to spell this out, but just in case you need it to be spelled out ---- of course the lonely people died more when the waves of covid were at their worst. Nobody said they predominantly died of loneliness, they died of not wanting to live when they were faced with a wave of covid.
Very plausible, but this might merely decrease the "Covid mortality" as a number. It does not decrease the overall "seriousness" of the pandemic, or reduce the reasons for e.g. requiring people to wear masks.
Scott's argument in the first place was, like, "remember the 1.2 million people who died because of Covid?" - and you can't deny that also those 10% of deaths were directly caused by Covid, can't you?
If you argue ".. but they didn't die of Covid, they died because our health system was unable to cope with the sudden outbreak of Covid" - yes, sure, but maybe quite a large part of those other 1 million US deaths (people who died FROM Covid) were due to the fact that the health system was not ready for the pandemic.
There are those who would quibble that it’s unfair to characterize the health system as “unready for the pandemic” as opposed to “unready for the panic stirred up for various non-medical reasons in the face of a mostly unremarkable disease”.
I don’t *think* I am one of those, but I’m apparently sympathetic enough to that quibble to point it out.
Do you remember when we were hearing about hospital beds filled and not enough ventilators? People were there because they were sick and having trouble breathing.
There was plenty of panic, too, but someone would have noticed if all the hospital patients had just been having panic attacks.
I don’t think anybody of note ever *denied* that there was a disease.
And I was talking about a societal panic, not claiming that anyone claiming to have Covid was actually having a literal panic attack instead.
As an example of the societal panic, I gather that in retrospect the consensus is that we used respirators rather too often.
But I’m not arguing with Scott’s overall point myself — I think it’s quite likely that American did have in the order of a million deaths that can plausibly be blamed on Covid itself. Whether the exact number is 750K or 1.3M is out of my wheelhouse.
A lot of areas kept beds open and reduced availability numbers *in case of* needing them, but ultimately never filled up. Source: family that works in hospitals.
The sheer volume of nurses recording Tiktoks while at work was taken as anecdotal evidence they really weren't that busy either, but it's not a great metric.
Several field hospitals set up in my vicinity, one of them a tent city in large parking lot. No patients ever arrived. On the other hand, hospitals were in financial distress and laid off employees because they didn't have enough patients.
There was a huge covid death toll in New York City in March-April 2020, with 15,000 deaths in April alone.
But the "lockdown" started in mid-March started to work about one week into April. I recall a late-April 2020 New York Times op-ed by a NYC ER doctor. He remarked that background noise all through March and the first week of April was the sirens of ambulances arriving with covid patients. Then, IIRC, about 1pm on April 7, 2020, the sirens mostly stopped.
The health system spent a ton of pointless effort saving, say, patients from the local hospice instead of instituting an age/condition-based triage instead. Example: survival rate of 70+ year olds in ICU in 2020 was <50% but there were still cases where 35 year olds from (say) a car accident had to wait for an ICU spot.
We have a hard time rationing kidneys and livers; denying an old person an ICU bed because they have a baseline survival rate less than. 50% is a pretty tall order.
I was in my residency training during the early pandemic, I would encourage you to really think about how you would go about getting someone, say 72 years old, out of the icu or hospital to open up space for someone else. You get patient transport to bring them to a van and them drop them at home… and then what? They come to the Ed and you tell them you will stabilize according to EMTALA and then… send them home to do it again the next day? Essentially telling patients they should die every day at work would be a psychologically devastating experience for providers. It sounds nice on paper but makes no practical sense.
Also small quibble but I do not think you understand what hospice is
I think there is a mistaken image here of people being “turned away”- the car accident victim is also taken care of here. The issue is that if the icu is full (say of older Covid patients), they would be stuck physically in the ED where care is not going to be as tightly managed. Obviously also a bad situation, and perhaps outcomes are worse, but very different from telling people you won’t try to provide standard of care
How about moving the old Covid patient back to the ED and moving the accident victim to the ICU then? I've no idea if that would have made sense; I don't really mean to make any assertion about what happened or what should've happened as I don't know much about it, I was just going off myst_05's and your comment.
I didn’t say it would’ve looked pretty. But step 1 would be to suspend EMTALA and all other such laws until the emergency is over.
I do know what it is and I also remember crystal clear that there’s been numerous reported instances of people on their deathbed being taken to the hospital after contracting the virus during the initial March-May chaos.
I suppose I still don’t have a sense of how you see this playing out on a patient by patient level. If an older patient refused to leave the hospital because they were still critically ill, does security pick them up and carry them outside? And if this were logically solved, wouldn’t removing EMTALA encourage hospitals to kick out people with no insurance or Medicaid/medicare rather than the elderly?
If someone’s in a critical state how are they “protesting” anything? You move them out of ICU into the regular hospital or if that’s full, into a field hospital (kind of like the ones setup in stadiums and whatnot in 2020). If they’re violently trying to stop the doctors… there’s a whole array of medications at the hospitals disposal to make a person calm down, no need to make a scene. But the vast majority of those who’d be removed would be frail 80 year olds who’d otherwise have another 3-4 years to live, slowly withering away in a retirement home.
> wouldn’t removing EMTALA encourage hospitals to kick out people with no insurance or Medicaid/medicare
Some unfairness would be present temporarily yes. Thats what happens in a crisis.
That was the huge fear justifying lockdowns: that hospitals would be so overwhelmed by covid patients that other patients would deny for lack of care.
A couple of things happened, though: to my surprise, hospital administrators turned out to be surprisingly good at deal with major surges in case loads. And public and private responses to lower risks of infection proved successful and blunting the worst saves of infection.
My great aunt died from cancer in 2021 and it was specifically blamed on her chemotherapy being delayed by two months during the start of Covid - but this was in Russia with a much more cavalier attitude towards healthcare, so I don’t know how prevalent this was in the U.S.
It is certainly possible that people died because of Covid even without actually contracting the disease, but Scott's analysis compares Covid deaths to the increased mortality, and the numbers line up. Unless the numbers are being faked, it is hard to see how deaths from Covid aren't really about 1.2 million.
But there are also various pandemic related behaviour changes things that would make excess death go down. Such as fewer road deaths. Heck even suicides went down during lockdowns.
My memory is that traffic related deaths did not decline which was a huge surprise given decline in ave miles driven? Meme key ain’t what it used to be though…
My memory is that traffic related deaths did not decline which was a huge surprise given decline in ave miles driven? Meme key ain’t what it used to be though…
Routine screening rarely prevents deaths on the timeline you're suggesting. Emergent care for heart attacks, strokes, gunshots... maybe. But not routine screening.
When trying to estimate the effects of NPIs on health outcomes during the COVID years, folks often fail to examine the counterfactual of "How do people react in a COVID world where there's no government effort to encourage or enforce NPIs?"
As a result, it's rarely clear whether folks who claim NPIs are responsible for excess deaths (rather than COVID directly) are saying, "The Government erred in enforcing NPIs", or "The Government erred in encouraging private organizations and individuals to use NPIs" or "People on an individual level reacted irrationally to COVID" or something else. It's not exactly a motte an bailey; it's more just a lack of a precisely stated hypothesis.
I don't think there's any government (or "elite") response that wouldn't have resulted in massive disruption to all kinds of services, including nursing homes, which will serve for a useful example. If half your staff is out sick because they have COVID, and someone dies because they're receiving a lower standard of care, is that death "caused by COVID"? What if people called out sick but wouldn't have absent strict Government guidelines that were in place because of the pandemic? What if the guidelines were put in place by the nursing home w/o government intervention because they were leery of litigation caused by staffing with sick nurses during a pandemic? What about if the nurses were just very worried about this new disease and weren't willing to risk giving it to their patients?
My point is that there's a lot of causal models you can consider for how people could die during a pandemic due to indirect effects from it. Some of those pathways became more or less likely due to the specific set of NPIs, elite message, and cultural circumstances we found ourselves in from 2021 - 2024. But those specific outcomes can't be investigated in a vacuum without considering the counterfactual of what pathways become more or less likely with some different posited set of NPIs, elite messaging, and cultural circumstances.
So even if 120,000 people died from 2021 - 2024 because they were waiting for care, that doesn't answer the question of whether those deaths should "count" as COVID deaths or not.
If we hadn't paid attention to covid at all, not much would have changed in our lives.
Some old and otherwise very sick people would have died more than usual but for a general person it is on the level that we barely happen to know such cases personally. Now we would be discussing and talking – there was a covid spreading around 5 years ago. Does anybody know who has died from it? Maybe an average person knew 1 or 2 old people from his family.
People working in healthcare would have noticed more people dying suddenly from a new type of respiratory virus. Some hospitals would have experienced overload, even more than they happen during nasty flu seasons. The number of deaths from covid as reported by Scott is real because people died indeed.
Provided that the government had allocated resources to deal with overload of hospitals and quick development of vaccine but without creating any panic, the timeline would be exactly the same except that we wouldn't have had any lockdowns, school closures or travel restrictions. We wouldn't have caused crisis in economy and the society would be more cohesive and less split about unnecessary restrictions.
Obviously, it mattered to people who died. Now I will saying something that will make blood boil to many people. Sweden's example showed that the number of those deaths would not be much greater in either case. In fact excess deaths would be lower all things considered.
If people were afraid of getting medical treatment and died of other things because of that, then wouldn't we see the excess deaths be less-closely related to reported covid deaths? Either it'd be correlated with lockdown measures (if people are afraid of being isolated in a hospital) or with news stories about how bad covid is (if people are afraid of catching covid).
I know someone who died of a heart attack during the pandemic, because he didn't seek medical attention due to fears of dying alone in a hospital due to isolation measures (so instead he made his wife wake up to him dead in bed beside her - not cool to do to a partner). So I'm not saying this didn't happen, but I doubt it was a significant fraction of excess deaths.
I have heard from more than one person who had to fight to keep COVID off a death certificate. "Oh, it was because of delayed treatment caused by COVID!"
Also, of course, there's nothing to prevent an original outbreak from being followed by a wave of deaths from the lockdown and other reactions to COVID.
My boss died from this. He had a heart condition that he had a surgery for that generally calls for three days of observation afterwards because of a known complication that sometimes occurs afterwards. Instead, he was sent home immediately after the surgery and died that night from the known complication. I guess maybe he would have died from it anyway, but presumably it's something that can be dealt with in the hospital or they would not normally keep them for monitoring.
As for iatrogenic deaths, I would expect more medical errors from overworked and overstressed docs, not fewer. Same with telemedicine, some things that have diagnostic value don't transfer across the screen (e.g. ease or difficulty of movement, various smells).
Spot on. While I understand there's some debate about the precise numbers of deaths (especially in countries with poorer record-keeping), it's frustrating to see people pretend that COVID wasn't a big deal.
I think that's a result of the overreaction and counterproductive measures (and sometimes just spiteful measures) put in place in response to the pandemic. The reality is that the pandemic was a really bad global tragedy that may have been completely avoidable depending on the source of the virus, and the response to the pandemic was a more or less global tragedy, maybe not as bad as COVID but maybe worse depending on how you count the economic impacts, that was completely avoidable.
We more or less aren't capable of having even slightly nuanced public debates, so I think to a lot of people conceding that COVID actually was really bad feels like conceding the argument before it begins.
I certainly wouldn't be surprised to learn, at some remote future time, that all the chaos that arose from Covid, in multiple fields, actually killed way more people than the condition itself.
Yeah we now all have Covid Rage Virus. It kills dialogue, fair-mindedness, goodwill towards others, common sense, friendliness, and some relationships. May be worse than covid itself
What's even more frustrating to me that even the contingent on board with it having been a big deal is lukewarm at best about ending gain-of-function. Sure, maybe this pandemic wasn't caused by a lab leak, but the next one very well might be, but nobody gives a damn.
People are allergic to nuance, is a lot of it. It’s not enough to criticize the response to covid, they have to take it to the extreme of pretending it was all fake news.
There's certainly room for debate about "with" versus "of."
There's the old meme, of course, of a guy halfway through his car's windshield, bleeding out, and someone with a microphone asking him "so when did you come down with COVID?"
But it's the same on the other side -- a lot of people on the "with" side seem to be doing the equivalent of "sure, an eight-ton rock fell on him ... but he had diabetes!"
My mother was in her late 80s in mid-2020. She had numerous health problems.
Then she got COVID and was dead within ten days.
Did those other health problems contribute to her dying instead of recovering? Absolutely ... and the death certificate says so, listing heart problems and high blood pressure as secondary causes of death. If I get crushed by a Mack truck while crossing the street, that was my main cause of death even if I was a walking heart attack / stroke risk.
I don't think such memes are actually worth bringing up. I haven't seen any actual evidence that traffic deaths were being miscategorized as caused by COVID, but people keep handwaving to that effect.
I, for one, find it frustrating to see the denial arguments equivocate between “covid statistics include deaths happening for unrelated reasons” and “clearly nobody is claiming this unrelated cause of death was being counted as covid, that would be silly and irrelevant”. I do realise it is different people saying the different things, but the overall combination is very draining.
And people also handwave that if there were any comorbidities, then it wasn't COVID that did the killing.
Those are the extremes. I agree that there's reasonable skepticism as to whether any particular death was "mostly" due to COVID or "mostly" due to something else. Some people disagree, on either end, and some of those disagreements aren't really very reasonable.
In my mother's case, there were comorbidities, but there hadn't been any recent emergencies with those comorbidities, so it's reasonable to think of COVID as the "tipping point" cause of death.
How many others got misidentified and not removed because they didn't receive mass media attention, who knows. But I would be highly skeptical that the "COVID traffic deaths" would be more than a couple thousand at absolute most, and even that's on the very high uncharitable end of estimates.
My understanding is that deaths attributed to traffic accidents (rather than COVID) also rose, and the common explanation was that with less traffic cars were driving faster, while Sailer argued that after George Floyd there was a decline in policing (Scott also argued that, he might have just focused on homicide). I was going to suggest look at the total number of traffic accidents and see if the fatality rate per crash changed, but if crashes were up because vehicles were driving faster, then you'd expect that to cause an increase in the fatality rate by itself and it could be hard to detect a fraction being misclassified as COVID deaths amidst that increase. I don't think it's impossible though. COVID deaths vary a lot by age, in a different patterns to traffic & homicide deaths. Sailer has also written about different patterns by race following Floyd, and while there was a time when COVID deaths varied by that as well, I don't think that persisted.
I don't really understand what you're saying here. Let me pick a concrete example to discuss: someone is 82, and has COPD (a complication of smoking that leads to chronically poor lungs). They come down with COVID and die 8 days later, because of respiratory failure (suppose they weren't an ECMO candidate). Suppose that, if they hadn't had COPD, then they wouldn't have died. Are you saying that they should be counted as a COPD death, rather than a COVID death, because they wouldn't have died if not for the COPD? That doesn't make sense, because they also wouldn't have died if not for the COVID infection. So you could say that they died because of COPD, and they died because of COVID, but you can't say that they didn't die of COVID without also saying they didn't die of COPD. If you say they died of neither COVID nor COPD, then I'm not sure how useful your notion of causation is.
They should be counted as so-and-so percent of a COPD death and statistically, X number of people dying this way should be counted as a certain number of Covid deaths and a certain number of COPD deaths, where each one is less than X.
We're trying to figure out how many people would have been saved if we were able to prevent COVID cases. So what we care about is the number of deaths that happened, but would not have happened if the person involved didn't have COVID - our hypothetical person is an example of this. It's also correct to say that our hypothetical person should be counted among the people who would have been saved if we were able to prevent COPD, but I don't think anyone here is talking about the costs of COPD. The discussion here is about the costs of COVID.
(To use the car crash analogy, someone walking across the street absentmindedly and being hit by a distracted driver could be said to have died from crossing the street absentmindedly and the driver being distracted. In a discussion of the societal cost of distracted driving, this death should be considered as 100% part of the social cost of distracted driving. A policy that prevents distracted driving would have prevented this death, for example, even though a policy to prevent absentminded street-crossing would also have prevented it.)
I think I need a little more than "someone somewhere profited from this pandemic" in order to entertain a theory where significant portions of the pandemic were faked. An obvious question is, for example, whether those selling the shots also faked the pandemic in Africa, which also saw significant excess deaths?
"...in order to entertain a theory where significant portions of the pandemic were faked."
And note that the faking and or conspiracy needs to be world-wide. Europe, China, Australia and New Zealand had to be in on it, too. As well as a lot of other countries. Maybe not Sudan.
Yeah. Did the weird lobbying corruption between pharmaceutical companies and politicians contribute to the rabid support in favor of vaccines and against literally any other possible cure? Almost certainly. Did it lead them to fake Covid or its primary attributes? Almost certainly not.
If I was a politician and had to choose between supporting either the methos of avoiding a disease or methods of curing the disease, I would, of course, prefer avoidance. Although I'd rather also support the cure, just somewhat less. Avoidance is better than cure because a disease, before it's cured, is likely to cause tissue damage in your body, much more than e.g. vaccines that also cause a little tissue damage.
this is really, really silly: vaccination is far and away the best approach humans have developed to addressing viral illness if you look at the track record of the past century. Anything other than prioritizing vaccines would be a very strange policy.
Anxiously awaiting sci fi where we live in a post scarcity utopia, but for a hundred powerful cartels that are constantly injecting planned obsolescence into everything to justify their trades. But the one man not in any secret society discovered their secret and the whole house of cards is about to come tumbling down!
"Excess deaths" is not a measured quantity, to compute it requires a fairly sophisticated model of expected deaths. This, perhaps, explains why a number of countries, including Sweden, reported negative excess deaths during the covid pandemic.
If a novel disease arises there will be a large incentive to treat it. Whoever develops that treatment will profit. This is not a problem. It's the way the system is designed.
Should people not be paid for developing life-saving medical treatments? I have my criticisms of pharma patent trolling, but developing a treatment that craters mortality from a widespread disease which kills 1-2% of those infected seems worthy of compensation.
Criticizing the pharmaceutical companies for producing life-saving medical treatments is a really bad idea. IMO, the vaccine makers should have gotten a lot *more* money, maybe a percentage of the GDP loss avoided.
Obviously we need to take care not to incentivize them to produce ineffective or harmful treatments, but that part of the system seems to have worked quite well -- though I do wonder if perhaps in a situation where hundreds of thousands are dying every month we should find a way to reduce the required testing time a little more. The COVID vaccines were created in January of 2020 but not deployed until December of 2020, mostly because it took ~8 months to test them. With the benefit of perfect hindsight we can wish that we'd put them into production immediately and started vaccinating medical personnel in maybe March or April, and mass vaccinations in June or July. Of course, that would be foolish because they might have been ineffective or dangerous; testing is important. But we can and should ask how much risk would have increased with shorter testing periods and whether we should have traded that risk against the ongoing deaths.
In the future, as we gain experience with mRNA vaccines, I'm hopeful that we can establish baseline safety standards for the mRNA carrier component so all we have to do is differential testing of the effectiveness of the encoded antigens. That probably wouldn't reduce the time it takes to test for effectiveness, but it might significantly reduce the time it takes to test for safety (which has to be done before effectiveness testing can really be done). Also, using a standardizable strategy like mRNA might enable standardization of production facilities, enabling production lines to be converted in days, maybe hours. Similarly, it should enable standardization of shipping and storage. Taken all together, it seems possible that mRNA could make it possible to begin delivering vaccines for a new virus in as little as two months after the virus is identified and isolated.
And we really want to make that sort of rapid development and delivery highly profitable for pharma, because it would be extremely beneficial for humanity. Imagine if every doctor's office and pharmacy in America had ample stocks of the vaccine in March 2020. We'd have had no lockdowns, no masking, no kids struggling to learn at home, no supply shock... most everyone would have just gotten a shot and gone on with their normal lives, avoiding a million deaths, and trillions in economic losses and federal deficit spending. Pipe dream? Maybe not.
I also believe that we could have started vaccinate people against covid in July or August, 2020.
It could be with less known safety, more experimental while other studies are being done but for the most vulnerable it could still be a big gain. The benefit/risk ratio would have been quite high.
It wasn't done mostly because covid wasn't such a big deal. It was causing more panic than it deserved.
The UK removed covid from High consequence infectious diseases list.
It is just formal decision but in practice experts in the field didn't think of covid as something that requires disregarding usual safety rules for vaccine trials etc. This highlights the disconnect of what the society including most politicians and rationalists thought about covid and what the highest experts in the field thought about covid.
Obviously, most people on this site disagrees with experts. But that's exactly the point, disagreement with experts due to emotions and not expert knowledge or thinking.
Very few people are antivax. Maybe 5% of parents don't vaccinate their children. We don't have antivax laws, yet. Who knows what will happen with Trump.
Whereas lockdowns were wrong and yet forced to all as the policy. Evidence deniers had won temporarily.
Instead of sounding like a literal strawman saying "DO YOUR OWN RESEARCH SHEEPLE", do you want to steelman yourself and provide some information about the royalties and how it goes beyond "pharma company makes new vaccine, gets paid"
'Go research this phrase' generally means that the point isn't for you to learn a true fact, the point is for you to encounter the same propaganda and go down the same rabbit holes and fall into the same community that the speaker did in order to get them where they are now.
(But seriously, when I read this comment I was confused because it took me a few moments to parse this as about a vaccine rather than a photograph. Neither set of royalties causes deaths though.)
Good analysis, and far more than I would have bothered to do, which is “excess death rate, QED.” I think when numbers get large it is very hard for people to have an intuitive sense of what they mean, so they latch onto poor heuristics like “do I know anyone who died of Covid?” Relevant: the average American only knows 600 people (according to the NYT).
The average American may know only 600 people, but each of those people has ~2 parents, ~1 sibling, ~4 cousins, ~1 spouse. What's driving my skepticism on these numbers is that nobody I know in my office reported that their mom died of COVID, or that they needed a day off for their aunt's funeral. I didn't observe an uptick in "R.I.P. Uncle Joe" posts from my 300 Facebook friends. I'm not going to argue that my experiences are better than hard data, but still, nothing in my experience is consistent with the data.
This is Forbes data from Jan. ~1% of used car sales in 2023 were for the Chevrolet Malibu. More than 90% of American households have a car, and nearly a quarter (~22%) have three.
It stands to reason that a buttload of those households have Chevy Malibus, and I know I don't have the math skills to just figure out how that shakes out but I do know I see plenty of Malibus on American roads with me.
Shifting gears, I can say from experience in my work that, where we look for (among many other things) size differences between the pupils of the eyes. It's anecdata, but around 25% of Americans have a substantial enough difference in pupillary size that we have to investigate it. It nearly always ends up being nothing, a subclinical data point that enters the chart and sleeps there.
The point I want to make is that people's lives are full of facts and statistics, much of which is publicly available. We nevertheless miss that data or discard them because it's not relevant, apparently useful, or doesn't serve some immediate purpose. But, those rare things happens *all the time*, because there are so vastly many of us.
Did you know that migratory birds have been observed having an average of 3 snails on them? Over the long epoch of several million years, snails accidentally hitching rides on (or inside of) birds could explain much of the diversity of snails on Hawaii, a set of islands far removed from, well, anything. How in the hell would you know something like that unless you were A) a malacologist or B) very bored this morning while waiting for patients to filter in?
We only know this stuff because we take the time to look. It doesn't rise to conscious awareness because there's no obvious need for it, unless you're motivated to look or so bored that you'll read about snails.
Okay, but whether my secretary needed to take a few days off to attend her sister's funeral is something that I actually do have reason to pay attention to, unlike her pupil diameter or the number of snails on her boobie.
Some years I go to four funerals and others I don't go to any. The claimed increase in the death rate is 20%, or maybe a little less. I'm not convinced that this would be noticeable at a personal level, at least not over a short period of time. And then, as I get older I expect I'll probably go to more funerals anyway... Population statistics are very different from personal experience.
I don't know what the mean, median, mode, and standard deviation of funeral attendance in your life is, but I bet you'd be skeptical of the 1.2 million excess death total in a year that you went to zero.
If you average 2 funerals a year, then the chance of a year with none is 13%. If an extra 500k people died in the USA in 2020 then that means the chance of no funeral is 9%. A difference that is far too small to draw any conclusions from, on a sample size of one.
In my own personal network. My kids' pediatrician died of covid. My husband's friend lost his father (not sure how old but I think under retirement age). Our family friends in Queens NY lost their grandfather to covid. I followed composer Sarah Duke on Instagram, her husband was young and healthy and was brutally hit by covid and she shared the whole journey until his death.
I remember vividly seeing pictures in March or April 2020 of a funeral home in Brooklyn with covered stacks of bodies because they couldn't keep up with the funerals fast enough. My social media feed was full of quiet small funerals outdoors from friends of friends who lost relatives. That was only the first few months of the pandemic really but it makes me baffled why people don't remember those days. It was publicly shared information.
I expect people know most of the 600 very tangentially, so that they are unlikely to hear about their family members dying. And of the people they know more closely, the 600 often already includes their family members.
The excess deaths were some 20% of the normal amount of deaths, so the uptick shouldn't have been obviously noticeable.
How many Facebook friends post about a death in any given year? It might be interesting to actually gather statistics on your feed, rather than going off a gut feeling about how many you think you see in a normal year, and how many you think you saw in 2020 and 2021.
There were entire subreddits devoted to posting the dying Facebook posts of “uncle Joe”. In a lot of cases the family would be completely unwilling to admit that their loved one died of Covid because they had for months been loudly proclaiming to their echo chamber how it was a hoax to begin with. No one wants egg on their face while grieving.
The fact is you probably do know folks tangentially related to you who had a loved one die of Covid, but the same social milieu that makes you likely to question the official figures would stigmatize advertising your loved ones death from the virus.
My observation was not that there were few people stating that Uncle Joe died of COVID. It's that I saw few reports of Uncle Joes dying. And given that my social milieu - Facebook in particular - comprises mostly educational (very liberal college) and professional (left leaning field) associates, it skews liberal beyond second-degree consanguinity. I feel like I was paying pretty close attention to whether anyone I knew had relatives dying because I was very concerned about my elderly parents and in-laws and my infant contacting COVID.
I'm not using my experience to dispute the data, but at the same time I'm not going to accept someone telling me that experience didn't happen the way I remember because of some aggregate data.
I’m not sure what your point is. If you don’t dispute that the deaths happened, then you need to accept that your experience is an artifact of reporting bias or exists as a statistical outlier.
Also, your Facebook newsfeed is not a random sample. Even if we grant that you were putting an insane amount of your attention into keeping track of the health status of your vague acquaintances, that information very well may not percolate its way to your newsfeed in the first place.
If my father had died of covid I certainly would not have told my "600 friends" because some percentage of them would have felt compelled to tell me covid was fake news and he couldn't have died of covid, then some other percentage would have felt compelled to scream at those people for reasons. Instead I would just have said my father passed away and kept the reason to myself and maybe privately told 2 or 3 really close friends that I trusted to accept it.
“I like these less because they’re class- and location- stratified, so your chance of knowing them goes up or down a lot depending on your own characteristics.”
But COVID deaths *were*, in fact, class- and location-stratified, certainly more than MS - further explaining how some people (think they) know nobody who died of COVID.
Example: my aunt worked at a nursing home in NY which had something like 16% mortality in the March - May 2020 wave. She knew like 30 people who died of COVID.
Most of the conspiracy theories at this point are elaborate and require a great deal of complexity. I am not sure there is much value in arguing against them.
One of the worst parts of this community is how people will, without providing any context or explain how it is relevant, lazily link to an old post from Scott. Both lazy and condescending.
The post in question is Scott arguing that conspiracy theories are worth arguing against rather than ignoring, even if they seem very wrong. The current post is an example of Scott doing exactly that, and the commenter above is (much more concisely) saying roughly what Chris Kavanaugh was before.
I think that conspiracy theories are worth arguing against, but only by people who are already experts on something related. If the topic is new for you, it is probably not worth your time researching everything from scratch. You would basically have to out-learn the conspiracy theorists, while the expert is already ahead of them.
So I am happy with Scott posting against conspiracy theories, but there is no way I would try researching the same thing to make my own informed opinion on the topic -- it would take literally years to get to a useful level of medical knowledge, and I don't have that much free time.
Just because "many of [you] in the audience" approve of lazily posting a link and giving no context or synopsis whatsoever, doesn't mean it's good practice to lazily post a link and giving no context or synopsis whatsoever. This place is supposed to have etiquette above that of your typical subreddit; and "Scott said this, QED" is below the level of thought this place is supposed to exhibit as well.
A related argument (to the died of covid / with covid controversy) is that the death of an 86 yr old is not the same as the death of a 34 yr old: in the first case covid merely hastened what was inevitable, depriving the old person of at most a few low quality-of-life years. In a sense an old person is always dying _with_ whatever condition is the proximate cause of their death. It would be interesting to see how many quality of life adjusted years we collectively lost due to covid and compare that to car accidents or smoking.
You seem to assume there is a social benefit from years of college, but Bryan Caplan's "The Case Against Education" argues otherwise. Our own host has discussed how wasteful education for doctors like himself is quite popular even though the example of Ireland shows undergrad isn't necessary for them https://slatestarcodex.com/2015/06/06/against-tulip-subsidies/
My one link was to Scott Alexander's post about whether doctors need to go to undergrad first. I merely referenced Bryan Caplan, who was making an argument against his own interest as a college professor. If you want actual studies on the social vs private benefit of education, here you go https://www.econlib.org/archives/2013/10/international_e.html
The fact that you enjoyed college is not evidence against that. You might enjoy it if the government provided you with hookers & blow, but that would just be a pecuniary transfer for your consumption. College is heavily subsidized on the assumption that there are spillovers to the public good.
TR wasn't even talking about "social benefit." He was talking about quality of life. You're the one who brought up social benefit and it was a complete non sequitur.
"You seem to assume there is a social benefit from years of college..."
The two things being (implicitly) compared here are not two years of college vs. two years doing something else such as working or being in the military. It is two years of college while either (a) being on campus and interacting with the people there (profs, fellow students, police ...) or (b) doing this all via Zoom from the student's childhood bedroom.
(a) is a much better way to spend two years of one's life than (b). Especially if the student has carefully chosen the college to be near, say, a beach (e.g. Santa Barbara Community College, UCSB).
College, and young adulthood more generally, is when people are (hopefully) building the foundations for the rest of their lives: getting qualifications, gaining experience of independent living, perhaps even meeting their future spouses. Even if you think we send too many people to college -- and personally I'm inclined to agree on that point -- majorly disrupting someone's life at this stage is going to potentially have far bigger ramifications than letting his grandmother die at the age of 82 instead of 86, callous as it sounds.
And of course, it's not just college students who are affected. Anecdotally, lots of teachers in both the UK and the US report a big uptick in mental health problems amongst their students, and whilst maybe this would have happened anyway (anxiety and depression were on the rise even before COVID), it's at least plausible that spending several of their most impressionable years being kept away from their peers and told that they might end up accidentally killing granny if they went outside could have exacerbated the problem. For older people, lockdowns might mean a two- or three-year delay in getting married, which in turn might mean they're now too old to have children, or too old to have as many children as they otherwise would. And so on. Life isn't like a computer; you can't just switch it off and switch it on again and expect it to keep working like before.
Not only that but grandma's QOL during the additional 4 years takes a bit as well: she can't see family, she is essentially a prisoner in an institution, and instead of seeing her grandkids enjoying college and building a foundation for their adult lives, she seems them holed up at home listening to Zoom lectures.
This is correct and massively under-discussed. Extended solitary confinement is a terrible punishment imposed on the worst prisoners. Yet not only is it dismissed from the calculation of NPI costs, people actually pretend that the elderly should be grateful for it. Someone who's 85 has probably 5-7 more years left. Locking them down for a year or two to supposedly avoid Covid not only plausibly hastens their demise but also turns a significant fraction of their time into a life hardly worth living.
I view the idea that letting someone older die 4 years sooner than they otherwise would have (an underestimate to be sure, but we can run with it) is not worth the in-person schooling experience of a 19 year old as monstrously callous. It's strange to me that you take it as self-evident.
If you were told that you had a choice between a 19 year old having to do freshman year remotely at a college or putting a bullet in a 80 year old's brain, you'd think the choice is obviously do the latter because the value of an old person's continued life is minimal?
That's not what I meant. I mean in practice in the real world, who would get to decide that without creating incredibly terrible decisions. Even if everyone could agree on a definition that covered all use cases (which is impossible) there would surely be abuses of that system.
I agree. And there's a lot of fulfillment in living in your 80's and continuing to be present for your loved ones compared to some restrictions on the range of experiences in college one might experience. Saying "quality" doesn't mean your views on the value of a life are correct.
You have it inverted. I was saying in addition of the value of life to the the older person who is alive and still able to have fulfilling experiences, their continued existence is also fulfilling to people who care about them. It's part of the overall benefit of their not dying that you are weighing against the value of a more vibrant college scene.
Grandma not dying is of significant interest to grandma, but it's also of interest to people who happen to care about grandma. That's part of the overall utility calculation you are trying to invite. Why, I'd be willing to bet that odd as you might find it, there are lots of people who would be willing to trade off attending college remotely rather than in person for a few years in order for their grandma to live 4 more. This might be a foreign idea to you - "insane" in your own words - but that's an ordinary way to order one's desires.
Agreed! Thank you for saying this! I have no problem with someone making the argument by posing the question of whether age should be taken into account when healthcare has to be rationed and under what conditions. but to claim that for anyone to say that "old people's lives matter" is insane and madness?
Plenty of old people live meaningful and fulfilling lives (my grandfather traveled the world throughout his 80s and was healthy enough to play baseball with my kids in the park at age 89, just saying).
But even if they didn't! At what age do human lives stop mattering? Why is a 79 year old more valuable than an 81 year old? A 75 year old? A 69 year old?
Should a 19 year old give up college to care for an elderly parent? People do that all the time. What if there's no one else able to care for them? Are they wrong? This just makes no sense.
This philosophy logically extended turns into a horrible dystopia very quickly.
I'm familiar with it thanks. You didn't address any of my points. I wonder if you're serious or trolling. Not because of your views but because you don't admit that it's your position which is outside the mainstream and requires a defense not the other way around.
Do you really believe that all 80 year olds automatically have bad quality of life?
Can you say something about how the repugnant conclusion relates to the debate of whether it is worth sacrificing in-person classes to delay the deaths of senior citizens?
* You compare "a 19 year old" to "an 80 year old," but in fact the 19 year olds in college way outnumber the octogenarians who died
* You assume that closing college is a 0%-100% choice in whether an 80 year old dies of Covid, when we don't even have any good evidence that it made any measurable difference at all
I'm following the comparison of the person who I replied to. I was not the one who introduced the direct comparison. I agree that it doesn't properly model the actual tradeoffs, but it's not my comparison to start with. The underlying issue is someone who decided that 80 year old lives aren't worth much and it is "insane" to think they are. They invite you to judge years of an elderly life against in-person attendance at a college and see that the latter is self-evidently of higher value.
We are not making a choice here between whether a 19 year old or an 80 year old gets to live. In reality, we were making a choice between a set of experiences young adults olds might have vs. some increased risk across the general public that they'll experience death or a significant reduction of quality in life that happens to increase significantly with age. There is lots of evidence that mitigative behavior collectively reduced said risk.
You should also reverse the scenario. If you were 80 years old, and you were going to die soon but you had a 1/6 chance (my upper-bound estimate of the ratio of college students to elderly people who counterfactually would have gotten and died from covid without lockdowns) of extending your life by a few years by making your grandchild do their freshman year of college remotely, would you do it? I would imagine probably not.
Of course, the college student who plays russian roulette with her grandfather's life so she can go party on campus for a year also comes out of this looking pretty bad.
I think this is just "selfishly imposing costs on others for your own benefit is bad".
The scenario already allowed for the reverse, since it was presented simply as a choice for a 3rd party. The participants are not the ones imposing the consequences. Kill the 80 year old who would live another 4 quality years is the meaning of "soon" and the choice was doing that or not compared to the offered tradeoff. This is not realistic, but it is the contrast that was set up by the person I initially responded to who was trying to communicate just how worthless years lived beyond a certain age are.
If we change the scenario entirely in the way you offer, someone extending their life four years by transforming a young person's in-person schooling to remote is defensible and is ultimately a vastly oversimplified version of what public health in the face of pandemic has long done with various forms of quarantine.
At 1/6 (my absolute upper bound on the estimate of the number of life years saved by lockdowns per year of remote schooling for university students) it is defensible, and approaching an absolute slam-dunk win from a public health perspective. At 1/1000 (my lower bound) it is not defensible.
I personally think the public health people made the wrong call with lockdowns but I don't think their call was *drastically* wrong (though separately I *do* think they made terrible calls with what particular things were allowed and forbidden).
Part of the problem that occurs in these discussions is people have ambiguous use of the term "lockdown." America never really had "lockdowns" in the formal sense at all, but the term was picked up early in the pandemic as it was describing the policy of significantly more restrictive nations. The closest the US came to lockdown-like behavior lasted for a brief period in the spring of 2020.
That said, we did have a series of both formal restrictions and informal recommended behavioral changes that were mitigative, but inconsistent with each other and suboptimal. This is what people usually somewhat unartfully call "lockdown." Slapped together, these collectively were effective enough to keep the seroprevalence down until vaccination rates climbed in a way that saved a very large number of lives and long-term health problems. At the same time, they were drastically inferior to what a nation like New Zealand was doing and the gaps likely cost a large number of lives and caused a lot of heartache among the still living. The US response sat in the zone of "could be a lot better" but also "could be a lot worse." It's sad when you think about the conventional wisdom prior to this was the US and its once vaunted CDC were among the best situated to weather a pandemic.
If you were to knock out any one pillar in the US strategy in suppressing community spread, in isolation it wouldn't necessarily have moved the needle a large amount. Even if all the legal restrictions were removed, the specter of death would've still had many people behaving with more caution. The problem with trying to isolate any one thing and estimate its contribution, however, is that mitigative recommendations were interlocking. They contributed to each others' effectiveness and communicated a sense of urgency and danger that unto itself also likely influenced propensity to safe behavior. An answer to any one particular "was this necessary?" is generally, "sort of, yes" with an added, "and you can only pull so many blocks out of the jenga tower."
Your imagination apparently works differently than mine! I usually try not to speed in residential neighborhoods, even if the people are old, and I assume even old residents would like me to do that.
... I also don't speed in residential neighborhoods. I don't think I am understanding how your analogy works - are you saying it's sufficiently selfish for old people to be against speeding that it should outweigh their own concern for their safety?
Or do you object to the more general idea that we tend to take a dim view of people choosing the side of the tradeoff that benefits themself, even if the benefit they get is larger than the costs to others (less so if the cost is trivial and the benefit is large, as in your speeding example but not as in the original example)?
Imagine telling an 18-year-old: “If you do your first two years of college remotely, you’ll live four years longer.” They say, “Yeah, that sounds like a fair trade.”
What mental illness are they suffering from, exactly?
"It would be interesting to see how many quality of life adjusted years we collectively lost due to covid and compare that to car accidents or smoking."
I did a back-of-the-envelope calculation for this a few years back with the intent of comparing Covid to other types of deaths that people are familiar with. I did this using the California excess deaths per age group and then combined it with life expectancy for that age group.
Going from memory, Covid (at the time, but we were fairly far into it) cost the average American about 5 weeks of life expectancy (more death amongst the elderly, but the 'cost' in years was lower ... interestingly the lost years were pretty flat between most age demographics [maybe within 2x]; fewer 25-34 year olds died, but they lost more years of expected life when they did; I didn't try to adjust for health ...]
I was unable to find a reasonable comparable, though. This also works out to something like a "double your miles driven per year for 10 - 20 years" risk, which I find to be unhelpful.
Smoking for a lifetime probably, on average, costs you more years of life than Covid. Driving fatalities per year are much less. I don't know of anything 'common' in the middle.
Ooo, I wonder about the lock downs vs quality of life years.
E.g China had strong lockdowns and only 5,272 listed deaths.
India has a similar population and 533,570 deaths.
So we could say that China saved about 500k lives.
I asked Perplexity to work out the lockdown duration in person years. E.g the population affected by the lockdowns with the duration applied, e.g 10 weeks in Wuhan.
It gave 14,431,091 + 573k (it didn't think it had strong data for).
So just over 15 million human years.
So to compensate the people who died would have had to have lived 30 years in average to have made it one to one.
However life under lock down isn't the same as years not lived.
Personally I'd just had a new child and work from home. So we weren't going to be going out anyway. My memory is that we had maybe a 5% quality of life reduction due to the intense lockdowns in the Philippines. It was worse for 2x two week lockdowns where I had to line up to get food from the supermarket and apart from that couldn't go outside.. I couldn't go for a run outside and had to do beep test style running back and forwards inside. But when partly opened up there was almost no car traffic and running on the streets was way nicer than now. Less traffic and pollution.
Wearing a face mask is something we were already doing outside. Wearing a face shield for a while was a little annoying.
However my wifey's Mum had a stroke and then caught COVID in the hospital and died 2 weeks later whilst still being ravaged by COVID (delta at the time, it hit us HARD).
Thankfully Perplexity is better at researching than me and suggests the China style lockdowns were worth a 30% Quality of Life reduction. Although the global average was 12-18%
So 30% of 15 million is 4.5 million years.
Now the 500k people should have lived at least 9 years or more before it turns out to be a net positive.
Perplexity suggests that the Quality is life years lost in India was 8 million.
Or to put that another way, the lockdowns in China were 56% (basically half) the list quality of life years as compared to that of the deaths in India.
But it's different compared to USA stats.
> The U.S. COVID-19 death toll of 1.2 million resulted in 9.7–13.2 million years of life lost, depending on adjustment criteria.
The USA lockdowns affected 27.9 million person-years.
The USA quality of life adjustments of the lockdowns are listed as 15%
So 27.9 x 0.15 = 4.185 million years of life adjustments.
So the death toll was over double that of the lockdown toll in terms of quality of life.
They could have certainly pushed lockdowns harder and faster (as part of the Hammer and Dance) to reduce the death toll related Quality of Life years.
It feels like to me that normalising to Quality of Life years for the lockdown vs death toll is a good way of post-hoc analysing the results and comparing.
There's of course going to be those people badly affected by COVID lockdowns (e.g Travel agencies or resorts) who probably didn't know anyone or almost no one who died.
The lockdowns also disproportionately affected lower income people and those in less developed nations, like Bangladesh. Whilst the virus more targeted the elderly.
However countries applying counter measures for this, like providing stimulus checks seems to have somewhat offset the cost imbalance. Or those like in Sweden were the least affected because of the good social support.
China and India have similar populations but very different demographics. ChatGPT thinks there are 280M vs 159M people over 60, and ~55% vs ~35% of 40+-year-olds. So I think the expected number of covid deaths in China with Indian levels of lockdowns should be at least 1.5-2 times higher than in India.
I remember a back of the envelope calculation that suggested it was pretty comparable in QALY impacts to the opioid/fentanyl epidemic (fewer deaths [though still 100's of thousands], but more QALY's lost per death) over a similar time frame. Not sure what to think of that in terms of our response to either of the problems.
"Bottom line: Using CDC opioid-involved overdose‐death counts for 2020-2024 and the most recent peer-reviewed estimate that each fatal opioid overdose removes ~38 healthy life-years on average, I estimate the U.S. lost ≈12.7 million quality-adjusted life years (QALYs) to the opioid epidemic in the five years 2020-2024. Even under conservative assumptions this burden remains above 11 million QALYs, underscoring that opioid toxicity has erased more healthy life in this short span than many leading chronic diseases."
That basically matches the estimated 12 million QALYs lost to COVID.
Interestingly that almost exactly matches the O3 estimate of QALYs lost to the opioid epidemic over the 2020-2024 period. Roughly 1/4 the deaths, but about 4x the QALYs lost per death.
"Bottom line: Using CDC opioid-involved overdose‐death counts for 2020-2024 and the most recent peer-reviewed estimate that each fatal opioid overdose removes ~38 healthy life-years on average, I estimate the U.S. lost ≈12.7 million quality-adjusted life years (QALYs) to the opioid epidemic in the five years 2020-2024. Even under conservative assumptions this burden remains above 11 million QALYs, underscoring that opioid toxicity has erased more healthy life in this short span than many leading chronic diseases.""
I rather doubt this number. Average age of US covid deaths was a little over 77. The average life expectancy of any 77 year old is around 10-12 years (depending on male or female). But covid deaths also disproportionately hit the sickest with many co-morbidities. I find it hard to believe that the average 77 year old who died of covid had an equal life expectancy to the average 77 year old in the general population.
It's probably within an order of magnitude (10x), but I'm not convinced it's particularly close like within a factor of 2. Just as an example, around 200,000 of covid deaths were nursing home residents. The average life expectancy for any nursing home resident is 1-2 years (all, not the sickest and weakest which you'd expect from the covid cohort). Over half don't make it 6 months. That's a lot to counterbalance with the remaining deaths.
That's why the average amount of life lost per COVID death has been estimated at around a decade. Certainly not as bad as Spanish Flu, which killed the young more than the old. Won't we be screwed if something like that happens again and people continue to be so stupid in their response.
It’s so weird to me that people see this as a case of leadership being stupid - it’s rather a case of transparency causing all the problems. Leadership did as well as in any crisis in all of human history, but because we see there were costs as well as benefits, people have decided that we don’t want leadership any more.
There is a basic immaturity to me in the view that doesn't accept:
1. the profound imperfection of human institutions across all history and places;
2. that uncertainty is a basic feature of existence that all the lovely leadership in the world cannot manage it away.
There is so much room for criticism, but the value of the criticism is so that we respond better then next time calamity hits.
Raging because the data is imperfect and complex or because the guidance in an emergency was contradictory or that some measures were counterproductive is to rage at reality as it is.
So many people worked so hard to take care of so many other people during this mess. It was traumatic and people are still recovering. From trauma, we can grow or we can freeze. We really do have some choice about that.
I get that this pandemic and the handling of it -- like 9/11 did for some other people or Watergate and Vietnam before them, and so on -- led a lot of people to lose faith in all kinds of things. And that losing faith (along with the willingness to hang in there with nuance) is a normal kind of grieving process.
But persistence in the belief that the world *should* have done more for me than it has is ultimately a childish view that needs to be outgrown in order to live the life we do have.
I think this is also the case (at least as Americans) where we really suffered from having 50 states as laboratories of democracy because people can point to ineffectual leadership on both sides to find reasons for their side rather than looking at the cases where the leadership pursued level head decisioning and made reasonable tradeoffs.
Interesting. I wonder how many life year-equivalents were lost due to lockdowns? 300+ million people locking down for a few weeks would surpass 13.5 pretty easily, but we would likely need to do some kind of adjustment for actual death verses the significant loss of [enjoyment, freedom, career advancement, etc.].
This is so strange. At literally NO point during Covid was I 'locked down,' and I lived in a liberal college town. The way people talk as though everyone was under house arrest is literally not true. It's so weird to me.
So in my state, entire industries were literally told they could not operate. Restaurants were either take out or closed. Other states forced the closure or churches and other public accommodations. Lots of people lost jobs over this, including a close family member of mine. People got arrested for going to the beach.
There were no people local to me that were getting arrested for going outside, but it was possible. My boss wrote me a letter that I kept in my vehicle saying I was an essential employee in case I did get pulled over. If the police were more supportive of the lockdowns they could (and often should) have been arresting people for violating the orders.
Maybe your location didn't involve such orders, or maybe the people around you just ignored them all.
I don’t see any example of lockdown in what you say - just paranoia that someone might be treated as though they were locked down. Is a snow day lockdown? What if you get to play outside in warm sunny weather during your snow day?
Okay, so China literally locked people in their homes, and the West didn't do that. I still don't have a problem with the term "lockdown" (mainly because it's what the leaders who implemented these programs and the journalists who wrote about them called them) to describe the significant restrictions on movement, business, recreation, and normal life that were mandated on ordinary people in a totalitarian fashion.
"I was still allowed outside of my home" is not a reasonable response to what anyone complaining about the lockdowns (outside of China) is worried about.
This is the sort of thing that GPT-o3 with search is usually great at. It gives an estimate of about 9 million QALY lost over 2020 to 2023. Note that QALY lost due to long Covid are explicitly omitted, since the evidence on that isn't great and the numbers are pretty up in the air. The sources seem reasonable, although I haven't checked them individually.
I think the argument can be stated as '1.2M Covid deaths did not lose nearly as much QUALYs as 1.2M totally random deaths would have, which is why it doesn't feel as bad as other events with similar body counts.'
I think that's true, but also it wasn't *all* old and sick people, lots of QUALYs were still lost here and there - even if we only lost the QUALYs we'd expect for 600,000 'random' deaths, that's still a lot, more than how we talk about it.
Also, of course, lots of QUALYs were lost to non-death... after a bout of Covid, I lost my ability to taste sweet, and it's never come back. Others have much more debilitating long-term effects.
My grandmother died a little over a year ago at the age of 99. The last few years of her life were certainly of low quality on several different dimensions, for a host of reasons I won't go into here.
She didn't die from or with COVID. But if she had died from or with COVID during the first year of the pandemic, she, I, and the rest of our family would have been deprived of the the opportunity for her to meet my son. And I know, for a fact, from conversations with her, that moments like that with her great grandchildren made her few remaining "low qualify of life years" worth it to her and her family.
This is what makes quality adjusted life years so fraught. I understand the need to make trade-offs within any system, and in healthcare that sometimes means prioritizing one life over another. But I also think it's pretty gross when people seem to write off elderly people's last few years of life as being "low quality". That's a value judgement that 1) you can't make in the aggregate and 2) you can't determine just by considering medical conditions.
The "quality adjusted" part is less important than the "life years" part, IMO. (And it's unfortunate that the whole conceptual space is now taken up by "QALYs", which imposes this inherently-subjective and extremely difficult-to-reckon concept of "quality adjustment" on what is otherwise a simple matter of arithmetic on actuarial tables.)
It's very reasonable not to discount the final years of an octogenarian's life on quality grounds, but the fact remains that if someone in their 80s dies of COVID, it's a loss of 5-10 expected years, depending. This is clearly less significant than the death of a young person who could live another 50, even without any dubious quality adjustment at all.
What accounts for the global discrepancy between officially 7 million (US makes up one sixth) to unofficially 20-30 million (US makes up something closer to its age-adjusted share of the global population). Is the gap mostly China or a lot of countries? Do death rates seem close in countries perceived to have reliable data? Should we expect undershoot of expected mortality going forward?
That's a good chart there. Russia was such a strange case, they were able to develop a perfectly good vaccine but almost nobody agreed to take it, so combined with an elderly population they had the worst excess mortality in the planet. I wonder if they've had any kind of post-COVID reckoning.
It truly is insane. Russian government managed to make one good thing in years, and yet that was the thing which provoked widespread "resistance" among all social groups. Russians would happily accept war, poverty, corruption, they would ask no questions about police violence or shutdown of free speech. But vaccines? That's where they draw the line.
I don't think 70% approval rating is a meaningful figure. It can unpack to something like "Do you prefer Putin to the criminal 90s or some other form of political instability?" or "Do you prefer Putin to the clowns he didn't remove from public politics in one way or the other?"
OTOH, Russia has a long history of "The tsar is good, it's the boyars who screw us up". And it wasn't Putin who personally developed the vaccines.
So I don't know why Russia had high levels of covid/vaccine skepticism but this is not a contradiction.
It literally _must _ be yes, mathematically. Everyone is eventually going to die. So if some people die early, that is fewer people available to die later. No matter what the cause for an increase in deaths, there must be a corresponding decrease in deaths at some point in the future to balance it out.
India (3x the population of the US) had less than 1/2 the total reported Covid deaths of the US (~550,000 vs 1.2 million), so maybe 1/10th the per-capita death rate.
India is poorer, but has fewer overweight and obese people.
Reporting is probably poorer.
I don't think anyone knows if the number is correct, or off by 10x.
This is *deaths* attributed to Covid. I'm confident about the death part :-)
But he US might well assign more deaths to Covid because testing was much more extensive and India might just see a lot more deaths but not score them as Covid.
One would need to look at overall Indian (and Latvian and Greek and ...) deaths over time. I'm sure the data is available. I don't know where to get it.
My guess is that it exists but too small a signal to measure vs noise. If 1.2 million undershoot is spread out over the next 20 years, it could be just a 2% decrease, and it bounces around about that much and will certainly be affected by other health trends and medical progress.
Given that everyone alive during Covid will die eventually, then yeah, any given person who died during Covid means some year is eventually going to have one less death than it would have otherwise.
"My guess is that it exists but too small a signal to measure vs noise. If 1.2 million undershoot is spread out over the next 20 years, it could be just a 2% decrease"
With the bulk of the deaths clustered in the 65+ age demographic (and that clustered in the older of *that* demographic) it should not spread out over 20 years. Lots of people who would have died within, say, 5 years of 2020-21 should have died in those two years. I'd expect that to show up soon-ish rather than spread out over 20 years.
Maybe an undershoot would be more visible if looked at in age-bins (so no obvious undershoot for the 25 year olds, but an obvious undershoot for the 85+ crowd)?
I think you are underestimating their life expectancy.
The life expectancy for a 65 year old male is 17 more years. The life expectancy for a 65 female is 20 additional years. Those are averages, so 50% will live longer.
Even the average 80 year old has 10 more years.
Counterintuitively for some, the longer you stay alive, the higher your predicted lifespan will be.
The average female newborn will live to 79. The average female 79 year old will live to 89.
I remember COVID skeptics being asked to bet if mortality would subsequently undershoot due to getting rid of the "dry tinder", but I don't recall anyone taking the bet.
The issue is the terms of the bet. Let's say 1.2 million dry tinder might have otherwise had a mean life expectancy of five more years in the no-pandemic counterfactual, so, to simplify the math, assume 120k undershoot of an average 3 million deaths, which is -4%. The difference between 2019 (before covid) and the average of 23 and estimate for 24 (which are pretty close and so plausibly "post-covid") according to CDC is around 230k or +8%. Are expectations of future mortality numbers good enough to know precisely whether that gap otherwise should have been 50% larger? I doubt it. And if the excess death undershoot is spread out even more and unevenly while other things are changing too, it just doesn't seem the kind of dispute easy to resolve with bets with terms that are both well-defined and epistemically reasonable.
All that being said, I think something to look for in the future may be claims of progress in mortality, longevity, etc. - where small percent changes are the norm - that could be largely mere artifacts of long slow "mortality dividends" of dry-tinder undershoot.
There are a lot of countries that just don't have the infrastructure to begin to put together a count because of unavailability of tests, or just simple nonexistence of a public health system to collect data. e.g. when a man dies of covid a village in South Africa he doesn't go to the hospital and there is no coroner to record the death.
Also there is the issue of age. Over 65 makes up 90% of covid deaths. The US has about twice the percentage of people over 65 as the world population, so our death rate is just going to be higher no matter what.
Then there is China who had such a draconian lockdown that they were able to actually halt the spread of the virus until after the vaccine and the much weaker but faster spreading omicron strain came out. So, even if they reported true statistics, which I doubt, their death rate would have been massively lower than the US. And China is 1/6 of the world population, so that's going to skew the numbers a lot.
It’s probably mostly low to middle income countries that don’t have great statistics. China likely had very little undercount until late 2023 when they ended their zero covid policy - though they probably have an undercount in the month or two after that.
Total death numbers are likely fairly accurate in most countries, though there are some (like Somalia and Venezuela) where I expect we have few if any meaningful demographic statistics.
It’s probably mostly low to middle income countries that don’t have great statistics. China likely had very little undercount until late 2023 when they ended their zero covid policy - though they probably have an undercount in the month or two after that.
Total death numbers are likely fairly accurate in most countries, though there are some (like Somalia and Venezuela) where I expect we have few if any meaningful demographic statistics.
It amazes me how the rightward shift of the tech community has motivated so many previously intelligent people to accept conspiratorial ideas (like Covid denialism and antivax) they previously would have dismissed as being for low-IQ nutters. Nothing has actually changed about how plausible these ideas are, only how normalized they’ve become on the online right.
It's crazy how anti-vax jumped from the left to the right so rapidly. Before COVID it was Marin county that had third world measles vaccine rates and West Virginia with the highest in the country. But I guess that's just the tips of the horseshoe.
The antivax community has always been an exercise in horseshoe theory. It's always been granola essential oil hippies and religious homeschoolers sharing health tips.
It feels to me like the New Age movement has finally found its natural home on the right. The antivax movement’s roots in crunchy hippy counterculture once kept it mostly restricted to the cultural left. But as the hippies have grown old and cranky and normie conservatives have grown increasingly conspiratorial and less respectful of traditional authority, the distance between these two groups has grown narrower and narrower. Covid was just the catalyst that finally fused them together.
Maybe just because I'm in another country, but I don't remember dishonesty.
I remember some mistakes that were corrected - and, in one case, the public corrections never got through (that the whole "wash your hands/clean surfaces" thing where the later correction that it was airborne never got through to the public). And it was certainly the case that if you were following experts online then they were ahead of the official communications. But that's true of just about anything; official advice always takes time to catch up with scientific knowledge, if for no other reason than that it requires consensus.
But we got things like "vaccines prevent spread", quickly followed by "no they don't; they reduce it" and then "but not by that much" as each variant made the vaccine less effective than the previous one. Sure, they were still saying "vaccines prevent spread" for three or four weeks after it had become obvious that they didn't, but that's not dishonesty, that's the process where scientific research becomes official communications being very slow.
The main example I recall of calculated dishonesty is an ironic one for people who're generally COVID denialist. The public messaging at the beginning was that masks were not effective, because they were worried about an undersupply for medical workers and didn't want regular people buying them up. They reversed course once the supply was secure. There were some cases of pragmatic (if poorly considered) dishonesty like this, but it's all consistent with public officials trying to cope with a serious pandemic where the public often responds in perverse and unhelpful ways.
If they were so good at thinking ahead and dealing with the unhelpful masses, they should have predicted that lying to them in such an obvious way would cause then to trust then even less, making their good-intentioned interventions less effective in the future.
This is the entire point of having simple heuristics like “just tell the truth” instead of galaxy-braining every specific situation: because you’re probably not smart enough to do it properly anyway.
I never said they were good at it, and I do think this cost them valuable trust, but given ordinary human behavior, I think a substantial contingent of people were going to distrust them anyway.
Was this actually a thing somewhere, or is it just something people say?
If it was actually a thing, how much of a thing? Official country-wide government policy, a local low-ranking official making an unofficial tweet, or somewhere in-between?
My bet is some random media worker mis-spoke, or something like that.
I'll have to wait until I have more time to delve into this more, but as I recall the messaging on this point at the time was concerted and coordinated, but t not very long-lasting.
Okay, so the linked article seems to suggest the tweet occurred at a point where the US system thought covid wasn't properly airborne ("The WHO and the CDC were saying the same thing that I was saying, so the science supported me.") Or perhaps attempting to make a more nuanced argument about how the public masking wouldn't change much at that point. (e.g. if the incidence in the public at large is very low, then each individual mask won't stop many instances of covid - because the user probably wouldn't have been exposed /anyway/).
So it looks to me like this doesn't support the interpretation that this was officially sanctioned, calculated dishonesty.
"Was this actually a thing somewhere, or is it just something people say?"
The US Surgeon General posted on Twitter in late March 2020:
"... my office have consistently recommended against the general public wearing masks as there is scant or conflicting evidence they benefit individual wearers in a meaningful way..."
I'm not an expert at the twitter interface. Your link goes to tweet one of three. Am I right in thinking the other two parts are the images below?
If not, you ought to cite them too, for completeness.
If the images are parts 2 & 3 - then the wider picture seems to be that they were asking people not to wear masks if they were in a low risk situation while supply was also limited. It's not really a nefarious pivot or underhand to change advice when the situation also changes.
I agree the text in isolation was misleading, but that's a problem with the format, and the problem looks to me like a failure to realise how it would be interpreted rather than some sort of lie.
This is anecdotal, of course, but I don't remember any official messaging at all about masks not working. But I distinctly remember messaging that we should hold off buying them to ensure medical professionals had access to supplies early on, which of course implies that they do work.
It was specifically that masks didn't work for limiting transmission when worn by normal untrained people in casual settings, so the general populace shouldn't buy them because 1) medical staff needs them more 2) it won't do anything for you anyways.
This was, indeed, a pretty odd combination of ideas to hold in one's head. Since when do you need special training to cover your face? But that's what was said for about a month before it was recommended for everyone. I do think it damaged trust somewhat and I wish the government had just instituted formal rationing on masks early on to reserve supply for hospitals.
EDIT: To clarify, public health officials were not deliberately lying to mislead the public and secure their own mask supplies. There was a pre-existing belief in medical science that masking did not slow the spread of pandemic respiratory illness, influenced by an over-reliance on RCTs for evidence and a backlash away from 1800s miasma theory: https://pmc.ncbi.nlm.nih.gov/articles/PMC8504883/
Of course, at the very beginning of the pandemic when nobody had COVID except in NY and a few other places, it was basically true that wearing masks provided no benefit in (most) "casual settings." There was nothing to protect against for the majority of people. I don't know, but can imagine, it being said off the cuff as something valid today without any thought about the guidance being very contextual.
I stand corrected, though I think you're overstating his initial claims. He didn't say drug store masks don't work, he said they "[don't] really do much to protect you", which implies they do do something to protect you. And you removed the two qualifiers in your quotation, where he said "Now, in the United States, there is absolutely no reason whatsoever to wear a mask." This implies that there might be a reason to wear a mask if circumstances change or in another location.
I completely agree with your final sentence though. With rare exception public officials should always tell the truth, even if they think doing so will result in some negative consequences. The long term erosion of trust has severe consequences, some of which we're now seeing.
I can confirm that this initial messaging on masks not being helpful for the public absolutely did occur. I think the concept they were trying to communicate was more subtle: “Masks are unlikely to help the public because their exposure is less concentrated than what healthcare workers are exposed to.” But it came off as more black and white. And in retrospect was just wrong (“aerosol generating procedures” being much more infective than coughing fits remains a nonsensical idea that healthcare is somehow still holding on to) on a number of levels and was absolutely an unforced error created by panic. They were terrified we would run out of PPE.
There was a lot of dishonesty. I'm not talking about things that could be charitably interpreted as mistakes or bad calls, like when the NYC health commissioner said early on that there was nothing to worry about and people should continue to take the subway and congregate for the holidays, I'm talking about flat-out lies. I'm talking about things like the head of the CDC going before a Senate committee and claiming that if you get vaccinated you can't spread COVID.
When was that said? The vaccines weren’t expected to be neutralizing, but the Pfizer and Moderna vaccines were so effective against the original wild-strain SARS-CoV-2 that they were neutralizing. The other vaccines (e.g. AstraZeneca, Janssen) were not neutralizing, but they weren’t initially available in the US (and only Janssen was ever approved). I think that “if you get vaccinated, you can’t spread COVID” was therefore true until the Beta variant.
The non-neutralizing nature of the mRNA vaccines with the Beta variant wasn’t confirmed until 10 April 2021 when data from Israel first showed that people with a Pfizer vaccination were able to spread Beta asymptomatically. And Beta struggled to spread widely outside of South Africa and Israel; Alpha mostly outcompeted it, so it was reasonable to refer to “you” (ie the typical American) as not being able to spread COVID if vaccinated until the Delta wave in late June/early July.
It wasn't true. It was never true. She made the claim that "vaccinated people do not carry the virus, don't get sick, and that is not just in the clinical trials but it's also in real world data," in March 2021 based on a CDC study that showed nothing of the sort. What it did show was that vaccinated people had a 90% lower risk of infection two weeks after their second dose, which is distinctly different from a 100% lower risk of infection, and distinctly different from sterilizing immunity.
She made her claim, the one I quoted above, to the Senate committee in May of 2021. There was no evidence for sterilizing immunity. To make a claim honestly, you need evidence for the claim, not just a lack of evidence that the claim is false.
There was so much blatant dishonesty, that it astounds me when people act like there wasn’t, or “can’t recall” any. Why did they NOT lie about.
- origin
- masks
- mortality
- transmission effectiveness
- vaccine safety
- prevention of all kinds
- herd immunity
- slow the spread
- lockdowns
- protesting government response is “white supremacy”
- contradictory response to protests based on what the protest was about (transmission is not a problem if you’re at a George Floyd protest, but you’re a “white suprematist super spreader” if you’re protesting government tyranny).
- children must get vaccinated with a novel mRNA transaction therapy or they’ll DIE. In reality, children were not at risk.
Who is your “they” here? Because I didn’t see the scientists lying, I saw the politicians lying about what the scientists said (mostly, because the politicians were too fucking stupid to understand what the scientists said and were trying to “simplify” for non-scientists and ended up taking away so many qualifications that they weren’t telling the truth).
Origin is still the same answer as ever “natural causes more likely, can’t completely rule out lab leak, can completely rule out intentional spread”
Masks: cloth masks went from “no evidence they work” (which was true; no-one had ever studied it) to “better than nothing” (also true) to “still not very good and we’ve actually got some supplies of N95s now, so why not use them instead” (also true).
Mortality: went from “appears to be really high, but we’re only seeing the worst cases” to “1% or so” - and it appears to be a population-wide 0.3%, and there are a lot of people who either never got COVID or only got it after vaccination which definitely reduced the risk of death.
Transmission: well, yeah, the vaccine effectiveness was way higher on wild-type in the original trials in July 2020 than on Delta in July 2021. So of course the story changed. That’s not a lie, that’s the facts changing.
vaccine safety: how has the story changed? There are some risks that are either rare or minor, the same as just about every other vaccine.
Prevention: now, there is a real scandal here, which is the refusal of the scientific establishment to accept that COVID is an airborne disease and the total determination to stick to the idea that it only spread in droplets. That was a scientific fuck-up. But it wasn’t a lie; it wasn’t dishonest. They were just wrong and stubborn and wouldn’t listen to the people who had the evidence.
Herd immunity: is real, but requires you to get the reproducibility number under 1; vaccines alone weren't enough after Delta. Delta changed a lot of things and was worse than expected; so any statement made before July 2021 wasn’t a lie, just wrong.
Slow the spread: Yeah, this was a stupid idea - you need to keep the R number under 1 until you reach eradication. Riding the wave up and down was correctly assessed as a risk by an Imperial College scientific report early on and yet just about every country did it. So can we blame the politicians and not the scientists, please?
Lockdowns: These worked when implemented properly. Sadly, the countries that did lock down properly generally got didn’t push vaccination aggressively enough. So Australia, Japan, Thailand, etc all had a massive wave in 2022 (or China in 2023). The only country that did get everyone vaccinated before unlocking was New Zealand, which had one of the lowest deathrates anywhere (the lowest of those with trustworthy statistics).
Floyd protests: yes this was bloody silly, though the version I recall was that the Floyd protesters were OK because they wearing masks and the anti-NPI protestors weren’t OK because they weren’t - but I admit I’m in a different country and that just never seemed like an especially big deal.
“They,” just to start off, would be Dr. Fauci (NIAID), Dr. Robert Redfield (CDC), Dr. Tedros Adhanom Ghebreyesus (WHO), Dr. Deborah Birx (White House Coronavirus Response Coordinator in 2020), Dr. Rochelle Walensky, (appointed CDC Director on January 20, 2021).
Just to name a few (the heads of major scientific institutions and groups managing public health).
All lied. All deliberately mislead. All were repeatedly, demonstrably wrong.
Those are the political leaders of scientists, they were not actively working scientists. Not one of them published a single peer-reviewed paper on COVID.
Then you weren't watching. The scientsts you refer to were telling each other things in their own personal emails which they were directly denying in public.
The people who don't remember it are absolutely sold on the idea of not believing it, and have (subconsciously) taken every effort to forget anything that isn't The Narrative. The rationalizing will abound as it does with our dear host.
As a foreigner you probably didn't see it. I kept a list.
• Covid will not be a deadly pandemic
• Closing the borders is racist
• Masks don’t work
• Lockdown two weeks to flatten the curve
• Trump didn’t close the borders fast enough
• Lockdown three months to flatten the curve
• Masks work
• Lockdown until Covid is gone
• Covid CANNOT be solved at the state level
• It was never about Herd Immunity
• Saying the virus came from a lab is racist
• One mask doesn’t work but two masks definitely work
• Lockdown because symptomatic spread will kill old people
• Asymptomatic cases can still have long term damage, ground glass lesions, and such
• Lockdown until a cure
• Every hospitalization with Covid must count as a Covid hospitalization regardless of the reason for admission.
• Don’t trust any vaccine announced under the Trump administration
• The vaccine announced the day after votes were counted is safe and effective
• Vaccinated people cannot transmit Covid-19
• We must mask children because science says so.
• The vaccine has no side effects
• We will not mandate the vaccine
• We must vaccinate as many people as possible to get Herd Immunity
• You can return to your life after getting the vaccine
• The virus probably came from a lab
• Only N95 masks work
• The NIH doesn’t fund Gain of Function research
• The vaccine was never about stopping transmission
• Vaccine side effects mean it’s working
• The vaccine has no effect on women’s menstrual cycles so stop being hysterical
• The Wuhan NIH research doesn’t count as Gain of Function research because some gain of function researchers decided those gains of function didn’t meet the legal definition
• Asymptomatic transmission among vaccinated people is fine, will not kill old people, and don’t worry about long term damage, ground glass lesions, and such
• You can return to your life after getting booster shots every six months for the rest of your life
• We will take your life from you if you don’t get the vaccine
• We will use the FBI to investigate parents who complain about school vaccine mandates
• The vaccine turns symptomatic illness asymptomatic, but don’t worry about the increase in asymptomatic spread
• Go ahead and get infected with the virus after your vaccine to improve your immunity.
• Peter Daszak, the guy who organized the “lab leak is a conspiracy theory” letter in the Lancet, proposed to build Covid-19 for DARPA in 2018.
• Vaccinated people can spread Covid-19 just as easily as non-vaccinated people can.
• The spike in heart attacks and vascular conditions isn’t from the vaccine, it’s from the lockdowns.
• Unvaccinated children will be removed from school.
• Covid MUST be solved at the state level.
• The vaccine actually does affect menstrual cycles but you should get it anyway.
• Hospitalizations with Covid don’t count as Covid hospitalizations if Covid wasn’t the reason for admission.
• The case for masking children has collapsed
• Leading scientists knew it came from a lab but hid their opinions to promote “international harmony.”
• Even though the case for masking children has collapsed we’re going to do it anyway.
• Even though mask mandates are illogical, all that matters is that they align with the “community’s goals.”
• Mask mandates didn’t make much of a difference anyway.
• Fauci: “Vaccines don’t protect overly well” from Covid infection.
• The vaccine was bad because it was Trump’s fault.
• A vital part of hurricane preparedness is to get vaccinated.
• When the FDA told people not to take ivermectin that was totally just an informal recommendation.
• Department of Energy: “The virus came from the lab.”
Some of the things in your list are bullshit, some of the things in your list are wild exaggerations, some of the things in your list are “some low-level state official may have said something like this.” Every single thing in your list needs a citation.
The main thing that I would describe as dishonesty was the very early anti masking instructions. The reasons for this one are obvious although from my point of view it would have been better to just tell the truth and seize the supplies.
In the US there was a massive campaign of censorship where the government put pressure on social media companies to censor anyone who went against the official narrative. They even threatened scientists careers to force them to sign on to a public declaration that contradicted what they actually believed https://en.wikipedia.org/wiki/Lancet_letter_(COVID-19)
On top of that the public health bureaucracies issued decrees that were clearly motivated by politics and not public health e.g. at a time when they had banned all large outdoor gatherings and closed public parks and beaches, they issued a declaration that BLM protests were not a risk to spread covid, and were in fact good for public health.
Official communication about COVID was often dishonest, but the current administration's official commissions are deeply and transparently dishonest, and it doesn't motivate people on the right to conclude "we should disregard or reverse everything the administration says." This clearly isn't a behavior people apply consistently across the board.
Right. If the politicians du jour is grossly dishonest, half the population will have absolutely no trust in their leadership, while the other half will be solidly on team "well, they're *our* lying liars and it's not really lying if it's what they have to do to beat *your* lying liars". This is catastrophically bad for society, no matter which half is in which position.
Maybe try electing politicians who aren't grossly dishonest lying liars at all?
I think this is the crux of it. If the science had been squeaky clean from Day 1, we wouldn't in the place we now are. Once the fundamental trust in professional objectivity is lost, a doorway to lunacy opens up that will not be easy to close again.
It exposed how fragile the knowledge base for epidemiology actually is, that is for sure. But you can’t just not make decisions in an environment where you have all the facts. I think it could certainly be executed better in the future but in practice it is very hard to both convey nuance and uncertainty to the public while also trying to push the policy you think needs to happen in the moment. How would you do it?
I agree this is a genuinely hard problem. I would even argue it is impossible to have a useful conversation online, because it requires so much good faith on the part of the interlocutors.
I think sometimes about all the “lies” that FDR told in the run up to WW2. Do we wish he didn’t do that?
I think there needs to be more appreciation of just how great an achievement objective science was. It took us maybe a million years to get there. You throw that out the window because of a media driven panic? That's not good. People in power need to recognise the importance of the average citizen fundamentally trusting science and act accordingly. But, post the arrival of social media, many areas of gov became convinced that narrative control was all that mattered. That is fundamentally untrue.
"it is very hard to both convey nuance and uncertainty to the public while also trying to push the policy you think needs to happen in the moment. How would you do it?"
Short answer: with humility and honesty.
Longer answer: I think you start off with a bad assumption, namely that one needs to push a policy. How would I do it?
Provide the best available information while being honest about your confidence in said information.
When your opinion changes, explain why.
Only advocate for or institute policies backed by force of law if you have a very high degree of confidence, otherwise focus on providing information and advice.
Trust people.
Grant them autonomy.
Own your mistakes.
If you egregiously mess up, or consistently demonstrate poor judgement, resign from your position of authority.
If you're worried that things will fall apart without you, recognize that you're not a very good leader, because good leaders build robust systems that can operate without any one specific person. Start making corrections now.
Don't try to expand your sphere of authority anytime you see someone doing something objectionable.
The thing is, in a crisis, where there is high threat and high uncertainty, when people are overwhelmed and exhausted, we humans are going to do all these things badly.
Which is why we need our systems not to depend so much on the individual character of leaders.
In the 1980s, I taught at UCBerkeley about the lack of pandemic preparedness despite the high likelihood, near certainty, that there would be one in the coming decades. There were books and articles and loads of research done on this back then. The work ahead was to improve infrastructure for preparedness, not to rely on the good character of individuals. A lot of that work never happened and so when this pandemic came were were not nearly as prepared as we could have been.
One of the reasons for that is that as a society and as a human race, we are not yet that good at long-term planning, or at making short-term sacrifices in service of long-term well-being. This human frailty exists at the individual level all the way up to big organizations.
I disagree. Populism and general anti-science and anti-elitist thinking spread via social media had been undermining trust in experts long before Covid. People were actively looking for any excuse to disbelieve the authorities from the very beginning, especially given the obvious political and social ramifications of something as disruptive as a pandemic. There was nothing new or unusual about how experts handled an event like this other than that it happened during the age of social media.
Well, I think we're fundamentally agreeing. The advent of social media and the unprecedented level of connectivity that it afforded, created an environment where previously repressed and marginalised conspiracy material could now achieve take off. Yet, the Big Gov response of attempted narrative control simply did not work.
It worked throughout most of the 20th century, but in the 21st century we raised our standards and got more honesty, and as a result the attempted narrative control no longer worked. We got more accurate and reliable messaging, and as a result people started to understand that it was not perfectly accurate and perfectly reliable (even though it was better than all the things they trusted without complaint in the past).
What you are asking for is the opposite of “squeaky clean”. What we got during Covid was much cleaner scientific information than at any other point in history, because scientists were talking directly to the public. But it turns out that seeing scientists speak made people realize that scientists aren’t an alien race of godlike intelligences who are perfect on everything, and they decided that if they’re not like this, then we might as well not listen to them at all.
We lost the fundamental trust in professional objectivity because we allowed transparency and honesty.
But they weren’t all being transparent and honest. Too few were honest about the level of genuine uncertainty, and too many were quick to declare the current best guess as “expert consensus” and label anyone who questioned it a science denier.
And then they started using their newly elevated platforms to push obvious politics rather than just communicating scientific fact.
It probably would be better if public pronouncements were even more honest about uncertainty. There are a lot of such ways I have endorsed. But I think it’s also true that experts during the COVID pandemic did at least as good by this measure as experts at nearly any time in history before then.
My contention is that it is precisely the visibility of uncertainty and disagreement that was a large factor in the loss of trust.
Can you think of a time when there was better honesty?
It’s funny how people on the right who complain about being lied to by the mainstream media/establishment, they never bring up the countless lies and misinformation coming from Trump in 2020.
Speaking as a right wing guy who complains about being lied to by the healthcare establishment during covid...
What do you want me to say about Trump lying? He does it all the time, he's very dishonest, and thats bad. People don't typically cite him as a reliable source so it doesn't come up much, but it is bad
I think that the objection is less specific. I suspect that the RW-ers I know, would say the same. They don't trust what DJT says. However, they parrot the general talking points of the administration (e.g. Tariffs are both paid by foreign countries while simultaneously boosting domestic production, DOGE is eliminating loads of corruption and saving $100s of Billions). At least to me, this seems to better articulate the [edit: +apparent] hypocrisy being alluded to.
The double standards are insane. The president, who people should be looking towards in times of national crisis, lies through his teeth about how big of a problem Covid is and peddles snake oil treatments. And all you have to say is, idk it’s bad.
But, the medical establishment that has been 99% correct on everything deserves all the vitriol because there was conflicting information about mask usages in the early parts of Covid and maybe Covid made out of a bio lab (which still isn’t 100% confirmed).
My understanding is that during the early days of Covid, the health department (under Trump btw) discouraged masks because there was a shortage of masks and they wanted to save them for medical workers/high risk people. Eventually, it was corrected. Deliberate or not, lies made with good intentions happen everywhere. If other parts of academia were as politicized as epidemiology had became, you would literally trust no one.
Is there a time in the past where you think the healthcare was more honest than it was in 2020 or 2021?
My claim is that they were at least as honest during the pandemic as they have ever been, and it is partially because of the honesty of admitting mistakes that people now think they should never be trusted.
But why believe people who were even more dishonest? During the Delta wave three antivax conservative radio hosts died of COVID in a single month. These people made a very lucrative income from partisanship.
It is surprising to me that so many of them are so stupid. I also was distrustful of various official communications, but by the power of reasoning was able to come to reasonable conclusions such as "taking the vaccine is good", "covid is not that dangerous to me personally, but low-cost measures that inhibit transmission are not a major imposition", and "mask recommendations are not part of a new world order scheme to extract your adrenochrome". It didn't take me that much mental effort to navigate between the conspiracy theories and rigid adherence to the CDC gospel.
This is such an ironclad argument. That's why the frequent dishonesty deployed by internet atheists gave me a prior that young earth creationism is probably correct
I remember many examples of dishonesty. The efficacy of masks (or not), risks to children, the need for the most draconian lockdown measures (e.g. closing outdoor spaces), double standards for which activities were allowable and which weren't (e.g. social justice protests)... I was cognizant of the dishonesty in real-time and it still makes me angry.
But I still don't traffic in conspiracy theories. I believe myself to be a savvy enough consumer of a variety of sources of information that I can differentiate between what's likely to be true and what's likely to be false. Of course I have my biases and blind spots, but I think, or at least hope, that they aren't too egregious, and that I'm self-aware enough to assign confidence levels and come up with a realistic world-view accordingly.
I understand why people have come to distrust elite individuals and institutions. I understand why certain conspiracy theories have emotional appeal. But I expect reasonably intelligent people to stay true to a core set of principles, navigate their distrust and their emotions, and not be pulled in by dishonest, biased anti-elites, simply because they stand in opposition to the dishonest, biased elites.
The problem is that the level of dishonesty and failure that you are observing here is likely comparable or lower to the levels of dishonesty and failure that have existed among medical authorities at any other time in history - you shouldn’t stop trusting someone just because they became more honest about the fact that they are only 80-90% reliable.
Interesting, I haven't particularly noticed that. Can you point to some examples of rationalist groups being crazily conspiracy-theory-ish? Of course, the rationalist community has always supported all kinds of wild thought experiments much weirder than these, I'm wondering about cases that are more than that.
I think part of the issue is that Scott's blog exists at the intersection of a few different communities: the rationalist community, the centrist classical liberal community, the anti-woke community... So I think a lot of the commenters here with the worst epistemic hygiene may not really be "rationalists", or at least, might be more centrally from some other community.
Not to say that there's no crossover with "core rationalism", and I don't want to do a no-true-rationalist thing here, but I think you see more of this stuff in Scott's comments than on LW, for example.
This was both six years ago, and before the "mainstreaming" via both the NYT article and move to substack. So most likely the rationalist ratio is likely far, far lower than 13% now. This at least matches my anecdotal experience, where people who I typically see making dumb mistakes that would be caught by skimming the sequences never end up identifying as rationalist. Same with using this as identifying known rationalists (from their shared handle on LW)
Yeah this has always been a property of Scott's blog, though maybe more pronounced in recent years?
I've definitely encountered more IRL people who know of ASX then I ever did (a grand total of 0, I'm pretty sure) in the SSC days, which suggests a broadening of audience.
Ah, yeah, I certainly don't think of Scott's commentariat as being very representative of rationalist discourse or the rationalist community these days. I assumed Swallow meant something different, but maybe not.
I think part of the issue is that Scott's blog exists at the intersection of a few different communities: the rationalist community, the centrist classical liberal community, the anti-woke community... So I think a lot of the commenters here with the worst epistemic hygiene may not really be "rationalists", or at least, might be more centrally from some other community.
Not to say that there's no crossover with "core rationalism", and I don't want to do a no-true-rationalist thing here, but I think you see more of this stuff in Scott's comments than on LW, for example.
'Here are largely-undisputed all-cause mortality statistics from CDC and census.gov' is a level of crazy conspiracy-theory-ness I'm pretty comfortable with :)
Oh c'mon, we all lived through a hodge podge of policies and random claims from scientists and officials. It may have not made us all conspiracy theorists, but if were paying any attention it sure changed how you perceived "experts."
It really shouldn’t have! This is how experts function, and function well! People have unreasonable expectations, which past expert communities lived up to by hiding their failures.
If you do trivia competitions, you’ll understand that informed guesses that are nowhere near perfect can still be the best option, and lead to you reliably winning over the competition.
Informed guesses are often necessary. Falsely claiming your informed guesses are actually indisputable scientific fact never is.
Flat out *making shit up* because the facts you know make your desire to virtue signal about a murdered fentanyl addict inconvenient is a whole other level of credibility-nuking.
I think it's valuable to separate scientists from spokespeople from politicians from policy wonks from corporate interests in thinking about this.
Science is, as it has ever been, messy and imperfect. And yet an incredible human endeavor and still improving decade over decade.
Politicians (and their spokespeople) and corporate interests have never been very truthful because the incentives don't run that way.
If you lived through Watergate, Vietnam, the two gulf wars (with for sure WMDs!), a couple of finance-sector-engineered economic crises, the lies of tobacco and pesticide companies, the extent of child abuse and coverup in the Catholic church (and elsewhere), and all the stuff in-between, or some fraction of those, the hodgepodge is not remotely surprising.
I think the pandemic hit a bunch of people with more limited life experience and many of those people are now understandably gobsmacked that the world doesn't work better than they thought it did.
You should consider, here, whether the previous dismissal of these ideas as "for low-IQ nutters" wasn't also influenced by political considerations. It's not a given that this was a shift that made people net less rational.
The growing institution of Alternative media has herded the cattle with Candace Owens snd Catturdism, and therefore it’s politically advantageous to pander to the cattle if you’re a tech bro trying to change the world or get rich.
I wonder if people coming from countries with a hyper-inflationary past have easier time dealing with big numbers like this. I bet to somebody from Turkey or Argentina 1.2 million deaths are more believable (in fact maybe too normal as in desensitized?) than somebody from say Brunei or Switzerland.
Contrast "I only actually know one person who died of Covid" with the breathless "expert" and media hyping at the time that we should expect to see hordes of dying people staggering the streets because emergency rooms and entire hospitals would be flooded, then the hordes' corpses to be stacked like cord wood awaiting a mass bulldozer burial. Then, poof, after the beaches and playgrounds were reopened, it seemed to have been more like a bad flu season (not to mention actual flu disappearing for some mysterious reason). The cognitive dissonance is massive.
For real. This discussion is honestly just darkly humorous as someone who worked the Covid floors from March 2020 through July 2021. Those of us who were there remember.
Anyone in healthcare would be agape at this stuff. You know, the people who were actually taking care of all the Covid patients. Who also got sick, who also died, burned out, left the field despite years of training and debt for that training. Because it was such a horror. And now to see people dismiss all of that because they didn't personally stand in the hospitals and see it with their own eyes. Like Paris doesn't exist if you haven't been there yourself.
I was a healthcare provider during Covid and it changed me and my career permanently. Me along with many many others I know directly.
Hospitals aren't equipped to deal with a large public spike above baseline use of medical services. Recall how in the beginning of the pandemic, supplies like toilet paper, and rubbing alcohol became nigh inaccessible. The supply chain wasn't equipped to quickly respond to a significant change in demand, even though these are cheap and abundant goods.
If an industry takes 20% of our GDP I would expect edge cases to be worked out and available. Handling unforeseen spikes is something expected of the tech industry, but somehow medical system needs all our money and can't be bothered to think on their feet and handle a slightly above normal spike.
It takes a lot more time and resources to scale up the operations of a hospital dealing with highly infectious patients than it does to scale up the production of toilet paper.
If you are trying to build a hospital with exactly the same equipment in the same configuration sure. In an emergency I'd expect other solutions (deputizing the public like EMTs and those with some medical training) and nurses to expand the number of workers and making beds out of things not hospital beds. Especially if it's taking 20% of our GDP.
20% of our GDP is what we're spending under business-as-usual scenarios.
If you don't have proper facilities and appropriately observed safety protocols, sticking a bunch of sick people together to care for them makes outcomes worse rather than better (because they're infecting other patients and workers.) Even under ordinary circumstances, hospital-induced comorbidities are a significant risk. There simply aren't enough regular people with adequate training to safely run hospital-like facilities who can be deputized in order to dramatically increase our capacity to provide medical care under pandemic conditions.
> deputizing the public like EMTs and those with some medical training
Why would you think that would work? Is EMT training just all for show and any Joe blow off the street can do it at the drop of the hat?
The reason medicine and tech don’t behave in the same way is that the human capital of medicine has a way longer lead time and the regulatory framework is far stricter. EMTs, nurses, phlebotomists, RTs, PTs, OTs, SLPs, NPs, PAs, MD/DOs can’t just be conjured out of the ether. You either need to dramatically over expand your resource pool for edge cases, or accept that trade offs will need to be accepted when the system shock occurs.
This is why you need a federal government in a capitalistic system. Capitalism ensures that if 1 medical provider voluntarily kept excess production available, then he would be put out of business by his competitors. This massively parallel computing machine of industrial competition is what is so brilliant about capitalism but it does mean that you need intelligent governance to deal with the shortcomings.
Having 20% excess capacity sitting around means building a lot of extra hospital wings to sit around empty and hiring a lot of extra doctors and nurses to do nothing. Nobody is willing to pay for that during business as normal, so we have what we have.
I mean, we did see a million deaths. I don't know if this qualifies as a "horde" of corpses, but you could build a pretty sweet skull pyramid out of it.
This was my whole point in the last post - people act as if this was "hyped" or "overestimated", whereas in fact it was the largest mass casualty event in American history and if anyone had given the true number at the beginning of the pandemic, people would have been even more scared than they were. It's just that people are able to ignore a million deaths because that's 1/300th of the population and dead people don't write thinkpieces about how angry they are to have died.
A lot of hospitals did get overwhelmed. You didn't hear about it because you weren't in those hospitals and, I predict, don't know any of the doctors or nurses who were on the front line.
I once worked in an actually flooded-by-hordes emergency room - not during COVID, but during a blizzard in Ireland that the locals were completely unprepared for (they almost never get snow). Every bed occupied, gurneys lining the hallways, some patients on the floor, long line of people waiting to get it, stretching out the door. What percent of people in Ireland do you think were in that emergency room, or know someone who was? Would the average person on the street have thought "Yeah, I am personally connected to the overcrowdedness of this emergency room?" How exactly do you see this working?
You say this, but 1-4% mortality for the entire population was floated early on. Everyone I know expected way higher than a million death and for it to be spread across demographics equally. I remember thinking 3 to 12 million would be a reasonable estimate when they were still using NY and Italy as the estimation.
And? Some people happen to be over 65, and still value their lives greatly. Even more people have relatives above the age of 65, and would greatly prefer if they didn't die a painful death years ahead of their time.
"Effective vaccine" that was not a vaccine. Did not stop people from getting Covid. Did not stop people from transmitting the virus. Did not build up any immunity keeping "vaccinated " people from testing positive for Covid over and over. The only thing left in defense of the "vaccine" was the weak, "Well, you would have been sicker without the vaccine" . . . and I would like to know how that would be proved.
You are aware that people who've had the MMR vaccine can still get Measles, right? And people with the flu vaccine can still get the flu? There is a range of efficacy in vaccines, like with all other things.
The vaccines, generally speaking, reduced the risk of dying of Covid by factor of ten or so, which is far better than nothing. For the early variants, they probably would have prevented the spread well enough to eradicate the disease if everyone had become vaccinated, but the virus evolved too fast, became to virulent and too good at evading the immunity conferred by the vaccination to achieve that. As far as I recall, that was a bit of a surprise even for the experts. Still, without the vaccines, things would probably have turned out way worse in terms of casualties.
Not trying to be political (really!), but the predictions (like you mention) and the actual results reminded me a lot of the AIDS epidemic in the early 1980s.
The AIDS deaths started rising sharply, folks started projecting the curve and warning that everyone was at risk, things DID get worse, but ... there was also clustering and lots (maybe most?) people didn't personally know anyone who got AIDS or maybe knew only one person.
Other areas (e.g. San Francisco) had lots of people who knew 10s or 100s of people who died of AIDS.
But the exponential stopped exponenting as AIDS burned through the super at-risk population. And the general public might be forgiven for wondering why if things were so bad they weren't noticing anything particularly unusual. And wondering why they were supposed to panic?
Not quite the same here, but the parallels are similar enough I think. Folks in the medical community really noticed this in their own lives. Folks who worked a lot with the elderly probably noticed this in their own lives. For a lot of the rest of us we needed the news to remind us how bad things were because we weren't seeing it.
AIDS has R0 less than 1 via heterosexual intercourse in places outside Africa. So it wasn't going to break out of some relatively small subpopulations. COVID was not like that, it was a respiratory virus that hit everyone who breathes.
Right. AIDS wasn't going to break out of some relatively small subpopulations.
But in 1985 we were getting articles saying things such as this:
"By early this year, most Americans had become aware of AIDS, conscious of a trickle of news about a disease that was threatening homosexuals and drug addicts. AIDS, the experts said, was spreading rapidly. The number of cases was increasing geometrically, doubling every ten months, and the threat to heterosexuals appeared to be growing."
The doubling didn't go on forever and the heterosexual population wasn't totally unaffected (e.g. Arthur Ashe, Isaac Asimov) but that wasn't the message ordinary Americans were getting in 1985. Without news coverage, most Americans in 1985 wouldn't have known about AIDS. And would not have though to be worried about it. Folks in SF didn't need the news.
I forget if Covid eventually surpassed HIV in deaths. But it seems to me that there is remarkable similarity in the entire process, but compressed into 2 years instead of 40. (Government denial, then government acceptance, then Anthony Fauci changing policy to allow better responses when demanded by patient advocates, then better treatments, and eventually a preventative that means we don’t have to worry too much.)
I agree large numbers were floated. Small numbers were also floated. I think if God came down and said "the true number will be 1.2 million, even with pretty extreme countermeasures like locking down the whole country for months", most people would have interpreted that as "the large number people were right", albeit not the absolute largest.
I never saw anyone say 4%. I hang out in groups who’ve been thinking about this for a long time, who range from totally mainstream to batshit crazy and their consensus at the start was 1%, moved lower very fast, as low at 0.1%, then rebounded to roughly whar actually happened. I would say we knew the likely scale by May or June 2020
Fairly early predictions of deaths sans countermeasures turned out to be accurate. You can project the mortality among those who *did* become infected and compare that against who didn't in 2020. It's harder to estimate what the loss of a functioning health care system would've done to mortality, but even that's something you can ballpark. The projected scenarios that seemed gloomy were on the money as far as that goes.
There were countermeasures - people aren't just going to behave normally as corpses pile up - but it's just innately hard to to predict what political will is going to look like, particularly so in the United States. Losing over a million people is staggering, and a significantly worse overall outcome than some global peers such as Japan.
I think a lot of this is missing concentration. New York City did, in fact, have piles of corpses in the street. Everyone I know from New York City knows lots of folks who died of Covid.
I know a few.
My family members who live in rural North Carolina know zero.
> My family members who live in rural North Carolina know zero.
It’s probably also an issue of New Yorkers being much more willing to say that someone they know died of Covid, whereas the rural south which got absolutely ravaged by the delta wave had folks who would have a bunch of egg on their face if they admitted that pawpaw died of that virus that up until yesterday they had been loudly proclaiming to be a hoax.
Actually living in the south and knowing a lot of Covid deniers... there's *some* of this, but not as much as you'd think. Most folks who have someone close to them die of Covid change their tunes pretty quickly. Often they're vocal advocates for the protections they argued against.
Generally people are people, not caricatures. You can show them all the facts and figures you want and they won't budge an inch, but when they have real life experiences with the things they adapt.
I lived and worked in healthcare in the rural south through Covid. At least in east TN there were plenty of people dying from small places in the mountains where their absence would be noted given how small the community was. Some church outbreaks had multiple congregants die from a pretty small population. It was surreal driving home from the hospital past the freaking freezer truck full of bodies and seeing TGI Fridays packed to the gills with old gomers just half a mile down the street.
People are people, and people the world around are extremely invested in saving face.
I got curious about the skull pyramid thing - assuming each skull is around 17.5cm wide and lies on its side, and that they pack 1, 4, 9, 16, 25 etc, you would get a pyramid roughly 25-30m tall, and probably quite a bit wider (because skulls are sort of irregularly shaped). Which is impressive but not, like, skyscraper-sized.
But the pyramid of Giza is something like 150m high and 200m wide.
For visual impact, I think you could instead just clad a giant pyramid in skulls, or make a bunch of smaller pyramids and put them all together in a large flat plain.
For your skull-clad pyramid, you put the skulls upright next to each other. We again use the 17.5cm width, but now each step is 25cm high. The number of skulls per layer goes 1, 4, 8, 12, 16, 20 etc. This one is a monster: 194m tall and 135m wide. I'd suggest making it look more like a Mayan pyramid, so that you can add a bit of width and make it a nice round 200x200x200m. Estimating cost is a challenge, ad building costs are so variable and depend on so many factors (land cost, labour costs, transportation costs, material costs etc). If built analogously to a skyscraper, then the cost per might be anything from $450 million to $1.8 billion (assuming a floor height of 4m and cost-per-floor at Dubai vs New York prices). Just as a sanity check, the bulk cost of the concrete alone should be something like $160 million (at around $134 per m3 and assuming you fill the whole pyramid up with it), so I suspect the higher price is more accurate. If you don't want to use real skulls, then I found ceramic skulls on sale for around $8. Assuming a bulk discount ($4 per skull), this makes them by far the cheapest part of the whole exercise ($4.8 million). The folk from construction physics can probably math this out a lot better than I can.
My final suggestion, since we've made a flat-topped pyramid, is to have a chamber of important covid-related figures' skulls in the heart of the pyramid, and put the inevitable gift shop and restaurant on the roof.
In re: the discussion above about the true rationalist commenters vs the riff raff at this-a-here blog, this comment is first ballot Hall of Fame True Rationalist material.
This comment alone makes me glad for this whole mess of a comments section.
Would like to express my agreement with this sentiment. Cry6Aa's comment was a delight to find (and will be getting relentlessly linked to my circle of friends).
And statistically both groups of people are entirely justified in their choices. As you note, medical personnel in hospitals were totally overwhelmed, because the healthcare system of every country on earth runs on fairly tight capacity margins, so a pandemic that increases the baseline rate of ER visits by (totally making up numbers, but approximately correct) 50% is going to totally swamp the system. It was like a never-ending mass casualty event.
But for the average person, who is younger than 65 with few to no comorbidities the approximate risk of dying from Covid was zero. Compared to the many other things that tend to kill people younger than 65, it didn't even make the top 10 causes (approximately 260k covid deaths in the under 65 cohort, and 80% of that is the 50-64 cohort). So you have this massively disruptive event which altered or destroyed the careers, families, and social circles of a large number of people who were at very low risk in the first place, and you can rather correctly say we locked down to save all of the grandmas at the expense of their granchildren.
Its only natural people would feel resentment in either group. But neither of them is wrong, and the entire discussion around covid is has to take one side or the other.
I know this is not the view you're arguing for, but this we/them frame in the face of a huge shared calamity speaks to the real dissolution of a functioning society to me.
Besides the young people resenting the old people, there are millions of people with chronic illness in any given year, millions undergoing cancer treatment with wrecked immune systems, bunches of newborn vulnerable babies, bunches of pregnant women who are also medically vulnerable. Bunches of people without great access to healthcare or who can't economically afford to miss work. *And then there are all the people who love them.*
And it makes me worry that these guys who are so resentful of the price they paid during a national calamity don't have anyone to love. Because if you love someones in your life, chances are one or more of those someones are vulnerable in some kind of way, and you would feel eager in the face of such a threat to have things to do that would make the ones you love be just a little less likely to get really sick or die.
The tradeoffs in some situations were absolutely terrible, no doubt. But all of life requires tradeoffs. And exceptional times require some really big ones. None of us is promised smooth sailing for any amount of time, ever.
Fine, life requires tradeoffs. The tradeoffs made during covid were terrible, maximizing suffering and minimizing flourishing. The people responsible for those decisions should not be taken seriously.
Like many commentors here, I do not have any direct acquaintances who died of Covid. Hopefully unlike many here, I know multiple people who chose to take their own lives during lockdown, and I am convinced within a reasonable margin of absolute certainty that the isolation is what killed them. So in the face of your medically vulnerable person, I offer the emotionally vulnerable person, who was forced by threat of government force to deal with their issues alone.
You frame the issue as if its a bunch of old white conservatives who were anti-lockdown. My experience is rather the opposite, and while I have sympathy for the medically vulnerable, I have none for the decision that society must revolve around their needs at the expense of everyone else.
That's a really profound kind of experience to have people you know commit suicide and it's really awful that that was part of your Covid time experience, that direct trauma of losing people that way during a time that was already horrible.
I'm sure you're right that at least part of the reason for the suicides we saw increasing during that time were due to social isolation. For the healthcare workers who committed suicide, it seems like those were maybe fueled by the moral injury and overwhelm of being a healthcare worker at that time.
I'm a psychologist who works with individuals and families and so I had a front seat view to the small, medium, and enormous mental health impacts of peak Covid time. In addition to the people who suffered from terrible isolation (young and old), I think the tradeoffs were also especially horrible for children and for parents of children.
I also had a prolonged non-Covid physical health threat that emerged for me during Covid (cancer) and so I had a front row seat to the impacts on hospital-based workers beyond those I know from my private practice. And I saw how various doctors and nurses and other staff carried the weight of tradeoffs, everyone carrying a slightly different mix of burdens and trauma. I also worked with quite a few patients who had severe social anxiety who experienced a couple of years of real relief from it for the first time in their lives as a result of the permission to stay home and the opening of working from home. So while Covid and all that came with it was a horror for mental health broadly speaking, there are also other stories in there as well.
I think the only difference between you and me on this topic maybe is that I took from my experience of that time the sense that most everyone is doing the best they can all the time. That humans and human systems have forever been very flawed. That there's a real limit on our capacity to make good decisions in an environment of high risk and high uncertainty and where the conditions are rapidly changing. I see those limits and tradeoffs as built into the human condition and into the reality of the world we live in.
And then I witnessed up close so so many people daily doing acts of generous and heroic kindness -- including to me personally hundreds of times -- that despite the losses I experienced and witnessed, I also feel a profound gratitude to be here, to be alive in this world, as messed up as it is. I feel a lot of grace is due to all the people who worked really hard to take care of others in a situation none of us has ever been in and for which there was no adequate road map (including because we failed to do the planning for this that had been mapped out decades ago). Long-term planning being another weakness of humans.
I also think criticism is important and it's key that we all in various ways reckon with the losses, the tradeoffs, and the costs of the different choices made in different times and places through the pandemic. I see the value of this not in assigning blame but in preparing us to do it better next time. To me that means wrestling with the nuance and complexity rather than drawing quick judgments and conclusions. Bringing curiosity, in other words.
There's a view I think that everyone who was in any way on the front lines of responding to the pandemic *should have known better* about a whole range of things, *should have communicated better* and should have overall *been better people*. And of course there were in there here and there some truly venal and opportunistic people as well as the usual mix of people of mediocre competency. To me, this is where humanity is as a whole at this moment. We would like everyone to be better and having those wishes for humanity is good. But we live in this world where people are how they are -- an incredible mixed bag. Glory and horror all wrapped up together. I have no confidence that if you'd dropped me into any other frontline role than the one I had that I would have done better than those people in those positions. I had my hands absolutely full with the position I had and was bending under the weight of it every day until it kinda broke me.
I guess I want to say there is relief in accepting this. It doesn't mean don't ever be angry or give up and don't do anything to make things better. But accepting all this as where we are as humans means that one's actions can move better from a place of calm and compassion and the choices we make from that place tend to be better ones than the choices made from rage. And also it feels a whole lot better.
I think (I don't know if I'm right) that a lot of the political discord we've seen since Covid (though it obviously was well along already) was fueled by the lack of official acknowledgement of the need for grieving, reckoning, criticism, and taking responsibility. I could write a huge proposal sitting right here with ways I think institutions and governments needed and still need to create official space for people to grieve and recover as well as to reckon and criticize.
Grief unprocessed can lead to both rage and depression (as well as a whole host of other health problems) and my country at least (US) has done a very poor job of making space for grief. Not because people suck so much as that historically as a culture and an economy, we have not prioritized emotional health and connection. Instead we have prioritized economic productivity, accumulation, and status. While at the same time, we didn't do the work over the past decades to set up our institutions to respond well to this specific kind of calamity, which then exposed a lot of people to the particular foibles and weaknesses of individual humans even more than would ordinarily happen in a crisis.
I think you are entitled to whatever feelings you have about what went on during Covid, how all of it impacted you and the people you cared about. If it leads you mainly to anger and blame, I guess I want to say have hope, there is a place on the other side of that that feels better. And it runs through you validating everything you feel and felt about the pandemic and not from minimizing it (which I don't hear you doing).
Well here's the thing. If there were a new shocking ear worm parasite that was 100% fatal but only infected people the day before they were going to die of something else, that ear worm parasite could kill millions of people in a year and cause the largest fatality event in US history, and it still wouldn't really matter because those people were going to die anyway.
Excess deaths that year would be close to 8,000 because that is the number of deaths in one day. If the parasite moved all deaths forward by two days then excess deaths would be close to 16,000. In order to get excess deaths up to 1 million you have to move deaths forward about 100 days.
I think by that time, you’re talking real sacrifice - people lose several months of time with literally everyone they love.
It’s not as bad as a war that kills tens of millions of people decades early and thus leads the average person to lose several years of life, but I think we get a sense of the magnitude.
I think it's reasonable to assume that a very large number of people who died from Covid had their deaths moved up one to two years, and a few moved up around five.
Emergency rooms and hospitals were flooded. I know a PA who pivoted to work directly with COVID patients pretty much from the start and it was an absolute nightmare for them.
It's not cognitive dissonance, it's survivorship bias. It was "only" 1.2 million deaths because various measures were taken. It's like saying the Y2K bug was overblown because nothing really happened come Jan 1st, 2000 - but that was because legions of programmers had been employed to actually fix the bug in affected systems. That is always the curse of "preventing bad things from happening" as opposed to "making good things happen" - if you prevent something bad from happening, it can look like you've done nothing at all.
Some one up above, about the harm of government-issued NPIs, pointed out that there would still be a lot of ad hoc NPIs done in the baseline.
That applies in the other direction, too. Even for a disease that genuinely kills 1% of a population acting like normal, people will react and restrict contact in a way that reduces the impact, even with no government-led interventions or vaccines.
Right. If no official measures had been in place, it probably would have been 1.5 million deaths. No official measures except vaccine development, 1.25 million deaths.
The observed death rate isn't that far off what you'd get if you postulate the disease running unchecked until the population reached herd immunity, and even in the absence of official countermeasures you'd have lots of people taking private ones. That 1.2 million people died *anyway*, means there isn't really much room for the official measures to have done much good.
And then there's the world where we imagine no official measures but also no official impediment to vaccine deployment and distribution, and I'm pretty sure the death toll is well under a million in that one.
I'm not sure how you are deriving these numbers (personal intuition?), but this isn't what you see in epidemiological publications at all.
Also, herd immunity wasn't ever achieved. We don't have herd immunity right now. Things are much better than they were from 2020-2022, but COVID is still circulating, acting like quite bad influenza waves crashing against us. The disease burden is substantial. Turns out our immune system's limitations + viral evolution is able to frustrate herd immunity for this disease. That hope was decisively dashed in late 2021 when it became clear that developing an immune response through vaccination / exposure was insufficient to produce lasting immunity in the general population.
Even assuming these numbers are true, without getting into details like your confidence in them and whether they would have been available at the time when they would have mattered: Would you have decided "Eh, only 300k additional deaths, that's not worth a lockdown"? Especially if it was a matter of public record and you had to expect someone eventually writing nasty headlines about your decision?
I'm fairly confident in those estimates, for reasons that I've discussed here too many times to want to repeat. But as to your question, oh hell yes. First off, if you read my statement carefully, that's 250,000 lives saved (US only) by vaccines, and 50,000 by mandatory lockdowns, masking, social distancing and all the rest. If anyone here is assuming that because I opposed lockdowns I'm also an antivaxxer, then please just shut up and go away.
And second, even setting aside concerns about pesky little things like civil liberties, and the economic cost of the lockdowns, and the effects on the mental and social health of American adults, if you ask me to trade fifty *million* children going through a year and a half of half-baked zoomschooling, against fifty thousand mostly septuagenarians dying five years earlier than they otherwise would have, then there's no contest.
Even if we assume 99% of the kids came through just fine, and the 1% that don't suffer only a 1% reduction in QALYs per year from the educational shortfall, the lockdowns fail the consequentialist math. And I'm not a consequentialist.
You haven't answered my question whether any of these numbers (1.2M, 1.25M, 1.5M) would have been available in early 2020 at the same confidence. Hindsight is always easy.
Oh, sorry, yes I missed that. These numbers would definitely have been available at the time, if someone had wanted them. I mostly took these numbers from my own analysis in 2020. and that of people like Zvi Mowshowitz, and the subsequent observation that the actual infection and death tolls were tracking pretty closely with the "we fuck it all up and our interventions are useless" predicts right up until the vaccine rollout. At which point there was a substantial drop in mortality, though less so with infections.
The thing that should *not* have been available with confidence in early 2020, was the belief that lockdowns, mask mandates, and "social distancing" were going to dramatically reduce the death toll. Those were unprecedented interventions, with no proper scientific studies to back them up and based on multiple dubious assumptions.
As best-guess "maybe this will work; let's put it out there as a recommendation" measures, sure, worth a try for all we knew in 2020. But an intervention that is intrusive, mandatory, and controversial, is one that really needs to be backed up by the sort of science we can be confident in.
The distortion of reality where people pretend that hospitals weren’t flooded is ridiculous. They were. It was insane. I’ve never seen anything like it before or since.
Also It’s pretty clear why the actual flu disappeared — enough people were wearing masks that it couldn’t spread. Note that flu is spread via droplet transmissions so simple masks are highly effective against it
I think a lot of the comments section is missing the concentration thing. New York City had mass graves. A random town in Iowa may have seen zero cases.
I don't know if it's clear any respiratory disease spreads via droplets rather than aerosols now. COVID is just more transmissible than the flu (measles being super-transmissible is why we're starting to see outbreaks now that vaccination has dropped a little).
Hospitals were either flooded or empty depending on where you were.
NY hospitals got demolished because Wuhan is where all the garments are made, and the places that got bombed are tied in with the fashion industry. Wuhan, Milan, London, NY. It spread after Fashion Week. One or several Chinese fashion industry execs were Patient Zero.
Statistically, nationwide, hospitals were not overwhelmed. In fact most were under-occupied because they were saving beds, or at least that was the argument they made when they claimed taxpayer bailouts. There were local anecdotes in certain places, New York most prominently. But even there, they brought in the hospital ships and built out extra capacity and nobody used it.
Meanwhile, against the anecdotes of people claiming hospital flooding, there were the three hospital visits I made during the Covid period, twice with my mother, once with a child. The places were ghost towns. I've never seen the ER so quiet and with so few patients. Like three people in a space that normally seated 200.
I'm not claiming those observations necessarily generalize. But neither do the anecdotes of flooding.
Hospital overload is like traffic congestion. 90% of the day, highways are empty. But at the time that everyone wants to drive, they get crowded. Hospitals were under capacity for most of the pandemic because there were plans to try to save capacity - but when a big wave hit anywhere, they got flooded. (See New York in March 2020, Houston in June 2020, etc)
COVID turned out to be significantly more contagious than influenza. The globes' attempts to mitigate COVID spread actually appears to have wiped out one of the major circulating influenza lineages, which has simplified the vaccination formulation going forward.
The excess deaths chart above does NOT look like "a bad flu season", unless you mean something like Spanish Flu. Actual flu disappeared because measures that failed to stop COVID reduced R0 enough for them, enough to kill off some old flu strains https://www.livescience.com/flu-virus-types-extinct-covid-19.html
During the Delta wave hospitals in my area -- an area with low vaccine uptake -- got slammed and had to bring in hundreds of medical personnel from out of state.
During this same period my local state representative was posting antivax crap on Facebook.
Why do you demand perfection from experts but ignore the political partisans who got people killed?
Delta hit red areas hard because the red areas dodged the earlier waves and had lower herd immunity than the northeast did. NYC fared very well during Delta because they all had immunity from earlier infections.
This needs to be analyzed regionally and without political bias.
I higher vaccination rate among people under age 35 would not have made a difference. A higher vaccination rate of people over age 65 would probably have made a tremendous difference. And neither rate matters much now that the vaccine is less than 40% effective at stopping spread anyway.
(After that made news I called him an complained because multiple tenants from my property in Branson were in the hospital with COVID. One of them was under 40 and overweight but not morbidly obese. Fortunately they all survived.)
Or maybe they were listening to these people right as the Delta wave was surging. This is literally at my local hospital.
I can retype "a higher vaccination rate of people over age 65 would probably have made a tremendous difference" if you like. Otherwise I'm not sure what your point is.
Remember that the “immunity” New York had during delta wasn’t just from higher vaccine uptake and earlier immune system encounters with the virus - it was also from behavioral changes triggered by living in a place that was hard hit. The brain is the biggest organ in the immune system, just like it’s the most sensitive erogenous zone.
Unless NYC shut the subway down during Delta then NYC has infinite vector potential no matter what other "behavioral changes" they tried to implement. No amount of separation is going to tamp down being stuck in an enclosed poorly ventilated space, given what we know about how the virus is transmitted by stale air.
You don’t have to shut down the subway to reduce spread in the subway - as long as fewer people are using the subway it will contribute less spread. (Presumably quadratically, because spread is proportional both to the number of infected people present and the number of uninfected people present).
I suppose a reduction in subway ridership may have had some impact, but the fact that basically the entire city all got Covid at the same time six months earlier probably had a dramatically larger effect. And that effect wasn't present in the red areas that had their first wave during Delta. These are areas, keep in mind, without anything remotely approximating a subway.
My point is that if we're talking about people's perceptions, the news stories that would tend to show that conditions were really bad were often balanced by news stories that showex that they were not nearly as bad as anticipated.
> we should expect to see hordes of dying people staggering the streets because emergency rooms and entire hospitals would be flooded, then the hordes' corpses to be stacked like cord wood awaiting a mass bulldozer burial
This more or less DID happen in New York City, so I don't think it was unreasonable for people at the time with imperfect information to think it may happen in other places. I was turned away from the emergency room and told not to come back unless I thought I was going to die.
> actual flu disappearing for some mysterious reason
This is such a perfect encapsulation of covid denier idiocy. "It's a mystery that with everyone staying home and wearing masks and trying to prevent the transmission of a disease that is more virulent than the flu, there were fewer flu infections this year! The NWO doesn't want you to think about this!" I don't know, there's one explanation that jumps out as pretty plausible to me!
In my city, the hospital where my stepmom works had a network of tents set up outside the building to handle the overcrowding, and the (Republican) governor sent us several freezer vans to hold corpses until they could be properly autopsied because the morgue couldn't hold them all (and because the medical examiner was actually not just declaring every death a covid death if there was a positive covid test, but was ensuring every corpse had a proper investigation into cause of death).
It makes sense that COVID killed a million or so in the US. That's a 0.3% mortality rate, or 3x the flu. It also made plenty of sense to have a strict lockdown as an initial response to the virus. The question is why did we keep it shut down after we figured out that it was going to be an endemic virus and all of those people would die of it eventually anyway. And most damningly, why did we keep it shut down for another year after the vaccine came out? All the conspiracy talk is just a distraction from discussing the very real authoritarian nightmare that was inflicted on us by the administrative state. If you had asked people in 2019 if we should shut down the economy for two years and print trillions of dollars causing massive inflation in response to a virus only 3x worse than the flu, they would have thought you were insane.
It was more than 3x as deadly. About 500,000 died in the US that first year. The worst flu season of the 21st century only had 80,000 deaths in the US and many had more like 20-30,000. So we're talking 6x-15x depending on the flu season. And that's with all the measures we took to prevent transmission.
The "3x the flu" number is an average ranging from the beginning of the pandemic (more lethal virus, no vaccines) to the end (much less lethal virus, lots of vaccinated people).
Comparing it to actual flu waves, you also have to take into account that Covid was, and became even more so, much more infectious than the flu, such that sooner or later everyone got it, whereas each flu wave only hits a few % of the population.
But, yes, initally it was much more deadly than the flu (10x as lethal as the nastiest flu strains IIRC).
If you average it over a long enough time, you can just say that it had exactly the same mortality as flu (it converges to 1x as t=>oo). This reduces my sympathy in justifying the 3x number, significantly.
You're using a very weird definition of mortality. My point is not that more people die the longer you wait. It's that the probability of dying of the disease decreased with time (because the virus mutated into strains with lower mortality, and vaccines were rolled out), but for the initial measures, it's justified to base them on the mortality for the variants that circulated then in the absence of vaccines, and at that time "10× more lethal and twice as infectious as a the nastiest strains of flu since the Spanish Flu" was a reasonable estimate.
My comment was not intended as a rebuke. You said "The "3x the flu" number is an average ranging from the beginning of the pandemic...to the end...." and "[but initially] 10x more lethal." My point was just that averaging over a period of time (3x from beginning to end) eventually converges to 1x since you are averaging a finite peak with an exponential decay back to baseline. In other words, you could pick any multiplier between 10x and 1x and there exists a time period for which this is the correct average. The ability to chose any number raises the question of why one might pick "3x." Again, my comment was at least half tongue in cheek.
That's part of it, but not nearly all. The Omikron variants had a significantly lower fatality rate, even among unvaccinated people, and by that time most people had not had the infection yet.
The difference was Covid-19 was NOVEL, not that it was that much more deadly. The difference was that unlike the flu, basically everyone caught it, all in the same year.
Flu x 3 is not a great way to look at it. If you were over 65, it was 7*[Flu], and if you were under 35 it was 0.5*[Flu] or lower.
Note that just saying "it turned out fine so the precautions were unnecessary" is not valid reasoning. What you *can* argue is something like that countries that didn't do a lockdown had similar death rates, which might or might not be true. But my point is that the argument needs to be a little more complex than what you said.
Note that I mentioned in my comment that it was reasonable as an initial reaction. Erring on the side of caution is a reasonable approach to an unknown disease. But by June 2020, those initial unknowns were pretty much all known. Also, it didn't "turn out fine." It was essentially the worst outcome possible. Containment was a complete failure and 100% of the population got the disease.
"What you *can* argue is something like that countries that didn't do a lockdown had similar death rates, which might or might not be true."
A reasonable pair for comparison (and *I* remember saying this in 2020/21 so this isn't hindsight bias, though I didn't write anything and you don't have to believe me ...) is Sweden and Denmark. Similar populations (though not identical) and Denmark locked down much harder than Sweden.
I have seen analysis that claim that Sweden did *better* than Denmark, but that analysis requires sophisticated math and I am skeptical.
Disappointing, because philosophically I much preferred the Swedish approach of providing guidance to the population and expecting them to behave reasonably. It would be nice if that ALSO resulted in fewer deaths. But I don't think that it did.
The UK is a popular country to compare against Sweden by those who opposed the lockdowns and the UK experienced 3,389 reported covid deaths per million, but that comp tended to arise afterwards and I think Denmark is better.
if you look at excess deaths as opposed to reported deaths I believe Sweden comes out pretty dang similar to other Nordics other the whole pandemic period.
I feel like you are cherry picking here. Sweden is right at the median for death rate of countries in Europe in your source. And Denmark is near the bottom. Unless you have some actual hard criteria here, like age structure of population, for using Denmark in particular I would say it's not really valid.
Buuut, you didn't keep it shut down. At what point did you have to stop getting permission from the local police to leave your house? That's when lockdown ended.
Needing permission from the police to leave your house is a lockdown. If you can just leave your house whenever you like, then you’re not locked down, so there’s no lock down.
We (England) had a lockdown from 26 March 2020 to 28 May 2020 and then again from 5 November 2020 to 8 March 2021.
Yeah, I don't understand where this meme comes from. We had school closures, restrictions on indoor commerce (usually fine w/ a mask), and were encouraged to get vaccines. I know other parts of the world had stricter interventions and a bunch of lefties were enormous hypocrites about SJ protests, but even in NYC at the height of it, I don't think people were locked in their houses.
You will justifiably find this highly annoying, but in the US, people often use "lockdowns" to refer to all the NPIs deployed here, collectively. This is true even though only a few of them could (arguably) qualify, like school closures.
I was in Italy at the time, and we were only permitted to leave our apartment to go to the grocery store for several months, and had to carry around papers certifying that we were going somewhere that was necessary for several more months after that. There wasn’t anywhere in the US that had this style of restriction, and it became common to use “lockdown” to mean “needing to wear a mask at the grocery store” in the US.
Tens of thousands of people lost their businesses. Tens of thousands lost their careers. Hundreds of thousands had to relocate their families.
What the hell are you talking about?
Is it a “request” to shut down your business while strip clubs and Costco remain open?
Is it a “request” to be targeted by police if you want to walk around without a mask?
Is it a “request” to be blocked from accessing facilities if you don’t comply with nonsensical policy they just invited 3 seconds prior?
Is it a “request” to have to quit your job (as opposed to being fired for not getting an experimental mRNA transfection therapy) to watch your kids during school hours since the school won’t open for them, but your still get the joyous opportunity to pay for the closed schools as though they’re open?
In California, it was illegal for anyone to leave their home for "nonessential" reasons from roughly March 2020 through September 2020. And if you did have an "essential" reason to leave, there was a very limited set of places to go.
No, they didn't literally have a policeman standing outside everyone's door to check. If you were willing to lie about what was "essential", you would almost certainly get away with it. I'd still consider that a lockdown.
0.3% is also consistent with the Diamond Princess, as close to a laboratory environment as possible, where the mortality rate was around 0.38% (~14 deaths, ~3700 people on board).
> "All the conspiracy talk is just a distraction from discussing the very real authoritarian nightmare that was inflicted on us by the administrative state."
I don't think it's a distraction. I think that given that ~1/5th of the ACX comment section seems to believe this false thing, it is a separate but also-serious problem, and it's worth talking about it. We can't have a sensible discussion about how bad the real facts are until we agree on them!
> The question is why did we keep it shut down after we figured out that it was going to be an endemic virus and all of those people would die of it eventually anyway.
> If you had asked people in 2019 if we should shut down the economy for two years
We didn't do that! Restaurants reopened before the end of 2020, despite it being obviously impossible to eat in a restaurant without spreading a respiratory disease. Real "shut down everything in-person" restrictions only lasted six weeks before economic reality and cabin fever forced people to reopen while pretending that viral particles couldn't move more than 6 feet indoors.
Scott should implement a rule: when discussing COVID lockdown everyone should have to preface each comment with the jurisdiction in which they lived so we can see what actual measures they lived under.
I will start this grand tradition and admit that I lived in Florida and mostly self-isolated the first three months. I remember how ghostly downtown was when it really hit people that this was happening. I also remember most people shrugging and moving on with their lives within around four months with maybe half masking consistently. Given Florida's place in history as the flagship of low-restriction state governance, my experience may not resemble that of blue staters.
We never once shut down the economy. As far as I can tell, we shut down 5% of the economy (GDP decreased by 2.2%, while in other recent years it had been growing by about 2.4% per year, so at most 5% of the economy was shut down.)
If you want to correct someone’s exaggeration, try not to exaggerate in the opposite direction.
It was a lot more than 5%. They just made up for it by printing trillions of dollars and handing it out. This, of course, caused the inflation mess. If Biden hadn't been such an idiot and lifted all the economy at the beginning of 2021 we probably would have had a much better recovery and less inflation. We probably wouldn't be dealing with Trump today.
What have excess deaths looked like in the years since Covid, 2023 and 2024?
If I weren’t so lazy, I would look this up. I can imagine a scenario in which this number is low or negative, with fewer deaths in these years than expected. And if that were the case, I would think of Covid as having accelerated the deaths of the difference. The people meant to die in 2023 and 2024 died in 2020 or 2021 instead.
And that might explain why it just doesn’t feel like a 1.2 million deaths tragedy.
This is the worst graph you could use to prove your point because it compares different cohorts of different sizes.
Death is strongly correlated with age (duh). If everyone dies at age 75 then you would expect a difference proportional to the difference in birth rate between 1940–1944 and 1945–1950, and those are really different periods because of World War 2 and the following baby boom. In reality it's more smoothed out of course but this still seems too big of a factor to ignore.
So from the graph alone I really cannot tell if the ~10% increase we are seeing in 2024 compared to 2019 is above or below expectation based on changes in demographics between those years.
And at least eyeballing it, the data from early 2025 seems higher compared with 2023 and 2024 which at least suggests a dip in those years.
If you look at the comment I’m responding to, this graph exactly addresses the question there. You can create scenarios to explain what the graph shows, but the comment asked if excess were down after Covid and the answer is clearly no.
The question was whether there were fewer deaths in 2023 and 2024 “than expected” (which would indicate a pull-forward effect of mostly older patients dying).
”Than expected” doesn't mean “than 5 years ago” if there are reasons to expect the death rate to be higher than 5 years ago regardless of COVID (i.e., the baby boom).
So no, it doesn't answer the question, at least not convincingly.
And you do need strong evidence if you want to deny there is a pull-forward effect, since we know that COVID deaths disproportionally affected the elderly (and secondarily the obese, who generally have lower life expectancy as well), which nobody denies, so purely statistically you would expect the death rate post-COVID to be lower “than expected”, everything else being equal.
It seems to me the "pull-forward" theory proponents should actually present some actual evidence their theory is accurate other that just saying it makes sense so it must be true. If you have a better graph I would love to look at it. Given the oldest boomers are around 80 years old, and their life expectancy is also around 80 (although really more given the life expectancy for anyone at age 75 is around 12 years), a 10% increase in deaths seems highly, highly unlikely without some other cause.
I've given you a perfectly reasonable explanation why the pull-forward theory is likely, you just chose to ignore it. It relies only on facts that I assume you already accept to be true: that COVID disproportionally killed people with shorter life expectancies, namely: the elderly, the obese, the infirm.
Then it logically follows that the survivors have higher than average life expectancy, resulting in a reduced death rate in the following years, if everything went back to normal.
Of course there is an “if” there. You could argue that the introduction of COVID has permanently increased the human death rate which offsets the expected short-term decline (which would mean life expectancy at birth is lower post-COVID). This is plausibly true at least to *some* extent, but if you want to make a specific claim about how strong that effect is the ball is back in your court. Also this is not actually an argument against the pull-forward effect per se; it's just an explanation why the death rate doesn't go below average levels *despite* the pull-forward effect.
Note that the effect on the death rate of the pull-forward effect is likely to be small because deaths don't very neatly occur at the end of life expectancy.
What, exactly, are you arguing against? The pull-forward theory is that people who died of covid would otherwise have died later, AKA, they would not have been immortal. Really the only alternative to this is immortality.
And why are people so excited about the pull-forward theory, anyway? There seems to be an implication that it makes things ok again, and yet, dying sooner is worse than dying later (assuming quality of life above 0). All sorts of humanity's most celebrated achievements - agricultural revolution, eradication of smallpox, antibiotics - only delayed death, rather than preventing it entirely.
I started writing this comment to Calvin, but then threw it out when I realized it didn't contradict him, but I guess I'll resurrect it now.
==
Let's say we have 10M elderly and we expect 5% to die each year, so 500K deaths. Also 500K people age into elderly, nice and even.
Then a virus comes along and kills an extra 10% in year N, so we're down to 9M next year. However, out of these 1M deaths, it was strongly correlated to the worst health, so 250K of them were going to die in year N+1 and 250K would die in year N+2 and 250K would die in the year N+3. The remaining 250K were from the otherwise healthy elderly.
Now, in year N+1, we would normally expect there to be a population of 10M and 500K deaths, but 250K of those deaths happened in year N. So we have a population of 9M and 250K deaths.
I'm a bit confused if you are agreeing or disagreeing with me.
What you describe is essentially what I'm arguing: that since COVID killed people with relatively low life expectancy, we should logically see a reduction in death rate post-COVID, barring other effects.
It sounds like you're arguing that if the pull-forward effect is significant, then we should see a stronger dip than we've seen, and that argues against a pull-forward effect. But I'm not sure we have accurate enough data to tell whether there is a dip, as I've explained above.
Note that you used exaggerated numbers to get the desired effect: a 200% increase in deaths during COVID while more realistic estimates are 20-25% (though over roughly a 2-year period). Additionally, you assume ¾ of victims would have a life expectancy of only 2 years; I think it's plausible the average was higher but still far below average overall (say, 5 or 10 years) which would make the dip shallower and wider.
I understand you were using the example to explain the principle, and didn't claim your numbers were accurate, but we do need to use accurate numbers to estimate the size of the dip.
Even with your numbers, the dip is significantly smaller than a peak: a 200% increase followed by 44% decrease over the next three years. If it turns out that e.g. a 25% peak results only in a 5% dip in the following years, that wouldn't necessarily be visible in a low-resolution graph that includes lots of other phenomena too.
By the way you have to account for the fact that the US has an uneven population pyramid, with many Baby Boomers who are going to die in the coming decades, which is why I suspect the numbers are trending up in the first place.
I'll grant you that there is no obvious decline to compensate for the COVID peak, but the graph is also conveniently missing the data from 2024, which is really important: does 2024 continue the sharp decline from 2021–2023 or continue the upward trend from 2010? That seems a really important question we cannot answer with this data.
And as I said, there is significant demographic change as well, so ideally I'd like to see this data controlled by age.
The idea you're referring to is often called the "pull-forward effect" or "mortality displacement," where a spike in deaths during an event like a pandemic reduces subsequent mortality because many of those who died were already near the end of their life expectancy. Let’s examine the available evidence to see if there’s support for this theory regarding COVID-19 deaths after 2021, using data and trends from reliable sources.
### Background from the Provided Link
The Census Bureau article you referenced confirms a significant spike in U.S. deaths during 2020 and 2021 due to COVID-19. It notes a 19% increase in deaths from 2019 to 2020 (from 2,854,838 to 3,390,029), the largest in 100 years, with deaths remaining elevated in 2021 (19.7% higher than 2019). The data highlights that older adults, particularly those aged 85 and older, accounted for a disproportionate share of deaths, suggesting that many fatalities were among those with pre-existing health vulnerabilities. However, the article’s data stops at 2021, so we need to look at more recent sources to assess post-2021 trends and the pull-forward effect.[](https://www.census.gov/library/stories/2022/03/united-states-deaths-spiked-as-covid-19-continued.html)
### Evidence on Post-2021 Mortality Trends
1. **Decline in Overall and COVID-19 Deaths Post-2021**:
- **2022 Data**: According to the CDC’s provisional mortality data, the total number of U.S. deaths in 2022 was approximately 3,273,705, with an age-adjusted death rate of 832.8 per 100,000, a 5.3% decrease from 879.7 in 2021. COVID-19-associated deaths dropped significantly to 244,986 (7.5% of total deaths) in 2022 from 462,193 in 2021, a 47% reduction. This indicates a notable decline in both overall and COVID-19-related mortality after the pandemic’s peak.[](https://www.cdc.gov/mmwr/volumes/72/wr/mm7218a3.htm)
- **2023 Data**: The decline continued into 2023, with provisional data reporting 3,090,582 total deaths and an age-adjusted death rate of 750.4 per 100,000, a 6.1% drop from 2022. COVID-19 deaths fell to 49,928 (1.6% of total deaths), a 68.9% decrease from 2022. This suggests a sustained reduction in mortality as the pandemic’s direct impact waned.[](https://www.cdc.gov/mmwr/volumes/73/wr/mm7331a1.htm)
2. **Natural Increase Rebounding**:
- A Census Bureau report indicates that natural increase (births minus deaths) began recovering post-2021. In 2021, natural increase was only 144,013 due to high mortality and a 3.5% drop in births. By 2022, it rose to 236,155, and by 2023, it reached 531,432, driven by fewer deaths and increased births. The number of states with more deaths than births (natural decrease) also dropped from 25 in 2022 to 19 in 2023, suggesting a return toward pre-pandemic mortality patterns. This rebound could be consistent with the pull-forward effect, as fewer deaths in later years might reflect the prior loss of vulnerable populations.[](https://www.census.gov/library/stories/2023/12/state-population-estimates.html)
3. **Excess Mortality Trends**:
- Excess mortality, defined as deaths above expected levels based on pre-pandemic trends, provides insight into whether deaths dropped below normal after the pandemic. A study published in 2024 found that excess deaths declined from 655,735 in the first pandemic year (March 2020–February 2021) to 586,505 in the second (March 2021–February 2022). However, for younger individuals (under 65), excess deaths increased in the second year, particularly for non-COVID causes like alcohol, drugs, and homicides, suggesting that the pull-forward effect may be more pronounced among older populations.[](https://pmc.ncbi.nlm.nih.gov/articles/PMC10988222/)
- Another analysis noted that excess mortality was highly concentrated among older age groups in 2020, with those aged 85 and older accounting for 34% of excess deaths despite being only 3% of the population aged 25 and older. This supports the idea that COVID-19 disproportionately affected those closer to natural life expectancy, potentially reducing subsequent mortality.[](https://www.census.gov/library/stories/2021/02/indirect-impact-of-covid-19-results-in-higher-pandemic-death-toll.html)
4. **Pull-Forward Effect Discussion**:
- A post on X by @Voltaire765 references the pull-forward effect, citing Alex Berenson’s Substack, which argues that after the initial COVID-19 death spikes, mortality should drop below normal as those who died early (primarily elderly and sick) were no longer at risk. The post suggests that five years post-COVID, mortality data in the U.S. and Britain might reflect this. While this aligns with the theory, X posts are not conclusive evidence, and Berenson’s claims require scrutiny against primary data.
- The significant drop in COVID-19 deaths from 2021 to 2023, alongside a decrease in overall death rates, aligns with the idea that the pandemic accelerated mortality among vulnerable groups, particularly the elderly. The high proportion of deaths among those aged 85+ (e.g., 307,169 COVID-19 deaths from January 2020 to June 2023) supports the notion that many who died were already at higher risk due to age or comorbidities.[](https://www.statista.com/statistics/1191568/reported-deaths-from-covid-by-age-us/)
- **Non-COVID Excess Deaths**: While COVID-19 deaths plummeted, excess deaths from other causes (e.g., drug overdoses, homicides) increased, particularly among younger populations, which offsets the expected drop in overall mortality. This suggests that the pull-forward effect may not fully explain post-2021 trends, as societal disruptions from the pandemic contributed to other mortality risks.[](https://pmc.ncbi.nlm.nih.gov/articles/PMC10988222/)
- **Vaccination and Immunity**: The sharp decline in COVID-19 deaths after 2021 is also attributed to widespread vaccination and prior infections, which reduced mortality risk. This confounds the pull-forward effect, as lower deaths may reflect improved immunity rather than solely the loss of vulnerable individuals.
### Conclusion
There is evidence supporting the pull-forward effect post-2021, particularly among older populations. The significant decline in overall and COVID-19 death rates from 2021 to 2023, combined with a rebound in natural increase, suggests that the high mortality during 2020–2021 may have reduced the number of vulnerable individuals, leading to fewer deaths in subsequent years. However, this effect is not universal across all groups, as younger populations saw increased non-COVID mortality, and factors like vaccination, undercounting, and data lags complicate the analysis. While the theory holds some weight, especially for the elderly, it’s not the sole driver of post-pandemic mortality trends.
For further details, you can explore the CDC’s National Vital Statistics System (https://www.cdc.gov/nchs/nvss/index.htm) for updated mortality data or the Census Bureau’s population estimates (https://www.census.gov) for natural change trends. If you’d like me to generate a chart comparing death rates or excess mortality across years, let me know!
Frankly, if people believing that this didn't happen at all is what it takes to keep things moving forward, so be it. This country can't afford to become more sclerotic than it is now. Lives come and go, and a few million are easily replacable when you have hundreds more to work with. It's time to move on.
I think this is a bad way to do this, and the fact people think it didn't happen is support for that argument - people don't believe this because the experts and public health officials who support it are liars who lie all the time. They don't...uh, lie in the way people think they lie, but they definitely massage the facts to fit their narrative.
And the result has been a loss of credibility and usefulness.
Giving false reasons to do good things turns out not to be reliable. It gives you goals other than getting good things done, and like a zany sitcom plot, those goals will eventually take up your whole attention, and your original goal will be tragically or comically abandoned.
(Also helpful to your goal, which I support, would be not speaking like a comic book supervillain. Lives matter. Of course they do. You can acknowledge that society can't be infinitely safe without denying that.)
The trust is lost regardless. They shouldn't have made statements that could be used against them, regardless of whether they were true or not. And, failing that, they should have thrown the people making the offending statements under the bus. But they didn't, and here we are. Still, there's some silver linings to this outcome, and it would be foolish not to take advantage of them. There's no need to make the situation worse by making mountains out of molehills.
I appreciate Scott doing this follow-up because I will be more skeptical of his commenters in the future. It's been sometimes challenging to assess the credibility of comments on topics that are novel to me (unlike this one), and this post gives me insight into how many complete morons are among his followers.
I am pulled in two directions: it is extremely frustrating sharing a comments section with a large number of people who have no interest in truthseeking whatsoever, yet who aggressively act like they are the smartest people in the world. On the other hand, it's a huge public service for Scott to be so patient and thorough at exposing so many people with, uh, rationalist skills still at an early stage of development, to what actual reasoning looks like.
It’s unfortunate that the response has been to destroy all of the institutions that are engaged in progress, because you’re mad that they made less progress in 2020 than you wanted! It’s time to move on and stop blocking vaccines, scientific research, new social technologies, etc!
I really don’t think that it matters what evidence you present about this, because Covid skeptics have a remarkable ability to ignore any facts that contradict their beliefs. Anyone getting their beliefs from people like RFK and Alex Berenson, who are really shameless in the way they twist facts, are not people who can be persuaded. In general, conservatives have decided reality is whatever they want it to be to a remarkable degree
I think this is true of some people but not others. All of these people were convinced because someone argued them into their position, and if people are constantly arguing for these ideas but refuse to argue against them because "nobody can be convinced", they win by default.
More realistically, I don't think I will change the mind of any hardcore conspiracy theorist. But I think some conspiracy theorists will refine their conspiracy theories to be more plausible (while learning some useful things about good thinking skills), and people who are not yet either explicitly conspiracy theorists nor anti-conspiracy-theorists will hone their thinking skills and be less likely to believe conspiracy theories in the future.
Maybe this is true of the people you know, but the conservatives I know (which are many; I was raised very conservative) have very insane views on Covid, most of which are not remotely based on fact and many of which are mutually exclusive (ie Covid can’t be both a leaked Chinese bioweapon AND designed as a plot to enrich Anthony Fauci).
If it changes the mind of conversatives he knows, but not the minds of conservatives you know, that's a victory. Because a world where some people change their minds and some don't is of course better than a world where nobody did.
I see somebody has never been to the Historic Downtown Anthony Fauci Bio-Thunderdome, which doubles as his mountaintop fortress. Shaped like a skull, naturally.
> All of these people were convinced because someone argued them into their position
Or maybe it's because they were never willing to believe the alternative in the first place. After all, what good does it do to believe such an unavoidable, senseless tragedy occured? It's not as if you have any workable plan to prevent this from occurring again either. Better to believe it never happened, or better yet, put the blame for it on the people you want gone. Both seem more useful than making yourself miserable for no good reason.
But it was avoidable. International travel could have been suspended to prevent the virus from entering, then with a lower baseline and non-endemic virus in other countries we could more easily isolate & contact-trace the few who brought it in.
But none of that happened, did it? In fact, no country was able to accomplish that, suggesting that this wasn't a localized issue at all, and that the root of the problem is beyond the scope of national politics. Whatever would have been necessary to prevent its spread, there would have been too many incentives standing in the way to get even close to accomplishing that. And given that this wasn't even the first global pandemic... what makes you think next time would be any different? If humanity still exists by then, and they have the same hindsight and knowledge of infectious diseases that we did right before Covid, they will make the exact same mistakes again. Why wouldn't they?
So yes, in practice, it was unavoidable. And the things that would need to change in order to prevent this would not require the consent of the rabble anyways, so I don't see any point in wasting time arguing with them.
Many pandemic preparedness plans have been written. If you truly care about preventing tragedies, you may wish to read one of them and perhaps convince people to implement it.
Mostly because that's what elites they look up to believe (or pretend to). So it's those elites that largely need to be convinced, or utterly discredited and overthrown. Both approaches prove to be pretty difficult in practice...
I think most True Believers were "convinced" of whatever conspiracy theory because
1. It flatters their egos to think that only THEY can see the truth, and that everyone else is just a sheep
2. It validates their rage at having to do something they consider inconvenient or scary. If it's a conspiracy from the NWO, the mask or vaccine requirement is not just an annoyance meant to save lives (which would not be socially acceptable to complain about), but a form of tyrannical oppression that MUST be resisted
3. It is a tool they can use to partisan ends. First Dems did this (Kamala's insistence during the primary that she wouldn't take a vaccine rushed out by Trump), then the valence flipped (afaict because the actual acting president changed, even though the vaccine really was made as quickly as possible under Trump's aegis)
I don't think these people were convinced rationally, they just accepted something they already wanted to believe. I still think it is good to make sure that true arguments for correct conclusions are ubiquitous, because otherwise low-information people end up believing conspiracy bullshit by default.
Fireflies flash in specific ways to find mates, birds pick their own song out of a squawking forest.
Somewhere out there there's a budding rationalist who just hasn't deeply weighed out the alternative views on <your topic here>. If you speak thoughtfully, it can cut through the noise and resonate where it matters, even if that's like 1% of people.
I believe this because I've often had 'conversion' experiences after reading arguments similar to this one. Though more often each argument moves me just a few degrees.
This one didn't radically flip my view, but reinforced it slightly, and better prepped me to help acquaintances and in-laws weigh key evidence in a calm and respectful way.
My most unsubstantiated conviction is that the talkorigins archive wasn't a failure at all and we should have doubled down on it and done extensive argument mapping for all controversial topics.
(Maybe AI has already quietly solved for this, "@grok is this true? @grok what are the flaws with this argument?")
I was raised christian conservative, then went to university as a left-leaning atheist, and am now a researcher (and still atheist). Unfortunately, I would say your last sentence applies cleanly across the entire political spectrum. Science is rife with it.
The problem is and was very simple; Covid was a time with very little confidence, yet the scientific establishment pretended high certainty about a long list of things that later turned out to be provably false (masks don't work, then suddenly they are obligatory; Trump is pushing the vaccine development dangerously fast and one month after Biden's election it's suddenly 100% safe; vaccines fully stop the spread so we can open up again oops actually they only protect the person taking it; Lockdowns are sad but necessary actually countries without lockdowns did just as fine; Lab leak is conspiracy theory well actually most intelligence agencies consider it the most likely reason, etc.). And social science already had been doing the same for social justice, so it was straightforward to extend that mistrust just one category further.
When people lose trust in the institutions, they start looking elsewhere. These other places however have little to no quality control, so it's easy for grifters to get a foothold. It's stupid, it's completely avoidable, but that's what happens if you burn your long-term credibility in the name of the short-term greater good(which actually turns out to be wrong later anyway).
I agree somewhat but also disagree. I agree a lot of our institutions and "establishments" failed very badly during covid (and continue to fail in many other areas), however, the widespread embrace of conspiracies and obvious misinformation started way before any of the events you mentioned. For example, I remember many conservatives I knew promoting the "Plandemic" movie almost as soon as the lockdowns started. I also remember at least one acquaintance trying to argue with me that the covid vaccines were dangerous before they were even invented.
Also, I remember some of those events differently than you do. I know some Democrats said Trump was rushing the vaccine approvals, but that wasn't the general consensus on the left. Similarly, after the vaccines rolled out I saw a number of articles warning that if covid mutated the vaccines would be much less effective, which is what happened.
Agreed: it's a big problem that authorities believe they can't admit uncertainty, frequently playing armchair psychologist when that isn't actually their area of expertise.
I don't think the scientific establishment "pretended high certainty" at all - I kept hearing scientists make weak qualified statements and then non-scientific officials taking all the qualifications away.
But it's probably unfair to expect the general public to know what scientists are saying, if they can't trust the people whose job is to transmit scientific knowledge to them. It's possible for scientists to behave well, while the "scientific establishment" behaves terribly.
Perhaps we should call it the "science-bureaucracy complex" to make it sound more sinister, and to emphasize that it's not just scientists running the show. It works for "military-industrial complex", after all.
The "scientific establishment" is more than just "scientists". If the scientists speak perfectly hedged, generally incomprehensible academese to each other and then communicate the results to the outside world through "non-scientific officials", then those non-scientific officials are part of the establishment. And what *they* say, is what the establishment says in the only forum that matters.
If it's just scientists talking to other scientists (or science-adjacent nerds), then there's no reason for anyone else to ever fund scientists to do anything; they can buy lab equipment with the tax-free tips they earn as Starbucks baristas or whatever. If they want to be relevant (and funded), the scientists need to take responsibility for how their work is transmitted. Because they will, properly, be held responsible for such.
The problem is that those non-scientific officials are generally the bosses of the scientists, so it’s hard to hold the scientists responsible for their bosses’ fuck-ups.
Also, this isn’t a failure of “the scientific establishment”. Professor Chris Whitty (UK Chief Medical Officer) was very clear that the vaccines would prevent transmission for the current version of COVID, but that couldn’t be guaranteed for future ones (which is, perhaps, why I’m less annoyed than Americans).
It’s a specifically-American failure of your “scientific establishment”. That is, it’s a failure of Americans who were managing scientists. Not of politicians, not of scientists, but of the middle layer in just one country. Other countries didn’t have the same failures. But “science” and “the scientific establishment” are global phenomena, not national ones.
The "bosses of the scientists" are, while they are acting in that capacity, part of the "scientific establishment". There is an edifice that consists of scientists, bosses of scientists, and mouthpieces of bosses of scientists. Some of whom also have other duties, but in *this* context they are part of an edifice that is spewing bullshit masquerading as science. Words that cause many people to believe false things about important subjects.
That is a huge failure of the scientific establishment, even if one scientist named Chris was very carefully pedantically correct in the words he spoke to other scientists within the establishment. Even if *every* scientist was carefully pedantically correct in all of their intra-scientist discussions.
We all don't fit into simple boxes. Many of us were rational and skeptical, and yes, based on factual evidence early on, such that outdoor transmission was extremely unlikely. This from PubMed in May 2020. None of the media and few experts clarified this.
That isn't true. Many experts and many media people did clarify the facts around outdoor transmission. It was not hard to find that stuff.
A LOT of the skeptics claim that the media never reported facts that they learned about from the media. I will agree though that while the facts were available, the most easily available information from the "establishment" was often incorrect or nonsensical.
Also a lot of what people seem to believe was official was just vaguely authoritative social media posts from people at the more scared and gullible end of the spectrum. There’s one I saw everywhere “explaining” why masks were to help others and not yourself which was not only wrong (that’s not really true) but also from an Australia anti-flu campaign
I mean, TONs of media and experts clarified this. I know this because I listened to them and behaved accordingly.
But the media ecosystem is sufficiently fragmented and attention spans are what they are that it was easy to go through the pandemic watching popular media without having that information internalized.
This is the problem of our time. It's in no one's financial interest to do accurate, factual mass-communication, so no one does it, and we are all worse as a result.
With regard to the underestimation of COVID, allow me to point out what I think is the largest elephant in the room of public discourse, in politics, mainstream media, and incredibly even in the scientific community outside of long COVID specialists: the ongoing risks of contracting long COVID even for vaccinated people (only partially protected against it) and children (who can suffer permanent consequences like diabetes too). To grasp the extent of the underestimation, I think it is sufficient to read the following recent review article from Nature Medicine: https://www.nature.com/articles/s41591-024-03173-6
(from its abstract: “Long COVID represents the constellation of post-acute and long-term health effects caused by SARS-CoV-2 infection; it is a complex, multisystem disorder that can affect nearly every organ system and can be severely disabling. The cumulative global incidence of long COVID is around 400 million individuals, which is estimated to have an annual economic impact of approximately $1 trillion—equivalent to about 1% of the global economy.”)
("A new study that my colleagues and I published in the New England Journal of Medicine on July 17, 2024, shows that the risk of long COVID declined over the course of the pandemic. In 2020, when the ancestral strain of SARS-CoV-2 was dominant and vaccines were not available, about 10.4% of adults who got COVID-19 developed long COVID. By early 2022, when the omicron family of variants predominated, that rate [of long COVID for infected people] declined to 7.7% among unvaccinated adults and 3.5% of vaccinated adults. In other words, unvaccinated people were more than twice as likely to develop long COVID.
While researchers like me do not yet have concrete numbers for the current rate in mid-2024 due to the time it takes for long COVID cases to be reflected in the data, the flow of new patients into long COVID clinics has been on par with 2022").
It is not coincidental, given their knowledge of the syndrome, that the already mentioned Dr. Al-Aly and another author of the Nature Medicine review article, Dr. Akiko Iwasaki ( https://medicine.yale.edu/profile/akiko-iwasaki/ ) still wear a mask virtually everywhere.
For people looking for concrete numbers, I like to point to this paper: [1]. In their sample from mid-2022, about 0.23% (from 1.67% * 14.12%) have apparent chronic fatigue syndrome due to COVID infection. If we assume that, on average, they'd had half a COVID case per person, that means that each COVID infection caused about a 0.5% risk of chronic fatigue syndrome. Divide that in half because of vaccines, and we can expect about a 0.25% risk of chronic fatigue syndrome every time you catch COVID, with a conservative assumption that your risk of catching chronic fatigue syndrome is independent each time. Given that people with chronic fatigue syndrome say, on surveys, that chronic fatigue syndrome is about half as bad as death, and that chronic fatigue syndrome is generally incurable, and that the average person gets COVID about once a year in my native Canada, that's really scary. (Remember that, since chronic fatigue syndrome is incurable, that number will keep going up over time. My guess is that the steady state is somewhere in the single-digit percent of the population having chronic fatigue syndrome, which will be a small but real drag on GDP growth.) It's why I personally wear a mask almost everywhere indoors: given my remaining life expectancy of about 60 years, that means I lose more than two weeks of quality-adjusted life expectancy each time I catch COVID. (Perhaps I should rate my risk even higher, since chronic fatigue syndrome is more common among young people.) I also get my COVID booster each fall, since it mRNA boosters in the fall reduce the risk of COVID during the winter season by more than half.
It's not clear to me if I should assume that the risk of chronic fatigue syndrome is independent between infections. It's also consistent with the data about chronic fatigue syndrome to believe that some people are born susceptible to chronic fatigue syndrome, and most of those people will go on to develop chronic fatigue syndrome at some point. But given that we don't know, I have to assign a pretty high probability to the risk of chronic fatigue syndrome being independent each time you get infected.
As far as I can tell, the risk of chronic fatigue syndrome is, for under-50s in otherwise good health, more serious than everything other risk from COVID combined (including the other long COVID stuff).
(The authors hesitate to say it's definitely chronic fatigue syndrome, presumably because nobody knows what chronic fatigue syndrome is or how it works, but I don't see any reason to think that COVID-induced chronic-fatigue-syndrome-like-illness is different from chronic fatigue syndrome as we know it.)
Scott wrote a post three years ago (now behind the pay wall) detailing the statistics that lead to his conclusion that long covid is a psychosomatic illness. His reasoning seemed reasonable, but I've had ME/CFS for 20 years with weirdly specific features that are common among people with the syndrome and/or long covid, so I have to dismiss his conclusions.
It's worth saying that "long COVID" is so poorly defined as to be essentially a useless category - generally including any symptoms that linger longer than a few weeks. I assume this is so that press releases about research can say "we studied a thing that affects 15% of people who get COVID, we're really important, please give us media interviews so we can get grant money that our institution will take a large chunk of to fund more press releases" or whatever - perhaps I'm a bit jaded. But almost all "long COVID" stuff goes away relatively quickly - presumably including whatever Scott was talking about then. A small (but still surprisingly large) portion of "long COVID", though, is actually quite serious stuff like chronic fatigue syndrome or consequences of having a heart attack or stroke (COVID can cause blood clots). I think a separate category - like "serious chronic COVID complications" or something - would be useful.
(Also, to anyone reading this who can direct grant money: please spend money researching chronic fatigue syndrome; it's a huge cause of disability and mortality but is very poorly understood.)
Man, I'm not sure if this is a case of "I'm a completely insane person that spent 17 hours a day on early covid twitter and knew how to counter this stuff by heart by April 2020" or "Political polarization and its consequences have been a disaster for the ACX comments section" but I'm surprised your commenters could be split on this. The evidence against the "with covid but not of covid" stuff is overwhelming just by excess deaths even before you get into estimates of IFR and the infected fraction.
I was hoping this would get to the (still utterly unpersuasive to reduce the magnitude of the tragedy, but certainly potentially making the comparison to say, the civil war unfair*) 'years of life left of covid victims', which given the old or pre-existing disease over-representation of deaths, probably makes the number seem less bad than it seems at first glance, even though still more years of life lost than events of substantial historical/cultural impact. But it seems that better argument was absent.
I feel bad singling out anyone, but the last comment in particular made me groan. I know several people who died or covid, including early on a former high school teacher than was only on her 50s and substantially scared away my parents, of similar age, that this could happen to them too. Would that convince any covid downplayer than the number was exactly 1.2 million? Anecdotes are simply worthless on statistical analysis at the national, several millions of people level
*tho with the lower life expectancy back then, even after child mortality, maybe it's closer than it may first seem
Good grief, this is quite the rounding up. I know that to young people everyone over thirty seems ancient, but a 50-year-old woman can expect to live over thirty more years, on average.
> doesn’t justify taking away the irreplaceable joy of the college years from all of my younger friends.
That's irrelevant to the point, which is a factual discussion of whether 1+ million people died of COVID, not a political discussion over whether specific COVID policy choices were worth the tradeoffs.
There are three different things I would say to answer this.
The first is to return your preamble. I'm sorry you had to lose the joy of your college years due to the unfortunate coincidence of them being during the pandemic. If it's any consolation, experience makes me find the 'irreplaceable joy' a bit overstated: being a bit older, I personally enjoyed the social life soon after college (specially since I had my own money) more enjoyable, but part of this was people I first met in college and to the extent you lost that chance, it is truly regrettable.
The second is that you're arguing a different, and wholly conceded by Scott in his previous blogpost, point. I quote:
"Usually I’m the one arguing that we have to do cost-benefit analysis, that it’s impractical and incoherent to value every life at infinity billion dollars. And indeed, most lockdown-type measures look marginal on a purely economic analysis, and utterly fail one that includes hedonic costs. Rejecting some safety measures even though they saved lives was probably the right call."
The hedonic costs you mention are indeed what makes this not close. I would add that people are not automatons blindly following the law: people naturally relaxed, met irl and partied a bit when case counts went down, then stayed home or tried to do stuff online as a new wave of deaths came. Analyses that take this into account are the ones that make lockdowns not worth it on a purely cost-benefit perspective. This doesn't change the main question of these two blogposts, which is about the huge number of deaths and the fact they are legit.
The third is that '50s is nearly elderly' is something I'm skeptical most would agree with outside this context. In Spain, a 54 year old woman that dies from covid was expecting 31-32 years of extra life, including several years of retirement and knowing the grandchildren. It is basically equivalent to a 25 year old male dying in the civil war in 1860 (who would expect just a bit more, around 35-36 years of life if he survived rather than died in the American Civil War)
I submit that people in their 50s are very much not done growing and producing things, and if you did not want to imply otherwise you should not have said as much. If you want to argue over the tradeoff between disruption of two years of college and shortening the lives of the elderly, I suggest you review the estimated overall life-year cost and put forth your conversion factor: https://www.cidrap.umn.edu/covid-19/study-covid-pandemic-stole-nearly-17-million-years-life-adults-18-european-countries
"Math can't speak here"? What blog do you think you're reading right now that you can just brush off statistics regarding human flourishing?
It's quite easy. Take an actual measure of how many years of human life were lost to COVID, how many years of college were disrupted, and compare how much value people get out of being alive versus how magical college years are and how much worse COVID made them. I should hope there is some level of tradeoff here that would give you pause.
Whether slightly, mostly, or fully trolling, what I want to get across is that the literal reading of it is probably more inaccurate than you imagine. Scott has received plenty of mockery from people not reading literally the next paragraph as the screenshots going around for conflating deaths of old people from covid to deaths in the American civil war, and I just wanted to press the point that qualify of life and life expectancy has changed so much in the last 140 years than a 54 year old covid victim and a 25 year old civil war victim are much closer than people realize, so that basic intuition does not hold
yes, I agree it's different. namely, I would rather be a Spanish 54 year old woman in 2020 over an American 25 year old man in 1860, and nothing about 'life stage' or 'personal trajectory' compensates for the massive difference in quality of life standards, but I decided not to mention this since it's further opposition to your point.
For how long? Most non-college young people are in service industry jobs like restaurant and store operation, which reopened almost entirely after a few months. Aside from that, very few people were ever prohibited from hanginb out in the local park or visiting friend's house. True "lockdown" only lasted about six weeks.
What is a "lockdown" by your standard, and where was it extended past the initial six weeks?
The kids flip burgers no matter what, the presence or absence of masks seems of limited impact. Getting laid up for two weeks from talking to your customers sure sucks though.
I was always under the impression that the lockdown was imposed not primarily to save lives but to prevent hospitals and hospital staff from collapsing under the weight of the gravely ill.
I'm extremely sympathetic to your complaints about lockdowns.
But my kids will be teenagers when I'm in my 50s. If I died in my 5th decade, would you tell my kids that the "irreplaceable joy" of your college years is more important than having their father around for some of the most formative years of their lives, to say nothing of their next few decades as adults (which I hope to be alive for)? What if I were already a widower and they'd be losing the only parent they have left?
Our lives are not entirely our own. When we die the loss is felt by family and friends. Surely they deserve some consideration, no?
I still get a bit mad everytime I remember some of these, particularly one article titled 'the bearer of good coronavirus news' focusing on Ioannidis and saying total US deaths would get to 40,000 or something. The tone was so much 'why do people resist good news?!?!?!' instead of modeling someone that truly believes the claims to be false for technical and empiricial reasons with high confidence, and thus sees people letting the guard down about the coming tragedy/disaster, very similar to such debates when it comes to climate change, AI risk, etc. I wish all of these were nothingburguers, but I was/am simply not convinced!
I realize I should probably meditate and get over it or something.
I should really reread his stuff on most published findings being false. That’s great stuff, but now I’m worried that he meant it to say that research is bad, rather than saying that research is good as a practice even though individual claims shouldn’t be believed.
I think it's the opposite. He made his career by being against the consensus and doubtless made some enemies that way. As I think we've seen, that takes a tole on folks and can make them become reflexively contrarian across the board. I view his COVID-era thinking in that light. Doesn't hurt that there was plenty of shoddy analysis being thrown around at the time. Back in May of 2021, every "data scientist" with a blog had a COVID tracker app and a SIR model projecting exactly how many of us would die.
It's strange to see the comments attract so many deniers given that Scott has been quite consistent on the dangers of pandemic disease in general and COVID in specific. I remember when the EA/rationalist community was a few months ahead of the curve in calling that COVID would be a big deal, or at least that the odds of it being a big deal were severely underestimated. I suspect the "skeptics" here are a significantly different crowd, but I don't know what would attract them other than the fact that Scott is unusually reluctant to ban anyone who believes themselves to be arguing in good faith.
Still, I think it's good that Scott bothers to write such detailed proofs of obvious things. It's good to have strong arguments spelled out in detail to point to when "that's obviously dumb" fails.
oh absolutely, to be clear I agree. it's more of a 'despair that this is necessary' more than 'Scott was wrong to post this'. while plainly wrong, I don't think the 'don't give creationists a platform by debating them' stuff applies for this sort of position
Yeah, I wouldn't argue with these guys on CNN with no moderator support, but that's not what Scott's doing. He's writing TalkOrigins for COVID, and that's something we really need.
>The evidence against the "with covid but not of covid" stuff is overwhelming just by excess deaths
The evidence for lab leak versus wet market should get you *at best* to a 50/50 shrug, and yet Scott spent way more time on that one.
Political polarization is certainly one reason this comment section is lower quality than back in the SSC Golden Age (ie, 2014-15), but not the only one, and not only for the reason you seem to imply.
I'm confused about both your first sentence and what you think I'm implying.
Most obviously for the former, if something is a controversial 50/50 shrug, *obviously* you will spend more time presenting the presumably strong case each side can make that results in such a close call? as opposed to simple things, which require less time? as seen in the fact this article here is short on wordcount for Scott's standards
edit: just for clarity (in case you don't check this in e-mail and do see the edit: I'm *wondering* if it's political polarization, the other option being that I'm the insane one. it's an unsure statement)
Short version: it was very unclear snark about the way Scott gives attention to covid. More clearly, I think he is stuck on wanting to write about COVID (not unreasonable) but also chooses the least interesting ways to write about it. Is this really what's worth his time?
Longer but quickly typed, inaccuracies likely abound:
I think the Rootclaim thing was a waste of time and Scott should've stuck at 50-50.
>For what it’s worth, I was close to 50-50 before the debate, and now I’m 90-10 in favor of zoonosis.
I also think that GoF research is risky whether or not covid was a result of GoF, and arguments in favor of zoonosis are sometimes (not necessarily by scott, but others) used to act like GoF isn't risky.
That's the kind of thing I think someone of his talents could spend more time on. There are all sorts of policy failures and social failures around COVID, and he's now spent *how much time* writing about one not very important thing (zoonosis vs gof is broadly irrelevant unless we're considering to destroy Wuhan specifically) and one not particularly interesting conspiracy that *also* will have no impact how to react to future pandemics.
I mean, his blog, his talents, his engagement farming! I just think he could do more, and gets stuck in these outrage-inducing but not really impactful cul de sacs instead.
I mean, I think I either disagree with you, or you advance possibilities you don't believe in but worry others will believe, but that do not apply to me. For example, I think gain of function research is risky and must be stopped and will believe that exactly as much if tomorrow they find the exact animal with one base pair difference to wild type covid on its lungs. To the extent some people use arguments in favor of zoonosis for that, they're my enemies.
I also do not think it's about destroying Wuhan specifically. Back in the pre-pandemic days, people complained about serial passage in ferrets to make a strain of avian influenza airborne, and that was within the US, UK and Netherlands if I recall correctly.
As for the reasons for Scott to spend his time, I'm afraid I'm less concerned to let him write whatever he wants. Perhaps this was just strategic considerations as he saw people around him over correct too much towards lab leak, perhaps he found it interesting and decided if he was going to sit down six hours of it he might as well take notes and make it into a blogpost. With the amount of writing output he can do I'm not overly concerned.
>I'm afraid I'm less concerned to let him write whatever he wants.
Well, yes, as I said
>I mean, his blog, his talents, his engagement farming!
Of course he can write what he wants, and he will. I am making a suggestion that he could have more impact on other topics.
And suggesting it should be unsurprising that the comment section is contrarian and split when the posts ultimately aren't... "substantive" isn't the right word. Pushing forward? Broaching unaddressed topics? Missing the forest for the trees?
Unless you're writing government lab policy, lab leak versus zoonosis is kind of a nerd snipe. Deaths with COVID versus deaths of COVID is sort of interesting, but it's also Scott picking a fight with a subset of his audience/some guys he finds irritating on twitter, rather than trying to communicate something novel.
Not every post can be a banger, not every reader has to love every post. But I think it's fair to observe a marked shift in the sort of topics of his during the pandemic posts, and these later ones.
I... have nothing to contribute to this debate. I just want to show up and be a person online who agrees with and supports things that are obviously true. In this case, I mean, the things that Scott is saying. Let no-one think that all online people believe in craziness. We believe the normal stuff, we just don't comment that often.
I have a friend whose both parents died of covid (both started having difficulty breathing, went to a hospital, died there). So the idea that covid is not real (and my friend probably just suffocated his parents using a pillow, and the nurses in the hospital finished them) is just absurd to me.
Of course, this is internet, so I expect people to argue for absurd things. And in absence of a downvote button, people can compensate for the absurdity of their opinion by saying it more loudly and more often, because that's how you win online debates.
I can’t believe you are taking people who question the mortality stats seriously. Not only was there unprecedented excess mortality, the daily mortality counts always peaked two or three weeks after clinical cases peaked.
Even in the comments section here, no one skeptical of the death counts is offering a causal model for the relationship you described. It's pretty telling...
>so if 1.2 million died of COVID, that’s about 1/300
Another way of making this more intuitive: "do you know 300 people? If not, it's not surprising that you personally don't know anyone who died from COVID"
It's actually 208 people! I'm not smart enough to automatically remember the math myself, but I'm smart enough to know this is one of those situations where you ask a mathematician before giving any specific figure (in my case I asked o3, and double-checked with Claude).
This comment made me check the math; it's correct. When I read the article I assumed you computed it yourself, and I wish you had said you got it from AIs in the article. I know they're getting good, and they did get it right — but I'd still like to know where the information I read is coming from. One reason why: as the companies do more reinforcement learning on usage-time/subscription-revenue/etc, the AIs will start trying to manipulate us more; they've been getting better at that too.
For the record, here's the math: Let p be the probability of death of covid. Either know >=1 people who died of covid, xor you know 0. It's easier to find the probability of the latter. The probability *surviving* covid is (1-p), so assuming independence the probability of all n people you know surviving covid is (1-p)^n — the more people you know, the less and less likely it is that everyone you know survived. We want to know when this reaches 50%; ie, how big does (n) have to be for the probability to fall all the way to (1-p)^n = 0.5. Solving, we get n = log_(1-p)(0.5). In our case, p=1/300, and with a calculator we get log_(299/300)(0.5) ≈ 207.6.
So if you take 207 people at random, there is a slightly higher than 50% chance that all of them survived. If you take 208 people at random, there is a slightly less than 50% that all of them survived.
The response will be that they know far more vaccine injured people, including people who died from myocarditis, and that at some points the VAERS numbers far outmatched the covid fatality rate because some antivax activist(s) spammed their followers to submit false reports to the database.
I will neither agree nor disagree, but could it not be the case that what COVID did if just front-load(*) deaths that would have occurred anyway due to putting stress on people in already frail and fragile condition? We do know that people with diabetes and obesity were more likely to be affected. I'd love to see how that first figure you presented continues. Does it slow down after that? For example, if you extend the line from pre-2020 trend through to 2025, does it fit? Was the bump transitory or a permanent shift upwards?
(*) A bit like incentives for people to have children; they don't work, they only make those who were going to have them anyway do it earlier. There's little to no permanent effect.
You can easily look this up yourself with a two minute search. The statistics don’t support that theory. Excess deaths are still higher than before COVID, even after 2022.
Since life expediency at birth is increasing at a lower rate than median age, excess deaths in a year on year comparison would need to be normalized by that before we could rule out this sort of front-loading. Otherwise its kind of pointless.
Now you could say that the people collecting that data are taking that into account but a quick search of https://pubmed.ncbi.nlm.nih.gov/34190045/ (the source for the excess deaths from your link) doesn't indicate that considered median age in their calculations or really much at all. Meanwhile the source for the European excess deaths (https://ec.europa.eu/eurostat/cache/metadata/en/demo_mexrt_esms.htm) is explicitly just a rolling average where the population (and therefore expected deaths) of a decade ago is assumed to be similar to the population of today.
Digging into the data more, you could point to the fact that you see such burn effects in the French/German data every winter (spike than fall), but that is potentially a bit of a reporting artifact, and there is no such effect ANYWHERE in the US results which raises some eyebrows. Could be the US medical system is failing, could be that people in the US are dying of a lot of things such that deaths due to cold and flu are being drowned out by other periodic sources of noise. Maybe its fentanyl or it could be that the books are cooked and the 'I think its all lies' people are correct. Who really can say? It would require a lot more digging than a 2 minute search to figure it out.
Taking a step back, I want to also bang the loss of trust drum here. Between the continued metastasization of the replication crisis*, and the politicization of every single truth finding institution its generally really hard to make ground level claims, and I see a lot of commentators (and everywhere else really, but this is a walled garden I care about) here just drifting closer to "I alone hold the truth of the world, and everyone else is too intellectually lazy to figure it out". I think actually making the claim that 1.2 million Americans died due to COVID(and would still be mostly alive in 2025 otherwise) on a solid enough level that you could reach across divides and establish it as a ground truth would be a painful adversarial collaboration, and probably come out as a long ass report. Since it seems like that's not going to happen, all this sort of article yields is more stuck priors and more evidence that the other side is dishonest.
* This isn't even getting into the weeds of the academic coding crisis, a separate but related problem where most academics can't code, most research is coding, and bugs are literally everywhere and being used to make decisions.
This is the long-ass report. Scott has written even longer-ass reports on Ivermectin and lab leak vs wet market origins. At this point, it's on you to explain why you don't think a clear chart of recorded COVID deaths track precisely with observed excess mortality. Aging populations alone wouldn't have done that.
It's all well and good to say that more transparency and open communication could have better preserved trust in the medical establishment on a broad level, but individuals have a responsibility to justify their own claims. All this "the CDC gave inconsistent early guidance on masks therefore we can never know for sure, no I won't present a superior body of evidence I'm just saying we don't know" only demonstrates a poor grasp of logic and epistemic cowardice. The skeptics are invited to make their case or go home.
Well I didn't mention mask mandates, or the CDC in my comment, so thanks for making my point about prior sticking so clearly :P.
To be more serious, its not like mask guidance was the only thing that happened over the past couple of years that degraded trust. There were things like the Ferguson Imperial Model and its subsequent code issues, vaccine safety concerns about Operation Warp Speed, and their social inversion post election. There were issues with protests vs social distancing, and issues with vaccine rollouts getting muddied by politics instead of purely by efficacy, and that's without getting into the mud of "were lockdowns effective" or "was it a lab leak or wet market origin".
Regarding the discussion here, I'm not sure what you mean by the skeptics are invited to make their case or go home. Martin Blais asked a question about if and how much is it possible that the 1.2 million people would have died anyway by 2025 if not for COVID. It's a fair question given COVID's known comorbidities with other serve health issues like age. Calvin Blick responded that excess deaths do not show the sort of effect that one would expect if COVID had a front loaded death rate and thus would have a lower excess death for the next couple of years until it normalized.
At which point I pointed out that national aging is not something considered in excess death statistics. Given the speed at which the west is currently aging, this is a potential massive cofounder for looking at things like deaths from COVID as a front loading issue (75% of the 1.2 million dead were older than 65 https://www.statista.com/statistics/1254488/us-share-of-total-covid-deaths-by-age-group/, but the number of new 65 year olds is also spiking up by year https://www.stlouisfed.org/on-the-economy/2019/may/how-many-people-will-be-retiring-in-the-years-to-come). I then pointed out the there is no seasonal effect for deaths in the excess statistics unlike France and Germany and note that that is weird and needed further analysis to figure out if we could rule out COVID front loading. Then I bemoaned the breakdown in trust and the fact that there isn't a clear neutral source that we can turn to and determine, "Hey why aren't old people dying in the winter more than the summer right in in America? That's kind of weird right?".
If I'm playing the role of skeptic here, I've made my case on this point. I disputed Calvin's point and brought evidence to back me up both that excess deaths are elevated because of population aging and that the US is statistically weird. Seriously, the US has been sitting constantly at ~8% excess mortality for 5 years now! It doesn't have a natural season cycle of death like France or Germany. Additionally both France and Germany look like what you'd expect with a front loading effect (drop down to 0 or negative % deaths on a trend line that's slowly sliding up https://ec.europa.eu/eurostat/statistics-explained/index.php?title=File:Mortality_statistics_Fig01_v2.png) If you subtract 8% from the figure the US does have a similar dry tinder cycle.
The ball is in the court of the ... (non-skeptics? skeptic-skeptics? ) now. They can pitch back something like "If there was really a dry tinder effect it would have been so obvious that it would have dwarfed aging effects beyond even what a trendline would expect, because xxxx)" or "Even if the people who died of COVID would have died of something else in their near future, it would have been lower and they would have had many years" or "some other argument that my biases are blinding me from". That is how civil truthfinding is meant to work. Not just drive by "You brought no proof", or "You must bring a full gish gallop of evidence for every inch of this debate or I'm going to dismiss you".
Finally, and this is entirely in the voice and energy of Crocodile Dundee. This is not a long-ass report. Both articles written here are 2500 words together. Scott's H5n1 more than you wanted to know sits at 4500 words. https://www.astralcodexten.com/p/h5n1-much-more-than-you-wanted-to
That extra 3-4x word count gives a lot of room to dig into assumptions like "What about the distribution of deaths by age? What does that fact that ~50% (https://www.statista.com/statistics/1254488/us-share-of-total-covid-deaths-by-age-group/) of Covid deaths occur in a group that is older than the 2018-2019 US life expectancy(78.8) mean for the categorization of them as COVID deaths? What about the fact that 27% of the deaths are significantly older than the life expectancy?". I mean when we are talking about 1.2 million, that's potentially 600,000 people who were statistically expected to die.
I guess you can fall back and say 600,000 is still a tragedy, and that's true it is but that must means we have to go back to quibbling about population aging and health issues.
You have gone to the effort of making a case, I'll give you credit for that. I do appreciate it. I don't know much epidemiology or all the relevant math, but I'll give it my fair shake.
To dig into the specifics, it's true that the US has been aging, but it hasn't been aging that quickly. (https://www.statista.com/statistics/241494/median-age-of-the-us-population/). In 2017, the country was 38. In 2022, it was 39. That's fast enough to be concerning for the long-term fate of social security, but it doesn't look like we should be seeing a %5 increase in overall mortality on the slowest day of 2022. This suggests to me that we're seeing broad-scale impacts of a virus known to be more lethal than most and to inflict longer-lasting residual symptoms on survivors than is typical. COVID is past the point where it can be treated as a crisis, but it's still a respiratory disease with a seasonal cycle that's stronger than the flu, and the flu already killed ~20,000 a year.
I'm not really sure what you're talking about with the US not having a seasonal death cycle? It definitely looks like excess mortality is higher in the winter where the flu and COVID are known to flourish.
I'm not familiar with the idea of a "dry tinder effect", but I gather that you're trying to argue that the longer-term mortality rate is unaffected because as many as half of all COVID victims would be dead by now anyways? This seems contradicted by the fact that excess mortality didn't dip after COVID. If those people were going to die anyways, then their earlier deaths would have "borrowed" against later deaths since they couldn't die a year later as they would have in the counterfactual.
> "What about the distribution of deaths by age? What does that fact that ~50% of Covid deaths occur in a group that is older than the 2018-2019 US life expectancy(78.8) mean for the categorization of them as COVID deaths? What about the fact that 27% of the deaths are significantly older than the life expectancy?"
I don't dispute any of this, but I don't think it means much of anything at all. Sure, just saying "1.2 million deaths" masks the fact that most of them were old, but everyone knows that. If somebody was going to die next year of heart failure but instead they die this year of COVID, they go down as a COVID death because that's what killed them. I don't think it's very complicated. Sure it impacts the overall life-years lost to disease, but that sum is still pretty high (https://www.cidrap.umn.edu/covid-19/study-covid-pandemic-stole-nearly-17-million-years-life-adults-18-european-countries).
Mostly, I find this whole line of questioning kind of confusing. Like, you could do this for basically every disease. Most diseases in developed countries mostly kill old people. That's why fighting diseases is our main method of increasing human lifespan and healthspan. You never see somebody show up to the heart disease fundraiser arguing that everyone is wasting their time because most who die of heart disease would have just died of cancer in a few years. The fact is, COVID deaths were counted more or less exactly the same way as every other death of disease, and the number dead is about 1.2 million. I don't know what you're looking to prove by digging through actuarial statistics.
My cousin's husband died from COVID after being discharged from the hospital. Just couldn't get enough air at home, even with eventual supplemental O2. And yet the state of TX-or perhaps merely the city of Dallas--would not list COVID as his cause of death. Chris, my cousin, had to fight for weeks to try to get this changed. Not sure of the result, but I know the original decision.
So we know some places labeled cod as COVID if you died of something while incidentally having COVID. And we know of places that suppressed the actual cod via COVID figures. Looks like a wash, statistical noise.
Comparing annual death rates from COVID vs. Flu in Japan. COVID killed about 1/3 as many 0-9 year olds, but 33% more 20-29 year olds, and >2x as many 80+ year olds.
The flu is more dangerous for children than for teenagers/young adults, while COVID is safer.
To me, the reason everyone is still busy with COVID is because they're's no sense of closure. The perceived authorities have failed at primary tasks incumbent upon them.
The problem is that the perceived authorities were given the task of transparency and admission of imperfection, which was not a task they took on before. This is a task that will inevitably lead to perceived failure.
Tangential, but as a formerly-suicidal person I think the reason we (sometimes) cope with disaster better than expected has something to do with the fact that I often felt most hopeless when my life was going well. After all, if there's nothing in particular to blame the bad feelings on, and nothing you can take action on in an attempt to feel better, you're left with the conclusion the problem is "I am inherently broken somehow and will never feel good no matter how well things go." On the flipside, I broke my femur a while back and found it remarkably clarifying. Yeah, it absolutely sucked, but there was a clear external reason for the suck and a process to go through to make the suck go away (surgery, physical therapy, etc). And emotionally that's noticeably more pleasant than interminable existential angst.
(That said, I was not doing well during the pandemic).
My experience of covid-related debate has been that those who are in the 'denialist' camp are also bad at handling statistics. They don't have competency to handle the concepts needed, and will misunderstand things and draw the wrong conclusion.
Obviously there's a broad range, from people who are a bit vaccine hesitant because of what they've heard second-hand, through to the conspiracy theorists. But unfortunately that extreme has a large advantage in uninformed discussion, because they're not beholden to facts, and can wilfully misinterpret any set of data.
Some people can, but some people can't. There's the phrase "pearls before swine", and of course a literal swine can't learn statistics no matter how much you try to teach it.
More specifically, you get better at understanding and applying statistics by reading lots of good statistical arguments of topics to interest to you, and working through the math on the stats that are strongly counterintuitive to you.
A wicked fast update on a specific strongly held prior after reading some specific statistical metric (and then the math) is the most effective teacher at the nature of non-intuitive realities in the statistical domain.
* gotta love the technical term for “wicked fast update,” the first derivative of acceleration: jerk
The vast majority of pro-COVID (you know what I mean) are also bad at handling statistics, and so will be ineffective, perhaps entirely counterproductive, at correcting the denialists.
On vaccine hesitancy, there's also an issue that there for a while (maybe still?) being opposed to *mandates* got you called an inhumane braindead anti-vaxxer. Words are supposed to mean things and "antivax" became an absurdly large tent during COVID when it didn't need to be.
Very effective but not perfectly so - you just need to compare death rates among unvaccinated to vaccinated (obviously correcting for the necessary factors).
This is one of the things I find so frustrating about this camp. We hear all this innuendo and vague gesturing towards supposed statistical anomalies, etc. "Hmm, that really begs the question, eh??" And then if you spend one minute googling this question you immediately find ten different lines of evidence that the vaccine was extremely effective and it probably saved hundreds of thousands of lives.
One othe point: to the extent that the Covid responses varied by state, those results are consistent with Covid being the problem. States that had higher levels of vaccination had lower mortality, states that were more open and less shut down had higher death rates.
Two states with high levels of elderly, Maine and Florida, had very different mortality (it was much safer to live in Maine), while Florida is the state that opened up more quickly and vaccinated less of its population.
Macedo and Lee, in their book In Covid’s Wake, find no association between non-pharmaceutical interventions and mortality in the pre-vaccine period. Some state/country comparisons show the inverse - high restrictions and high mortality pre-vaccine, and the studies they cite accounted for all manner of demographic and other factors. (Ebook version allows you to pull up footnotes, which link directly to the cited research).
I've never been able to take such studies seriously, because they ignore that places that are more susceptible to COVID (such as by being more urban) are going to use more strict NPIs. In other words, the reason there's no association is because NPIs are caused by high death rates, and are maintained at a level where they lead to a similar death rate compared to the rest of the country.
Statistically, the public health evidence against NPIs both pre-COVID and post-COVID is very strong. After the pandemic runs its course they seem to make no difference to deaths or case loads and of course they do economic damage. If you compare Sweden, or even Florida (pre vaccine) with California, their outcomes were about the same even though California had much stronger NPIs.
But when you look at specific smaller areas it’s not so clear. The Bay Area did better than California as a whole and it’s hard not to see that as being because NPI compliance was better. New Zealand and Australia had strict travel restrictions and even though their quarantine failed they had very low deaths and case loads even in the end.
I appreciate the effort that you put into proving things from first principles anyway but this is just so bizarre holy hell. As if flat-earthers suddenly decided to heavily comment on a rationalist blog
I don’t comment here much, but the overall level of comment on this thread is far, far better than most discussions of Covid on social media (or, i suspect, most other places). Which isn’t to say it’s ideal, or even good, but much better than the usual.
You mention: "What about the alternative claim - that it wasn’t COVID that caused the extra deaths, but various treatments - ventilators, remdesevir, vaccines? We know it wasn’t a specific single one of those treatments, because the treatments were only used during certain subsets of the pandemic, but the excess mortality was a constant function of COVID cases."
I suspect you're saying total US excess mortality divided by total counted US COVID cases on a per-week basis looks flat.
Can you link to strong data regarding that? I'm really skeptical.
I assume that data is far too noisy to "prove" that statement. And, I think that part of the noise (COVID case count itself), is not just suffering from random noise, but time-dependent noise (I don't know when we'd be over our undercounting, but I'm sure the amount that was over or undercounted was strongly varying over time).
Restated, it'd be really interesting to state and test two hypotheses:
Hypothesis 1 (wildly held by the mainstream?): Mortality rate dropped with Omicron.
Hypothesis 2 (held by skeptics?): Mortality rate increased with ventilator use, dropped with Omicron, and increased with vaccination.
I suspect we can't confirm either.
Instead, what I think we do, is simply bolster priors. For "establishment supporters," the "relative" flatness of a mortality rate will bolster thinking that COVID was deadly, and vaccines were not. For skeptics, the "relative" flatness of a mortality rate will be insufficient given the noise, and the null hypothesis that fatality didn't increase with vaccination won't be confirmed, and that'll bolster their priors.
I'm making the much weaker and less-trouble-finding-statistics claim that about the same number of people died of COVID in 2020 as in 2021, so if all the deaths were from vaccines, the 2020 deaths would be unexplained. I agree that you would need more sophisticated statistics to detect a 10% or something effect on death rates.
I don't see the numbers clearly enough to even really believe that 2020 vs 2021 death rates are constant (maybe, maybe not - I just don't see them). I suspect that case rate reporting may be wildly off over time (example guess: 4x underreporting in one year, 2x in another). If so, then the excess mortality rate year over year has a 50% error. There's a time gap between when people test vs. report death, etc. It's just a morass.
Regardless: if your claim is something akin to -- we can definitively confirm that less than 100% of excess deaths are due solely to vaccines -- that's so weak it begs why to write the article and leaves both sides just doubling down on their priors with any whiffs of data provided?
Pro-establishment: "Indeed, excess deaths were large, and we can definitively rule out that less than 100% of excess deaths were due to vaccine. We suspect that 0% were due to vaccine, but can't really prove that."
Pro-skeptic: "Indeed, excess deaths were large, but at least as many of them were due to vaccine as due to virus. We can prove that excess deaths must be >0% due to vaccine (how could weird heart problems not be correlated with at least some deaths?). We suspect that >50% of were due to vaccine, but can't really prove that."
Your best posts tend to show data/logic that makes one or the other side recheck their priors. Here, I think what's being demonstrated (at least to me, convincingly) there are excess deaths. I think only a small fraction believed there weren't excess deaths (and your data is useful with respect to that, thanks - and thanks, as always, for good faith posts). However, I think a more important fraction think that the fraction attributable to care (vaccine, treatments, lockdowns, etc.) may be comparable or greater than the fraction attributable to the virus itself. Both sides are just doubling down on their priors after that, right?
Thx. That sort of data might have helped Scott's point here, but it's not going to be particularly convincing to the "other side." Regression analyses in observational work off of foundationally limited data are obviously interesting, but face profound limits.
My recollection is that most data pools treat vaccination status oddly. I think such data sets bin an individual as unvaccinated until two weeks after their second dose. So if a person got 2 shots, separated by 2 weeks, any occurring to that individual in the period from their first shot up until 27 days are registered as to an unvaccinated person. Let's assume for the moment that the vaccine itself did cause some deaths - if any of them happened in a time period of up to 27 days after the first shot, it's a really interesting problem.
First, in most analyses (including I suspect Silver's) that person's death is not added to the vaccinated bin (where it belongs). Second, it is added to the unvaccinated bin (where it doesn't belong). The observational stats need to be far richer to sway that side here.
The second link might help address that by separating out status, but I assume still has the two week delay issue (i.e. it has a pair of 13 day problems, rather than a single 27 day problem) that I described above.
Further, both make an implicit assumption that the vaccinated and unvaccinated groups are otherwise demographically equivalent. Perhaps they are, perhaps they aren't- I don't know. Maybe vaccines are always given to sick people, so excess deaths would be terrible there (and dramatically understate vaccine utility).
I understand why a pro-vaccine camp would look at these data and revel in their apparent conclusiveness.
I also understand why the skeptics camp seem them as garbage-in-garbage-out data.
My point - all these somewhat-foundationally-limited data sets aren't moving the needle. It takes something new and possibly sophisticated to move priors that are this deeply established at this point.
> Further, both make an implicit assumption that the vaccinated and unvaccinated groups are otherwise demographically equivalent.
No, Silver makes the point that there weren't big differences in mortality between states based on partisanship until the vaccines became available. If pre-existing differences between states were driving the difference in deaths, it should have shown up earlier.
It still relies on the assumption that vaccine decisions cut cleanly across state/party lines rather than some other confounder-- which they clearly don't; even the least vaccine-friendly states still got up to like 60% uptake as I recall.
> 500,000 - 700,000 excess deaths in each of 2020 and 2021, which adds up to most of the 1.2 million (although I think the full number might include some residual deaths during 2023)
On this parenthetical in footnote 2: "also, suicides didn’t rise during the strictest part of lockdown, and may even have paradoxically fallen. I agree this is mysterious, but it’s only one of the many mysteries of suicides, which often fall during bad weather or social disasters - see my article here. Drug overdoses rose, but by a tiny fraction of the COVID death toll."
This is just speculation, but given that suicide seems like it would happen when people feel disconnected from others. Perhaps the lockdowns, 'social disasters' (guessing something like terrorist attack), or bad weather (hurricane, tornado, etc) are events that, although scary or 'depressing' also make us feel more connected with other people, more part of a community, even temporarily. That may be enough to depress suicide rates to some degree. Again, just wild speculation on my part.
There is some evidence for this - suicide rates are highest in the happiest countries. If misery is common and temporary people just kind of deal. COVID lockdowns gave a "maybe it'll get better if I tough it out" vibe.
All the accounts I've seen from suicidal people is that it's not the misery, it's the isolation and the sense that there is no potential end to the misery. "Only two years of lockdown" indicates participation in a collective effort with a clear end goal.
I think that you're on the right side of this one Scott, but I want to ask, because it always is overlooked in this discussion for some reason, what share of the 1.2 million excess deaths should we attribute to death's caused by the lockdowns? E.g. weren't deaths from overdoses and suicide higher than they otherwise would have been? And has anyone tried to quantify deaths due to cancer or heart disease that was caught too late because of missed appointments?
Edit: shame on me, I should have read to the footnotes!
I'm sure it was not a very large share of the excess deaths, but I think about this every time people say something like " well if there were excess deaths above baseline, it means people died of covid". I think lockdowns definitely caused some excess deaths too!
I would also trust data from Sweden more than from the US. Less panic means more precise reporting.
The US was totally tribal – Trum vs. the rest. I talked to some doctors from the US and somehow mentioned one thing and the doctor rejected that by saying that Trump had supported that. I wondered why is that a problem? “Even a blind chicken happen to get grain.” Or even Trump can sometimes be right by accident. He rejected that notion and maintained that everything should be done to fight him. They even delayed vaccine because of Trump. Terrible tribalism that permeates the whole discourse in the US.
I think that the US failed to get proper statistics and other countries got this right. Yes, a lot of people died from covid but also a lot of people died because of lockdowns. And it is not surprising because that was expected.
What other countries statistics showed that people died from lockdown rather than COVID?
Trump was right to do Operation Warp Speed and get the vaccine out as fast as possible. But he can't take credit for it because so many of his fans hate the vaccine, and he wound up appointing RFK Jr despite denouncing him earlier.
And you immediately didn't answer which statistics show deaths from lockdowns over COVID. I will remind you that you are commenting under an article that very clearly shows COVID deaths tracking excess deaths overall and both lines moving in a way that does not seem to track the severity of pandemic restrictions, which were highest during the first six weeks of response starting in March.
Death rate is back on trend in 2023, and it's easy to see that the total excess deaths over trend are ~1.2 million. 2.84 in 2019, 3.38 in 2020 (+0.54), 3.49 in 2021 (+0.65 over 2019, +0.51 over 2023), 3.19 in 2022 (+0.21 over 2023), and 2.98 in 2023.
I think you're right because I attended the funeral of my wife's great grandmother in the municipal cemetery of Sofia, Bulgaria and I saw the new fields the cemetery had to expand into. Hundreds of new graves with temporary wooden crosses that all said 2020.
Very insightful information. The # of excess deaths during this period certainly show a real and honest trend that makes me less skeptical of the large # of deaths ascribed to COVID-19.
Where I still have a lingering bad taste in my mouth is the general COVID response hand the blame being thrust on the virus itself, and not the growing number of comorbidities that we as a global society have let metastasize in the general public that greatly exasperated the devastating effect that the virus had on public health.
When we speak about forest fires we have grown intelligent enough to recognize that it is not the nature of the initial spark that is the most determinant factor in how widespread the damage can be, but rather the long term accumulation of underlying pre-conditions that led up to the moment of ignition (i.e. the growing abundance of dry, dead tinder from poor forest maintenance; the expansion of high voltage transmission lines in these area; and increasingly dry/hot weather from localized climate change).
When we speak about COVID-19 I feel we really never addressed these underlying health issues. There was no big response to getting people in better shape to prevent the growing number of cases and deaths, rather we had everyone drastically slow down their lives to "bend the curve" which if anything had a counter-beneficial effect in actually making these comorbidities worse as people become less active.
I believe this is what has caused such a vague understanding of the total death count. Was it 1.2 million healthy people who all got hit with the worse global virus in a century? Or 1.2 million unhealthy people who caught a slightly more infectious and deadly virus? Or somewhere in the middle. And if its the third option, how do we move forward in preventing such devastating results from the inevitable next pandemic?
Most of the 1.2 million were unhealthy because they were quite old, and old age is a “comorbidity” that not much can be done about. Yes, our health care system’s priorities are wildly skewed when it comes to promoting overall health and well being. Yes, there are a lot of unhealthy people. But Covid mortality is so closely tied to age that it doesn’t make much sense to try to sort deaths into “healthy” vs “unhealthy” buckets
It really was caused by the virus, and the main way we're addressing chronic major health issues now isn't due to COVID at all but instead semaglutides coincidentally becoming available a few years after. After the "dry tinder" is exhausted in a forest fire you shouldn't see a fire there again, but we aren't seeing excess mortality undershoot in subsequent years the way it overshot previously.
Maybe we should try clearing out some of the old people so that we don’t have so many vulnerable next time a big disease comes about.
/s
I also sometimes people say the medical establishment has been ignoring obesity - I sometimes wonder if these people have ever been to a doctor or read anything from the medical establishment, because basically all of it tells people to lose weight.
I wonder if we can calibrate the worldwide data with the data from the US. 1.2 million is about 0.3% of the American population. If a similar percentage of people died around the world from COVID, that would be 28 million worldwide, assuming America dealt with the pandemic no better or worse than average.
Given the known age profile of the disease and of the US, you'd expect US deaths to be higher than average. And we don't have good data from countries with poor health-care systems where many people will have just died without getting any sort of COVID diagnosis.
But if we look at Europe (EU+UK+Switzerland+Norway+Iceland) then 0.3% is pretty much normal here (some other European countries report low and untrustworthy numbers, e.g. Belarus reports 0.075%)
Rich Asian countries (ie those with healthcare systems capable of producing quality statistics) tended to have more effective lockdowns and lower death rates as a result - some had huge surges when they finally unlocked, others managed to vaccinate widely and then unlock carefully (New Zealand, famously, had a 0.05% death rate, which is pretty much a best-case scenario).
This issue is a good illustration of how people hate hypocrisy 1000x more than dishonesty. The official public health community made some bad calls to say what they thought would save the most lives in the long run rather than expressing the full uncertainty and changing weight of evidence as more facts rolled in, as well documented here and in other places. And yet they probably were still among the top places to get reliable information! People recoiled from a source that claims to tell the truth and tries very hard to do so but got its priorities twisted and leapt into an alternative information world of grifters and liars who will never admit anything.
I largely agree with this comment and anyone who reads my comment as doing otherwise is wrong.
But this is also an excellent argument for, uh, just telling the truth. Lying didn't work, and permanently crippled our ability to react to crises.
In the years leading up to COVID arguments about "voting against their interests" and people being too stupid to know what's good for them gained full penetration in elite circles. As a result, we leapt directly to simplifying the narrative and keeping back information. And everyone noticed.
I think you have it backwards - the medical establishment used to lie all the time, and people trusted them as a result. But by 2020, they had adopted an ideology of admitting their failures and being more transparent than they used to be, and now people hate them.
> People recoiled from a source that claims to tell the truth and tries very hard to do so but got its priorities twisted and leapt into an alternative information world of grifters and liars who will never admit anything.
False dichotomy. You can fall back to your priors, and first-hand evidence, and heavily discount everything else. Which for most people would tell you something like "COVID isn't very dangerous, unless you're old or with comorbidities, but vaccines are pretty safe in general, if not always highly efficacious. Also the authorities are going to keep overreacting and keep claiming they're having a significant positive impact regardless of the truth, for political incentive reasons."
It worked pretty well for me. Including predicting (5-6 months in advance of the shot being available to people in my age group) that the government (this was in Ontario Canada) would impose vaccine passports. I made sure to save all the paperwork because at the time I didn't have a health card, and I figured I'd later need to prove I was vaccinated to access public places or travel. Which is exactly what happened. (Okay, I was a little too paranoid: I expected their computer system that recorded who got which shots to flat out fail as IT systems often do, so I literally had video recordings of getting vaccinated as a fallback proof. That turned out to be unnecessary: Recovering my vaccination records to get the fancy QR certificate only required multiple hours on the phone).
Other people had bad priors (mostly starting from "lots of random internet people agreeing with each other is a useful signal"). That sucks but the mistake wasn't in reducing trust in authorities.
Ok, more precisely my point is that people reacted to the authorities’ behavior by going from 95% trust to 5% trust (or whatever), leaving much less reputable sources as rivals or winners for many people. Sounds like this doesn’t describe you. Me either.
I'm reminded of Bounded Distrust (https://www.astralcodexten.com/p/bounded-distrust) though I'm struggling to phrase the exact connection, and rereading Scott's Lying Sequence will not be sufficiently edifying.
Anyways. If you want to avoid ever suffering from a pandemic, you listen to the people that predicted 10 of the last 1 pandemics. If you're a normal person that doesn't want to spend your entire life in a panic or building a survivalist bunker, you listen to the people that only accepted the reality of the last pandemic three months after it was too late. The side effect of the latter is the substantial failure mode that collapses consensus reality ("very rarely lies" can be more dangerous than "constantly lies") and leads people into epistemic nihilism.
I personally know zero people who died from Covid and one relative of mine who died from the secondary effects of the lockdown (delayed cancer treatment in April-May 2020 => plausibly an earlier death). However, I do agree that 1.2M is a completely plausible figure. I know several people who were *close* to dying from the virus and many of them are still alive 5 years later, so it clearly wasn't "just the flu" for the elderly.
However, an arguably much more interesting question is:
- Was the ACX/LW/Tech Twitter crowd right about taking additional precautions in 2020, assuming they were under the age of 65, had a BMI under 35, and had no major preexisting conditions?
- Was the same crowd right about taking additional precautions after getting their vaccines in spring 2021, such as the insistence on testing during various LW/Rationality events that persisted until summer 2023?
I've mentioned this in a few Open Threads over the years and I still feel like the story hasn't been set straight.
I agree that the crowd I was a part in took too many precautions in early 2020. I still remember all the copper tape on doorknobs!
In https://slatestarcodex.com/2020/03/02/coronavirus-links-speculation-open-thread/, I included age-specific mortality tables and hospitalization tables. In retrospect, the mortality numbers were inflated by about 10x and the hospitalization numbers by >100x, because only severe cases had been detected and so the government was underestimating the number of cases (and, since they detected most deaths, overestimating death conditional on case). The mortality rates were low enough that even the 10x inflated ones weren't so bad, and I said that "the good news is that it’s pretty unlikely to kill young people". But the inflated hospitalization rates looked awful - 15% - 20% of young people who got COVID were hospitalized! I was suspicious at the time, saying:
> "This is a weird pattern – why are so many young people getting hospitalized if almost none of them die? ... Are these an overestimate? Maybe most cases never come to the government’s attention? There’s some evidence for this ... So hopefully the 20% hospitalization rate will prove to be a worst case scenario, and the real number will be less."
...but absent 100% proof that this was the explanation, I elected to stay inside rather than get a disease that had some chance of a 20% case hospitalization rate for young people. I'm not sure how I could have done better without being able to be confident in the rate at which the government was underestimating cases, something that's beyond my epidemiological knowledge (and apparently everyone else's). Also, later in the pandemic lots of people made the opposite mistake, assumed that the government was still underestimating cases when in fact by that time it had gotten pretty good, and confidently predicted that COVID would end near-immediately. So a heuristic of just always assuming detected cases are underestimates wouldn't have worked.
> Was the ACX/LW/Tech Twitter crowd right about taking additional precautions in 2020, assuming they were under the age of 65, had a BMI under 35, and had no major preexisting conditions?
For me, the main motivation to avoid covid was to avoid accidentally bringing the virus to my older relatives. If I wanted to stay in contact with them, I had to protect myself, too.
Some of the things I did were probably excessive, but that is hindsight; and frankly I found most of the precautions relatively easy. For example, at the beginning I bought a bottle of disinfectant, and whenever I bought something in a shop, I wiped most things with a paper towel soaked in the disinfectant. Probably useless, but it took me one minute of time, so no problem.
> the insistence on testing during various LW/Rationality events that persisted until summer 2023?
In Vienna, during covid we had LW meetups under the open sky. No one checked anyone's tests.
The part about covid denialism that really irritates me is that it implies a belief that actually huge numbers of people in this country can't do their jobs at all.
Like: epidemiology and statistics has much much more sophisticated methods for attributing causality than just comparing graphs and looking at if they spike together. They know what they're doing. Likewise for developing vaccines, confirming they worked, and tracking the data on proving they're working.
And every single doctor and medical worker went through living hell for most of covid because of the spike in cases of people with clearly life-threatening illnesses of the type they are deeply familiar with. There are no doubts whatsoever if you worked in a hospital or nursing home.
And my favorite example: I lived on the island of Kauai in Hawaii for three months during Covid, which managed to almost completely prevent covid from showing up by implemented a strict quarantine (you had to quarantine for two weeks and the national guard would show up and make sure you were there). And people on the island would simply not believe covid existed because they hadn't seen it! Because their government had done a *good job*.
Outside of the conspiracy-theorist's low-trust life there is a huge world of mostly competent and well-intentioned people doing their best-obviously! But all they notice are the rare examples of incompetence or deception which they continue to harp on for years. Most of which make a lot of sense in context if you bother to read about what was going on.
Like, get over it. Reality is sitting there existing *even if* you are pissed about the media or cdc or whatever trying to manipulate you a couple times. You can be mad about that without turning delusional just to prove your point. We get it, you're mad about being powerless to prevent the bad stuff from happening. But having a tantrum and refusing to listen to everyone is not fixing that at all, it's just making it worse. If you want society to handle disasters better then maybe go contribute to fixing its general incompetence in some material way, like helping make society more high-trust, or contributing to the competence of your local government or something.
I disagree that those working with data to address pressing policy issues always know what they are doing. People are familiar with the replication crisis in psychology. I have first hand experience of research in physics and second hand experience of research in economics through my wife. Even if outright fraud is uncommon, there is plenty of work that is sloppy, rests on shaky assumptions, or is presented in a misleading way. I suspect epidemiology cannot be too different.
Psychology has been comfortable with BS for a long time. Freud & Jung were both pseudoscientists in the Popperian sense, but big deals in psychiatry. Robyn Dawes wrote "House of Cards" about the irrationality of his colleagues in psychology. Paul Meehl was writing about the things that gave rise to the replication crisis back in the 60s. For all the problems with physical medicine, they did at least switch from harming to helping people once germ theory, vaccines & antibiotics became prevalent. Dead bodies are an obvious outcome which psychology doesn't really have compared to physical medicine.
Yeah, there’s even a bit in one of Feynman’s autobiographical books about behaviorist experiments on mice and all the things you actually have to do to control for outside influence. To this day no one does any of those things. I remember this every time I hear about a social psychology finding
“ always know what they are doing” is very different from “ mostly competent and well-intentioned people doing their best”. People demand the former, and have trouble accepting that the latter is better than populism.
Highlighting the incompetence of specific people is contributing to fixing the problem. For example, you seem to put the epidemiologists and the vaccine researchers on equal footing. AFAICT the latter are much more competent in that the vaccines worked and the clinical trials, while slightly overstating efficacy, got numbers in the right ballpark.
In contrast epidemiologists continuously overestimated the mortality rate despite having access to a good source of true IFR data early on (Diamond Princess), insisted on using SIR models in all the curve-prediction efforts they made despite it fitting extremely poorly.
In the UK, where lockdowns happened semi-regularly based on epidemiologist predictions, when the 4th one was called off due to a backbencher revolt (largely driven by an open admission by one modeling team head openly admitting they don't bother modelling and presenting probable scenarios that don't require the government to intervene: https://www.spectator.co.uk/article/my-twitter-conversation-with-the-chairman-of-the-sage-covid-modelling-committee/ ), the predicted deaths failed to occur.
And the other famous UK model, the non-SIR model from Imperial College London, whose predictions are what made the UK decide to act much more aggressively, turned out to both reliably and dramatically overestimate deaths ( https://thedailyeconomy.org/article/the-failure-of-imperial-college-modeling-is-far-worse-than-we-knew/ ), trying to justify that the interventions suggested by their models saved lives by using their own model again, and when the source code became available it became obvious the entire thing is heavily non deterministic (even for the same parameters, including the random seed) and riddled with bugs and ad-hoc decisions.
Just because two groups of people are superficially similar (PhD degrees, peer-reviewed articles, etc.) doesn't actually get at the substance of who's actually doing science and who's cargo-culting it. Being able to sniff out the latter and stop rewarding them should be part of the conversation if we want anything to improve long term.
I think it’s wrong to say epidemiology is just cargo cult statistics. It’s less reliable than vaccine research, but only in the same way that cosmology is less reliable than planetary science, and nutrition research is less reliable than diabetes research. Some things are harder to study than others (and particularly harder to do controlled experiments with). That means we should put less weight on specific predictions from those fields - but it doesn’t mean we should cast out the field, because they’re still doing better than anyone else at addressing those particular questions.
> still doing better than anyone else at addressing those particular questions
That's begging the question.
In my example above, JP Morgan (not epidemiologists) were able to notice the obvious fact that the Omicron variant was less lethal than Delta, and that this made a huge difference in number of deaths expected in the UK when it started spreading there and what kinds of measures were appropriate to take.
When confronted with this fact, the epidemiologist running the UK advisory committee said such low-lethality scenarios, even though likely, were not modeled because that's not what the decisionmakers were interested in.
I mean, that's almost certainly bullshit. Epidemiology is a sophisticated and evidence-based science. A couple examples of malfeasance doesn't invalidate the field as a whole: it invalidates those people, sure. But more generously, they were in complicated uncharted territory and almost certainly trying to do their best in spite of all the politics around them. But the way people talk it's like, one epidemiologist is wrong once and now you think it's possible their covid death numbers are off by two orders of magnitude. Like.. Lol? No way.
The heads of two of the biggest epidemiological forecasting groups in the UK whose work drove actual policy. And in one case not just wrong, but openly saying to anyone who can read that they weren't even trying to make accurate predictions because scenarios that don't lead to lots of deaths don't have policy implications. That's the guy who was the head of SAGE, the advisory board for the government on COVID policy in the UK (until this incident and growing fatigue caused a backbencher revolt and Boris Johnson ignored the advice to go into a 4th lockdown). I'm not nutpicking here, these are pretty central people, elevated as the experts among experts.
> and now you think it's possible their covid death numbers are off by two orders of magnitude. Like.. Lol? No way.
(Assuming you're talking about retrospective numbers, not prospective numbers): That some people manage to be even more wrong is not a ringing endorsement of the experts. You don't have to like how people react to competence crises by falling back to even worse methods such as believing the whole thing's made up. But when you decide that therefore
> huge numbers of people in this country can't do their jobs at all
is false, that's a transparent appeal to consequences.
My point is that the data people use to formulate their dismissal of 'the entire expert establishment' is massively underpowered for that conclusion. They see a few failures, don't measure the successes at all, and lazily assume the whole thing is bunk.
I think the whole issue is that epidemiology (especially for COVID) is not about scientific observation and experimentation (what the layman would view as science). It is using models and statistical methods to predict the future which is prone to motivated reasoning.
Another poster claimed that SAGE couldn't show that their models prevented deaths. That may have been too hard a level. But I have a more fundamental critique. Has SAGE release their original code and data? That would give me more confidence that SAGE just erred in their predictions and it was an honest mistake.
I really don't know and my quick search didn't turn up anything.
I forgot to respond to this, by the way, but I wanted to mention that you seem to know nothing about epidemiology. It is very very much a science that is based on observation, experimentation, and sophisticated statistics. Skim... any paper... if you want examples. (Not an epidemiologist myself, but my dad and ex both were, so I've interacted with a lot of their work, enough to believe it's very serious.) Any science, of course, is going to falter in a situation that has never been seen before. But you seem to be dismissing the field out of hand, and I cannot imagine why.
I remember that incident. In front of a parliamentary committee one of the epidemiologists defended their models saying, <paraphrased> "Well, we didn't intend them to dictate public policy." And one of the MPs responded, "Well, why are we paying you then?"
The SIR models absolutely sucked at predicting COVID deaths. A few epidemiologists raised this issue and proposed solutions, but SIR models are the bread and butter of epidemiology (when it comes to grants). No one really wanted to dump their precious models.
If the powerful institutions of the media, CDC etc. regularly try to manipulate people, increasing trust them is bad, and you're really complaining about "conspiracy theorists" seeing them for what they are.
Trust is a difficult thing to calibrate, and unless you're confidently savvy to distinguish between good and glorious harvest announcements (https://www.astralcodexten.com/p/bounded-distrust), not trusting them at all *probably* is your "best" (a word which here means prevents the failure mode of inappropriate trust in accordance with your abilities) options.
The problem is that instead of retreating to well-tested traditions, or even some trustless epistemic nihilism, people in that situation tend to resort to whatever talking head is saying acceptable things regardless of correctness. Everyone likes a talking head that tells them what they want to hear, as was amply expressed during the pandemic.
Using your post as a starting point and extrapolating. Please correct me if I have taken the extrapolation too far. Taking your claim to the next level. Every claim by an authority should be evaluated independently and the people should trust if they believe in it 90%.
What I think most people do and what I do much of the time. I think they take a more Bayesian approach to making decisions (ie previous observations do matter) and whether to trust is binary (trust or don't trust).
So, people were trusting the authorities at the beginning of COVID. But they then started loosing trust based on the actions of those authorities. I am not going to rehash it all (others have done it in this comment section). I am referring to the mask issue and also the political issue where the authorities at least gave tacit approval to some protests.
The person assigns a 90% trust score to the specific claim based on the facts they have. But based on priors, the person assigns a 50% trust score to the authority making the claim. They discount that 90% by 50% so the score is now 45% which is now in the not trust territory. They assign the binary score of not trust and conduct their life with that belief.
No, I mean acting in trustworthy ways and helping other people feel like the world is generally high trust. That doesn't mean you should trust those media companies in particular.
"implies a belief that actually huge numbers of people in this country can't do their jobs at all."
I think it more implies a belief that huge numbers of people don't actually have the jobs their job title would naively suggest.
If someone's job title is "academic research scientist", for example, we might naively expect that their job is to learn new things about some aspect of science, things that are real and true and important, and then to report them accurately. But that's not their job, because that's not what we pay them for. Their *job*, is to get research grants.
And while research grants are correlated with doing good science, the correlation is often quite weak. Some research subjects are more likely to get grants than others, even if they are less valuable to the world at large. Some *results* are more likely to lead to future grants than others, even if they are false and the others are true, And how the result is presented, matters too.
If you're a research scientist and some journalist calls you asking "Is This Thing in your area of expertise a Major Crisis?", then if you say "no" or "who knows?", it doesn't bleed so it doesn't lead so you go back to your desk full of long-shot grant proposals and wonder if you're going to be able to pay your grad students. If you say *yes*, you get your face on TV and people will recognize you as the Expert on That Thing. And That Thing will be elevated in priority, with more research grants going into the field. You've got grad students to feed. What is your real job here?
And maybe you have the intellectual integrity to not outright lie about it, but now you're in a situation where your salary kind of depends on not understanding that the profitable answer is a lie, and motivated reasoning is a powerful thing.
People in nursing homes and with chronic debilitating conditions tend to have small social circles and low visibility. My memory of the AIDS epidemic of the 1980's is of friends dying. What most people will remember of Covid will be lockdowns, masks, and social distancing. (excluding critical care docs and nurses).
There is maybe no greater argument that this (excellent) article is an exercise in futility than the existence of the article itself. It’s a reflection of over five years of mass delusion, despite consistent, excellent analysis and advisement.
Instead of reading the article I'm going to assume this means putting a bunch epidemiologists and public health experts into a dome for a deathmatch, and whoever survives get to set WHO policy.
That graph of actual country's performance plotted against the efficient frontier people achieved in the game is really interesting, do you remember where you got it from?
I think it would be interesting to see more countries, with an extra few years of data, to get a sense of what the space of possible outcomes looked like.
My answer is "what New Zealand did", but that (shut down all international and interstate travel entirely, yes that means that NZ citizens abroad couldn't go home for over a year) wasn't even close to the Overton Window.
I think the evidence is that the NPIs were not that effective, and the vaccine was extremely effective. So my recommendation with hindsight would be fewer to no NPIs but much stricter vaccine mandates and requirements. I would also have tried to increase production of Paxlovid and increase access to early mABs.
It’s very hard to measure the effectiveness of distancing behaviors - most of the measurements were of distancing regulations. (I also think we shouldn’t have bike helmet laws even though bike helmets are probably a good idea.)
"People still don't think it happened" was my first thought reading yesterday's post.
I think all this, from ignoring covid deaths to denying covid deaths to claiming charities are bad, is better explained with the Biggest Logical Error Of Our Age: Things that prove us wrong are never true.
You might think, "That's not even a logical error." But it is. The process works like this:
1) You have priors for how the world works.
2) You inculcate a philosophy of skeptical rationalisim - there is no higher authority that defines truth. It's something you discover yourself. And you follow the "rationalist, scientific" method where you don't update without good evidence.
3) No evidence is good evidence. Experts are corrupt, information sources have agendas, anecdote can always provide a counter-example to any statistic.
4) You are never wrong, and you know this because you did a stringent analysis of the evidence that you were wrong and it's bad.
I want to be clear that I'm not saying "other people sure do this a lot." I think I am probably the world's biggest perpetrator of this fraud. I don't know how to clear it without giving up on rational skepticism, and I don't know how you determine truth without using rational skepticism. This community in particular has lots of tools for this, which is why I'm so interested in what you write. But I still think giving people these tools mostly just gives them more ways to rationalize their priors.
But if in 2019 you thought this whole thing was an exaggeration, would you rather believe that you were wrong and your wrongness killed a million people, or that you were right all along and the number is made up.
Postscript: Deaths from Covid were extremely concentrated, especially in big cities and especially on the <corrected, said west> east coast. A huge percentage of those deaths happened in New York City alone. If you're a rural republican who doesn't want to believe it happened anyway, it's very likely the pandemic never even really reached your sparse town...
Do we know why there was a such a mess in New York City? Yes, they were the first place in the US to get COVID, but shouldn't other places have had the same problem once the pandemic reached them?
I'd have to look at it more, and I might be doing exactly the thing I complain about above...
But if I recall correctly, our early preparation was really bad, and then after New York the other major cities went "oh crap it's going to be really bad" and took precautions like increasing hospital capacity and expanding staff.
(Instead of writing ten thousand qualifiers I should do the rationalist thing and just say "I only have 55% confidence in this because I haven't read up on it" lol)
One thing a lot of people have pointed to is infected people being sent into nursing homes to infect more elderly people, but I don't actually know how what impact such policies had.
First place hit by (probably?) the worst strain and rapid dense spread during the period when most "mainstream" types were still skeptical, large elderly population, and the stupid nursing home policy.
My personal conspiracy theory is that the 'rona was circulating in the US much more rapidly or earlier than people were paying attention, but I have no good explanation for why it wouldn't show up in the excess deaths sooner because of that (even if they're not labeled covid deaths). It's mostly rooted in my personal experience. *Maybe* I happened to get the worse flu of my life and lasting lung damage early March 2020 when theoretically it wasn't in my state yet, but it's a heck of a coincidence. I hadn't been to China so my negative flu test was treated as evidence it was probably flu anyways.
I wonder how much is due to intuitive distancing behavior. In march 2020, everyone still treated staying home from work when sick as optional. But once you knew in your gut that there was a pandemic, you were going to be a bit more cautious. Everyone adding that bit of caution would slow the spread and make the waves crash more gently.
As an ICU doctor who personally has seen over 200 people die directly from Covid (guess the famous CT scans were just fake!) and saw firsthand how hospitals got overwhelmed, the comments here are making me more angry than anything I think I’ve ever read.
I feel deeply betrayed and disappointed. Humanity together was at war with a common enemy, and people found any level of discomfort or inconvenience too much to bear. Easier to practice denial instead, and leave those fighting on the front lines, like you, to carry the weight alone.
I always imagined that when the rubber hit the road, people would take a stand to fight together, but it turns out most people are self dealing spineless cowards.
Thank you for fighting on the front lines, and know that some of us do understand and remember what you did for us, and were doing everything we could to fight along side you.
I know a nurse who got completely burned out, not in 2020 but in 2021 when people who refused to vaccinate flooded the same hospitals. Again. She had a sense of purpose in 2020, but when the unvaccinated waves came she just couldn't do it anymore.
So yeah, very little patience for this crap. That's why I stay away from this whole disaster of a debate, it just makes me want to punch people.
I don't think anyone here is arguing that "no one died from COVID". The debate is over "I think fewer than 1.2 million people died from COVID". Why would that make you angry?
Some people are saying the whole thing was vastly overblown. For example, "after the beaches and playgrounds were reopened, it seemed to have been more like a bad flu season"
This has been a fun experiment in different factions of Scott's audience being radically uncharitable to each other and substituting the comments actually here with the most ridiculous caricature they can imagine.
I kind of hope that was his plan all along and this will feed into a future banger describing some interesting social phenomenon.
My ex was an MICU NP in NYC throughout the pandemic and, reading through these comments, I keep thinking she would be absolutely livid to see what some people are saying here.
The claim is not that physicians were lazy and/or dishonest when they classified deaths as "COVID deaths." The claim is that they were following protocol. On March 24, 2020, the CDC, through the National Vital Statistics System, changed the way that the underlying cause of death was to be determined for COVID, and only for COVID.
"Will COVID-19 be the underlying cause [stated on the death certificate]?
The underlying cause depends upon what and where conditions are reported on the death certificate. However, the rules for coding and selection of the underlying cause of death [on death certificates] are expected to result in COVID19 being the underlying cause more often than not."
Further:
"Should “COVID-19” be reported on the death certificate only with a confirmed test?
COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death. Certifiers should include as much detail as possible based on their knowledge of the case, medical records, laboratory testing, etc. If the decedent had other chronic conditions such as COPD or asthma that may have also contributed, these conditions can be reported in Part II."
No other disease or medical condition has ever been treated this way. The result is that any time someone either tested positive for COVID or had respiratory symptoms, they were required to be classified as a "COVID death," regardless of whatever else may have been going on with them.
No laziness or dishonesty was required on the part of the certifying physician.
Exactly this. No lab test was necessary to say that the person dying had covid. It would be interesting to get statistics by reviewing randomly selected records to see how many were actually announced to be dying from (or with) covid.
“Assume” in panic times should be taken with a big grain of salt.
But one other thing is that during peak times hospitals became a breeding place for covid infection. If you came in with, let's say UTI, you would get covid. During peak times no one could ensure good separation of patients, isolation etc. My mother went to the hospital with UTI and came back with covid, then had to go back to the hospital again to treat covid. Scott assumes that the background rate of covid infection for those in the hospital is the same but it is clearly not possible.
Both this comment and the parent comment miss the main point of the post. Lots (millions) of extra people died over the normal rate (“excess mortality”). Why, if not from COVID?
Yes, from covid but also from lockdowns or whatever it was.
Every single data point shows this. Most countries with lockdowns have clearly diverging lines.
Opioid deaths increased greatly during covid in the US. This graph provided by Scott seems outlier. As it comes from official statistics, I have to conclude that the most likely explanation (Ocam's razor) is that the statistics are poisoned. People who died from other causes were often counted as covid victims.
I don't know the reason for those excess deaths. It is the thing that needs to be explored. But USA is very much influenced by tribal politics and both parties don't want to investigate that it becomes almost impossible.
This paper from BC [1] suggests an excess mortality of about 6.5 per 10,000 of population in 2020 among people who tested negative for COVID (and didn't subsequently test positive any time in 2020). I would expect this to be an overestimate, given that people who are tested for COVID are likely to be otherwise unwell (early in the pandemic, you _needed_, IIRC, to have some other risk factor, or be hospitalized, to get a test).
Extrapolating that to the United States, that gives 195,000 excess deaths from non-COVID causes in 2020. (BC had very few COVID cases in 2020, but I see no reason to think that non-COVID excess deaths should be proportional to COVID cases, unless some of those non-COVID excess deaths are caused by true COVID cases in some way (people dying of heartbreak after losing a partner, for example?), in which case we should count them anyway).
So excess deaths in COVID-negative people don't explain most of the excess deaths observed in the US in 2020.
Deaths were going up even prior to any lockdowns. The lockdowns were in response to deaths. Early on we had officials downplaying things and encouraging people to go to NYC's Chinatown.
An outlier compared to what other statistics that you find more credible?
Looking how Sweden didn't have a typical peak and trough curve but more level, I find it more credible that peaks were created by lockdowns. They caused the very thing they were supposed to prevent.
I'd certainly be interested in finding that answer. But one is under no obligation to simply accept the explanation proffered by institutions that were deliberately juking the stats. Particularly when said institutions have gone out of their way to smear and stigmatize any who would question their credibility.
All-cause mortality numbers are, one assumes, still pretty hard to fake. In the U.S., anyway. China doesn't even know what its actual population is to within a few hundred million. But that, in and of itself, doesn't tell one anything about what caused those deaths.
Maybe people were obsessive and at the same time completely negligent. Testing made no difference in controlling covid spread at all. So much money was wasted.
My own mother in 2022 went to the hospital for UTI and got covid. They tested her at admittance but not at the discharge. No one cared. She came back home, tested positive and soon had to go back to the hospital again due to breathing difficulties. (Imagine people were banned from twitter for saying that it was possible for fully vaccinated to get covid.)
My father got infected from my mother, died a week later at home. No doctor even looked at his body, wrote a death certificate that doesn't mention covid. It was good because the crazy pastor would not have allowed the open casket funeral otherwise. See, all those people were unreasonable. Had strange beliefs, didn't do their jobs properly, and at the same time moralized over others.
That's why I am skeptical about these statistics because they don't align with the total picture. I don't want to end like Scott who trusted the statistics that 20% of young people get hospitalized (https://www.astralcodexten.com/p/the-evidence-that-a-million-americans/comment/119233511). He could not explain how it is possible if all other data disagreed. Why don't be brave and say – the data is most likely corrupt and should not be trusted?
See, I am not covid denier or vaccine denier. In fact, I predicted the total number of deaths quite precisely and yet I saw that children and young people are very little affected. I saw it better than Scott.
> Testing made no difference in controlling covid spread at all
What makes you say this? In the case of other illnesses, contact-tracing can be very effective. I suppose you could argue that we should have been contract-tracing and collecting statistics by clinically diagnosed cases (that is, people who a doctor diagnoses based on their symptoms rather than a test), but that seems like a weird argument (tests are cheaper than doctors' time).
From first principles, it seems like it should work: isolating people who might have been infected can prevent them from spreading the illness further. It's hard to get statistics on exactly how effective it was on a population level, but that doesn't mean we should assume it doesn't work; rather, we should try reasoning about it in other ways, such as reasoning from first principles. It's true that it might have been a waste of money, but it might also not have been a waste of money. We can't really be certain either way, so it makes sense to have funded it even though it wasn't clear how effective it should be.
Contact tracing works when you've only got a small number of people spreading the disease. If we had locked down the borders to prevent entry, we could have relied on it. But instead we had enough infections for it to become endemic.
I’m really sorry to hear your story. My spouse also clearly caught COVID while working in the hospital and using precautions such as surgical masks and hand hygiene, so believe me I understand that these precautions were far from perfect. But I don’t really think it is clear they did not have some degree of efficacy. The pandemic really highlighted the methodological limitations in most studies regarding these interventions.
I mentioned our frequent testing to point out that we did not observe frequent nosocomial spread, either at my institution or in reports from others. We did not have many patients who were discharged then readmitted with COVID either. It happened but was not frequent. Nosocomial spread was something the medical community was highly concerned about. It was a major driver of the initial push for early intubation that turned out to be the incorrect approach. Estimates vary, but for example this study from the UK, where shared hospital rooms are more common than the US (https://www.nature.com/articles/s41586-023-06634-z) suggests a 1-2% nosocomial spread rate.
As for the efficacy of test and trace, studies vary on their estimates and certainly have a lot of systematic bias due to their observational nature, but most reviews do suggest some benefit, such as this one (https://link.springer.com/article/10.1007/s10654-023-00963-z). I am not certain of the precise magnitude of efficacy, but I don’t think anyone can clearly say it doesn’t have some degree of impact.
These interventions are just really hard to assess in the context of a changing pandemic with changing population immunity, while multiple different interventions are being used in different ways in different locales. Sure you can do some clever “natural experiments” leveraging these differences but there is always so much residual systematic bias it is hard to really know. Nothing beats good prospective experimental methodology.
I would really love to see more robust research on this. We should have adaptive platform randomized trials, leveraging the amazing power of Bayes theorem to test multiple NPIs in combination, geared up and ready to be implemented when H5N1 or the next pandemic hits. Unfortunately, there is no political will for this in the US currently but the EU is working on it which is encouraging.
By the time vaccine became available most healthcare workers had already got covid infection. It means that any attempts to limit the spread of covid in hospitals were useless.
Ok, maybe not completely useless but no one showed a graph in advance how the rate of covid infection among hospital workers were expected to rise reaching 100% in a given period. Basically it means we have no idea how effective those measures were. You are trying to quote studies but why should I trust them?
Similarly I saw a lot of studies trying to measure how effective covid vaccines are in limiting the spread. Until at some point seroprevalence studies showed that nearly 100% of population had been infected with covid. Apparently all this modelling was completely wrong and didn't have any effect whatsoever on attempts to change the course of pandemic. I was better following the statement that original covid vaccine trials did not attempt to measure its effectiveness in stopping transmission therefore this is not guaranteed or in other words, no evidence that it stops the spread. (Ironic that Scott wrote an article that experts misuse “no evidence”=“do not work”, to assert that masks actually help but at the end it turned out that masks actually do not work).
I don't want to look into more research. I think that most people trying to do that, including Scott, are clearly over their heads. The reason is that it requires a lot of time and man hours to review and evaluate those studies. Only big groups can do that, like Cochrane or NICE but not a single person.
Scott tried several times: he reviewed masks and got it wrong. He defended lockdowns and got it wrong. He even believed that 20% of young people get hospitalized despite his strong analytical skills. What hope do we have?
Instead, I take more humble position. I assemble a puzzle from trusted evidence that has high confidence. For example, death rates are trustable. whereas average life years lost by covid death is not. “A person was expected to live 10 more years” is a modelling, not something that is measured. Simple fact that the average age of death from covid was around 84 years and average life expectancy is 83 years makes it suspect. It could be if there was great variance in age but we know that risk of death was exponentially increasing by age. It means that 74 year old person was about 10 times less likely to die from covid than 84 year old. It becomes mathematical impossibility to have 10 years lost on average. Maybe you can make a model where it seems possible but why should I trust such a model if so many other models turned out to be wrong.
I would wait for Cochrane or NICE to make a better estimate but they are probably too busy with other things. But a lot of people won't wait and will believe any model they are being presented. I am ok with keeping uncertainty about this issue.
You said, "The result is that any time someone either tested positive for COVID or had respiratory symptoms, they were required to be classified as a 'COVID death'."
What you quoted, however, says no such thing (or anything remotely similar). It says, "COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death." (sensible guidance) Is the reason for your confusion that you are misreading the first part? That just says that a lot of people are dying of COVID and we expect that a lot of the death certificates will reflect that. It is not an instruction. It is an observation.
Agreed. Since the parent comment claimed that “No other disease or medical condition has ever been treated this way,” I checked the CDC handbook on certification of death. It contains no requirement that any diagnosis be supported by a laboratory test.
You're missing the point, and misunderstanding the way death certificates work.
"Cause of death" is not a unitary concept, but instead contemplates both parallel causation and chains of causation. It also recognizes that not all causes are equal in either their contribution to a particular death or their clinical and broader significance.
On death certificates, "cause of death" is divided into two parts, Part I and Part II. Part I is used for those conditions/symptoms that most immediately and significantly caused/contributed to death. Part II is used for conditions/symptoms that were present, and while they probably contributed to death, they weren't deciding factors. So, for instance, a 90-year old patient with stage IV lung cancer who caught a cold and died might have just "respiratory arrest" and "lung cancer" listed on their death certificate, but another physician might also include "UTRI" as well. But we all recognize that it wasn't really the cold that killed them, even though the cold did indeed contribute to their death. It was really the lung cancer that did it. So respiratory arrest and cancer go in Part I, but the cold, if it's going to be included at all, goes in Part II.
So first off, "caused or is assumed to have caused or contributed to death" is an incredibly, incredibly broad net and normally subject to some kind of discretion. If a condition is present, it's basically impossible to say it didn't contribute at least a little bit, regardless of what else is going on. Even so, an illness or condition that is otherwise trivial but just happened to tip someone over the edge might, in the certifying physician's discretion, be omitted, particularly when we're talking about frail, elderly people who are one stiff breeze from keeling over anyway. But what the CDC guidance says is that COVID should always be listed, if present, no matter the circumstances of any given patient's death. Doesn't matter how much COVID contributed. This serves to artificially increase the number of death certificates on which COVID was listed by removing certifier's discretion about whether it was serious enough to be worth recording.
Second, the CDC guidance says that COVID always goes in Part I, exaggerating the overall seriousness of COVID's contribution and making it basically impossible to distinguish between "dying with" and "dying from". It's a nondiscretionary inclusion, and it has to be put in the most serious causation section.
RE: Commenters saying they didn’t know anyone who died so it wasn’t real.
I worked briefly as a security guard at a hospital around mid-late 2020. There were bodies everywhere! Every single day we had a covid patient die, and the ICU was at max capacity for months. The death toll was never unbelievable to me.
And yet at the same time health care professionals on average were underutilized during pandemic. How do you explain that?
Also, last year NHS had flu season and overflowing hospitals and many reported, it is worse than it was during covid? How could it be?
Yes, the death numbers are real and yet it is not the whole story.
The art is not selecting one piece of evidence and make a story. It is by collecting a puzzle.
First were videos from China where people on streets were dropping dead from covid. What those videos mean? Were they real or showing what we were told they mean? Obviously not. Some pieces are poisoned. If you believe everything, you will not be able to make a puzzle because they won't fit.
Health care professionals were on average underutilized during the pandemic, in the same way that I-35 through Austin is on average underutilized. Sure, during rush hour it can take 45 minutes to drive a couple miles, but 80% of the day it has less than one car per lane per five seconds. We unfortunately don’t have a good way to Target capacity to the moments when it is most needed, and during the pandemic we tried to make up for that by reducing usage generally.
We actually had. By not introducing lockdowns. Somehow, those countries who didn't introduce them generally had higher but flat curve which led to more stable usage of healthcare services.
That's because it's arguably incorrect to list a death as being simply from COVID. Rather, death certificates would say something like "acute respiratory distress syndrome, caused by COVID". That's still a death caused by COVID.
Death certificate would not read ARDS caused by covid. Certificate would read primary cause ARDS. Other conditions etc. If ARDS caused by covid should read Covid primary with etc. CDC hides the most interesting data in technical reports and footnotes. Turns out they knew the vaccine was causing myocarditis and deaths but suppressed it. Oops.
This is a California death certificate form [1]. If section 107 would read "(a) ARDS, (b) COVID", then that should be counted as a COVID death, right? Those are the death certificates the paper is talking about (the 91.5% (94.5% * 97%) of certificates that list a "plausible chain-of-event condition").
I recently learned that MS was very convincingly linked to Epstein-Barr virus a couple years ago. I haven't heard if there were any new treatment recommendations based on that but it seems like taking drugs to keep EBV inactive might be a decent idea.
I still can't understand my fellow Americans reactions to anti-covid measures. I thought we were a country that was proud of our resilience, and strength throughout history.
We were at war with something literally killing millions- as far as we knew the greatest threat we ever faced, especially early on when information was unclear, and some reports were suggesting 10% mortality rates. We had a chance to stand together and fight, but people were cowards, and the mild discomfort of masks and distancing was too much to bear, it was easier to pretend it wasn't happening, even if the pretending killed your grandparents, and your elderly neighbor.
Then these same traitors had the gaul to invoke masculinity and say stuff like "you're not a real man if you're cowering behind a mask." I got this from people literally spitting in my face and invading my personal space, while I was at the store volunteering to buy groceries for the immunocompromised people at the retirement home next to my house. I'm young and healthy, I am wearing the mask to protect you and the guy with failed kidneys I'm buying these groceries for, not to protect me.
I wouldn't have put this in these terms but I agree on how weird it was to see people claim masking was a sign of weakness and fragility when it was an inconvenience I was taking on for the benefit of others - something I'd think of as quintessentially masculine behavior. Also weird was people talking about personal choice literally spitting on me for making a choice, personally.
The language is important- we all had a once in a lifetime opportunity to rise to the occasion, and practice leadership in our own communities, to protect the weakest among us from being killed and disabled. Some of us showed a total lack of character and strength - and then try to hide their shame with the type of language I am using, painting themselves as some type of masculine warrior fighting for freedom.
I was followed, spit on, and mocked in public for being out with my 2 year old son in a mask. I was following strict distancing and laboratory grade sterility measures in public, because I was the only person bringing food to immunocompromised elderly people that would die if I didn't do so. I bought their groceries in full lab PPE, and dropped them off at an agreed upon location so they never came near me. I will never forget the cowards that tried to actively thwart my efforts to keep my neighbors, friends, and loved ones alive and mocked and betrayed me in the process.
Not a lie, early on there was no good information yet, all anyone really knew is that hospitals around the world were overwhelmed, and the bodies were piling up faster than anyone could deal with.
I’ve also encountered this stuff, as have many I know.
It predates COVID and masks, though. TBH I’ve never understood people who initiate hostile confrontations with complete strangers in public for clothing choices.
> you're not a real man if you're cowering behind a mask
…yup, you’re only truly brave if you cowardly conform to my ingroup’s expectations instead of just wearing whatever the hell you want.
It’s just football hooliganism bleeding over. I don’t think there’s anything new or complex or special here.
It may have helped if more people calling for lockdowns weren’t at the same time minimizing the level of disruption and sacrifice required to have a lockdown. Lots of “oh why do you care about the *economy*, grandma killer?” at people genuinely afraid of losing their livelihoods. “Oh just stay home, use FaceTime” as if missing births, deaths, church, and everyday socialization was all zero value stuff that only a selfish idiot could dare weigh against The Virus.
Crazy conspiracies on one side and toxic, condescending virtue signaling were two sides of the polarization coin.
> Crazy conspiracies on one side and toxic, condescending virtue signaling were two sides of the polarization coin
I agree with that, and see how they drive each other as a system. People want to believe the conspiracy theories to avoid the guilt being thrown at them from the other side. And then the virtue signaling is doubled down on in response to the conspiracy theories.
The lockdowns were a good idea for a short time when there was so much uncertainty, but they went on way too long after we discovered that masks worked, and so did good ventilation and being outdoors.
My theory is that, in much of the country, we “threw away our shot” on lockdowns too early. As soon as New York got bad, everybody locked down hard right away, anticipating an immediate huge spike that in many places didn’t come. I’m in Arizona, and people were fairly compliant with the initial orders. But then the massive emergency hospitals were empty, the wave of death was more of a ripple, and people were absolutely sick of being home by Memorial Day. At which point everyone went bonkers at the first lifting of restrictions, and THEN we had a huge wave.
Covid in America was more of a rolling wave and I think we may have been better off with keeping areas open until cases started to spike. Keep medical systems from being overwhelmed, but not tried too hard beyond that point because the returns diminished.
An additional "thumb-on-the-scale" for why you didn't hear about elderly people dying was that a large number of elderly people are socially isolated -
Meaning people are less likely to hear about an extended elderly family member dying, especially with with everything else going on during the pandemic
> The yearly mortality rate for 70 year olds is 2%, so the risk-during-the-time-they-incidentally-have-COVID is 2%/(52/2) =~ 0.08%.
This is wrong reasoning. Hospitals were the place where covid infections spread the most. The rate of testing positive in hospital was probably 10 times the background level. Even if you had no covid before hospitalization, you most likely got infected in the hospital. People, even doctors were so naively unaware about those risks. They trusted masks and hand washing which did almost nothing.
Even more, no covid lab test was actually necessary to report a person to be dying from covid or at least with covid. Did the person have cold symptoms, cough and difficulty breathing? Yes (who doesn't before death?) Let's write that this was covid. Scott clearly underestimates how much incentives can play role and misreporting cause of death is a normal thing.
Do you think anyone was worried that a death is reported as covid without testing and/or sufficient evidence? If you do, then you have forgotten the panic that existed during covid pandemic. Any talks that lockdowns are unnecessary were met with abuse. Experts and lay persons were censored for saying the correct thing, banned from social networks. Even most doctors online said all kinds of absurd things.
I can easily imagine that during outbreaks panic becomes contagious. Once people start dying from covid in the hospital, the staff starts suspecting that everyone who is dying has covid and therefore I don't trust any statistics from those times.
A good start would be reading the article. Scott addresses this - his argument is that all-cause mortality is higher during that period, not that there were a lot of confirmed deaths from Covid.
As most countries had similar rates of deaths relative to Covid infection, this requires a global conspiracy to doctor (sorry) the causes of death to attribute hundreds of thousands of deaths cause by lockdowns (?) to Covid. Why and how? Why would the Chinese government do this, for example? Why the government of New South Wales? Tough one for Oxfam’s razor
We used a host of lab tests to diagnose covid, not just a positive PCR. The symptoms are actually fairly distinct, and the blood work is quite distinct. One example, flu, RSV, rhinovirus and other common colds do not cause d-dimer levels to spike up. Another, ferritin levels spike up. Another, blood clots form in the lungs. Another, quite distinct patterns on lung x-rays and CT scans. The first couple of months were quite confusing since we had never seen any disease with the pattern covid showed. Now it's routine.
Yes, but this is about what was put on death certificate. My father's cause of death on death certificate has no relation to reality. I am not going to believe that this was an a rare exception. They didn't seem to care much, to be honest.
So, according to you hospitals were clusters that helped COVID to spread, but lockdowns to prevent clusters were a bad idea ? Don't you see a contradiction.
Lockdowns actually created big peaks. See data how Sweden and Belarus avoided them.
The biggest problem with pandemic is the amount of disinformation that was distributed by the authorities. I cannot blame people that still believe these lies but these lockdown policies were bunkers and extremely damaging.
Do you have a theory with predicting power that explains how locking people at home alone or with few relatives causes clusters and helps virus diffusion ?
Because my predicting theory is straightforward : virus has dramatically less opportunities to reach new hosts, so I expect that Rt crumbles after incubation delay.
Locking people at home happened in China not in the west.
You have to be precise when talking about these things.
People had to stay at home during lockdowns, they obeyed voluntarily or out of fear being fined or sanctioned but they were not locked. And they had to leave the house to buy food or do some other urgent things. Many people had to go to work.
Rt might be reduced but we don't really understand this that well. All models about pandemic failed, turned out to be wrong. Some places had very little lockdowns and virus didn't spread in the first year. Basically we had little information how and when any lockdowns could be introduced to achieve effective suppression.
Sweden decided not to bother. They did better than any other country.
It says something when Trump was the incumbent at the outset, that Operation Warp Speed is widely regarded as a success, yet COVID policy is an albatross for the Democrats (despite it probably costing Trump the 2020 election, and lockdowns being less of a thing as time went on).
I think it's mostly uncontroversial that there were surplus deaths from Covid. What Democrats need to contend with is that lockdown policy and the like are now regarded skeptically. This has tarnished their image and doubling-down on unpopular things (see: Biden's laundry list) is not helping.
It's not enough to be right about some technicalities. Pick the right battles.
Lockdowns lasted way longer in blue states, and while OWS is regarded as successful (once Trump couldn't take full credit), vaccine mandates were hilariously unpopular and tied strictly to the Democrats.
For me I think it is useful to zoom out a bit and ask: What do we do with overwhelming evidence when it points to something we don't want to be true?
The level of evidence here is at a similar level to "Men are significantly better at most sports than women," which is a fact that I don't want to be true but none the less is.
It's painful to see the mental gymnastics that people go through to deny the sports differences. It was similarly painful to read the comments section of a rationalist blog twisting and turning into anecdote and conspiracy over the simple fact that Covid killed quite a lot of people.
Perhaps this tends to happen when a fact comes into conflict with a core moral belief. In the sports case this might be the moral principle of "equality and the evils of discrimination." In the Covid case it might be "individualism and the evils of forced conformity."
What we do is bet against the idiots and take their money until they can't afford to bet anymore, then shame anyone who continues to BS while unwilling to bet.
Somebody managed to set up a refereed COVID origins debate with a $100,000 prize and was able to collect on it. That could be done for similar arguments.
Yeah, I was asking whether he was proposing something like that. Because in that case, you'd be betting on the outcome of a debate, an entirely different thing from what is true.
I think it went great for getting the best arguments out of the lab-leak and wet market crowds. The lab-leakers' pivot from professed confidence in the format to complaints that their representative only lost from being bad at presentation settled any doubts I had about COVID origins.
What's the problem with men being better at most sports? Men are bigger and have an easier time building muscle mass. Being significantly better at physical sports should totally be expected.
The one "oops" is chess where it's possible the brain size plays some role and it would be a hard pill to swallow. But I think fewer women finding motivation (whether because they're less competitive or because chess in particular aligns with their natural or socially conditioned inclinations worse) might be a good enough explanation in itself?
I deeply appreciate all your work in keeping these discussions open, sensible, and sensitive. This sort of analysis of what happened, how did we respond, why, and how to be better oriented in the future, is important. The commenters are appreciated and invited to respond further. We all learn and grow in this manner. Thanks.
Excellent coverage of the subject but as seen by the comments there are still some people that you will never convince with data because as they bathe in the soothing sounds of right wing talk radion and Newsmax, their "feelings" about Covid both being a hoax and a real virus that was engineered by the Chinese to kill Americans will outweigh any real data that can be presented.
As I'm deep into our $1,000,000 debate on vaccine efficacy, all this is fresh in memory. So here are the quick answers, using 'trust me bro' as a source, with the full documents to be released after the debate is done.
1. Covid caused around 1.2M US deaths
2. Covid deaths were under-reported, not over-reported. 'Deaths with covid' is a myth. There are around 10% where covid was a major cause of death but wasn't reported as such.
3. There were many frail people who died of covid. 10-20% would have died within a year or so.
4. There are other causes of death that increased due to the pandemic, especially overdoses, totaling 100-200,000.
5. Vaccines saved almost 1,000,000 in the US, and caused near zero deaths.
2. Even Scott's graphs doesn't show this. Maybe if you think that lockdowns saved some people who then died from covid. Not impossible but unlikely.
3. Most of them were frail.
4. very likely
5. Sounds right. 0.5% projected death rate -> 0.3% died from covid, 0.2% saved by vaccine. Less than it was worth. What did we spend on vaccine development and manufacturing? Probably many billions. Saving a life for, let's say $5000/QALY is within normal expenditure limits but for a vaccine that is quite a poor result.
2. CDC data can be quite rubbish. GIGO. Remember, this organization recommended covid jabs to kids until very recently.
3. I mentioned somewhere else, 10 years are not believable at all. Apparently you have too optimistic assumptions.
5. I mean, in the US each saved life was done at the expense of $5000/QALY. Obviously you are ready to spend up to $100,000 but overall we don't spend that for every person. If vaccine costs are $5000/QALY saved that is very bad.
2. there's a difference between bad decisions and raw data manipulation. do you have any evidence of cdc ever manipulating raw data?
3. we didn't research it heavily as it is not central to our thesis. we just used the consensus from studies. would be interesting if you can substantiate your claim that is far lower.
5. note there's probably 10x more benefit from preventing non-lethal damage and related costs.
2. Raw data does not exist for that. Even in ideal conditions there is a big uncertainty about how much covid could contribute in each death.
3. I already showed that it is mathematical impossibility. To overturn it, a lot of our understanding about pandemic needs to change. Not impossible but unlikely. If you are taking studies at their face value and not being experts in reading studies, you arrive at wrong conclusions similarly to “masks work” etc.
5. Actually it only increases the cost, since all those people who survived will require a lot of healthcare. I am not saying it is a bad thing, just that it is unusually costly compared what we generally expect that vaccines are cheap interventions to save lives as compared to costly treatments when vaccines are not available.
On the "knowing someone who died of COVID" point, I suspect the distribution of people who died (and who know someone who died) is actually much more similar to the categories in footnote 3 than to MS. As acknowledged in footnote 4, if you have certain characteristics--you're elderly, or you worked in a nursing home or hospital, etc.--you may know loads of people who died of COVID, like a police officer knows loads of police officers or a Rhode Islander knows loads of Rhode Islanders. The corollary of course is that more people without those characteristics, like relatively young people who work in tech and read blogs, would be less likely to know someone who died of COVID than they would be to know someone with MS, or with some other hypothetical randomly-distributed 1/300 trait. So I don't think COVID deaths feeling thin on the ground should actually surprise too many people, if they really consider how their own characteristics map onto those that would make one more or less likely to know COVID victims.
I unironically really appreciate Scott doing the work to persuasively argue the truth of something very obviously true to better inform his commenters, though am kind of bummed that there were enough commenters skeptical about this that he felt it was necessary.
Well, RN here. We had lots and lots of people die during that period. Die horribly, gasping for breath. I work in a 26-bed ICU. We expect a few people to die each week. We don't expect a few people to die each day, for 2 years running. We put people on ventilators, one or two a day. We don't expect to have 20 of 26 beds with ventilated patients.
We put people on ventilators only if there is no other choice, if they would die otherwise, because they need AIR. Not because some lab test says they have covid. Not because some wacky doctor wants to try out some new treatment. It was ventilate or die, after everything else had been tried.
Why is there no age-standardized excess mortality in Sweden? Why do so many African countries show no excess mortality?
If you are not considering disruption, panic and fear as significant drivers of excess mortality, you aren't aware of the power psychosomatic illness. Consider this study from 1980:
"Fifty-two volunteers were given experimental colds by nasal inoculation with rhinoviruses during the course of a 10-day residential stay at the Common Cold Research Unit, Salisbury. Prior to their inoculation with virus, they were assessed on five different measures of recent life stress and they also completed the E.P.I. Clear evidence of a psychosomatic component in colds was obtained. Introverts developed significantly worse symptoms and infections than Extraverts. And life events, when they involved change in the person's general level of activity, were significantly related to magnitude of infection."
Now consider that we, essentially, locked people in their houses and told them that the worst virus in 100 years was circulating.
The studies on stress and cancer survival are ubiquitous.
It hard to overstate just how disruptive COVID mitigations were to literally every person in the developed world. If you factor in delayed/missed health care, profound economic disruption and an incredible stress/terror burden, it's not that hard to get to a 5% excess mortality.
I didn't research the topic but it's a claim I've heard fairly often. Just checked, ChatGPT confirms it, too. Is it the standard of discussion here to do more serious research before asking a question?
It's a common belief, but not because of evidence behind it (hence Barry Marshall's skepticism). Similarly, many people concluded that placebo effects have a big effect on health without first checking whether regression to the mean could explain those outcomes instead.
> then total all-cause mortality during the COVID pandemic would be the same as always
No, the correct comparison would be to the deaths if governments acted as they did and instituted lockdowns and other "pandemic measures" even though the virus isn't particularly harmful.
As others have noted: people were afraid to go to the hospital for a while. I know family members who opted out of seeking care during the covid era.
Also: I don't claim to have any special knowledge but shouldn't you also look at the trailing excess death numbers AFTER the pandemic? Something like (this is from memory! sorry if it's not entirely correct!) 97.5% of covid deaths were people over 40 and 93% wfere over 50. And about 80% were over 65, I think. My expectation would be that when you look at the data, you'd find that covid largely accelerated the deaths of a cohort of people who were older and already in bad health and were going to die in the next few years anyway. Now that's still a tragedy- no doubt about it. But the headline of "COVID KILLED 1.2 MILLION PEOPLE!!!!" seems a little bit misleading, to me, when you consider that the vast majority of those people probably would be dead by now, in 2025, anyway.
So I'd be interested in knowing if the high excess deaths during the covid era were followed by a lack of deaths afterward for a period of time, when all the people who would have died from other causes, but had already died, aren't... well, dying.
Excess mortality is the number of deaths above (or below) the number predicted by a model. At a minimum the model includes the demographics of the population. This means that if, for example, an extra 20% of all those over 80 die, then fewer of that cohort are available to die later. If the mortality rate for the now slightly older cohort return to the baseline then their contribution to excess mortality should return to zero, not become negative.
There is no reason to expect the integral of excess mortality over time to tend to zero, but that seems to be widely assumed in the comments here.
Has the number of excess deaths become negative the last two years? If there's a million people who would have died in the next 10 years but instead died with/of COVID during the pandemic, isn't that 100 000 fewer deaths (on average) each year? Or maybe that should be 50 000, if the average remaining life expectancy was 10 years so half would live longer than 10 years. And perhaps it's even lower since more of the deaths will be concentrated toward the end. Regardless, wouldn't we expect fewer than expected since the weaker were killed off?
For those who are reluctant to blame covid, the first thing to think of is whether there were excess deaths compared to trend or not. If there were then its really not important whether covid was the primary reason or not.
All those excessive deaths would not have happened without the pandemic and our response.
This sounds deranged. Of course it matters whether the people who died were killed by the virus or by the "experts" entrusted with immense power to combat it, if for no other reason than to decide what to do the NEXT time a novel pathogen is discovered.
What is funny is that people find all arguments to minimize COVID deaths : all men must die anyway, those who died where mostly old and/or already sick, it's not the virus but the treatment, the hospital, the mask, the ventilator, the vaccine, the lockdown, the immune response, whatever, anything but not COVID.
However the same people are less likely to contest statistics concerning cancer, heart attack or any other cause of death.
I'm really glad the quoted commenter in III didn't have the personal experience of someone they know closely dying from COVID - but I knew at least four, including my father and two grandparents. So, I acknowledge that I come to this specific topic with a fair bit of emotion attached.
Given it's a communicable disease, and that there was a stratified response in terms of precautions, vaccine adoption, etc, it would make sense to see patterns where some people know one person, if any, and others know countless others. If you're primarily participating in conscientious, wealthy, and educated circles, I think you'd expect to know fewer dead than those who aren't.
Regardless of the merits of the ongoing debates on lockdowns, school closures, etc (which I think are 100% worth having), I do think the attitude of "well, I didn't know anyone who died of it" shows a certain incuriosity and spiritual sclerosis.
I agree with the numbers strongly indicating the excess mortality is from COVID and not from other sources.
Assuming that is true, and that around 80% of COVID deaths were among the elderly (65+) and especially concentrated among those with comorbidities, we should expect to see a significant drop in excess deaths starting around 2022 when those people would have been dying otherwise. Instead what we see is that although the numbers go down from the 2020/2021 peak, it's actually still significantly higher than the 2019 baseline trend. So why did people keep dying despite many of the most likely to die in 2023-2024 having died early?
That I don't understand, and would love some thoughts about.
I don't know the answer, but there's a fundamental vulnerability to this analysis that I haven't seen addressed so far -- while "total deaths" is a pretty clearcut and reliable number, *excess deaths* is not so much; it relies on one's chosen baseline for expected deaths.
While I expect the people who calculate excess deaths are clever folks and take such things into account, some people on here seem to want to use a linear trend for YOY death rates -- which given the combination of life expectancy vs. age being non-linear with the well known spike in birth rates starting about 75 years prior to 2020 seems unreasonable even in normal times.
I'm glad I read far enough to see someone ask this, as it was going to be my comment. I'm not disputing that the excess deaths appear in Scott's graphs, but at some point they either (a) have to be reflected in a pull-forward effect and thus show some reduction later, or (b) have to be caused by something other than Covid, or (c) Covid is permanently increasing the baseline death rate by 5-10% (which it isn't, that would be hundreds of thousands of deaths per year and I don't think any side of the debate is claiming that many). At some point, one has to ask why the excess deaths are continuing when the virus has mostly burnt out. That could include things that EA / rationalist types usually care about, like the economic fallout of lockdown induced recession, stimulus induced inflation, and reduced educational opportunity.
A non-NPI thought: 10-20 million undocumented people likely arrived in these same years, some estimate even more. Could it be a modification of the denominator, that we're measuring deaths for circa 330 million when it should be closer to 350-360 million? Or to put it another way, extras added in? If the average illegal immigrant is 31 and mostly male (which Google suggests) then the death rate is about 250 per 100,000 for that demographic - but might be higher for this subset, since they're mostly from the Third World and work more dangerous jobs. If the mortality rate was, say, 400 per 100,000 and there are 25 million, that would be an extra 100,000 deaths per year. I don't know if that would quite match the increase in baseline, but maybe in conjunction with the other economic effects it could explain the increased death trend line.
While I don't think it contributes that much, covid or its damage persist in an infected organism sometimes. So it might increase mortality for much longer than it takes to clear respiratory symptoms.
I'm not interested in disputing that this many people died from COVID, but I think you should engage the distinction between what is seen and what is unseen. If you're going to say the COVID interventions were warranted because 1.2 million dead from COVID, then you need to make some accounting of the number who died because of the COVID interventions. You've got ~163,000 non-COVID excess deaths during the pandemic itself, and then there's going to be a long tail afterwards as cancers that went undetected and untreated during the pandemic make themselves known and take their toll, liver damage accumulates in people who turned into alcoholics because they didn't have anything to do but sit at home and drink, etc. You don't seem to be taking any of that into account at all.
I agree with everything Scott has written, but the original point (why aren't we talking about the 1.2M deaths?) seems...consistent with society's other choices? After all, the Spanish flu killed more people than WWI, possibly more than WWII, and hit younger people harder, yet there is an order of magnitude more discussion, media, memorials, etc. for the wars than for the flu.
The point is denying that COVID was really a health issue rather than a government conspiracy to make wear masks so we'd be more compliant or something. Evidence doesn't change minds on this.
Don't you sort of answer your own question here? 1 in 300 people having multiple sclerosis is really bad but most people don't go around thinking about it because they likely don't even have a close relative or friend that has multiple sclerosis and may not even have any acquaintance with multiple sclerosis.
1 in 300 people dying from COVID is really bad but most people don't go around thinking about it because they likely don't even have a close relative or friend that died from it and may not even have an acquaintance that died from it. Further, to the extent they know anybody that died from COVID, it is overwhelmingly likely that they were really old or had some very serious comorbidities, and often times both.
In contrast, basically everybody had their life negatively impacted by the reaction to COVID. Even if they liked some of the results (work from home; extra time with family; slower pace of life, etc), overall it was likely negative, especially financially if they did not get any PPP or ERC money.
There’s something that seems a bit contrived about the from Covid/with Covid distinction. I lost my aunt in 2020, my father in 2021, and my uncle earlier this year. They all died in the hospital, none from Covid, but cause of death obviously wasn’t a clear-cut checked box on a piece of paper. Once s*** started slipping, keeping them alive was a gauntlet of tradeoffs – stopping treatment of one thing to fight off some other thing. Especially my aunt went from seemingly quite healthy, going in for relatively low-risk surgery, to dying surprisingly fast. No one thought it was her time yet. Would they have been able to save her life, and give her another 10 years, if Covid hadn’t been part of the mix, messing up her recovery? I don’t know. But all this is just to say that excess deaths seems like a much better way of tracking this than what anyone reported on some form in a million different cases.
"This number - 1/300 - is also the prevalence of multiple sclerosis3. Do you know someone with multiple sclerosis? No? Then it’s not surprising that you also don’t know someone who died of COVID"
I have no particular reason to know that my paralegal's sister has MS. I would definitely know if my paralegal's sister just died.
Would be interesting to compare US covid death rates with that of other countries, particularly those with even better record keeping than the US to see if the US is an outlier. Then figure out why. Then figure out what interventions and restrictions actually made a difference. This reconning needs to be done. It is OK to panic initially, but then cooler heads and evidence based practices should take over collectively. The tribal hostility that became the public manifestation of the pandemic in the US was a disgrace, and people wanting to move on may want to do so to avoid facing that part of themselves.
There is a lot of data available on that, e.g. on worldometer, and analysis on excess death and other things on our world in data. USA seem to be on the high end of covid deaths (0,36%), but not really an outlier.
Thanks for the input. Compared (on Worldometer) to the countries I would expect to have reliable tracking, like Canada, Europe: Scandinavia, Netherlands, Germany, Austria, Portugal, Spain, France, most countries seem per my math to have significantly lower rates, UK, Poland and Italy (which was so hard hit) are closer but lower. Didn't on that limited search find any comparable country reporting higher rates. So definitely "on the high side", as you say, which is a less dramatic way of putting it.
I hear a lot of people say they don't know anybody who died from covid, meanwhile two people in my family died from it within a month or so of each other (these are people who i've met later in my life and don't even know i lost people to covid).
In my case, my grandmother was sick (she had tested negative once but a day later was still feeling bad), when my mom went over to her home to bring her some things to help her feel better, my grandmother had collapsed in her bathroom after using the toilet (i think my mom might have had fecal matter exposure which could be a very very high viral load, or so I have heard something about that). That was how my mom was infected as well. My grandma was dead about 10 days later, and my mom about a month and a half after that. In my mom's case, she was on a ventilator for a bit, and when she didn't improve, and then worsened, put on ECMO for about a month. After worsening, stabilizing and then improving, they took her off ECMO but after the ECMO was removed she just kind of... never came back. She was gone, so we pulled life support a couple days later.
Anyway, I didn't post to share too many details, but I wonder if this is perhaps an explanatory reason many people don't know someone who died from covid. That is, the nature of how it spreads and our lives at the time means only very close family members may have caught it from one another, and some genetic/environmental weaknesses could have led to some families being devastated while others went relatively unscathed.
> Leveraging county variation in the timing of reopening, we find that returning to in-person instruction increased youth suicides by 12–18 percent. Analysis of Google search data suggests that bullying is a likely mechanism.
1 in 8 suicide deaths are peopled aged 15 to 24, how many were going to school? Even if it's 10%, that's only about 1.5% drop of suicide rates explained.
Between 2015 and 2019 the average rate was ~13.8 per 100k per year, from 2018 to 2023 excluding 2020 it was ~14.1. 2020 was the only year with noticeably lower rates, 13.5. So we need to explain the difference between 0.3 and 0.6 out of ~14 in 3/4 of a year (from Jan to March 2020 there was nothing unusual). That's 2.9-5.7%. Indeed, 1.5% is a big chunk of that.
>Is it possible that the CDC and National Center For Health Statistics are lying about all-cause mortality patterns? Seems unlikely, because individual states reporting separately found similar patterns, and so did the other countries that reported data. This would take a truly global conspiracy.
I *don't* think these numbers were misreported or that any of your conclusions are wrong, but just as a general note:
It could also happen if all of those institutes had similarly strong incentives to lie in the same direction.
Lots of toddlers point at the dog when a parent demands to know who broke their vase. but this isn't an international conspiracy to libel dogs, it's just lots of people reacting in the obvious way to similar incentives in similar situations.
Heavily correlated actions can look like coordination if you suspect coordination going in, or don't think about the parallel incentives producing the behaviors.
Every single personal point of data that I encountered ran counter to the public narrative.
1. My son was born in the Autumn of 2020. Due to some complications, we had to stay in the hospital for three days. It was an absolute ghost town, with virtually no cars in the parking lot and no one walking around the hospital. I was watching the news on the TV about hospital overflows while sitting in a hospital in a major metro that appeared to be completely empty.
2. I have 2 close family members that work in hospice care. They both indicated that every single person who tested positive for covid in their facilities was marked as a covid death, even ones that entered hospice with stage 4 cancer.
3. I knew two people who died of Covid: one was morbidly obese (400+ pounds), and one suffered from Alzheimer's and was in his 90's.
It's not so much that I doubt the data on excess mortality, I just think that the counts don't tell the full story. The hospital system stress never materialized, the masks were a non factor, and the deaths were generally in very sick or elderly people. If we had the data to convert "excess deaths" to "missing life years" by collecting this data on an individual level, the whole thing would look much less important.
This post does a good job of reminding me why 2020-2022 did more damage my belief in humanity's ability to think and act rationally than any other period in my life. The people who continue to obsess over conspiracies surrounding the vaccine, the number of deaths, the "lockdowns" are simply so stupid I don't know how they are able to draw breath.
I've never analyzed the covid conspiracies so deeply before, but now that you point it out, it's absolutely hilarious that two mutually exclusive conspiracies have such a high rate of correlation in the population.
1. Covid was started by a group of Chinese scientists trying to engineer a deadly virus, potentially intentionally to disrupt the world economy.
2. There's a secret plot by the governments of the world to lie about the damage of covid when in reality almost nobody died.
The probability of either one of these being true on its own is low. The probability of them both being true feels nearly impossible given that they are contradictory.
You do realize that the people who were paid extra to report COVID deaths were also paid extra to report COVID cases?
Also that if the original burst, prior to the lockdown, was COVID, this does not mean that the drastic reaction to it did not, in itself, cause excess deaths.
Yes, did you look at the graphs? Do you think there was some other factor that caused monthly death numbers to swing double-digit percentages in the exact same pattern in every Western jurisdiction on the planet over the course of two years?
Really? You think there could be some other unspecified thing which caused millions of excess deaths worldwide that somehow evaded the notice of every government and respectable intelectual organization on the planet?
I know we're discussing this elsewhere, but, any chance you have a link for this claim? Your argument rests pretty heavily on it, so I think it's fair to show us the strength of this claim: who was paid, how much, by whom, when? All that good stuff.
2. With a source that provides an order of magnitude estimate for the total amount paid
3. And that total amount, distributed among a proportion of hospitals sufficient to account for at least 10% of reported COVID deaths in that country, accounts for more then 10% of the annual budget of those hospitals
I will concede that there is reason to believe that COVID death counts were incentivized upward sufficiently to make a material difference. How material, I reserve the right to judge depending on the actual countries and numbers.
In case you want to hear my reasoning: my case rests partly on international comparison, so I want enough countries to be sure you can't just be explaining the US. I want to make sure these payments can have materially affected COVID death counts, so I want to see the money going to places where large numbers of the COVID dead came from. And finally, I want to make sure the payment amounts are enough to incentivized real change in behaviour: if you find me $1M spent total, I don't think that's enough to buy a material change in number of COVID deaths on the scale of a country.
I also acknowledge that these criteria are just my first pass at trying to capture those concerns, and especially as I know nothing about hospital budgeting, I am willing to accept arguments that I should modify some of these criteria due to misunderstandings or mistakes; I just want to make sure whatever evidence you provide shows that paying for more COVID deaths happened widely enough that it can actually explain a large fraction of COVID deaths.
What goalposts? I asked you to provide evidence, I'm now telling you what I think counts as good evidence. I've explicitly said I'm willing to change these specific conditions so long as the evidence you provide is sufficient to show that this is widespread enough to materially affect reported COVID death rates in multiple countries.
If you really want to prove I'm arguing in bad faith, just post your evidence and if it's any good, surely everyone else will see so, and make up their minds about my good faith based on my response.
Seems to me he's actually trying to work something out with you. He's clear about his criteria, and offering to modify them if someone points out misunderstandings or mistakes. I think you are the person being unreasonable about goalposts. He asks for a link to your claim. Instead of just giving it -- and jeez, why on earth *not* give it, it's a url on the internet, not your wallet -- you demand that he commit to saying you are right if you hand over the url.
How would you like this?: I say I know of a website that offers excellent evidence that you are wrong. But I'm not going to tell you what it is unless you state in advance that unless there's nothing there but a dirty limerick you will agree that I am wrong. You don't get to point out ways you think my data is inaccurate or does not prove what I think it does. You just kneel down and kill my ass.
For one thing, because it's a public discussion and you can provide useful information not just to the person who asked it but to all the people who read your discussion without participating.
Hey, here's a URL proving you are wrong. I was going to insist that you agree in advance to admit I am right after you see it, but I am so confident you will admit it that I'm just gonna give you the url no strings attached: https://tinyurl.com/224cnwbh
It doesn't also not *not* mean it. You'd have to bring vastly more data to actually debunk what seems the rather obvious proposition that a communicable disease going around the population at the time actually did kill people in large amounts. The skeptics who are denying the covid=deaths link generally don't bring in that data and just go "well, there might be an alternative explanation, no?"
No, we bring in the data that shows that the data was falsified on a large scale.
Offering bounties for COVID cases and promising that no one would ask questions is all any rational person needs to doubt the data. (New York went so far as to state that anyone found dead without an immediately obvious other cause of death was a COVID case.)
Why on earth would they have to do something that would obviously inflate the numbers if the numbers were sufficient for whatever purpose they had?
Other countries did Covid reporting too you know? The conclusions were, in general, similar - so the fake reporting you are talking about is unlikely to have had such a very large effect. There may have been underreporting as well. That's why it's good to compare to excess deaths. Maybe the true number is 1M, maybe 1,4M, but how much does that really matter?
In a great many ways I would imagine. If you want to believe that there is a world wide conspiracy to inflate the deadliness of covid, while at the same time a lot of excess death happened for other reasons, I don't think there is much I can say to change your mind.
And if you think the openly admitted fraud about COVID deaths is -- what? Doctors accidentally diagnosing deaths as other causes when they were caused by COVID and accidentally putting down the right cause while trying to commit fraud?
There were, even at the time, many admissions that there was massive fraud.
Incidentally, we had the lockdowns. Lockdowns cause many problems, including mental. Doctors had a motive to declare suicides were COVID.
I mean that I am not sure that it's possible for someone to "know" than a specific acquaintance with cancer died "significantly earlier because he couldn't access the treatments he needed because of COVID."
It seems that the strain of virus at the time it erupted in NYC was particularly potent. I am glad that as it morphed, it weakened. People in their 40s died within two weeks. This, and not having effective treatment for those who had severe disease, is what freaked people in medical care, out. So you’re saying, hey, as long as my circle of loved ones and acquaintances didn’t die, it’s all fake? I am surprised that this comes from you. Did you volunteer to work in a NYC hospital in March 2020 until March 2021? Do you know for certain people could not die from this infection? How do you know? People often have many chronic diseases simultaneously. It’s the kicker that gets them.
"I like these less because they’re class- and location- stratified, so your chance of knowing them goes up or down a lot depending on your own characteristics."
COVID deaths were also very class and location stratified! Poor people in New York are a lot more likely to know someone who died of COVID than a rich person in Seattle or Minnesota or whatever.
There's an argument to be made on ventilators making COVID much more *lethal* in the early phases and that being part of why New York especially was hit so hard before treatment protocols evolved. I don't think there's any argument that would have shifted deaths by more than a high 4 digit/low 5 digit number however.
Yes some people died from disruption to the health care system. You can still attribute that to COVID since it was the disruptor.
And folks who ignored practices which could slow the infection rate, such as isolation and masks, added to the disruption. You could see at the time clearly different rates and deaths in countries with better behavior patterns. Or worse patterns.
I think I may have said this before, but there’s a huge difference in experience between people who, say, worked in hospitals or with elderly or as EMS workers in 2020 vs. people who experienced the pandemic from a distance.
I think that difference underlies a lot of the polarization.
As a professional debugger who deals in problems with tough unknown solutions often, I think a lot of people talking about the COVID response as though the people in charge had a lot of certainty about events / decisions made while the pandemic was unfolding.
In reality, the people guiding the pandemic response were dealing with messy logic of an unfolding pandemic that might cause an extreme amount of deaths, where best practices weren’t totally known at the time, and measuring effectiveness of lockdowns versus the damage to the economy / etc was hard to do.
In good tech companies, after the “end” of a bad outage / problem we come together and do a big blameless postmortem, where we talk about all the correct and incorrect actions we undertook during the outage, and try to come up with better practices for the next time something similar happens. This both acknowledges that we are acting with a lot of uncertainty (and are likely to make mistakes acting under uncertainty), and that we were trying really hard to fix the thing. And it provides a sense of closure.
America / the world needs a big post COVID postmortem, but this would be impossible to have the desired effect due to all the reasons in the comments here. I think the best thing we can do is make a big memorial to honor the hospital staff that took the brunt of this in stress / burnout / etc and add a national holiday of remembrance for the suffering of health / doctors / hospital workers.
TBH adding a national COVID Remembrance Day would help us all cope with the fact that it damaged everyone in some ways, and we are all upset about that and still coping with the aftermath.
This article is sort of boring because this isn't being framed right by either side. The honest question shouldn't be whether 1.2 million people died of Covid-19, it should be how many of those weren't going to die of something else in the next couple of years anyway.
The main reason we don't personally know anybody who died of Covid is because Covid cleared out the nursing homes.
A much more interesting article would be to correct the Covid death numbers by removing everyone over age 80 and everyone who already had a terminal disease, and see what's left.
You could even look at the lower-than-average death rates post-Covid as part of the analysis, which is an artifact in the numbers left over from a bunch of soon-to-die people all dying at once in 2020.
Indeed, excluding the surrounding years is a flaw here. Moreover, the excess mortality in other countries with different policies is relevant, with Sweden as the oft-discussed outlier.
What would really help is a study of co-morbidities. Assuming that the people dishonest enough to list “COVID” as a co-morbidity for an 86-year-old in hospice who coughed once before succumbing to stage 4 cancer were still honest enough to ALSO list cancer on his death certificate, we might learn something from seeing how the rates of other causes of death changed over this time period.
I don't think this is quite right either - it definitely killed a lot of <80s as well. A lot of them were more morbid than average, but nonetheless their deaths were real.
The question is how much could we have influenced the death tolls with the NPIs were implemented and did those pass basic tradeoff math. I think the answer is clearly "No" for everything other than vaccines, and would like to see the Democratic party have some sort of public acknowledgement that they were bad so we don't have to worry about going to remote school for several years the next time a pandemic rolls around.
The real tragedy is that if we get Measles-19 out of a secret lab and the lockdowns ARE warranted, nobody is going to do them because 2020 was a giant exercise in crying wolf.
well I think people will do things even absent Government intervention, because people are sorta pseudo-rational about some of this stuff. Like restaurant reservations collapsed in NYC before any lockdowns because people were like "maybe I'll wait a few weeks to see what happens with this whole Wuhan-flu thing"
If kids started dropping in droves from Measles-26 than I expect there would be pretty dramatic changes in behavior that would be largely justified.
BJ. You said: “The main reason we don't personally know anybody who died of Covid is because Covid cleared out the nursing homes.”
I do. My sister was 39. She lay in a coma for six weeks and died in April 2020 FROM COVID. She left a husband and an 8 year old adopted son. She had adopted him after his mother, her best friend, had died in a car accident a few years earlier.
My Uncle was 81, he was NOT in a nursing home and died from an aortic aneuyrism - he lay at home for five hours waiting for an ambulance because none were available, whilst I was on the phone with him - he was over 400 miles from me and I could not get to him, in May 2020. He had tested positive for Covid after his death. He had been a merchant seaman and I made a donation to the RNLI towards a new lifeboat in his memory.
My mother died in August 2020 from cancer that had not been spotted because she was unable to get to a GP because of Covid shutdown. She was 78. She was NOT in a care home. She had only been diagnosed in May 2020.
Please be careful of your comments on this issue. None of my family were ‘chronically sick’ prior to going into lockdown. Their deaths were not marked with funerals. I was unable to say goodbye to any of them. They were just gone. It has been five years and it still hurts that they died alone, in severe pain, without the help they should have had.
We do know the people who died. We remember them, even if the country doesn’t.
People who post in this and similar communities often presume that everyone involved in the discussion is thinking in a stochastic, Bayesian way where there are statistical distributions and outliers buried in all statements.
I apologize for not clarifying that "we" did not mean "every single person in the country," and instead meant "we who exhibit quality X which is clearly stated in the sentence, and who vastly but not universally have the same statistical experience."
There is a constant tension between lingual sensitivity and lingual brevity.
Don’t be so cruel. Turning this around on my ‘ignorance’?! Stop it. These were human beings not some philosophical experiment. You can’t even bring yourself to apologise properly, can you? I hope you never have to suffer what we have endured. You couldn’t take it, you really couldn’t.
Half an year ago, my mother died. At the point where it was confirmed that this was going to happen for sure it was obvious that she'd only have a few days to week until she was gone. I would still have been angry as shit if someone had broken in and killed her during that time.
The Brooklyn neighborhood and Jewish community I live in was one of the first NYC areas to be heavily affected by COVID. We have a community website that shares obituaries when someone passes, and pre COVID, we averaged 1-2 a week. During COVID, we had dozens. My grandfather survived a few bouts of COVID, but literally every single one of his weekly study partners died. Until he passed earlier this year, all his new study partners were young men because almost all the men of his generation were dead.
We have a local volunteer EMS service that was doing dozens of hospital transportations a day, and the Jewish cemetery was doing only one car allowed in at a time burial services, with burials going from sunrise to sunset wall to wall. Multiple new sections of the cemetery were inaugurated in a six month span.
So in short, the 1.2 million number makes total sense to me.
I think the argument that "you can't trust the experts" is not equivalent to "Covid is not a thing." Pretty clearly, something happened. The issue with trust is that there was never a mature discussion about what occurred at any acceptable level of seriousness, which lead to distrust.
That's why we can't point to hard numbers on Covid deaths and, when challenged, people point to excess deaths. I agree that something clearly happened in 2021 and it's pretty obvious it was Covid but technically we're not even establishing causality here. We just see a spike in deaths, that's where the level of discourse is.
And allow me to clarify this for some of the medical professionals here, because Covid has become very partisan. Let's imagine an ideal "Jane Doe" patient who might be misclassified as a Covid death. Our imaginary Jane Doe is a 79 year old female with a BMI of 38.1, diabetes, and COPD, who passed away in September of 2020 with Covid as the listed cause of death. I think everyone can agree it's fair to say that this patient's cause of death is...less clear than we might like and also not unrepresentative of the average person who passed from Covid.
And, to the best of my knowledge, these kinds of cases were never identified or investigated, certainly not in April-June of 2020, and overwhelmingly for political reasons. Because there are, to a rounding error, no technical challenges that would prevent this information from being gathered and summarized.
Take a trivial example. Say we wanted to know what % of patients who passed away from Covid at Kaiser Permanente facilities in 2020-2021 had 2+ high risk diagnosis, like diabetes or chronic lower respiratory disease. Well, I guarantee Kaiser has a SQL database somewhere in the backend to manage all the data and I guarantee there either a PatientDeath table or a PatientEncounter table listing every patient who died and there's also got to be a PatientDiagnosis table listing every active diagnosis for every patient, or at least a historical Diagnosis table with start and end dates.
And presuming your hospital has the most minimal technical infrastructure imaginable, the query to get this data would rhyme with:
SELECT
pd.UniquePatientIdentifier
, COUNT(DISTINCT(apd.DiagnosisName)) AS NumberOfHighRiskDiagnosis
FROM PatientDeath pd
LEFT JOIN AllPatientDiagnsosis apd
ON pd.UniquePatientIdentifier=apd.UniquePatientIdentifier
AND apd.HighRiskDiagnosisFlag=1
WHERE pd.DateOfDeath BETWEEN "2020-03-01" AND "2023-01-01"
GROUP BY pd.UniquePatientIdentifier
HAVING COUNT(DISTINCT(apd.DiagnosisName))>=2
We're talking about 10 lines of code. To be fair, not 10 lines of code I would expect a casual user to come up with but even the most junior BID should be able to knock this out in a few hours. It is simply not credible that no one at Kaiser, or any other major health system or hospital, was unable to provide those numbers for technical reasons. No one is credibly that incompetent.
This information was not provided, to the best of my knowledge, to either the public or to medical experts for obviously political reasons driven at the local level by hospital administration.
And, again, if you're a medical professional, can you tell me what % of Covid patients had 2 or more diagnosis from the top 10 causes of death (excluding Covid) at your hospital from 2020-2022? If not, why do you pretend that this is an acceptable level of discourse? Why do you pretend that Covid was managed and investigated to an appropriate standard? Why do you trust reports generated by a medical administration that has never treated you terribly well?
Because that's why we're all mucking around with excess deaths and other garbage metrics; because an intentional decision was made by hospital administrators under political pressure to not provide clearly relevant data, regardless of its impact on public discourse or, more importantly, patient outcomes.
100% Exactly my feelings on the matter. It seems like the individual level data that could clear this up must exist, so I find it incredibly frustrating that we have to talk about things like "excess deaths".
The truth is that legal and privacy requirements mean that the table you reference probably doesn't actually exist. I run into this problem all of the time in adtech: easy access to information to answer useful questions is a legal liability, so information is impossible to access (or simply not tabulated) unless it done in the context of a product that is making the company more money than the legal liability of exposing the information.
This is well stated. The difficulty in believing the official numbers is not solely a function of a priori distrust in the relevant institutions. It is that the relevant institutions deliberately set up the data collection system to make this information opaque and difficult to assess. Skeptics are naturally going to see deliberate choices by CDC et al. which actively made it harder to get reliable information as evidence that there's something nefarious going on.
Let's suppose we had all this data for every single death, would it let us come up with a better estimate than the excess death data? How would you write that SQL query?
Obviously we can't just say that anyone with any other health conditions apart from covid doesn't count as a covid death. Maybe we could do some maths with the actuarial tables for 79 year old female diabetics with COPD in 2019 vs 2020 and then decide to attribute Jane's death as 30% of a covid death?
But I suspect that if you did all this maths then you'd just wind up getting your excess death numbers back again.
2. Pretty much all non-vaccine policy choices had low efficacy and should not have been mandated (people would have chosen to change behavior some on their own)
3. Vaccines should have been sped up even faster
4. Vaccine should never have been mandated
People end up in this weird spot where they think either:
NPIs were unjustified, so COVID was not actually bad
Or
COVID was really bad, so NPIs should have been even harder!
Whereas the actual answer is sometimes you just have an illness pop up and there's only so much you can do. Focus on origins, vaccines, and everything else is roughly de minimis (especially on a QALY impact basis). None of this was controversial pre-2020 - the types of NPIs we rolled out had basically no scholarly basis in pandemic literature as responses to flu pandemics. Go back and read the Pandemic response papers by the guy who eliminated smallpox (DA Henderson). Suggesting closing schools for years would have been anathema.
Excess mortality during Covid was concentrated in older people, if your circle includes fewer of those, then you're less likely to know someone who dies of Covid. People in their 30s and 40s were likely to have already lost their contacts among their grandparents' generation, while their parents are not yet old enough to be at high risk. I'm almost 40, but my parents have older siblings; two of them (late 70s/early 80s) died during Covid. Among my social circle I probably have 20-30 people over 75, mostly family members, family friends and a smaller number of parents of friends that I'm closer to.
I would be interested in a study on how different age groups form social connections with one another, particularly outside of blood connections.
All of this Monday morning quarterbacking about COVID, its source, and arguing about how many folks died from it is absurd in the extreme.
Nature has been in the deadly viral and bacterial pathogen competition game with all forms of life on planet earth for billions of years. Mankind has only been around for maybe a few hundred thousand years, and written “history” for maybe a couple of thousand years. All religions are part of mankind’s search for meaning, understanding and finding a way to get along without killing each other. But, no religion or political belief has any linkage to nature’s pathogen game, and never has. Only mankind’s hubris allows it to think it does.
Religious zealots and politicians are mere observers and actually help the pathogens spread by denying their existence and impacts.
Pandemic viruses and bacteria are not political or religious. They don’t care about religion or politics, either way, but will exploit weaknesses wherever it finds them, to mankind’s and all life-form’s detriment, and occasionally to their benefit.
Thus far, nothing can change that tension, but vaccines can slow them down.
The Covid facts are: 1, It did exist; 2, It was deadly; 3, Millions died from it worldwide; 4, Approximately 1-1.2 millions of Americans died from it; and, 5, The various vaccines slowed it down and saved millions of lives.
Period.
So, please stop with the yabut’s, woulda shoulda coulda’s, and accounting nonsense and face the above facts.
Ditto about the Covid vaccines. They worked, otherwise millions more would have died.
Just stop with all of this political and religion driven naysaying nonsense. It’s irrelevant and deadly. That includes RFKJ and his pack of pseudo scientific “advisors.”
Pumba said it best: “put your past in your behind.”
In a sane and sensible world, there's a sane and sensible discussion to be had about how the public health response to covid went (really the hundreds of separate public health responses aroud the world) and what we can learn for next time.
Obviously the correct thing to do in hindsight would have been to lock down earlier and harder to eradicate the virus completely. But what would the second best response have been, once we'd missed the deadline for that?
It's still a pretty open question to what extent any particular NPI was worthwhile.
The case is a little stronger than simply tracking reported COVID deaths as indicated by cause of death and excess deaths because the rise and fall of excess deaths also tracks with a slight lag positive COVID tests and wastewater data where available. Excess deaths were rising and falling with *infection rates.* Excessive deaths were appearing where and when COVID was. The excess death numbers are consistently modestly worse than the reported deaths, which also tracks gaps in attribution you'd expect to see.
I know or know of 5 people who died of Covid: my mom's pastor and two of his congregants, and two inlaws--mother and son. All of them dedicated anti-precaution, all of them old or unhealthy or both.
I imagine there are some people just as distant from clusters of deaths as my connection to two clusters has me closer than average.
Fwiw I also know two people vaccine-injured and suspect a few more, with at least one possible death as well--but that one could have been post-infection damage too. I'm not close enough to have any idea, just that they "died unexpectedly" ie not while actively sick.
The spike is dangerous no matter how it's introduced.
But what's more interesting that no one talks about is that nearly all the COVID deaths (93%!!!) are people over age 50. Less than 100k people under 50 died according to this data! This actually explains a lot of the intuitive discrepancies we keep encountering.
Also this might mean that we could have just locked down all the people over 50 and let everyone else function as normal and it would have possibly been fine?
But also shouldn't this be a bigger deal? Am I missing something?
I mean, if people over 50 all lived in a different city from people under 50 that might have been possible, but...
In 2023 the internet suggests that 36% of Americans were 18-44, and 24.6 were 44-65. Assume that those account for all "working age" Americans, then over 50s are ~40% of working Americans; locking down 40% of workers doesn't strike me as all that much less disruptive than locking down ... whatever percent were actually locked down.
Well the percentage that was locked down was technically all of them at any time where there was a lockdown, so I don't see how having less people is less disruptive than having no people.
I just mean that a workplace where half of it's employees are on lockdown presumably functions at much less than half capacity; I'm imagining some non linear phase change type behaviour where having 5% fewer employees makes you 5% less effective, but at some point you pass through a critical value and suddenly at 50 (or whatever) percent, you're now at like 20% effectiveness.
But yeah, I don't totally discount that this idea might have worked (I think it's basically the Great Barrington Declaration); I just think "lock down only the fifty year olds" isn't logistically trivial and probably still has pretty big economic downsides.
I'm not saying it's easy or that it doesn't have down sides, just that it could have had less downsides than locking down everyone. Also it's not necessarily about work, you could have still had young people do stuff outside and go places and possibly congregate, etc.
Yeah, sure, I'm not taking a hard stand against this.
Fwiw, in my city we were allowed to go outside and congregate much of the time; I held my birthday in a public park in July 2020, was doing regular bike rides with one of my friends that whole summer; and definitely was eating in restaurants again (albeit very rarely) by August 2020 so how different your scenario is from what actually happened depends on jurisdiction.
Sure but if you're arguing that knowing that it's mostly people over 50 who died from COVID wouldn't have changed much in a practical sense then that's possible, but I disagree with that, or think it's at least it's worth considering what we could have done with this information
Some people, at least in hard-hit areas such as New York, may have died because overload of the healthcare system and/or fear of coming to healthcare facilities prevented or delayed treatment of conditions that would have been cures without delay. This includes prompt treatment of primary cancer. Also prompt thrombolytic treatment of acute cardiovascular disease.
These are not direct Covid deaths, but victims of system overload are just as dead as those who died directly from Covid. In my opinion, excess deaths constitute the most meaningful indicator of the health effects of the Covid pandemic, and will remain so unless the national invests in highly redundant capacity.
In continuing from yesterday, I think I would quibble with your tally along the 5-10% range, for the reasons you listed in your article. Which is to say I’d go along with an estimate of approx 1 million died “of covid”….which is still a very large number….but maybe not 1.2 million, but which doesn’t detract from the thrust of your original post.
Great point about the 1/300 ratio and the value of personal experience of actually knowing anyone who died. I fall into the category of your example: I didn’t know anyone directly, but I heard a friend of a friend did.
My personal prior is coloured by the fact that the hospital I admit to kept covid tallies based on “with covid” and not “of covid”…which is why I question some of the official numbers (at the margins).
We used to believe government statistics. Now (and especially after the last four months and DOGE), even if they tell us the truth, we won't believe it. That's what happens when you set the public trust on fire, pour gasoline over it, add a few sticks of dynamite, fan the flames, stir the ashes, and then salt the ground after for good measure.
No one trusts official spokesmen anymore. And even less official spokeswomen.
Agree 100% here, but I think we should show some sympathy to those who don't trust government statistics. Not because they are right, but because we have given them ample reason to be distrustful. The last time we had a genuine national security crisis before Covid was 9-11, and the public reacted with almost complete trust in the government to take the proper steps to fix the situation. To say they didn't deserve our trust is the understatement of the century.
Most people have absolutely no ability to make decisions on what to believe based on the facts involved, because there aren't enough indisputably facts to make a decision. It's all vibes and trust. The implementation details of vaccines, lockdowns, masks etc. doesn't matter at all, because a substantial fraction of people didn't believe anything they were being told.
Had a friend high up in the CDC during the first outbreak of West Nile. He said that stats and reporting just became completely politically captured *at that point*. Bad information went out and the doctors just fell in line like the government dupes they have turned out to be. It's disappointing. We are on our own.
Who is the "you" here? Scott isn't the Government.
The Government has lied in the past, so we should throw out normal truth-seeking behaviour and just believe the opposite of whatever the Government says in the future? This doesn't seem like it's going to get us closer to the truth. The untrustworthiness of official sources is a good reason to sharpen our truth-discernment abilities, not to throw them out completely.
I meant to be addressing the mouthpieces of the government. Sorry for the confusion. I don't necessarily believe the *opposite* but the opposite is probably closer to the truth than the narrative of the previous 50 years.
We know that covid incidence is highly age- and co-morbidity stratified. So, we would find the excess deaths in the old and folks with such comorbidities.
Conversely, excess deaths outside these groups (especially the younger than 50) would not be prima facie evidence of covid, but of other related causes, be they lockdown related (e.g. overdose deaths, missed "elective" procedures), treatment related (ventilators), or pharmaceutical (vaccine related).
I don't have sourcing, but I would hope someone has done this. But even allowing for these effects only in theory and in plausible numbers would suggest that the true deaths FROM covid are a lot lower than 1.2 million.
Try this one: "COVID killed a million people" is neither /wrong/, nor /semantically coherent/. Let's go:
Collapse "died with-but-maybe/probably-contributed" and "died from" COVID. Gander at the mortality rates by age (I think the risk-doubles-every-seven-years-of-age is still good?), plus the mortality by comorbidity (general ill/fragile health, lung weakness by any cause). All together, we can see COVID as a major driver of death in any cohort vulnerable to exactly what you'd expect from (especially the early, more aggressive variants of) COVID's lung-inflammation pathology on top of general flu symptoms.
To abstract, summarize the above as "a particularly potent and transmissible strain of influenza ripped through the population; the expected cohort died". This recalls that /the flu/ is a major contributor-and-cause of death, in exactly the cohort that is in general and pneumatic ill health.
This gives us a bit of Science: we predict from this that all-cause mortality would spike at viral introduction, stay high until everybody's had it, and then return to normal levels. If society were to, say, grind itself into worse health with lockdowns, ever-worsening chronic illness, and generally higher population, we'd see the raw numbers level back down a little higher than previous (don't forget population still rose at least ~3% over the span, beware raw figures).
So what do we see? Deaths by millions, starting 2019: 2.85, 3.38, 3.46, 3.28, 3.09, 3.07 (provisional 2024). This all makes sense, if COVID largely (largely! again beware the states are only meaningful epidemically) contributed-or-caused death in those who were, sorry to say but this is life, already going to bite it for some reason soon enough. Pneumonia, pre-COVID flus, or whatever other big systemic shock would have gotten them next, in the COVID-free alternate universe.
Semantic coherence breaks down if you simplify this into "COVID killed a million people", because COVID didn't "kill a million people" the way an big asteroid hitting Austin would kill a million people. Thus, if you want endless talking-past-each-other internet flamewars, keeps saying "COVID killed a million people", and you'll never be at want of gnashing teeth, without anybody in the conversation "being wrong". Everyone gets to be more-or-less-uselessly "right".
Thus the semantic incoherence of using "pandemic" for COVID: the death tolls matches our expectation from an introduction of a novel influenza, which finished off lots of people in a brief span, and then faded back into the more general tableau of holy-crap-flu-is-surprisingly-deadly statistics, which are now a smidge higher than their previously puissant death toll. This is different than an ebola or bubonic outbreak, which will rip through a population pandemically, but don't reflect a previous endemic problem, or simmer down into a chronic condition.
Semantically, "ebola killed a million people" coheres with the usual notion of "X killed Y total people", like a meteor or an airplane crash. It does not cohere with "a particularly bad flu season, lasting about two years, and then returning to near-normal". YES YOU CAN STILL BE RIGHT every time you say "COVID killed a million people", but as soon that gets used to justify pandemic/lockdown policy, you're semantically incoherent to the contexts where we might, say, go into incredible debt and immiserate hundreds of millions in order to "prevent the killing of a million people", who will go on to have healthy normal life expectancy afterwards. I figure there's still quite a lot of hot neurons on that connection, prompting the hissing and spitting every time "X killed Y people" gets applied to COVID.
You mean to tell me that there is a lack of critical thinking in the US. Or a majority of our citizens whom are totally dismissive of data driven decisions. Just trust your gut, huh? What would it take to convince the deniers of the 1.2M COVID deaths? My guess is there is no possible argument that will ever change their minds.
There were also medical establishments which intentionally underreported Covid deaths. A friend's mother, who was in her 70s, was admitted to her local hospital in Alabama with Covid. She died ~8 days later. Cause of death: pneumonia. My friend asked if her mother died of Covid and was told: no one dies from Covid. My friend didn't pursue it because she doesn't care what her mother died from.
That's just one anecdote. Were the underreporting and overreporting equal? I have no idea but the data which Scott refers to indicates that ~1.2mm Covid deaths is approximately accurate. Was the actual number 1.0mm or 1.4 mm? Maybe.
I believe covid easily caused the excess deaths mentioned. That said is there any factoring of the increase in all cause mortality caused by loneliness? As most people in lockdown were much more lonely.
Teen suicide went *down* during remote instruction! Whatever effect loneliness & isolation may have had, it was swamped by the reduction of in-person bullying.
Let's just put it directly: arguments are soldiers, and there's a lot of people who are so committed to the idea that the most important thing about Covid was that restrictions were bad that they are just going to never accept that Covid was actually deadly, since someone might use that as an argument that restrictions were good. They're going to just keep throwing up an endless amount of arguments of what *might* have happened (and that "might have happened" is always implied though generally not directly claimed to be equivalent of "did happen") instead of accepting the most obvious, Occam's-razor explanation that, indeed, it was the disease that mostly caused the excess death rates.
"Yes, Covid did kill all those people and yes, all restrictions were still immoral" is a rigorous and in many ways correct position (though at this point, this implicitly also is already close to the mainstream position - Covid still goes on and excess deaths still have larger-than-normal-flu-season spikes in winter, and no-one important is indicating that even larger death spikes would bring back the restrictions), but at this point even refusing to consider the idea that Covid did indeed kill pretty much the amount of people generally claimed to have been killed by Covid is just mendacious.
Furthermore, what burns me is seeing various and conspiracy theorists basically running victory laps in the sense of "everything the "conspiracy theorists" said was true! Everything!" whenever there's any indication of the mainstream accepting previously not-so-mainstream theories, like lab leak, when the common Covid-era conspiracy theory discourse was replete with predictions that didn't turn out to be true, like the vaccines basically being poison that will kill or sterilize hundreds of millions of people, or the restrictions and NPIs going on forever and ever as a tool of NWO slavery, or Covid being used as a reason for cancelling elections or so on.
> There are 340 million Americans, so if 1.2 million died of COVID, that’s about 1/300. This number - 1/300 - is also the prevalence of multiple sclerosis3. Do you know someone with multiple sclerosis?
I feel like you buried the lead here. Isn't the above the larger point, more important than the 1.2 million number? The per capita number is the meaningful stat, not the raw aggregate number. It just turns out that lots of things produce a death toll of 1/300, and it's not such a huge deal. Sure, if you multiply 1/300 by 340 million people, you get a big number, but that doesn't and shouldn't reflect our intuitions about seriousness.
Oh sure if some disease kills a million people then it's not a big deal. But if _I_ kill a million people then I go to prison. Where's the justice in that?
The whole point is that, given a large enough base population, *lots* of things kill a million people. Scott himself noted this, giving multiple sclerosis as one example of many. Are you freaking out about multiple sclerosis?
The question is not "Is a million a big number?" It's "How does COVID-19 rank among the various risks we're familiar with?"
Look, I'm not going to argue that people in general are particularly good at how they prioritise various different causes of death.
Certainly far more column inches have been spent on the untimely end of George Floyd than on the 72 people killed by bees or 20 people killed by cows in the US in the same year.
But a million excess deaths is a lot however you slice it.
But yes, Covid obviously did kill a ton of people. And I'm not sure why anybody finds that hard to believe. Most of us got it; as a young and healthy man who suffered through an episode, I'm not sure I would've survived the same infection as an elderly man with other health issues.
Wow, that does not look good. Any statistics from other countries? I checked my own country, and there is no similar trend there (raising, but continuing trend from before covid).
edit: Wait, what is this saying - there are more people who are registered with a disability or more people with a disability are able to work (due to more remote work?), seems like it's both? Is this just for St. Louis?
I think the context in which they offered it was in traditional academic papers looking very formally at effectiveness data.
Although I've heard skeptics cite it frequently, I've not heard any mainstream (pro-vax) folks tear into it.
My interest - I suspect that if that logic/paper were appropriate, it would apply to the conversation. If people die in a ~2 week window after vaccination (or anytime before 2nd vax, or 2 week after second vax, depending on study), those deaths would be attributed to "not-vax" group status because the vaccine hadn't been assumed to be at full power yet. That would undercount vax deaths, and overcount non-vax deaths.
For the folks who think there are attributable vax deaths, this is important because it doubly-changes the numbers (reduces vax efficacy, and increases non-vax excess mortality).
Is this paper not valid? Or is the effect assumed (by pro-vax) folks so small that it doesn't move the needle (either on vax-efficacy or on overall excess mortality) so the paper is effectively ignored by one side yet used by the other as a prime support?
I remember hearing the paper eons ago, but don't recall ever hearing a refutation.
It seems to me that few people give a shit about the numbers. Most are angrily, personally invested in their familiar set of beliefs about covid and about the stupidity and character flaws of people who disagree.
"Drug overdoses rose, but by a tiny fraction of the COVID death toll." Hrm, but do we know that? I personally had someone die during the pandemic but whether it was because of covid or poor health or a prescription drug overdose was never determined. The pandemic was very hard on people and all deaths does include those deaths that may not have had the floridity of a gun to the head.
I know people with MS and I know people who died of COVID. In 2021 I attended a dinner with the administrators of a large American university. The black women all had several relatives, including close ones, who had died of COVID. All anecdotal, but made me wonder at the time if African-American deaths were actually underreported.
Anyone questioning excess deaths directly caused by COVID needs to explain the first chart on that page.
None of the alternative proposed causes make any damn sense with the raw data.
This is the first chart that should be in the above post (as opposed to the above "excess" death chart referenced by Scott, which naturally creates questions about assumptions) - just show the raw data.
I guess one question I have that I think is the real reason people are so mad - well, one of them - about all COVID things is the sense that not only were the measures taken to combat the virus mostly useless in the end, but they exacerbated excess deaths. That's what I'm most interested in. How many of those rolling excess death rates are attributable to deaths of despair, alcoholism, suicides, etc., in timing with lockdowns or the like? Could be misreading the graphs I'm a tired dude.
At one of the Ivies (which I won't name because it will just make this post irritating) there was a debate club that kept track not only of people's wins by vote, but of their knockouts, so to speak, and also of their being "knocked out." The debate equivalent of being knocked out was a debater's admitting, on the debate floor, that he could not refute his opponent's argument, and that his opponent had in fact convinced him of his point of view. This public admission of defeat was called 'being broken.' Members who had broken opponents and had also been broken were most admired.
I think the norms that group had were admirable, and would have fostered excellent training of both minds and egos. I wish we had similar norms here.
Is there anyone who here was strongly committed to the idea that there were far fewer than 1.2 million covid deaths, and is now at least considering the possibility that they were wrong? In fact has anyone reading and posting about either or these related posts actually tried on the idea that they were wrong about *anything at all* covid-related? If so, I sure haven't seen many signs of it.
If practically nobody is changing their views in either direction? Then we are all so fucked.
As a way of modeling the capacity to be influenced, I'll disclose a way I have changed my view a bit. Various people have pointed out here that while the total deaths is a high number, the US population is a number 300 times as big. In other words, covid killed about one person in 300, which, as somebody pointed out, is about the fraction of people MS kills. Marinating in that info, I did move in the direction of thinking that in the big picture, covid was not all that lethal. In fact I'd say it was probably on the darker side of a gray zone when it comes to decisions of whether to sacrifice a lot of money, time and human well-being in an effort to mitigate its effects. (But I also think the US could have slowed it and mitigated its effects with far less collateral harms if it had approached the problem more sensibly, and been honest with the public.)
And for those who don't know: Despite all the chaos and suffering efforts to mitigate covids harms inflicted on its citizens, the US did a rat-shit job with covid. Only 16 out of 200+ countries had higher per capita death counts than ours. I am sure that some countries reporting very low death counts simply did not have the resources to keep track of covid deaths, and some are flat-out lying. But even after you take that into account by ignoring the data from those countries, we still look terrible.
MS is responsible for 0.04% of deaths globally (2017 data from wikipedia). Do you sum them over the years? But covid is not over.
Out of 200+ countries, most are nowhere near the US's league when it comes to the number of elderly people. Not saying the US did a good job but any rankings should take demography into account.
(Tangential: when it comes to maternal mortality, the US are now well behind most of Europe. And if you look at black women specifically, they're IIRC some 2.5-3 times worse off, on par with Kyrgyzstan or Mexico. So something about the US healthcare system does make it subpar for some tasks, compared to the countries of comparable income/HDI. Edit: Much of the elevated maternal mortality could be due to obesity. ChatGPT says severe obesity increases it x2-x5.)
You won't see any minds changed in the comments. People making comments are full enough of ourselves that we think other people should actually take the time to read what we're saying. The changed minds occur in lurkers. And they do happen.
If the guys in the Yale Political Union can enjoy debates in a setting where both breaking and being broken are admired, we can. Egos don’t come any bigger than the ones those guys had.
No one I know personally died of COVID. A friend of mine lost her father, she says to COVID, but she is a scientist so she has reason to lie about it. I was not with him when he died. I never met him. So how can I even know that her father died, let alone from COVID. Or that she even has a father.
My mother lives in a retirement community since 2022. The whole time she's been there, no one there died of COVID. She's still alive.
I got COVID and I was fine. I mean I am not even sure I had COVID. You can't believe those tests. I got the vaccine and over the curse of the next three years, my hair got grey, I gained 5 lbs, I stubbed my toe, I got in a fight with my husband, my alcohol tolerance went down, I had one episode of a-fib back in February (2025), I had trouble falling asleep one night in 2024, and some days I am really tired. Also my dog sometimes ignores me.
How'd I do? Do I sound like some of the other posts?
I’ve been through two pandemics, HIV and Covid. The rhetoric was identical, the conspiracies identical, the “where are the dead people” identical, the downplaying of disease spread identical. NEJM published articles by 1987/8 that HIV would be gone by 1990.
Humans are incapable of simple reasoning about anything with changing values except in linear patterns. Human eyes can track something moving but not something changing acceleration. We have no feeling for viral growth therefore it all is “fake”, made-up or false.
Compound growth whether personal debt or viral spread is not understood.
Always expect challenges to non-linear patterns and you will never be disappointed.
And I remember right-wing talkback hosts in the early 90s complaining about how anti-discrimination laws would prevent people from setting up AIDS-free towns and that we'd all be dying as a result when the virus mutates to become airborne or something.
Basically I think epidemiology is hard, people are stupid, public health messaging is optimised for something other than truth, and the errors aren't necessarily all on one side.
I wouldn’t know in the 90’s. My last close friend died when I lived in LA in late 1989. I decided to move on and got a place in Paris January 2000 and gradually people I met stopped dying.
I found out two decades later that I was immune (CCR5 Delta 32, homozygous; I have other strange mutations I’ve slowly realized).
The last person I knew with HIV died two years ago.
Fortunately for most people behaviors change very abruptly then slowly assume a new dynamic when a major disease hits.
Doctors don’t really understand the behavior of new viruses. They are new. Goes with the territory.
My colleague Billy Quilty had a thoughtful piece in 2020 when the UK reached the point at which most people would, on average, know someone who’d died from COVID. The key point, of course, is that affected communities aren’t evenly distributed - some were far more affected by others (and often those without the loudest voices): https://www.theguardian.com/commentisfree/2021/jan/22/silence-died-covid-figures-grief-pandemic
Orthogonal thoughts after reading some of the comments:
I think it's important to be able to separate being personally angry at something and thinking that something was unjust. I wasn't really able to visit my mother in hospital while she was dying (liver issues, not COVID-19), but I think it was generally right not to let me go there, it was the time before we had vaccines generally available. I'm deeply sad about it and I still haven't processed her death even though it's been multiple years; but I honestly also don't really know if I would have more closure if I'd been with her more often in those final days, we were on the phone every day until the end. But I was definitely upset that I couldn't see her - it felt 'unfair', but I had a lot of flexibility in deciding where to point my anger, and I pointed it at COVID-19 and the timing of it all, not at the people who were (to my perception) trying to do their best choices with limited data.
I'm still largely happy with the way my government handled things (Germany). I am pretty sure it wasn't ideal (Scott has posted other analytical articles, such as https://www.astralcodexten.com/p/lockdown-effectiveness-much-more, and generally you shouldn't listen to me about any of this anyway, I'm not even trying to convince you that Germany did a good job, which would be secondary to my point even if I believed it exactly like that, I'm just trying to convince you that it's possible to think people made wrong decisions and also not get angry at them for it), but I think honest efforts were made.
I realise some of the commenters who are saying "lockdowns were a big awful thing" are not necessarily also saying "and therefore the people who decided to do lockdowns are personally responsible in a morally reprehensible way," the claim can be much weaker, it can be "and therefore we can tell that the people who decided to do lockdowns were working with incomplete data." But I do see that this is sometimes lumped together, and I'd like to make an argument to be less angry at the people with the incomplete data trying to make decisions under pressure.
Caveat: "What about the lying?" As someone who habitually avoids the news I honestly can't say much about this (how much it was happening, etc). I especially can't say much about what was being said in the US. I got most of my information from either my biochemist girlfriend (who lives in Australia), or through three layers of indirection and refinement online (e.g. this blog). Sorry, I realise this is a point that has a lot of people really angry, but I just can't judge it at all. The closest I can come to this is that there was a ton of discussion about whether masking works and doesn't work, which through those sources looked like "things are changing as actual hard data emerges." I was super annoyed at the changes at the time (I wanted to have solid advice), but I cannot get worked up about it in hindsight.
(Maybe you're tempted to throw tweets and news articles at me to convince me that it was really bad. I'd personally advise against it, because my general attitude is "lying is indeed bad," but we'll probably get caught up in a lot of discussion about whether something is lying, strategic misrepresentation for political gain, strategic misrepresentation for damage reduction, confused misrepresentation (i.e. fumbling an explanation the person technically understands, but is simplifying in a way they're not cognisant is eliding something important), sincere misunderstanding (i.e. scientists say one thing, but you misunderstood what they say, and now you're telling other people what you (mis)understood), or some other category I'm forgetting about right now. If you think there was lying, take the short-cut and understand that I think this is pretty damn bad - but I also think you can be deeply angry at someone for lying about one thing but still be OK with how they handled other parts of the response, so I would probably continue to be OK with how people here handled other parts of the response.)
Also, as an SRE, +1 to people saying we could really use a blameless COVID-19 postmortem to learn from all the mistakes.
> I'd like to make an argument to be less angry at the people with the incomplete data trying to make decisions under pressure.
There's an associational problem where the people *informing* the decisions (the public health agencies, one assumes) get conflated with the people *finalizing* the decisions (in the US, mostly politicians), and then many of the latter going on to be absolute hypocrites- Gavin Newsom quite famously throwing a party while the rest of California was shut down and going to the beach, alone and socially distanced, might get you arrested.
Also from the US perspective, I think Fauci enjoyed being in the spotlight way too much and this contributed to that association. Among other messaging failures that had little to do with incomplete data and everything to do with incompatibly alien value sets.
> I think this disproves the claim that it was just normal deaths being reported as COVID deaths. What about the alternative claim - that it wasn’t COVID that caused the extra deaths, but various treatments - ventilators, remdesevir, vaccines?
I think that this weak claim, and the refutation thereof, masks a much stronger claim that goes entirely unaddressed in this post: what gets politely called "deaths of despair," or more bluntly, "people killed by the lockdowns."
We saw significant spikes in suicide and self-harm, in depression, and in drug use, and that's just the big obvious effects. Then there are the more insidious ones. For example, my grandmother died during Covid, not *of* Covid, but because she couldn't get into the hospital for her cancer treatments. (The hospital was not over capacity; they just didn't want to let her in because she might catch Covid!) She's by no means the only one.
So, the US suicide rate is about 0,014% per year. That is about an order of magnitude to small to account for excess deaths - and that is the total rate, the change in suicide has to be a much smaller number.
I think you will find that an analysis of increases in drug and alcohol related deaths, and reduced medical service will similarly amount to numbers that are much to small to account for the excess deaths.
> They’re pretty much the same, maybe margin of error of 10% or so but not consistently in any direction.
How are you getting this? Looks to me like reported Covid deaths is consistently under excess deaths for the entire time period, just timeshifted later by a few weeks in some areas.
> Do you know someone with multiple sclerosis? No? Then it’s not surprising that you also don’t know someone who died of COVID.
This seems like a bad argument. Multiple sclerosis is not necessarily visible, and people have a tendency to not publicly discuss their illnesses. It's a lot harder to hide "death" from the people you know.
Best steelman of the "excess deaths not due to COVID itself but the response would probably be something like:
1. Most died of COVID cases were died with COVID cases.
2. Excess mortality is due to lack of access/willingness to get health care, which tracked COVID waves due to correlated-in-time rises in distancing/fear or medical capacity shortages.
Even if this somehow worked out, it would probably require that the (2) involve almost entirely acute lack of care causing death (e.g. heart attacks) rather than lack of chronic care (e.g. chemo). Otherwise there would be a big lag in the excess deaths relative to the cases.
It's just so much more parsimonious to accept that a bad respiratory virus caused a lot of deaths directly.
Fwiw, I buy the death numbers. My biggest problem with the government response is the several steps along the way where they burned their credibility in response to public madness. Not a fan of Fauci. I was decent friends with a guy who died of COVID. He didn’t get the jab. He was Nigerian and just felt like they were lying because of the mask guidance. Couldn’t talk him into it.
I think Scott has inadvertently exposed the reason why the world has such trouble mourning the Covid dead and moving on. It’s because no one is sure what actually happened during the Covid years. It really destroyed faith in mainstream institutions. I hope we get some clear answers going forward.
> If people died “with” COVID, ie of normal causes like flu, cancer, or car accidents while only incidentally having a positive COVID test, then total all-cause mortality during the COVID pandemic would be the same as always. If people died “of” COVID, then total all-cause mortality during the COVID pandemic would be higher than usual.
No, that's not true. Or at best misleading.
Basically, all-cause mortality could have gone up because of the lock-downs (eg from lack of exercise or whatever).
You could say that COVID caused the lockdowns, and thus caused the excess deaths. But that seems misleading.
I think we have to work a bit harder than just looking at excess deaths to figure out how many people died of COVID.
(I don't know whether the lockdowns caused excessive deaths. This is just an example of indirect causation of deaths that we shouldn't attribute directly to COVID. Reality might differ.)
EDIT: ok, you seem to address this in the second section.
> But is this really so mysterious? There are 340 million Americans, so if 1.2 million died of COVID, that’s about 1/300. This number - 1/300 - is also the prevalence of multiple sclerosis3. Do you know someone with multiple sclerosis? No? Then it’s not surprising that you also don’t know someone who died of COVID.4
I suspect many people who died of COVID were elderly and didn't have many friends.
So there's clear selection bias. It's the usual result that most of your friends have more friends than you do.
I think it would be more useful to have data up until the present day.
I do not question that there were many more deaths due to COVID during the period, but this doesn't give us any information about how many quality life years were lost.
As everyone has to die at some point, there will be fewer deaths than there would have been otherwise in the following years.
If there is an extremely sharp drop that would indicate most of the people who died from COVID would have died from other causes in a short space of time anyway. If deaths are just slightly below where the expected trend would have been without COVID it would indicate that many of the people that died with COVID would have gone on to live considerably longer.
You're saying that it's not a given that in a counterfactual universe without covid, that the people that died of/with covid in 2020-2021 wouldn't have died at some point further into the future?
Or are you saying that somehow the interventions (quarantines, masks, school and business closures etc) somehow increased the mortality rate in the subsequent years to such an extent that it cancelled these out?
I.e. Person A died of Covid in 2021. In non covid Universe they would have died in 2022, so we'd expect to see a +1 in '21 and a -1 in '22. But person B dies in '22 from "Covid disruption", who would have otherwise gone on to live a long and healthy life, so overall we see a +1 in '21 and no change in '22?
That sounds very improbable to me. Does anyone have any actual data for all cause mortality from 22 to present?
In alternative universe Trump comes to power in 2021 and not being senile yet he turns America into communist country with devastating effects on people's life expectancy (like in North Korea).
See, this type of reasoning doesn't work.
>> Or are you saying that somehow the interventions (quarantines, masks, school and business closures etc) somehow increased the mortality rate in the subsequent years to such an extent that it cancelled these out?
Yes, I am saying that it is a very likely possibility. Why would you say it is improbable? It is strange that you are asking for this data because the data shows exactly that – greatly increase mortality after covid for unclear reasons.
Most people asking those questions have never seen or analysed anything. When I tell them exactly what are the data, they just are shocked in disbelief. I tell them no evidence that masks work and they are in so much disbelief (didn't Scott provide evidence that masks work? No, silly, Cochrane report clearly shows that evidence is of poor quality and even that shows no or little effect).
Another point – healthcare services were underutilized during covid. Yes, in some places and times certain hospitals were overwhelmed. But in general it was underutilized. It could explained increased mortality after covid because these chronic cases of non-covid were not properly seen and treated.
Supporting my statement that underutilization of healthcare during covid could lead to many more non-covid related deaths is the wonderful thread today by Crémieux:
In short: oversupply of doctors don't improve life expectancy but undersupply make causes people to die.
If someone had told me before covid that there will be pandemic, I could believe it. If he had said that during this pandemic, healthcare services will remain underutilized, I would call him nuts. And he had said that even after pandemic people would refuse to believe that such underutilization caused more deaths, I would think I am dealing with a crazy persons.
And here I am, in the world where all these 3 statements are true.
"In alternative universe Trump comes to power in 2021 and not being senile yet he turns America into communist country with devastating effects on people's life expectancy (like in North Korea).
See, this type of reasoning doesn't work."
>> My point was that everyone has to die at some point of something.
You can argue about numbers, but it is incontrovertible that A some people did die of COVID in 2020 and B had COVID not happened those people would have died from a different cause at some point.
> Yes, I am saying that it is a very likely possibility. Why would you say it is improbable?
>> I agree there were some tragic deaths that could have been avoided - missed cancer diagnoses in young people etc.
What I think is highly improbable is that the number of these deaths was anywhere near as high as you are suggesting.
> It is strange that you are asking for this data because the data shows exactly that
>> It's strange that I'm asking for the data? Why is that strange?
Around a quarter of all seniors in the US suffer from social isolation. There was also a clustering of deaths in retirement homes, where many people are known to receive next to no visitors. This, to me, explains some of the "but I don't know anyone who died from covid!" argument, which is already a poor one from a probabilistic perspective.
I actually know three people who were confirmed Covid deaths, so I guess that puts me above the trend for a 1/300 situation.
I wonder if folks just find it hard to believe that 1.2 million people could die of this thing, and society would not just move on as normal, but a huge contingent of folks would pretend it never happened.
>but a huge contingent of folks would pretend it never happened.
Is it that surprising? Public health decided to torch institutional trust. Society spent three months on semi-house-arrest before going insane. The groups that can make things stay in a news cycle were ultimately embarrassed in hindsight, and everyone else was so tired of the whole problem they'd rather just ignore it.
Lots of meat for interesting social observations by a talented writer known for interesting social observations, but no one is taking up the charge.
I know someone from rural Missouri who lost five family members between his and his wife's family – some deaths coming after the vaccine was available. All were morbidly obese, all refused to take the vaccine for political identity reasons.
My point is that the people posting on this forum might not have less contact than average with the socio-economic groups who got hit the hardest by COVID.
My dad's friend was the first reported death in Utah. He lived up the street from me. My friend's husband died a month before vaccines rolled out. His daughter is friends with mine. My grandma died from COVID. Granted it was in 2022, she was vaccinated, and she was in her 90s. My in-laws lost a number of friends in their church community. My high school friends lost their dad. He was so genuine and kind.
Also, funerals were not the same then. That made it even harder. People dying in isolation in the hospital, then friends and family not being able to gather the same to grieve the loss. Let's not do this again, mmkay?
I guess I post this to say that it amazes me people don't know a single person who died from it. Perhaps this is more about how we're not as connected to community anymore?
"One commenter gives a plausible argument for doing this: '1.2m dead is just not believable. I only know 1 person who died from catching covid'."
This is literally the argument of a toddler (lacking object permanence): Roughly 700,000 people in the US die of heart disease every year. I haven't personally known someone who died of cardiovascular disease in several years. By that logic, the CVD death stats are inflated or false, which is obviously nonsensical.
The effect of increasing birth rate 1940-1950 seems like an important part of the story. US Births increased by ~400k/y from 1940-1945 and then another ~800k/y from 1945-1950. The peak inflection of the Baby Boom was 1945-1946 at ~1M/y..so that's 75 years before 2020-21.
Very well presented and I think all the arguments you addressed you did so well. I would be interested in getting your take on how one could determine if the lockdowns themselves were the cause of the excess death. The argument being that the drastic and draconian lockdown measure contribute to increases of stress both physical and mental on the entire population. And doing so across a large enough cross section of the population would inevitably result in the increased mortality rate we see as it would essentially take anyone who was already in a borderline situation and "push them over the edge" so to say. This excess death is then attribute to Covid through whatever means and presto you have a non-virus explanation.
I don't know if the above explanation would be persuasive to anyone but it's a novel hypothesis I didn't see directly addressed above. I think what one could do is try and correlate the intensity of lockdowns in various states to the reported excess death one way or the other. That likely wouldn't be perfectly useful one way or the other but is a good starting point.
By my narrative, presented for reference without the expectation of being persuasive, the analysis of excess deaths compared to covid effects, which is the central pillar of the argument, is a reversal of cause and effect.
That is, excess deaths caused reported covid deaths.
The Wuhan Institute of Virology did exactly what it set out to do, which is to create a virus that would kill large amounts of old and infirm people who would naturally have died in short order anyway, and which could be used as cover for statist programs which would kill large numbers of people, and cause massive harm to the general populace.
I find that a lot of details are provable in a sense that the other side denies. Covid vaccines are deadly. Masking is deadly. Covid death counts are wildly inflated. Lockdowns and the accompanying economic damage are deadly. A proper analysis must start from these confirmed facts, and work out the remainder.
The reasons for these beliefs are out of scope, and available elsewhere. I claim that if you follow the evidence from the other direction, you reach a very different destination, thus the crux is whether physical claims and direct experiments regarding, say, mask efficacy are more trustworthy, or whether government statistics are more reliable. That is, internal bias versus external bias, aka, manipulative liars.
I do think here that, the local ingroup is confused about the outgroup. I see a lot of things that are treated here as reliable data, which to me, sound like pronouncements of a local priesthood. I can't confidently assert that my positions are entirely typical of the outgroup here referenced, but I hold that they're a lot more common than the other side expects, and treating "science" or institutional claims by states and associated credentialist entities like universities which can be easily falsified, as gospel truth, will continue to generate surprise when outgroup members, aka, conspiracy theorists, don't take such arguments seriously.
I'm late on this one, but I was an internal medicine intern at a large US academic hospital during covid. I did see many patients die of covid in the ICU during this time, and I think the characteristics of patients who died somewhat explain your phenomenon of not knowing anyone who died of covid.
1) They were mostly elderly, and skewed older even within that group. Like, picture 80 year olds not 65 year olds. I realize you and the commenter you quoted do know some elderly people, but I bet it's not very many. Most people who are not elderly themselves really only know their own elderly family members, but likely have many non-family contacts in other age ranges.
2) They were also mostly very unhealthy. I'm sure everyone has heard the association between things like obesity, diabetes, heart failure, etc and covid deaths already. But even among people with their diseases these patients were on the extreme end of doing-badly. Think a1c of 13, BMI of 45, or EF of 20%. Of the non-elderly patients I saw die they were all in this very unhealthy category. And I suspect that like the elderly, the chronically ill do not make up much of the average person's usual contacts.
3) A lot of them came from nursing homes. Should not be a surprise that there's a lot of overlap between the chronically ill, elderly, and nursing home patients. But it is another place where these people were essentially siloed off from the rest of society. They go back and forth from nursing home to hospital until eventually they don't go back to the nursing home. And unless you're in healthcare or they're your close family member you're probably not aware of them.
Every report, my personal experience and experience of others is in the same direction – that those who died covid where already quite close to death.
Also, the fact that risk of severe covid was exponentially increasing by age is in the same direction. The older or more sick someone is, the more risk he had from covid. But the same was true for risk of death from all causes.
It was as if mathematically the risk of death was suddenly doubled. Young people who has low risk, this doubling meant nothing, covid was less risky than driving a motorcycle or riding a horse.
For elderly people who had high risk of dying suddenly it was a big deal. If you had 10% chance of dying in one year, that would suddenly became 20%.
If your chance of dying within a year was 50%, it meant that you most likely died from covid.
People paid too much attention to ventilators. Great majority died without ventilators. They are not used that often at all. Even if we had no ventilators at all in hospitals, the death rate would not be any different. Even several thousand more deaths are rounding error in the million.
But we had ventilators and survival rate from using them was about 20%. It shows that we needed a better triage – using them only for those who are more likely to survive to increase it at least up to 50%. Maybe it was impossible, we had no diagnostic tools to understand who would benefit from ventilators more than others. I am just saying that creating panic that we ran out of ventilators was unnecessary.
This post missed two effects that I think are very important. First, lockdowns killed old people by keeping young people from becoming immune earlier, so old people were _more_ exposed to the disease before less deadly variants arose. Second, lockdowns will also kill young-during-Covid people later in life because lockdowns deprived those people of education and opportunities during a critical period of their lives, which will lead to crappier, shorter lives.
None of this is to say that it's not awful that 1.2 million people died, or even to question that number. I'm saying that it's facile to say that lockdowns weren't a major contributor just because the proximate cause was the virus, and we haven't seen all the damage of the lockdowns yet. Lockdowns were a stupid, deadly response to the pandemic.
One interesting thing was that Sweden had about the same amount of people dying from covid than in the UK (but much lower excess death) but also that their covid deaths did not form strong peaks, they were more equally distributed.
It shows that stricter lockdowns actually cause more peaks that tend to overload healthcare system. Lockdowns were sold with the promise to protect hospitals from overloading but they achieved the opposite effect. Hammer and dance indeed become a hammer and dance and more deaths.
Yes; the thing about lockdowns is that they *end*. Either officially, or when people stop caring. And then the disease spreads from its remaining reservoirs worse than it otherwise would have, because you've got a mostly immunologically-naive population all going out and engaging in higher-than-normal levels of social interaction to make up for lost time, all at the same time.
"Flatten the curve" was never going to work; that was I think known to be mathematically nonsensical from the start.
"Hammer and Dance", theoretically could have worked but in the way communism could theoretically work - the math is way too complex, the data is unreliable, and the system isn't responsive enough.
Locking down until either a reliable vaccine is developed, or the disease is literally extinct so there are no reservoirs from which it can rebound, takes too long - nobody in the West had the discipline or social cohesion to enforce a really strong lockdown for more than three months or so.
What *might* have done some good, is to realize that we probably could maintain a strict lockdown for two months without too much trouble, *once*, and that the best time to use that silver bullet would have been two months before we were ready to roll out the vaccine.
Perhaps the reason many people posting on blogs don't know anyone who died is that COVID mortality was highly correlated with other mortality causes which were already correlated with low socioeconomic status, diabetes, obesity, and so on. Based on my dear SO who worked on a COVID care floor in a hospital during 2020 and 2021, a typical COVID patient might be an overweight Hispanic man in his 50s, not a 39-year-old college graduate typing on a MacBook.
> but it’s only one of the many mysteries of suicides, which often fall during bad weather or social disasters
I don't know if this is original but now I have a theory. Suicidal tendencies correlates with how far you're different (below) from the current average. If the current average is rising (through industrialization or rising technology), you're more likely to commit suicide. But if the current average suffers (through disaster or just seasonal trend), you aren't. I'm pretty sure there'll be an easy data that immediately debunks this utterly, but I'm open for discussion.
I think the most plausible contrarian argument is that most of the people who died were old and fragile, so they would have died some time later anyway. We would need some "lost life years" measurement to measure the impact adequately.
COVID resulted in mortality indirectly as well as directly. It might, for example directly kill someone on whom another person depended, indirectly causing the death of this other person, even though the latter did not himself contract COVID. Mainly, it produced a remarkable social response, with people altering their behavior to avoid the risk of dying directly from COVID. The measures taken to reduce the risk directly from COVID probably increased the risk from other possible causes of death: thus there were extra deaths over and above the deaths directly from COVID.
This social response was likely suboptimal; still, these extra indirect deaths were *caused by COVID*.
The baseline excess deaths is 0 - excess deaths is how many more people died than in an average year. More people died in 2020 and 2021, and when they died it was marked as due to covid. What could possible be a non-Covid mechanism for this?
I'd strongly advise people to look at the CDC's graph of excess deaths per week due to all causes from 2018 thru 2023:
https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm
Most of the more popular theories are difficult to reconcile with this graph. I've been challenging people to come up with an explanation for over 3 years, and I haven't seen any rationalizations that I found persuasive.
You seem to be reading a lot into this post that isn't there.
The point of this post is that there is strong evidence that ~1 million people died in the US of Covid. It is useful to answer questions like this independent of surrounding context.
I think there is a subsection of the COVID truthers who say "it was just the flu" but really mean "there are trade offs between life and what makes life worth living, and we fell on the wrong side of the trade off" but I can't really respect the dishonesty. They are knowingly stringing along a lot of people who definitely don't feel the same way.
I don't think I defended the "general pro lockdown pov". I wrote:
> "Usually I’m the one arguing that we have to do cost-benefit analysis, that it’s impractical and incoherent to value every life at infinity billion dollars. And indeed, most lockdown-type measures look marginal on a purely economic analysis, and utterly fail one that includes hedonic costs. Rejecting some safety measures even though they saved lives was probably the right call. Still, I didn’t want to win this hard. People are saying things like “COVID taught us that scientists will always exaggerate how bad things will be.” I think if we’d known at the beginning of COVID that it would kill 1.2 million Americans, people would have thought that whatever warnings they were getting, or panicky responses were being proposed, were - if anything - understated."
I don't know how better to get across that the kind of lockdowns we did were probably an overreaction, but also it's insane to dismiss something that killed a million people as an overreaction that failed to pan out.
> Who cares if “1 mil is a big number” when divorced from population size?
I think it's useful to put it in the context of other problems. For example, are mass shootings a big problem? They kill about 100 people per year. So COVID killed 10,000x that. Is immigrant crime a big problem? A quick Fermi estimate suggests immigrants commit about 100,000 violent crimes per year. So COVID killed 10x more people than are affected by immigrant violent crime. Is cancer a big problem? About 600,000 people die of cancer per year, so COVID killed as many people as cancer during the two years it was most active.
Divorced from everything else in the comment thread here, I just wanted to say thanks for walking back on something in a public comment. It takes strength and I appreciate that you took the time to do that.
I generally wish we could praise and normalise this kind of thing more ('this kind of thing' being things like changing one's mind, admitting missing something, etc), so, in the interest of being the change I want to see in the world, I wrote this comment. :)
Man, now I'm kind of sad that whoever that commenter was, they wiped out the record of whatever they said and changed their mind about.
1.2 million deaths is about 0.36% of the US population of 330 million in 2020.
So, my new approach is to figure out how many years of lifespan covid cost Americans. Assume the average American is 38.7 years old and would live another 40 years and that there were 330 million people in the US in 2020. So that would imply the population at the beginning of the covid epidemic had 13,200,000,000 more years to live.
The highest estimate I’ve seen for how many Quality-Adjusted Life Years (QALYs discount years spent in comas, dementia, or horrifying pain, although not by all that much) the average covid fatality cost is 12.4 years. For 1.2 million fatalities, that would be 14.9 million years of life lost.
So, that would be 0.12% percent of U.S. lifespans lost due to covid.
Of course, the next question is: What would the toll have been if we didn't do X, Y, or Z?
But does "not doing X," such as not going to crowded apres-ski discos (big super-spreader sites in the late winter of 2020), means governmental decisions to shut apres-ski discos? Or does it mean people privately deciding to not go to discos?
And how big of a real-life difference is there between crowded discos shutting down by government order and discos shutting down because so few people are showing up its not worth paying the staff?
What it appears is that many people's preferred counterfactual is: What if nobody had noticed and/or nobody had cared about covid and thus kept behaving exactly as they had in 2019?
My guess is the death toll in 2020 would have been considerably higher.
But it's hard to realistically imagine a world where nobody notices or cares about a global pandemic.
Your guess that the death toll would have been higher is not based on reality, you are just echoing people who think they are smarter than everyone else. There was no chance that nobody would notice and no chance people would act in 2020 as they did in 2019. To be intellectually honest, you have to focus on things like this ---- People did notice that Cuomo's policies - brutal, penal isolation of old people as if they were criminals assigned to life imprisonment in solitary confinement, but close - within the same structure - to people who were dying from an infectious disease - killed many people who would have lived to see another day under a more compassionate government (and remember that Cuomo's government used threats of violence to impose its will). People also noticed that other people were being arrested for being alone with their family on a beach. This was an unimaginable disgrace to the people who ordered such arrests. I know you are a smart guy, Steve, but sometimes you look at the big picture too much, because that is what your pals look at (see, e.g., poor Murray and his attempts to count artistic value by the number of times names show up in mediocre art history textbooks). Yes there are stupid people who say there were no excess deaths and you can laugh at them all day but you are no better than them if you just focus on false and falsifying Leninist rhetoric about 'in the long run we were probably right'. It might be intellectually fun for you and others like you to brag about how much you understand numbers, and the big picture, but this was not a numbers game, and not a big picture game, from the point of view of the people who made absurdly bad decisions ---- and, because every problem in history is a problem that is based on particular details, and because we will never know how the counterfactuals would have turned out in detail, it is important to treat with respect those who disagree with you about the bad decisions made by people who were in the government and who backed up their bad decision with threats of violence, such as arrests, prison sentences, and unprecedentedly in the United States, threats of denial of basic decent medical support to those who dared to disagree with clowns like Cuomo and others like him.
I remember during the worst of it in late 2020/early 2021, the hospitals here in Utah were getting a bunch of Covid patients from Idaho and such because their hospitals had been completely swamped. Now I imagine what that would have been like if it had happened nation-wide over a week or two once the Delta strain got going (no more super-spreader events required).
I bet you'd have at least double the 1.2 million in excess deaths. Many of them wouldn't be dead from Covid itself (although there would be more of them with no free ICUs anywhere nor staff to surge locally from elsewhere in the US), but still dead.
Fwiw, I don't think the first article of the series was pro lockdown either. The point of *that* article seemed to be that 1.2 million American deaths is a lot of deaths and that it is an important and under-discussed fact. Sure, I suppose one could say that emphasizing the number of deaths is an argument a pro-lockdown person could use to support their case, but that isn't what Scott is doing.
In fact, he said "And indeed, most lockdown-type measures look marginal on a purely economic analysis, and utterly fail one that includes hedonic costs. Rejecting some safety measures even though they saved lives was probably the right call."
If you want to read something focused entirely on the lab leak hypothesis, I would recommend Scott's excellent review of the Rootclaim lab leak debate.
https://www.astralcodexten.com/p/practically-a-book-review-rootclaim
While I am no covid denier, I think you should really take into account all the people who died of other causes (heart attacks, cancer), because they could not or were afraid to access life saving care, including routine screening.
See also this, albeit anecdotal, comment: https://www.astralcodexten.com/p/the-other-covid-reckoning/comment/119032511
I can easily see, say, 10 per cent of the total excess deaths being caused by this, especially by the end of the period.
This. I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases.
I know a guy who died from a vaccine after white genocide.
That was me!!! I died from a vaccine shortly after they genocided me.
I’m the vaccine, all of the above is lies and russian propaganda!
Not sure this is knowable
Well, there was a strain on healthcare providers more generally. And that strain was caused by COVID. How many of the deaths caused by that were COVID deaths we don't know for sure but it was probably a lot and could well have been a million in the US.
If it's not knowable, even to a degree of uncertainty of approximation, then a meaningful cost/benefit analysis can't be made and the attempt to do so is misguided.
But it is knowable. Not only is it knowable, but the knowledge is easily available to just about anyone who's motivated to get at it. Here is a nifty chart I created tracking the top 8 major causes of death from Jan 2020 through May 2023 (first link below). (I've had problems with Google docs before, let me know if you have trouble accessing it.) If you need a basic tutorial on how death certificates work in the US, I've put together a very high-level description below the links.
If you open my chart, you'll definitely see that deaths due to diseases of the heart, malignant neoplasms (cancers), and cerebrovascular disease (strokes and such) would rise during peak COVID. IIRCC, COVID could only be listed as a contributing factor if the patient had been *diagnosed* with it within the 30 days leading up to death. Either these were due to COVID as a contributing factor outside the 30-day window, asymptomatic or perisymptomatic COVID as a contributing factor, or they were deaths because people couldn't get treatment.
HHS tracks ICU usage, and there were definitely people with non-COVID life-threatening illnesses who couldn't access an ICU during peak COVID waves. The CDC estimated that about 100K people died during the first 2 years of the pandemic because of insufficient ICU beds. We can sort of see this in their excess data. Second link...
Major causes of death 2020-2023
https://drive.google.com/file/d/1cFMYy2-_UYZz_2fWvxDynVfIAp52-WFX/view?usp=sharing
Excess Deaths 2018 to beginning of 2023...
https://drive.google.com/file/d/153SObMIzH4WBI-0cY0eA6Ars5B6pjL4e/view?usp=sharing
And here's how death certificates work in the USA...
1. Drs and coroners are required to issue death certificates. There's a standardized way of filling out a death certificates. On the certificate they list the immediate reason of death and the contributing reason of death. If someone dies of pneumonia while in the hospital being treated for COVID, pneumonia will be the immediate cause of death and COVID will be the contributing cause of death. While some jurisdictions allow a "cause of death unknown" option, that's rarely used.
2. There are 3,244 counties in the US. Each county has a health department. Hospitals, MDs, and coroners, are required to file their death certificates with their county health departments. County health departments are required to upload the data from the death certificates to the CDC. So, probably 99.99% of the deaths in the US have a death certificate with an immediate and contributing causes on it.
3. The CDC tracks this data. The CDC puts all this data online via their Wonder app, which anyone can access to download the data and/or create nifty charts from the data.
It's... hopeless, isn't it? No matter how clear, no matter how easy-to-access, no matter how often trotted out - the data are just ignored. After all, why know, when one can pontificate endlessly, make up clever hypotheses, and "just ask questions".
I CAN'T FUCKING STAND IT!
Thank you for your service!
Great comment!
nit:
>So, probably 99.99% of the deaths in the US have a death certificate with an immediate and contributing causes on it.
That sounds high. E.g. about 1% of the population goes missing at some point in their lives and are never found. Some may have left on their own steam, but I suspect that a large fraction are killed and not found (since that would be the same order of magnitude as homicides) - and these don't get death certificates.
We have a population of about 340 million people. You're saying 3.4 million will go missing and never be located during the course of their lives? Well, provisionally, if we accept that, and we take the average lifespan of a US person to be 77 years, that gives us about 44,000 people who will go missing every year — given the current US population. That's still a lot of people! That would be about 3,700/month, so I guess I could accept as a reasonable number. But...
It appears that the 1% number originates from the DOJ's NAMUS database, but it's currently offline. So I can't really evaluate its validity.
I see numerous media quotes surrounding that factoid, though. This one was interesting...
https://www.wvnstv.com/digital-desk/how-many-missing-persons-are-found-in-the-u-s-yearly/
> An NPR interview with Todd Matthews, director of communications for the National Missing and Unidentified Persons System in 2013 delved into the nature of missing persons reports and found the same thing.
> “You know, the missing touches everybody, I think. In 2012, we had 661,000 cases of missing persons; and that’s just from that one year. Very quickly, 659,000 of those were canceled. So that means those persons either come back; in some cases, located as deceased persons, maybe never an unidentified person; or just a total misunderstanding. So at the end of 2012, of those 661,000 minus the canceled, we had 2,079 cases that remained at the end of the year as unresolved.”
> As of December 31, 2021, the National Crime Information Center’s (NCIC) database contained 93,718 active missing person records.
> Children under the age of 18 account for 30,400 (32 percent) of the national recorded 93,718 and 39,114 (42 percent) recorded were children under 21 years of age.
According to the DoJ website these missing person records are never purged. They are retained indefinitely until the individual is located or the record is canceled by the agency that entered it. Some of those number date back to when the database was created. And if Todd Matthews is right, then we're only losing between two and three thousand people each year who never turn up. That's a lot less than 44,000/year.
Approx 2.85 million US citizens died last year. And say we had 3,000 unresolved missing peeps last year. That gives 0.1%. OK, so 99.9% instead of 99.99% of people end up with a death certificate at some point. ;-)
Thanks for keeping me honest! But color me doubtful about that 1% figure.
The comment to which I was replying was, "I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases."
The comment to which I was replying was, "I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases."
We can know a few things.
For example, this hypothesis suggests that the excess deaths should track COVID hysteria (and lockdowns etc). And it's unlikely that COVID hysteria exactly tracked COVID cases.
But the graph of excess deaths seem to track the graph of COVID cases fairly well. That's some evidence against the hypothesis. (Even if it's not enough to completely rule it out.)
The comment to which I was replying was, "I know a guy who died of cancer significantly earlier because he couldn't access the treatments he needed because of COVID. There are probably many such cases."
My sister died of leukemia last month. It becomes very apparent to those of us who have walked the cancer journey of the extremely long supply chains that are required to support cancer care. Example: In the fall of 2024, one of her chemo cycles was pushed a week, because the hospital had clamped down on “non-essential” fluid usage because of an extreme shortage due to a hurricane in Florida (we are located in Denver) destroying the primary distribution center for IV fluids, and fluids are required to flush the chemo to prevent organ damage (the irony here is apparent to me). Many such examples for just the supply chain, let alone the supplies and available hospital staff being diverted because of hospital strain.
I would rephrase that, since as stated, one who wants a high number for death toll could argue that those SHOULD be counted as "covid deaths." I'd instead attribute those deaths to "pandemic" measures, coördinated by global "health agencies."
Absent the NPIs, the number of such deaths (and associated morbidity) could have been worse. You can't meaningfully assess the casualty absent a fuller context. If someone strangles from a seatbelt in a car accident, how do you attribute the death?
For a more apt metaphor, if people were getting strangled while driving normally due to some new kind of seatbelt mandated due to OTHER people getting into car accidents, I would say attributing those deaths to car accidents is bullshit.
I don't see how thar would be an apt metaphor. Absent NPIs, there's a non-zero possibility that hospitals would have been overwhelmed and people would have been afraid to make visits to the doctor. Think of Bergamo, Italy, early in in the pandemic. The pandemic wasn't a function of "other people."
Nonsense. The data does *not* support your assertions. Read my response to Phanatic above.
Or people could just be dead honest and attribute every single goddam death,past present and future to "fools who don't share my opinions."
Let’s take out the “should” and just try to count these things in the specific groups, and worry about the “shoulds” later, when we produce several different interpretations of the same data. Do one where you attribute every single excess cardiac fatality to Covid, and one where you attribute every single excess cardiac fatality to NPIs.
I so agree.
We live in these complex interlocking systems and we're actually not bad at dealing with this complexity... when we're not insisting on reality being black and white for our own psychological comfort.
In fact, to get better at modeling complex systems (like a pandemic rolling through healthcare systems and societies across geographies and time), we need to accept uncertainty and some amount of approximation without freaking out and saying it's hopeless to understand anything.
Science has ever proceeded in this way. Imperfectly and unevenly. But that has never been justification to ditch science and the collection and analysis of data and just start speculating.
Who's the "we" you're talking about? "We" as in the denizens of ACX seem to have problems with complexity and "we" keep coming back to simplistic answers that satisfy our prejudices. "We" as in the scientific community (which does not include me), were actually quite good at modeling complex systems — especially when it came to healthcare outcomes of various NPI regimes and various different healthcare systems. Sure there's a lot of arguments about how to interpret the data, but the data is there.
And amen to this...
> Science has ever proceeded in this way. Imperfectly and unevenly. But that has never been justification to ditch science and the collection and analysis of data and just start speculating.
We as in humans as a whole.
I know one person who died from covid, but he was in a nursing home. A lack of visitors in the nursing home is, in my experience, very bad for health and mortality. Not only does one miss the benefits of social contact, but the nursing home personnel get the idea that no one cares about you, then you die. This was true for all nursing home patients during covid.
I also know one person, my neighbor, who died of cancer during covid. She was taking chemotherapy at home under the care of her elderly and overburdened husband, and almost certainly died sooner than she would have with proper hospital care.
For what it is worth, this is pretty much the conclusion that the article at one of Scott's links (https://mistybeach.com/mark/Covid.html) drew:
"The total 'excess' deaths roughly matches the reported Covid deaths. California reported around 50,000 Covid deaths in 2021 and saw a bit under total 55,000 deaths above those expected by the model. "
Note: I am the author of the document.
I think the entire argument is because of overly simplistic models. There's rarely one cause of death. It's usually a this + that + the other thing kind of happening, and if any one of them were missing, death wouldn't have happened (at that particular time).
So. COVID put a lot of stress on the system. Thus people died who wouldn't otherwise have died (at that time). I believe that it was extremely rare that it was the STRONGLY predominant cause of death (say over 70% of the reason). But I suspect that it was frequently the reason that (say) 40% of those who died, died at the particular time that they did. (The particular figures are just numbers pulled out of the air. Don't pay much attention to them. They are just there to paint the shape of this argument.)
Note that being shot in the heart is not sufficient to be the single cause of death. Andrew Jackson was shot in the heart in a duel before being elected president. (Well, it *was* with a black powder dueling pistol...small bullet, low power. But he lived through it.) Causation is almost alway complex. Sharpening things down to a single cause almost always cuts away crucial factors. (That's why designing good experiments is so hard.)
But why are you pulling numbers out of the air, when you can dig them up yourself? I refer you to <"Ta Da!">, CDC Wonder.
https://wonder.cdc.gov/
> I think the entire argument is because of overly simplistic models. There's rarely one cause of death. It's usually a this + that + the other thing kind of happening, and if any one of them were missing, death wouldn't have happened (at that particular time).
See my very high-level overview of death certificates in the US of A (above). Also, the models are *not* overly simplistic. While I have issues with standard SIR models, the signals in the data that were gathered as the pandemic progressed was pretty clear. And better yet, we can compare to the outcomes of different countries with different NPI regimes, because many other countries put their data online, too.
You (well, I) *CAN'T* pull them from the published data, because they only list a single cause of death. The documents always want to list a single cause, and that's explicitly the thing that I'm denying. (Admittedly, there's often a major contributor, but I'm asserting that a single cause is the exception rather than the most common event, or even *a* common event.)
Sorry for the delayed response. I poked around CDC Wonder last night, for the year 2020, I found that 378,048 deaths had an ICD code of U07.1 (which is code for COVID). Of these, about 94.5% included at least one additional ICD-10 code indicating other conditions, while another 5.5% listed only U07.1 without any other contributing conditions.
I'll admit it took me well over an hour to dig that data out of wonder. Their user interface is a little clunky. Does that answer your concerns? Because I admit I'm all Wondered out.
So, yes, just about no one *died* of COVID without some other ICD code on their death certificate. And, yes, that's to be expected. Especially since the frequency of comorbidities rise with age in the population. Yes, a perfectly healthy person is less likely to die from a pathogen than a person with comorbidities. But perfectly healthy people are less common than one would expect. According to a 2010 study a hair less than fifty percent of the US population has *no* comorbidities. And one quarter of the US population has multiple comorbidity conditions (MCC).
> In 2010, 49.1% of civilian, noninstitutionalized US adults had no chronic conditions, and 24.8% had only 1 chronic condition. One-quarter of US adults had MCC (Table 2); 21.1% had 2 to 3 MCC and 4.9% had 4 or more MCC. For both sexes, older adults had a higher prevalence of MCC than younger adults. Among adults aged 18 to 44 years, men were less likely to have 2 to 3 MCC compared with women (P = .003). For adults aged 18 to 44 years and 45 to 64 years, men were less likely than women to have 4 or more MCC. In contrast, among adults 65 years or older, men were more likely than women to have 4 or more MCC.
https://www.cdc.gov/pcd/issues/2013/12_0203.htm
Thank you. That's about what I expected, but I also expect that the original data (that the CDC data was compiled from) oversimplified the circumstances of death. If for no other reason, then it would take a HUGE amount of effort to dig those things out. (How do you detect arterial blockage in a corpse? I don't really think that's possible. Blood tends to coagulate after death. [I'm no expert, so I could be wrong here, but that's what I think.])
I also think that "no chronic condition" is a VAST overestimate. Try "no recorded chronic condition". Most of my life I've noticed chronic conditions that weren't serious enough to ask a doctor about. The same seems to have been true (my evaluation, I didn't ask) of everyone I've known well enough to have an opinion about on the subject. (So perhaps 8 to 10 people, but a 100% positive. [OTOH, it's easier to notice something like that if someone complains about it, so it's a highly biased sample.])
That's the most plausible counterargument I've seen, but it's worth noting that even shaving off 10 percent of 1.2 million leaves us with over a million.
any other drivers of excess death in the US in that period? fentanyl and immigration?
Immigration didn't surge in 2020.
no doubt the excess death models have that assumption
But why would we care about those when we could rely on your vague hunches?
Why know if you can hypothesize and leave clever one-liners? /s
The curves are similar for other countries, too.
Mortality due to drug overdoses is also tracked by the CDC. Deaths due to drug overdoses during the pandemic (including Fentanyl) happened at the rate of between 6000 and 6500 per month. Drug overdose deaths climbed until 2022, and started dropping off a bit beginning 2023. Last time I checked they were still falling.
Undocumented immigrants, tourists, etc., all get death certificates. So their deaths were also tracked by the CDC. But they're lumped in with everyone else. Additionally, death certificates do not include a field indicating whether the deceased was a U.S. citizen or of another nationality. Only "place of birth."
But there are studies that showed the undocumented immigrants died at twice the rate of people legally here. Given that roughly 0.35 percent of the US population died during the pandemic, and given that there were likely 10.5-11.5 million undocumented immigrants in the US during the COVID years, that suggests that approximately 77,000 of those 1.2 million deaths were undocumented immigrants. Roughly 6%.
Not likely. Most undocumented immigrants could be younger, healthier, working population and less of them would be old people in care homes. Even if they were dying double rate compared to people legally in the US, that would not be a big number.
Don't forget about exponential risk increase by age.
<Post exchange analysis of this thread. Substack sends comments to my email, so I have his pre-edited response. NoPie originally wrote: "I cannot believe that you still forget about exponential risk increase by age." NoPie was aggressively rude, I responded with his "I cannot believe with my own "I cannot believe.">
I cannot believe you didn't check your beliefs against the data before expressing your opinion.
I assumed the studies I checked would be normalizing their data, because all good demographers should age-adjust their datasets. The Minnesota study (third link below), shows that your expectations would be true for all ICD categories *except infectious diseases.* I found that to be interesting.
> Theoretically, there may be expectation of lower COVID-19 mortality among immigrants than US-born residents. Literature documents a healthy immigrant effect—where immigrants tend to be healthier than US-born residents (Kennedy et al., 2015).... In addition, in pre-pandemic times, better health outcomes among immigrants were consistent across all ICD-10 disease categories except infectious diseases and external causes (Aldridge et al., 2018).
Here are links to the studies. If you've got anything other than your supposition to counter with, I'd be interested in the some links to support your claim.
This is the study that I based my claim that undocumented immigrants died at 2x the rate of native-born US residents...
https://pubmed.ncbi.nlm.nih.gov/38703491/
This gives some different numbers but agrees that immigrants died at significantly higher rates...
https://pubmed.ncbi.nlm.nih.gov/38018488/
Here's the Minnesota study I quoted from...
https://link.springer.com/article/10.1007/s11113-021-09668-1
I don't argue against that undocumented immigrants (UI) might have higher rates of dying from covid (compared to legal residents in their age group).
But you didn't address the issue about UI age groups at all.
You simply did 2 × percentage of UI × 1.2 million.
Murder quite famously went up, but probably not enough to change excess death statistics.
Traffic deaths, on the other hand, probably did change enough. They dropped during the initial lockdown period then spiked, and haven't come back down all that much (while murder did drop back down, I think by late 2022).
From 2016 to present, every year has had between 36,355 and 43,230 traffic fatalities in the United States. The lowest year is 2019 and the highest year is 2021. 7,000 deaths is not enough to make a noticeable difference in the total excess mortality number.
Overdoses had been soaring, especially since fentanyl arrived c. 2015. Overdose deaths during the covid years were very high, although it's not clear what they would have reached without covid.
Murders and traffic fatalities shot up after May 25, 2020, especially among African Americans, but that was probably mostly due to the temporary triumph of BLM over law enforcement.
Note that these three causes of death hit younger people on average than covid, so they take away much larger expected lifespans per death. On the other hand, their total numbers are quite a lot smaller than covid deaths.
I will trust in your analysis, Steve - was the excess death caused by a coronavirus pandemic?
OD deaths were at approx 6K/month at the beginning of the pandemic (first months of 2020, and they rose to 6.5K/month by the end of 2022. They began to fall off in 2023.
The murder rate spiked in 2020 at 6.42 per 100,000 people, up from 4.99 in 2019. Approx 21,500 people. Meanwhile in the same year <checks the NIH website real quick> COVID-19 accounted 697,000 deaths.
Since most murders occur between people who know each other, we can assume lockdowns had a significant role in the rise in the murder rate. I haven't gone over to the FBI website to check that assumption, though.
So roughly 4800 extra murders happened in 2020.
And a quick Google shows that traffic fatalities went up from ~36,000 in 2019 to ~38,800 in 2020. About 2,800 more traffic fatalities during lockdowns.
These numbers are chump change compared to 697,000 deaths from COVID.
CORRECTION! There were ~375,000 COVID deaths in 2020. The NIH website gave me the numbers for a 20-month period from March 2020 through October 2021. It seemed too high. I should have read the report more carefully. But ODs, homicides, and traffic fatalities are still chump change, against total COVID deaths in 2020, but twice as valuable chump change.
Domestic violence murders weren't up much during covid, surprisingly enough. Instead, there was a huge surge in mass shootings at black social events after George Floyd's death unleashed the racial reckoning. Cops retreated to the donut shop, and so people felt more confident carrying their illegal handguns to parties.
I don't see that in the DoJ data. But I didn't dig very deeply. If you've got some DoJ data or data from other sources, I'd be interested in seeing it. There are hella lot of reports on domestic violence up on the DoJ website, though.
ChatGPT sez:
2019 Data:
Total female victims: 1,948
Killed by someone they knew: 1,773 (91%)
Killed by male strangers: 175 (9%)
2020 Data:
Total female victims: 2,059
Killed by someone they knew: 1,604 (89%)
Killed by male strangers: 197 (9.6%)
I didn't check into male-on-male murders, and what percentage where the victim knew the perp. But yes, you're right, it looks like 4K extra murders in 2020 were not due to domestic violence. But still, we're talking drop in the bucket to total COVID deaths.
Also, you forgot to mention suicides. At one point, suicides from the lockdowns were being blamed as a significant contributor to excess deaths in 2020. But that idea didn't pan out. The suicide rate actually dropped in 2020 from 14.5/100,000 in 2019 down to 13.5/100,000 in 2020.
so far we have: overdoses, traffic, violent crime, healthcare access, suicide, big increase in illegal immigration - anything else?
The answer is no, there were no other significant drivers of death during the peak pandemic.
Sure, but comorbidity was an issue. Americans have never been this unhealthy. And 0.3% hardly matches the nightmare scenario we were lead to believe. Moreover, it's probably why you use the word millions instead of 0.3%.
To be fair, 0.3% was exactly in line with many early estimates I saw. Most of the panic was because the broad CI included >1% estimates. It was always odd to me that people continued to implicitly rely on the broad early CI estimates long after better information came along.
Overall, I agree that not only are there a lot of risk factors in line ahead of COVID, but that improvements in those other health factors are probably a better focus for preventing deaths in the next pandemic.
I used million because that's the title of Scott's post. And more nightmarish scenarios were avoided in part because of the protective measures out in place. I agree some of those measures were overprotective in hindsight - but that was hard to know at first, and the main point is that this pandemic was very real and deadly and worth taking some precautions to minimize.
I was surprised to learn that "comorbidity" doesn't mean "another cause of death," it means "another medical issue," including high blood pressure.
Yes, if you portray COVID deaths as 0.3% of the population, it all looks rather innocuous compared to <gasp> 1.2 million deaths. But if we frame it another way, during the peak of the worst wave, we were losing the equivalent of eleven 747s full of passengers each day. We shut down US air traffic, brought the country to highest alert, and invoked our NATO treaty to get fighters from European members to help patrol our skies, all because four jets were downed and 3,000 people died. If you look at COVID deaths in units of jumbo jets, the pandemic was pretty frigging significant. But most people just shrugged their collective shoulders and complained about their freedumbs being curtailed.
As for comorbidities, it's important not to look at the US as a singular island of health patterns.
For instance, France, because of its higher rates of smoking (roughly 2.5x US), has a higher average rate of COPD, asthma, and lung cancer, related diseases than the US. Asthma, at least, was one of those comorbidities that increased the risk of developing pneumonia from COVID.
And in Germany, two-thirds of men (67%) and half of women (53%) are overweight. Hard to get an apples-to-apples comparison of obesity, though, because the EU defines it slightly differently from the US. But roughly 25% of their population is seriously overweight. Not US rates, but very high.
The most interesting comparison I took from the pandemic is that Finland has a very high overall comorbidity rate, but during the pandemic, it did better than Denmark, which has a population with lower comorbidities. Obviously, comorbidities are an important factor, but they were far from the only factor determining which populations were able to pull through the pandemic with fewer deaths.
Why not consider that pandemic restrictions may have reduced mortality from other causes. This also seems plausible. Less miles driven. Less trips taken. Less exposure to other illnesses. It’s not just a one way street.
Looked into this and it was a mixed bag unlikely to affect numbers much either way.
You could have figured this out with the article's central argument: excess deaths over normal, non-pandemic conditions. If you had enough lives saved by the lack of activity that it counter-acted a substantive proportion of COVID deaths, the excess deaths count would be materially different from the COVID deaths count.
Not quite. Since the article is using the excess death count to establish the veracity of the Covid death count, this would still need to be explored. But yes, assuming we know those two variables, no need to.
Among old people in retirement homes, access to family and social networks is, in effect, life saving care, and access to family and social networks was brutally denied to millions of such people for a very long time (in terms of their expected future lifespan, two years could have easily been more than half their expected future lifespan). This effect could easily explain hundreds of thousands of the excess deaths.
How is it life saving care?
Frankly, because no one fights for your life as much as your family. The other people are just paid to be there. If you take the most charitable view, they’ll do their best but they have a million other concerns. If your elderly family are in a period of prolonged medical vulnerability, you absolutely need to have a member of your family be their advocate in conversation with the system, because otherwise the system is often laggardly and obtuse.
I've heard that for end-of-life care distant relatives who didn't visit often and feel guilty are the most insistent on prolonging past the point doctors think is wise. But since these people are actually near the end of their lives (not like most people who died of COVID prior to being infected), there isn't actually that much time that the medical system can give the patient.
Not only that, but contact with family is literally what has many old folks going. Cut that off, and they have nothing to look forward to any longer and shuffle off the moral coil. Anecdotal observation but I have seen it enough to trust there is some validity to the claim.
But even more immediately, yes, lack of care. At the start of the pandemic in Canada some nursing home patients were literally abandoned by staff that fled. Found several days later covered in their own feces etc. Given their pre-existing conditions, such abandonment would cause a sharp decline in their overall health leading to much higher death rates.
Not all nursing homes experienced such drama but I'm sure care was compromised in a very high percentage, resulting in many additional deaths.
I'm sure a lot of this kind of second order effects happened, but again it's kind of ridiculous to suggest that it explains more of the surge in mortality than the respiratory virus that looks like the kind of respiratory virus that kills a lot of people. To me it mostly seems like people are unable to conceive a middle ground between the sniffles and the Black Death; since it wasn't the former, it must be the latter.
A close relative of mine died in a nursing home because there was not enough staff to ensure he was getting his meds (his condition was making it really hard for him to swallow). Previously, this was taken care of by his wife, but, when COVID started, she was denied entrance.
> when COVID started, she was denied entrance.
See, COVID was responsible for his death. Add that to the tally.
Yes in one sense, but we are trying to separately allocate deaths caused by our response to COVID.
Sorry, that was a joke. When I posted that, I thought it would so patently stupid that no one would actually say things like that unironically, but then I read other comments here, and no longer believe that.
It's amazing how much the "will to live" impacts whether you can recover from illnesses later in life. I wouldn't have believed it until I witnessed this phenomenon first hand.
My grandmother struggled with the sudden isolation during the pandemic. Her husband had recently died and she couldn't take care of herself on her own. The isolation in care facilities caused clear health deterioration for her. We made the decision that the increased COVID risk of taking her into our home to care for her was better than her dying alone. Within days of coming to live with us she became physically better. Her ability to walk and her general gait improved dramatically, her incontinence disappeared, she was happier, she reported less weakness and a greater ability to perform basic daily tasks. Her memory improved.
Measurable, tangible health improvements.
When circumstances prevented us from keeping her at home, she went back to a care facility. Her health rapidly deteriorated, she was quickly moved to a wheelchair, and she died soon after. She told me, "I don't want to live anymore," both when she was in the facility before coming to our home and afterward.
A friend of mine is a mortician. In December a few years back she mentioned she was going into their post-holiday 'busy season'. This was before my experience with my grandmother, so I was shocked that people didn't die at regular intervals.
She said, "No, a lot of people hang on until after the holidays, and then they just lose the will to live." Remember that when you're old you tend to have a lot of old friends, too. If all your friends die, but you don't see other old friends at the funeral because they discontinued funerals, that's a highly significant social factor that's being withdrawn.
I think Scott implied in this post that the 1.2M number must either be fully explained by indirect non-infectious factors, or we must accept the 1.2M number as a result of infectious spread. I doubt he would maintain that stance if pressed, but the defense he gives above strongly implies it. I think it's possible that a non-trivial number of deaths were the result of COVID policies that weren't necessary, and not the result of the virus itself. Those deaths may track with reporting because when we saw deaths go up we pushed harder on the harmful social policies. If the policy-related death number is, say, 200k that's a pretty big number that deserves its own reckoning.
Virtually all of the medical literature indicates having a social life and avoiding loneliness is the most important *health* factor for oldsters.
Is that really the most important? I guess other things pointed to as big causes of mortality (smoking, diabetes, hypertension, obesity) are likely to have started before you're elderly and there's less to be done about it now (although I suppose nowadays semaglutides actually can treat the chronically obese).
Once you're in your eighties, it really is isolation.
It's never the cause of death on the paperwork, of course. But isolate oldsters, and lots of them will die shortly thereafter.
Just think about the cliché of a person dying shortly after s/he loses a spouse.
My understanding is that men's lives shorten when their wives die, but for widows it's the other way around (sometimes phrased as men sucking the life out of women).
When you say, "hundreds of thousands", are you talking 200k or 800k? The former might be in the noise, but in the latter, you wouldn't see the clear correlative pattern in the data the Scott cites.
If your causal model is "intervention -> lack of access to families -> old folks die 'cause sad a la Padme", then you still need a model for the cause of the interventions. This was pretty clearly COVID cases and deaths, especially in the early years. But now COVID deaths are causing such a ruckus that people do a lot of NPIs, which brings us back around to "COVID is actually doing most of the people killing".
This would predict more deaths in retirement homes, while plain old viral disease would predict fewer deaths in retirement homes, so this hypothesis is testable
Yes, this is one of several testable hypotheses. For the record, 1.2 million deaths (and hundreds of millions of survivors) is a lot, but a good researcher can review the details of about 100,000 deaths - after spending a few days determining how to randomize the events leading to the deaths in question - in about a month of research. 'Covid deaths in the western world' is not an impossibly obscure historical event, and I trust that eventually medical researchers will understand what went wrong.
The way test the size of the two effects is by looking at when the deaths in old age homes happened: did they short follow the five main waves of covid or did they happen randomly throughout periods when they were denied visitation.
I can see either argument, so somebody who feels strongly in favor of one side or the other should test.
That is not a good test, because of course those who lost the will to fight Covid were not going to die from that lack of will when the waves of Covid were at minima, rather than maxima.
Why did loneliness kill in rhythm with reported covid deaths?
Right.
Steve, because, as you know (I know you know who Bruce Charlton is, and you probably have read his discussion of how flu was considered, fifty years ago, to be the 'friend' of the elderly), loneliness on its own did not kill, but loneliness combined with an attitude of "I don't care anymore" is what killed when old people whose endorphins were near zero got flu symptoms that they did not anymore have a great desire to struggle against (this is simple understanding of how humans work and I am surprised you said "right" to an obviously falsely posed question --- why did loneliness kill in rhythm with reported covid deaths).
I mean , you are too smart to need me to spell this out, but just in case you need it to be spelled out ---- of course the lonely people died more when the waves of covid were at their worst. Nobody said they predominantly died of loneliness, they died of not wanting to live when they were faced with a wave of covid.
Very plausible, but this might merely decrease the "Covid mortality" as a number. It does not decrease the overall "seriousness" of the pandemic, or reduce the reasons for e.g. requiring people to wear masks.
Scott's argument in the first place was, like, "remember the 1.2 million people who died because of Covid?" - and you can't deny that also those 10% of deaths were directly caused by Covid, can't you?
If you argue ".. but they didn't die of Covid, they died because our health system was unable to cope with the sudden outbreak of Covid" - yes, sure, but maybe quite a large part of those other 1 million US deaths (people who died FROM Covid) were due to the fact that the health system was not ready for the pandemic.
There are those who would quibble that it’s unfair to characterize the health system as “unready for the pandemic” as opposed to “unready for the panic stirred up for various non-medical reasons in the face of a mostly unremarkable disease”.
I don’t *think* I am one of those, but I’m apparently sympathetic enough to that quibble to point it out.
Do you remember when we were hearing about hospital beds filled and not enough ventilators? People were there because they were sick and having trouble breathing.
There was plenty of panic, too, but someone would have noticed if all the hospital patients had just been having panic attacks.
I don’t think anybody of note ever *denied* that there was a disease.
And I was talking about a societal panic, not claiming that anyone claiming to have Covid was actually having a literal panic attack instead.
As an example of the societal panic, I gather that in retrospect the consensus is that we used respirators rather too often.
But I’m not arguing with Scott’s overall point myself — I think it’s quite likely that American did have in the order of a million deaths that can plausibly be blamed on Covid itself. Whether the exact number is 750K or 1.3M is out of my wheelhouse.
I remember people being concerned about that. I also remember it not being an issue in a lot of areas (outside of NYC's initial outbreak), and that ultimately ventilators didn't matter in most cases. Most field hospitals that got stood up didn't treat a single patient: https://www.npr.org/2020/05/07/851712311/u-s-field-hospitals-stand-down-most-without-treating-any-covid-19-patients
A lot of areas kept beds open and reduced availability numbers *in case of* needing them, but ultimately never filled up. Source: family that works in hospitals.
The sheer volume of nurses recording Tiktoks while at work was taken as anecdotal evidence they really weren't that busy either, but it's not a great metric.
Several field hospitals set up in my vicinity, one of them a tent city in large parking lot. No patients ever arrived. On the other hand, hospitals were in financial distress and laid off employees because they didn't have enough patients.
I was surprised, but hospital administrators turned out to be surprisingly good at expanding capacity and dealing with the challenges of covid.
There was a huge covid death toll in New York City in March-April 2020, with 15,000 deaths in April alone.
But the "lockdown" started in mid-March started to work about one week into April. I recall a late-April 2020 New York Times op-ed by a NYC ER doctor. He remarked that background noise all through March and the first week of April was the sirens of ambulances arriving with covid patients. Then, IIRC, about 1pm on April 7, 2020, the sirens mostly stopped.
The health system spent a ton of pointless effort saving, say, patients from the local hospice instead of instituting an age/condition-based triage instead. Example: survival rate of 70+ year olds in ICU in 2020 was <50% but there were still cases where 35 year olds from (say) a car accident had to wait for an ICU spot.
We have a hard time rationing kidneys and livers; denying an old person an ICU bed because they have a baseline survival rate less than. 50% is a pretty tall order.
I was in my residency training during the early pandemic, I would encourage you to really think about how you would go about getting someone, say 72 years old, out of the icu or hospital to open up space for someone else. You get patient transport to bring them to a van and them drop them at home… and then what? They come to the Ed and you tell them you will stabilize according to EMTALA and then… send them home to do it again the next day? Essentially telling patients they should die every day at work would be a psychologically devastating experience for providers. It sounds nice on paper but makes no practical sense.
Also small quibble but I do not think you understand what hospice is
Is it less devastating to turn away the car accident victim in the example?
I think there is a mistaken image here of people being “turned away”- the car accident victim is also taken care of here. The issue is that if the icu is full (say of older Covid patients), they would be stuck physically in the ED where care is not going to be as tightly managed. Obviously also a bad situation, and perhaps outcomes are worse, but very different from telling people you won’t try to provide standard of care
How about moving the old Covid patient back to the ED and moving the accident victim to the ICU then? I've no idea if that would have made sense; I don't really mean to make any assertion about what happened or what should've happened as I don't know much about it, I was just going off myst_05's and your comment.
I didn’t say it would’ve looked pretty. But step 1 would be to suspend EMTALA and all other such laws until the emergency is over.
I do know what it is and I also remember crystal clear that there’s been numerous reported instances of people on their deathbed being taken to the hospital after contracting the virus during the initial March-May chaos.
I suppose I still don’t have a sense of how you see this playing out on a patient by patient level. If an older patient refused to leave the hospital because they were still critically ill, does security pick them up and carry them outside? And if this were logically solved, wouldn’t removing EMTALA encourage hospitals to kick out people with no insurance or Medicaid/medicare rather than the elderly?
If someone’s in a critical state how are they “protesting” anything? You move them out of ICU into the regular hospital or if that’s full, into a field hospital (kind of like the ones setup in stadiums and whatnot in 2020). If they’re violently trying to stop the doctors… there’s a whole array of medications at the hospitals disposal to make a person calm down, no need to make a scene. But the vast majority of those who’d be removed would be frail 80 year olds who’d otherwise have another 3-4 years to live, slowly withering away in a retirement home.
> wouldn’t removing EMTALA encourage hospitals to kick out people with no insurance or Medicaid/medicare
Some unfairness would be present temporarily yes. Thats what happens in a crisis.
That was the huge fear justifying lockdowns: that hospitals would be so overwhelmed by covid patients that other patients would deny for lack of care.
A couple of things happened, though: to my surprise, hospital administrators turned out to be surprisingly good at deal with major surges in case loads. And public and private responses to lower risks of infection proved successful and blunting the worst saves of infection.
My great aunt died from cancer in 2021 and it was specifically blamed on her chemotherapy being delayed by two months during the start of Covid - but this was in Russia with a much more cavalier attitude towards healthcare, so I don’t know how prevalent this was in the U.S.
It is certainly possible that people died because of Covid even without actually contracting the disease, but Scott's analysis compares Covid deaths to the increased mortality, and the numbers line up. Unless the numbers are being faked, it is hard to see how deaths from Covid aren't really about 1.2 million.
That's assuming normal medicine does save that many lives, which Scott has debated Robin Hanson about https://www.overcomingbias.com/p/response-to-scott-alexander-on-medical and on the specific subject of screening see https://www.overcomingbias.com/p/beware-cancer-screenshtml If hospitals are engaging in triage by cutting back on the least essential care, that could have minimal effect given how inessential so much medical care is in the US.
But there are also various pandemic related behaviour changes things that would make excess death go down. Such as fewer road deaths. Heck even suicides went down during lockdowns.
My memory is that traffic related deaths did not decline which was a huge surprise given decline in ave miles driven? Meme key ain’t what it used to be though…
Yes, the fewer people on the road were driving like maniacs.
Or they simply weren't stuck in traffic. Tough to have a fatal accident at 15 mph.
I just mentioned that potential cause in another comment and now I see you beat me to it.
Being the only car on 30 miles of four-lane is a really tempting experience.
My memory is that traffic related deaths did not decline which was a huge surprise given decline in ave miles driven? Meme key ain’t what it used to be though…
Routine screening rarely prevents deaths on the timeline you're suggesting. Emergent care for heart attacks, strokes, gunshots... maybe. But not routine screening.
When trying to estimate the effects of NPIs on health outcomes during the COVID years, folks often fail to examine the counterfactual of "How do people react in a COVID world where there's no government effort to encourage or enforce NPIs?"
As a result, it's rarely clear whether folks who claim NPIs are responsible for excess deaths (rather than COVID directly) are saying, "The Government erred in enforcing NPIs", or "The Government erred in encouraging private organizations and individuals to use NPIs" or "People on an individual level reacted irrationally to COVID" or something else. It's not exactly a motte an bailey; it's more just a lack of a precisely stated hypothesis.
I don't think there's any government (or "elite") response that wouldn't have resulted in massive disruption to all kinds of services, including nursing homes, which will serve for a useful example. If half your staff is out sick because they have COVID, and someone dies because they're receiving a lower standard of care, is that death "caused by COVID"? What if people called out sick but wouldn't have absent strict Government guidelines that were in place because of the pandemic? What if the guidelines were put in place by the nursing home w/o government intervention because they were leery of litigation caused by staffing with sick nurses during a pandemic? What about if the nurses were just very worried about this new disease and weren't willing to risk giving it to their patients?
My point is that there's a lot of causal models you can consider for how people could die during a pandemic due to indirect effects from it. Some of those pathways became more or less likely due to the specific set of NPIs, elite message, and cultural circumstances we found ourselves in from 2021 - 2024. But those specific outcomes can't be investigated in a vacuum without considering the counterfactual of what pathways become more or less likely with some different posited set of NPIs, elite messaging, and cultural circumstances.
So even if 120,000 people died from 2021 - 2024 because they were waiting for care, that doesn't answer the question of whether those deaths should "count" as COVID deaths or not.
Thank you. The lack of consideration for this counterfactual scenario blows my mind.
If we hadn't paid attention to covid at all, not much would have changed in our lives.
Some old and otherwise very sick people would have died more than usual but for a general person it is on the level that we barely happen to know such cases personally. Now we would be discussing and talking – there was a covid spreading around 5 years ago. Does anybody know who has died from it? Maybe an average person knew 1 or 2 old people from his family.
People working in healthcare would have noticed more people dying suddenly from a new type of respiratory virus. Some hospitals would have experienced overload, even more than they happen during nasty flu seasons. The number of deaths from covid as reported by Scott is real because people died indeed.
Provided that the government had allocated resources to deal with overload of hospitals and quick development of vaccine but without creating any panic, the timeline would be exactly the same except that we wouldn't have had any lockdowns, school closures or travel restrictions. We wouldn't have caused crisis in economy and the society would be more cohesive and less split about unnecessary restrictions.
Obviously, it mattered to people who died. Now I will saying something that will make blood boil to many people. Sweden's example showed that the number of those deaths would not be much greater in either case. In fact excess deaths would be lower all things considered.
If people were afraid of getting medical treatment and died of other things because of that, then wouldn't we see the excess deaths be less-closely related to reported covid deaths? Either it'd be correlated with lockdown measures (if people are afraid of being isolated in a hospital) or with news stories about how bad covid is (if people are afraid of catching covid).
I know someone who died of a heart attack during the pandemic, because he didn't seek medical attention due to fears of dying alone in a hospital due to isolation measures (so instead he made his wife wake up to him dead in bed beside her - not cool to do to a partner). So I'm not saying this didn't happen, but I doubt it was a significant fraction of excess deaths.
I have heard from more than one person who had to fight to keep COVID off a death certificate. "Oh, it was because of delayed treatment caused by COVID!"
Also, of course, there's nothing to prevent an original outbreak from being followed by a wave of deaths from the lockdown and other reactions to COVID.
Just like you "know" that a motorcycle death was attributed to COVID. But you don't know that lockdowns themselves caused any deaths at all.
Right here in this very thread we have counter-examples to your claim. So I don't see at all what you intend in posting such a falsehood.
But then lockdowns might have saved lives as well. We can't just go on a few anecdotes.
The claim was that it had not killed. It killed.
What counter-examples?
It would certainly be cool to see someone figure out:
* how many people died as a result of actually having Covid
* how many people died as a result of all the chaos that Covid created in multiple fields
Personally, I'd guess the second figure would be higher than the first.
My boss died from this. He had a heart condition that he had a surgery for that generally calls for three days of observation afterwards because of a known complication that sometimes occurs afterwards. Instead, he was sent home immediately after the surgery and died that night from the known complication. I guess maybe he would have died from it anyway, but presumably it's something that can be dealt with in the hospital or they would not normally keep them for monitoring.
It also presumably prevented some iatrogenic deaths, too. I have no idea how you decided on your 10%, so let's call it even?
"I can see" !== "I decided". A ballpark estimate.
As for iatrogenic deaths, I would expect more medical errors from overworked and overstressed docs, not fewer. Same with telemedicine, some things that have diagnostic value don't transfer across the screen (e.g. ease or difficulty of movement, various smells).
Keep in mind, absent NPIs those numbers could have been significantly higher. The point being they're still pandemic-caused deaths.
Spot on. While I understand there's some debate about the precise numbers of deaths (especially in countries with poorer record-keeping), it's frustrating to see people pretend that COVID wasn't a big deal.
I think that's a result of the overreaction and counterproductive measures (and sometimes just spiteful measures) put in place in response to the pandemic. The reality is that the pandemic was a really bad global tragedy that may have been completely avoidable depending on the source of the virus, and the response to the pandemic was a more or less global tragedy, maybe not as bad as COVID but maybe worse depending on how you count the economic impacts, that was completely avoidable.
We more or less aren't capable of having even slightly nuanced public debates, so I think to a lot of people conceding that COVID actually was really bad feels like conceding the argument before it begins.
I certainly wouldn't be surprised to learn, at some remote future time, that all the chaos that arose from Covid, in multiple fields, actually killed way more people than the condition itself.
Yeah we now all have Covid Rage Virus. It kills dialogue, fair-mindedness, goodwill towards others, common sense, friendliness, and some relationships. May be worse than covid itself
What's even more frustrating to me that even the contingent on board with it having been a big deal is lukewarm at best about ending gain-of-function. Sure, maybe this pandemic wasn't caused by a lab leak, but the next one very well might be, but nobody gives a damn.
People are allergic to nuance, is a lot of it. It’s not enough to criticize the response to covid, they have to take it to the extreme of pretending it was all fake news.
There's certainly room for debate about "with" versus "of."
There's the old meme, of course, of a guy halfway through his car's windshield, bleeding out, and someone with a microphone asking him "so when did you come down with COVID?"
But it's the same on the other side -- a lot of people on the "with" side seem to be doing the equivalent of "sure, an eight-ton rock fell on him ... but he had diabetes!"
My mother was in her late 80s in mid-2020. She had numerous health problems.
Then she got COVID and was dead within ten days.
Did those other health problems contribute to her dying instead of recovering? Absolutely ... and the death certificate says so, listing heart problems and high blood pressure as secondary causes of death. If I get crushed by a Mack truck while crossing the street, that was my main cause of death even if I was a walking heart attack / stroke risk.
Car accidents examples just aren't helpful here. It isn't relevant to the point people are arguing.
I was pointing to the silly extremes some people stake out on both sides of the debate.
If you don't find something relevant, feel free to ignore it.
I don't think such memes are actually worth bringing up. I haven't seen any actual evidence that traffic deaths were being miscategorized as caused by COVID, but people keep handwaving to that effect.
I, for one, find it frustrating to see the denial arguments equivocate between “covid statistics include deaths happening for unrelated reasons” and “clearly nobody is claiming this unrelated cause of death was being counted as covid, that would be silly and irrelevant”. I do realise it is different people saying the different things, but the overall combination is very draining.
And people also handwave that if there were any comorbidities, then it wasn't COVID that did the killing.
Those are the extremes. I agree that there's reasonable skepticism as to whether any particular death was "mostly" due to COVID or "mostly" due to something else. Some people disagree, on either end, and some of those disagreements aren't really very reasonable.
In my mother's case, there were comorbidities, but there hadn't been any recent emergencies with those comorbidities, so it's reasonable to think of COVID as the "tipping point" cause of death.
Excess deaths serve as a check on misattribution.
Yes, they do.
But, as with many issues, some people base their conclusions on rumor or assumption rather than on evidence.
https://www.fox35orlando.com/news/fox-35-investigates-questions-raised-after-fatal-motorcycle-crash-listed-as-covid-19-death
Thanks for the link, this was indeed documented.
You're welcome. Glad I could help.
There was at least one, and probably the most famous, but it was later removed: https://www.fox35orlando.com/news/fox-35-investigates-questions-raised-after-fatal-motorcycle-crash-listed-as-covid-19-death
How many others got misidentified and not removed because they didn't receive mass media attention, who knows. But I would be highly skeptical that the "COVID traffic deaths" would be more than a couple thousand at absolute most, and even that's on the very high uncharitable end of estimates.
My understanding is that deaths attributed to traffic accidents (rather than COVID) also rose, and the common explanation was that with less traffic cars were driving faster, while Sailer argued that after George Floyd there was a decline in policing (Scott also argued that, he might have just focused on homicide). I was going to suggest look at the total number of traffic accidents and see if the fatality rate per crash changed, but if crashes were up because vehicles were driving faster, then you'd expect that to cause an increase in the fatality rate by itself and it could be hard to detect a fraction being misclassified as COVID deaths amidst that increase. I don't think it's impossible though. COVID deaths vary a lot by age, in a different patterns to traffic & homicide deaths. Sailer has also written about different patterns by race following Floyd, and while there was a time when COVID deaths varied by that as well, I don't think that persisted.
I don't really understand what you're saying here. Let me pick a concrete example to discuss: someone is 82, and has COPD (a complication of smoking that leads to chronically poor lungs). They come down with COVID and die 8 days later, because of respiratory failure (suppose they weren't an ECMO candidate). Suppose that, if they hadn't had COPD, then they wouldn't have died. Are you saying that they should be counted as a COPD death, rather than a COVID death, because they wouldn't have died if not for the COPD? That doesn't make sense, because they also wouldn't have died if not for the COVID infection. So you could say that they died because of COPD, and they died because of COVID, but you can't say that they didn't die of COVID without also saying they didn't die of COPD. If you say they died of neither COVID nor COPD, then I'm not sure how useful your notion of causation is.
They should be counted as so-and-so percent of a COPD death and statistically, X number of people dying this way should be counted as a certain number of Covid deaths and a certain number of COPD deaths, where each one is less than X.
We're trying to figure out how many people would have been saved if we were able to prevent COVID cases. So what we care about is the number of deaths that happened, but would not have happened if the person involved didn't have COVID - our hypothetical person is an example of this. It's also correct to say that our hypothetical person should be counted among the people who would have been saved if we were able to prevent COPD, but I don't think anyone here is talking about the costs of COPD. The discussion here is about the costs of COVID.
(To use the car crash analogy, someone walking across the street absentmindedly and being hit by a distracted driver could be said to have died from crossing the street absentmindedly and the driver being distracted. In a discussion of the societal cost of distracted driving, this death should be considered as 100% part of the social cost of distracted driving. A policy that prevents distracted driving would have prevented this death, for example, even though a policy to prevent absentminded street-crossing would also have prevented it.)
A ridiculous comparison.
Track the royalties from the shot.
I think I need a little more than "someone somewhere profited from this pandemic" in order to entertain a theory where significant portions of the pandemic were faked. An obvious question is, for example, whether those selling the shots also faked the pandemic in Africa, which also saw significant excess deaths?
"...in order to entertain a theory where significant portions of the pandemic were faked."
And note that the faking and or conspiracy needs to be world-wide. Europe, China, Australia and New Zealand had to be in on it, too. As well as a lot of other countries. Maybe not Sudan.
Right? This argument not only proves too much: Is cancer a psyop by insurance companies?
Also disproves capitalism. The invisible hand of the market doesn't exist, it's shadow cabals all the way down.
Yeah. Did the weird lobbying corruption between pharmaceutical companies and politicians contribute to the rabid support in favor of vaccines and against literally any other possible cure? Almost certainly. Did it lead them to fake Covid or its primary attributes? Almost certainly not.
If I was a politician and had to choose between supporting either the methos of avoiding a disease or methods of curing the disease, I would, of course, prefer avoidance. Although I'd rather also support the cure, just somewhat less. Avoidance is better than cure because a disease, before it's cured, is likely to cause tissue damage in your body, much more than e.g. vaccines that also cause a little tissue damage.
An ounce of prevention etc etc...
this is really, really silly: vaccination is far and away the best approach humans have developed to addressing viral illness if you look at the track record of the past century. Anything other than prioritizing vaccines would be a very strange policy.
What other possible cures? Paxlovid?
“Against literally any other cure”. Like Paxlovid, developed by the same company that made one of the most popular vaccines?
Reminds me of The Man in the White Suit.
https://en.wikipedia.org/wiki/The_Man_in_the_White_Suit
But expanded to every market all at once.
Anxiously awaiting sci fi where we live in a post scarcity utopia, but for a hundred powerful cartels that are constantly injecting planned obsolescence into everything to justify their trades. But the one man not in any secret society discovered their secret and the whole house of cards is about to come tumbling down!
"Excess deaths" is not a measured quantity, to compute it requires a fairly sophisticated model of expected deaths. This, perhaps, explains why a number of countries, including Sweden, reported negative excess deaths during the covid pandemic.
If a novel disease arises there will be a large incentive to treat it. Whoever develops that treatment will profit. This is not a problem. It's the way the system is designed.
Should people not be paid for developing life-saving medical treatments? I have my criticisms of pharma patent trolling, but developing a treatment that craters mortality from a widespread disease which kills 1-2% of those infected seems worthy of compensation.
Criticizing the pharmaceutical companies for producing life-saving medical treatments is a really bad idea. IMO, the vaccine makers should have gotten a lot *more* money, maybe a percentage of the GDP loss avoided.
Obviously we need to take care not to incentivize them to produce ineffective or harmful treatments, but that part of the system seems to have worked quite well -- though I do wonder if perhaps in a situation where hundreds of thousands are dying every month we should find a way to reduce the required testing time a little more. The COVID vaccines were created in January of 2020 but not deployed until December of 2020, mostly because it took ~8 months to test them. With the benefit of perfect hindsight we can wish that we'd put them into production immediately and started vaccinating medical personnel in maybe March or April, and mass vaccinations in June or July. Of course, that would be foolish because they might have been ineffective or dangerous; testing is important. But we can and should ask how much risk would have increased with shorter testing periods and whether we should have traded that risk against the ongoing deaths.
In the future, as we gain experience with mRNA vaccines, I'm hopeful that we can establish baseline safety standards for the mRNA carrier component so all we have to do is differential testing of the effectiveness of the encoded antigens. That probably wouldn't reduce the time it takes to test for effectiveness, but it might significantly reduce the time it takes to test for safety (which has to be done before effectiveness testing can really be done). Also, using a standardizable strategy like mRNA might enable standardization of production facilities, enabling production lines to be converted in days, maybe hours. Similarly, it should enable standardization of shipping and storage. Taken all together, it seems possible that mRNA could make it possible to begin delivering vaccines for a new virus in as little as two months after the virus is identified and isolated.
And we really want to make that sort of rapid development and delivery highly profitable for pharma, because it would be extremely beneficial for humanity. Imagine if every doctor's office and pharmacy in America had ample stocks of the vaccine in March 2020. We'd have had no lockdowns, no masking, no kids struggling to learn at home, no supply shock... most everyone would have just gotten a shot and gone on with their normal lives, avoiding a million deaths, and trillions in economic losses and federal deficit spending. Pipe dream? Maybe not.
I also believe that we could have started vaccinate people against covid in July or August, 2020.
It could be with less known safety, more experimental while other studies are being done but for the most vulnerable it could still be a big gain. The benefit/risk ratio would have been quite high.
It wasn't done mostly because covid wasn't such a big deal. It was causing more panic than it deserved.
The UK removed covid from High consequence infectious diseases list.
https://www.gov.uk/guidance/high-consequence-infectious-diseases-hcid#status-of-covid-19
It is just formal decision but in practice experts in the field didn't think of covid as something that requires disregarding usual safety rules for vaccine trials etc. This highlights the disconnect of what the society including most politicians and rationalists thought about covid and what the highest experts in the field thought about covid.
Obviously, most people on this site disagrees with experts. But that's exactly the point, disagreement with experts due to emotions and not expert knowledge or thinking.
While you're at it, track revenues for vaccine deniers like Bret Weinstein and Russell Brand.
People can be stupid without being paid to be so. Comments sections & twitter are full of them.
Yes, these people are often known as fans. They are the ones paying Weinstein and Brand.
Very few people are antivax. Maybe 5% of parents don't vaccinate their children. We don't have antivax laws, yet. Who knows what will happen with Trump.
Whereas lockdowns were wrong and yet forced to all as the policy. Evidence deniers had won temporarily.
Instead of sounding like a literal strawman saying "DO YOUR OWN RESEARCH SHEEPLE", do you want to steelman yourself and provide some information about the royalties and how it goes beyond "pharma company makes new vaccine, gets paid"
Also, wouldn't the "covid is not real" argument apply to all other diseases equally? I mean, pharma companies get paid for those, too.
When people know a useful fact, they say it.
'Go research this phrase' generally means that the point isn't for you to learn a true fact, the point is for you to encounter the same propaganda and go down the same rabbit holes and fall into the same community that the speaker did in order to get them where they are now.
It's not a truth claim. It's a recruitment pitch.
Do you mean the Zapruder film, or the bullet itself that killed Kennedy? One of those must be the shot you are concerned with, right?
(But seriously, when I read this comment I was confused because it took me a few moments to parse this as about a vaccine rather than a photograph. Neither set of royalties causes deaths though.)
Good analysis, and far more than I would have bothered to do, which is “excess death rate, QED.” I think when numbers get large it is very hard for people to have an intuitive sense of what they mean, so they latch onto poor heuristics like “do I know anyone who died of Covid?” Relevant: the average American only knows 600 people (according to the NYT).
The average American may know only 600 people, but each of those people has ~2 parents, ~1 sibling, ~4 cousins, ~1 spouse. What's driving my skepticism on these numbers is that nobody I know in my office reported that their mom died of COVID, or that they needed a day off for their aunt's funeral. I didn't observe an uptick in "R.I.P. Uncle Joe" posts from my 300 Facebook friends. I'm not going to argue that my experiences are better than hard data, but still, nothing in my experience is consistent with the data.
How many people do you know that own a modern Chevy Malibu (ca. 2020+)?
I have no idea because I have almost no reason to know what my acquaintances drive.
This is Forbes data from Jan. ~1% of used car sales in 2023 were for the Chevrolet Malibu. More than 90% of American households have a car, and nearly a quarter (~22%) have three.
It stands to reason that a buttload of those households have Chevy Malibus, and I know I don't have the math skills to just figure out how that shakes out but I do know I see plenty of Malibus on American roads with me.
Shifting gears, I can say from experience in my work that, where we look for (among many other things) size differences between the pupils of the eyes. It's anecdata, but around 25% of Americans have a substantial enough difference in pupillary size that we have to investigate it. It nearly always ends up being nothing, a subclinical data point that enters the chart and sleeps there.
The point I want to make is that people's lives are full of facts and statistics, much of which is publicly available. We nevertheless miss that data or discard them because it's not relevant, apparently useful, or doesn't serve some immediate purpose. But, those rare things happens *all the time*, because there are so vastly many of us.
Did you know that migratory birds have been observed having an average of 3 snails on them? Over the long epoch of several million years, snails accidentally hitching rides on (or inside of) birds could explain much of the diversity of snails on Hawaii, a set of islands far removed from, well, anything. How in the hell would you know something like that unless you were A) a malacologist or B) very bored this morning while waiting for patients to filter in?
We only know this stuff because we take the time to look. It doesn't rise to conscious awareness because there's no obvious need for it, unless you're motivated to look or so bored that you'll read about snails.
Okay, but whether my secretary needed to take a few days off to attend her sister's funeral is something that I actually do have reason to pay attention to, unlike her pupil diameter or the number of snails on her boobie.
Some years I go to four funerals and others I don't go to any. The claimed increase in the death rate is 20%, or maybe a little less. I'm not convinced that this would be noticeable at a personal level, at least not over a short period of time. And then, as I get older I expect I'll probably go to more funerals anyway... Population statistics are very different from personal experience.
I don't know what the mean, median, mode, and standard deviation of funeral attendance in your life is, but I bet you'd be skeptical of the 1.2 million excess death total in a year that you went to zero.
If you average 2 funerals a year, then the chance of a year with none is 13%. If an extra 500k people died in the USA in 2020 then that means the chance of no funeral is 9%. A difference that is far too small to draw any conclusions from, on a sample size of one.
I agree with this entirely. Statistics that sound scary often don't translate well into real life.
In my own personal network. My kids' pediatrician died of covid. My husband's friend lost his father (not sure how old but I think under retirement age). Our family friends in Queens NY lost their grandfather to covid. I followed composer Sarah Duke on Instagram, her husband was young and healthy and was brutally hit by covid and she shared the whole journey until his death.
I remember vividly seeing pictures in March or April 2020 of a funeral home in Brooklyn with covered stacks of bodies because they couldn't keep up with the funerals fast enough. My social media feed was full of quiet small funerals outdoors from friends of friends who lost relatives. That was only the first few months of the pandemic really but it makes me baffled why people don't remember those days. It was publicly shared information.
I expect people know most of the 600 very tangentially, so that they are unlikely to hear about their family members dying. And of the people they know more closely, the 600 often already includes their family members.
The excess deaths were some 20% of the normal amount of deaths, so the uptick shouldn't have been obviously noticeable.
How many Facebook friends post about a death in any given year? It might be interesting to actually gather statistics on your feed, rather than going off a gut feeling about how many you think you see in a normal year, and how many you think you saw in 2020 and 2021.
There were entire subreddits devoted to posting the dying Facebook posts of “uncle Joe”. In a lot of cases the family would be completely unwilling to admit that their loved one died of Covid because they had for months been loudly proclaiming to their echo chamber how it was a hoax to begin with. No one wants egg on their face while grieving.
The fact is you probably do know folks tangentially related to you who had a loved one die of Covid, but the same social milieu that makes you likely to question the official figures would stigmatize advertising your loved ones death from the virus.
My observation was not that there were few people stating that Uncle Joe died of COVID. It's that I saw few reports of Uncle Joes dying. And given that my social milieu - Facebook in particular - comprises mostly educational (very liberal college) and professional (left leaning field) associates, it skews liberal beyond second-degree consanguinity. I feel like I was paying pretty close attention to whether anyone I knew had relatives dying because I was very concerned about my elderly parents and in-laws and my infant contacting COVID.
I'm not using my experience to dispute the data, but at the same time I'm not going to accept someone telling me that experience didn't happen the way I remember because of some aggregate data.
I’m not sure what your point is. If you don’t dispute that the deaths happened, then you need to accept that your experience is an artifact of reporting bias or exists as a statistical outlier.
Also, your Facebook newsfeed is not a random sample. Even if we grant that you were putting an insane amount of your attention into keeping track of the health status of your vague acquaintances, that information very well may not percolate its way to your newsfeed in the first place.
If my father had died of covid I certainly would not have told my "600 friends" because some percentage of them would have felt compelled to tell me covid was fake news and he couldn't have died of covid, then some other percentage would have felt compelled to scream at those people for reasons. Instead I would just have said my father passed away and kept the reason to myself and maybe privately told 2 or 3 really close friends that I trusted to accept it.
“I like these less because they’re class- and location- stratified, so your chance of knowing them goes up or down a lot depending on your own characteristics.”
But COVID deaths *were*, in fact, class- and location-stratified, certainly more than MS - further explaining how some people (think they) know nobody who died of COVID.
Example: my aunt worked at a nursing home in NY which had something like 16% mortality in the March - May 2020 wave. She knew like 30 people who died of COVID.
Most of the conspiracy theories at this point are elaborate and require a great deal of complexity. I am not sure there is much value in arguing against them.
https://www.astralcodexten.com/p/contra-kavanaugh-on-fideism
One of the worst parts of this community is how people will, without providing any context or explain how it is relevant, lazily link to an old post from Scott. Both lazy and condescending.
The post in question is Scott arguing that conspiracy theories are worth arguing against rather than ignoring, even if they seem very wrong. The current post is an example of Scott doing exactly that, and the commenter above is (much more concisely) saying roughly what Chris Kavanaugh was before.
I think that conspiracy theories are worth arguing against, but only by people who are already experts on something related. If the topic is new for you, it is probably not worth your time researching everything from scratch. You would basically have to out-learn the conspiracy theorists, while the expert is already ahead of them.
So I am happy with Scott posting against conspiracy theories, but there is no way I would try researching the same thing to make my own informed opinion on the topic -- it would take literally years to get to a useful level of medical knowledge, and I don't have that much free time.
Many of us in the audience both charitably grant the likelihood of relevance *and* don't find reading an old Scott post to be that much of a hardship.
Just because "many of [you] in the audience" approve of lazily posting a link and giving no context or synopsis whatsoever, doesn't mean it's good practice to lazily post a link and giving no context or synopsis whatsoever. This place is supposed to have etiquette above that of your typical subreddit; and "Scott said this, QED" is below the level of thought this place is supposed to exhibit as well.
uugr explained why it was on-point. I didn't say "QED", just as I don't say that for most of my comments.
A related argument (to the died of covid / with covid controversy) is that the death of an 86 yr old is not the same as the death of a 34 yr old: in the first case covid merely hastened what was inevitable, depriving the old person of at most a few low quality-of-life years. In a sense an old person is always dying _with_ whatever condition is the proximate cause of their death. It would be interesting to see how many quality of life adjusted years we collectively lost due to covid and compare that to car accidents or smoking.
You seem to assume there is a social benefit from years of college, but Bryan Caplan's "The Case Against Education" argues otherwise. Our own host has discussed how wasteful education for doctors like himself is quite popular even though the example of Ireland shows undergrad isn't necessary for them https://slatestarcodex.com/2015/06/06/against-tulip-subsidies/
My one link was to Scott Alexander's post about whether doctors need to go to undergrad first. I merely referenced Bryan Caplan, who was making an argument against his own interest as a college professor. If you want actual studies on the social vs private benefit of education, here you go https://www.econlib.org/archives/2013/10/international_e.html
The fact that you enjoyed college is not evidence against that. You might enjoy it if the government provided you with hookers & blow, but that would just be a pecuniary transfer for your consumption. College is heavily subsidized on the assumption that there are spillovers to the public good.
TR wasn't even talking about "social benefit." He was talking about quality of life. You're the one who brought up social benefit and it was a complete non sequitur.
Of course the social benefit is relevant to public policy.
"You seem to assume there is a social benefit from years of college..."
The two things being (implicitly) compared here are not two years of college vs. two years doing something else such as working or being in the military. It is two years of college while either (a) being on campus and interacting with the people there (profs, fellow students, police ...) or (b) doing this all via Zoom from the student's childhood bedroom.
(a) is a much better way to spend two years of one's life than (b). Especially if the student has carefully chosen the college to be near, say, a beach (e.g. Santa Barbara Community College, UCSB).
Fair enough, holding college constant but using Zoom could be similarly wasteful.
College, and young adulthood more generally, is when people are (hopefully) building the foundations for the rest of their lives: getting qualifications, gaining experience of independent living, perhaps even meeting their future spouses. Even if you think we send too many people to college -- and personally I'm inclined to agree on that point -- majorly disrupting someone's life at this stage is going to potentially have far bigger ramifications than letting his grandmother die at the age of 82 instead of 86, callous as it sounds.
And of course, it's not just college students who are affected. Anecdotally, lots of teachers in both the UK and the US report a big uptick in mental health problems amongst their students, and whilst maybe this would have happened anyway (anxiety and depression were on the rise even before COVID), it's at least plausible that spending several of their most impressionable years being kept away from their peers and told that they might end up accidentally killing granny if they went outside could have exacerbated the problem. For older people, lockdowns might mean a two- or three-year delay in getting married, which in turn might mean they're now too old to have children, or too old to have as many children as they otherwise would. And so on. Life isn't like a computer; you can't just switch it off and switch it on again and expect it to keep working like before.
Not only that but grandma's QOL during the additional 4 years takes a bit as well: she can't see family, she is essentially a prisoner in an institution, and instead of seeing her grandkids enjoying college and building a foundation for their adult lives, she seems them holed up at home listening to Zoom lectures.
This is correct and massively under-discussed. Extended solitary confinement is a terrible punishment imposed on the worst prisoners. Yet not only is it dismissed from the calculation of NPI costs, people actually pretend that the elderly should be grateful for it. Someone who's 85 has probably 5-7 more years left. Locking them down for a year or two to supposedly avoid Covid not only plausibly hastens their demise but also turns a significant fraction of their time into a life hardly worth living.
I view the idea that letting someone older die 4 years sooner than they otherwise would have (an underestimate to be sure, but we can run with it) is not worth the in-person schooling experience of a 19 year old as monstrously callous. It's strange to me that you take it as self-evident.
If you were told that you had a choice between a 19 year old having to do freshman year remotely at a college or putting a bullet in a 80 year old's brain, you'd think the choice is obviously do the latter because the value of an old person's continued life is minimal?
And who gets to define quality?
That's not what I meant. I mean in practice in the real world, who would get to decide that without creating incredibly terrible decisions. Even if everyone could agree on a definition that covered all use cases (which is impossible) there would surely be abuses of that system.
I agree. And there's a lot of fulfillment in living in your 80's and continuing to be present for your loved ones compared to some restrictions on the range of experiences in college one might experience. Saying "quality" doesn't mean your views on the value of a life are correct.
You have it inverted. I was saying in addition of the value of life to the the older person who is alive and still able to have fulfilling experiences, their continued existence is also fulfilling to people who care about them. It's part of the overall benefit of their not dying that you are weighing against the value of a more vibrant college scene.
Grandma not dying is of significant interest to grandma, but it's also of interest to people who happen to care about grandma. That's part of the overall utility calculation you are trying to invite. Why, I'd be willing to bet that odd as you might find it, there are lots of people who would be willing to trade off attending college remotely rather than in person for a few years in order for their grandma to live 4 more. This might be a foreign idea to you - "insane" in your own words - but that's an ordinary way to order one's desires.
I agree. It’s ridiculous that you think one year in college has more quality than four years in one’s 80s.
Agreed! Thank you for saying this! I have no problem with someone making the argument by posing the question of whether age should be taken into account when healthcare has to be rationed and under what conditions. but to claim that for anyone to say that "old people's lives matter" is insane and madness?
Plenty of old people live meaningful and fulfilling lives (my grandfather traveled the world throughout his 80s and was healthy enough to play baseball with my kids in the park at age 89, just saying).
But even if they didn't! At what age do human lives stop mattering? Why is a 79 year old more valuable than an 81 year old? A 75 year old? A 69 year old?
Should a 19 year old give up college to care for an elderly parent? People do that all the time. What if there's no one else able to care for them? Are they wrong? This just makes no sense.
This philosophy logically extended turns into a horrible dystopia very quickly.
I'm familiar with it thanks. You didn't address any of my points. I wonder if you're serious or trolling. Not because of your views but because you don't admit that it's your position which is outside the mainstream and requires a defense not the other way around.
Do you really believe that all 80 year olds automatically have bad quality of life?
Can you say something about how the repugnant conclusion relates to the debate of whether it is worth sacrificing in-person classes to delay the deaths of senior citizens?
Problems here:
* You compare "a 19 year old" to "an 80 year old," but in fact the 19 year olds in college way outnumber the octogenarians who died
* You assume that closing college is a 0%-100% choice in whether an 80 year old dies of Covid, when we don't even have any good evidence that it made any measurable difference at all
I'm following the comparison of the person who I replied to. I was not the one who introduced the direct comparison. I agree that it doesn't properly model the actual tradeoffs, but it's not my comparison to start with. The underlying issue is someone who decided that 80 year old lives aren't worth much and it is "insane" to think they are. They invite you to judge years of an elderly life against in-person attendance at a college and see that the latter is self-evidently of higher value.
We are not making a choice here between whether a 19 year old or an 80 year old gets to live. In reality, we were making a choice between a set of experiences young adults olds might have vs. some increased risk across the general public that they'll experience death or a significant reduction of quality in life that happens to increase significantly with age. There is lots of evidence that mitigative behavior collectively reduced said risk.
You should also reverse the scenario. If you were 80 years old, and you were going to die soon but you had a 1/6 chance (my upper-bound estimate of the ratio of college students to elderly people who counterfactually would have gotten and died from covid without lockdowns) of extending your life by a few years by making your grandchild do their freshman year of college remotely, would you do it? I would imagine probably not.
Of course, the college student who plays russian roulette with her grandfather's life so she can go party on campus for a year also comes out of this looking pretty bad.
I think this is just "selfishly imposing costs on others for your own benefit is bad".
The scenario already allowed for the reverse, since it was presented simply as a choice for a 3rd party. The participants are not the ones imposing the consequences. Kill the 80 year old who would live another 4 quality years is the meaning of "soon" and the choice was doing that or not compared to the offered tradeoff. This is not realistic, but it is the contrast that was set up by the person I initially responded to who was trying to communicate just how worthless years lived beyond a certain age are.
If we change the scenario entirely in the way you offer, someone extending their life four years by transforming a young person's in-person schooling to remote is defensible and is ultimately a vastly oversimplified version of what public health in the face of pandemic has long done with various forms of quarantine.
At 1/6 (my absolute upper bound on the estimate of the number of life years saved by lockdowns per year of remote schooling for university students) it is defensible, and approaching an absolute slam-dunk win from a public health perspective. At 1/1000 (my lower bound) it is not defensible.
I personally think the public health people made the wrong call with lockdowns but I don't think their call was *drastically* wrong (though separately I *do* think they made terrible calls with what particular things were allowed and forbidden).
Part of the problem that occurs in these discussions is people have ambiguous use of the term "lockdown." America never really had "lockdowns" in the formal sense at all, but the term was picked up early in the pandemic as it was describing the policy of significantly more restrictive nations. The closest the US came to lockdown-like behavior lasted for a brief period in the spring of 2020.
That said, we did have a series of both formal restrictions and informal recommended behavioral changes that were mitigative, but inconsistent with each other and suboptimal. This is what people usually somewhat unartfully call "lockdown." Slapped together, these collectively were effective enough to keep the seroprevalence down until vaccination rates climbed in a way that saved a very large number of lives and long-term health problems. At the same time, they were drastically inferior to what a nation like New Zealand was doing and the gaps likely cost a large number of lives and caused a lot of heartache among the still living. The US response sat in the zone of "could be a lot better" but also "could be a lot worse." It's sad when you think about the conventional wisdom prior to this was the US and its once vaunted CDC were among the best situated to weather a pandemic.
If you were to knock out any one pillar in the US strategy in suppressing community spread, in isolation it wouldn't necessarily have moved the needle a large amount. Even if all the legal restrictions were removed, the specter of death would've still had many people behaving with more caution. The problem with trying to isolate any one thing and estimate its contribution, however, is that mitigative recommendations were interlocking. They contributed to each others' effectiveness and communicated a sense of urgency and danger that unto itself also likely influenced propensity to safe behavior. An answer to any one particular "was this necessary?" is generally, "sort of, yes" with an added, "and you can only pull so many blocks out of the jenga tower."
> I would imagine probably not.
Your imagination apparently works differently than mine! I usually try not to speed in residential neighborhoods, even if the people are old, and I assume even old residents would like me to do that.
... I also don't speed in residential neighborhoods. I don't think I am understanding how your analogy works - are you saying it's sufficiently selfish for old people to be against speeding that it should outweigh their own concern for their safety?
Or do you object to the more general idea that we tend to take a dim view of people choosing the side of the tradeoff that benefits themself, even if the benefit they get is larger than the costs to others (less so if the cost is trivial and the benefit is large, as in your speeding example but not as in the original example)?
Imagine telling an 18-year-old: “If you do your first two years of college remotely, you’ll live four years longer.” They say, “Yeah, that sounds like a fair trade.”
What mental illness are they suffering from, exactly?
"It would be interesting to see how many quality of life adjusted years we collectively lost due to covid and compare that to car accidents or smoking."
I did a back-of-the-envelope calculation for this a few years back with the intent of comparing Covid to other types of deaths that people are familiar with. I did this using the California excess deaths per age group and then combined it with life expectancy for that age group.
Going from memory, Covid (at the time, but we were fairly far into it) cost the average American about 5 weeks of life expectancy (more death amongst the elderly, but the 'cost' in years was lower ... interestingly the lost years were pretty flat between most age demographics [maybe within 2x]; fewer 25-34 year olds died, but they lost more years of expected life when they did; I didn't try to adjust for health ...]
I was unable to find a reasonable comparable, though. This also works out to something like a "double your miles driven per year for 10 - 20 years" risk, which I find to be unhelpful.
Smoking for a lifetime probably, on average, costs you more years of life than Covid. Driving fatalities per year are much less. I don't know of anything 'common' in the middle.
Ooo, I wonder about the lock downs vs quality of life years.
E.g China had strong lockdowns and only 5,272 listed deaths.
India has a similar population and 533,570 deaths.
So we could say that China saved about 500k lives.
I asked Perplexity to work out the lockdown duration in person years. E.g the population affected by the lockdowns with the duration applied, e.g 10 weeks in Wuhan.
It gave 14,431,091 + 573k (it didn't think it had strong data for).
So just over 15 million human years.
So to compensate the people who died would have had to have lived 30 years in average to have made it one to one.
However life under lock down isn't the same as years not lived.
Personally I'd just had a new child and work from home. So we weren't going to be going out anyway. My memory is that we had maybe a 5% quality of life reduction due to the intense lockdowns in the Philippines. It was worse for 2x two week lockdowns where I had to line up to get food from the supermarket and apart from that couldn't go outside.. I couldn't go for a run outside and had to do beep test style running back and forwards inside. But when partly opened up there was almost no car traffic and running on the streets was way nicer than now. Less traffic and pollution.
Wearing a face mask is something we were already doing outside. Wearing a face shield for a while was a little annoying.
However my wifey's Mum had a stroke and then caught COVID in the hospital and died 2 weeks later whilst still being ravaged by COVID (delta at the time, it hit us HARD).
Thankfully Perplexity is better at researching than me and suggests the China style lockdowns were worth a 30% Quality of Life reduction. Although the global average was 12-18%
So 30% of 15 million is 4.5 million years.
Now the 500k people should have lived at least 9 years or more before it turns out to be a net positive.
Perplexity suggests that the Quality is life years lost in India was 8 million.
Or to put that another way, the lockdowns in China were 56% (basically half) the list quality of life years as compared to that of the deaths in India.
But it's different compared to USA stats.
> The U.S. COVID-19 death toll of 1.2 million resulted in 9.7–13.2 million years of life lost, depending on adjustment criteria.
The USA lockdowns affected 27.9 million person-years.
The USA quality of life adjustments of the lockdowns are listed as 15%
So 27.9 x 0.15 = 4.185 million years of life adjustments.
So the death toll was over double that of the lockdown toll in terms of quality of life.
They could have certainly pushed lockdowns harder and faster (as part of the Hammer and Dance) to reduce the death toll related Quality of Life years.
It feels like to me that normalising to Quality of Life years for the lockdown vs death toll is a good way of post-hoc analysing the results and comparing.
There's of course going to be those people badly affected by COVID lockdowns (e.g Travel agencies or resorts) who probably didn't know anyone or almost no one who died.
The lockdowns also disproportionately affected lower income people and those in less developed nations, like Bangladesh. Whilst the virus more targeted the elderly.
However countries applying counter measures for this, like providing stimulus checks seems to have somewhat offset the cost imbalance. Or those like in Sweden were the least affected because of the good social support.
My Perplexity thread of research:
https://www.perplexity.ai/search/how-much-time-in-weeks-was-chi-rAMs0_uQQnyLuza10MuJXQ
China and India have similar populations but very different demographics. ChatGPT thinks there are 280M vs 159M people over 60, and ~55% vs ~35% of 40+-year-olds. So I think the expected number of covid deaths in China with Indian levels of lockdowns should be at least 1.5-2 times higher than in India.
I remember a back of the envelope calculation that suggested it was pretty comparable in QALY impacts to the opioid/fentanyl epidemic (fewer deaths [though still 100's of thousands], but more QALY's lost per death) over a similar time frame. Not sure what to think of that in terms of our response to either of the problems.
I asked 03 and it said:
"Bottom line: Using CDC opioid-involved overdose‐death counts for 2020-2024 and the most recent peer-reviewed estimate that each fatal opioid overdose removes ~38 healthy life-years on average, I estimate the U.S. lost ≈12.7 million quality-adjusted life years (QALYs) to the opioid epidemic in the five years 2020-2024. Even under conservative assumptions this burden remains above 11 million QALYs, underscoring that opioid toxicity has erased more healthy life in this short span than many leading chronic diseases."
That basically matches the estimated 12 million QALYs lost to COVID.
O3 estimates 12 +- 1 million QALYs lost to Covid. That’s 10 QALY per death, which seems reasonable if it was mostly old people dying.
Interestingly that almost exactly matches the O3 estimate of QALYs lost to the opioid epidemic over the 2020-2024 period. Roughly 1/4 the deaths, but about 4x the QALYs lost per death.
"Bottom line: Using CDC opioid-involved overdose‐death counts for 2020-2024 and the most recent peer-reviewed estimate that each fatal opioid overdose removes ~38 healthy life-years on average, I estimate the U.S. lost ≈12.7 million quality-adjusted life years (QALYs) to the opioid epidemic in the five years 2020-2024. Even under conservative assumptions this burden remains above 11 million QALYs, underscoring that opioid toxicity has erased more healthy life in this short span than many leading chronic diseases.""
And the opioid epidemic was/is a very big deal! I wrote about it challenging my libertarian-leaning perspective https://entitledtoanopinion.wordpress.com/2020/08/11/a-contrary-perspective-on-the-opioid-epidemic/
I rather doubt this number. Average age of US covid deaths was a little over 77. The average life expectancy of any 77 year old is around 10-12 years (depending on male or female). But covid deaths also disproportionately hit the sickest with many co-morbidities. I find it hard to believe that the average 77 year old who died of covid had an equal life expectancy to the average 77 year old in the general population.
Perhaps the QALYs lost by people who did not die make up for the difference? At any rate the order of magnitude seems correct.
It's probably within an order of magnitude (10x), but I'm not convinced it's particularly close like within a factor of 2. Just as an example, around 200,000 of covid deaths were nursing home residents. The average life expectancy for any nursing home resident is 1-2 years (all, not the sickest and weakest which you'd expect from the covid cohort). Over half don't make it 6 months. That's a lot to counterbalance with the remaining deaths.
That's why the average amount of life lost per COVID death has been estimated at around a decade. Certainly not as bad as Spanish Flu, which killed the young more than the old. Won't we be screwed if something like that happens again and people continue to be so stupid in their response.
It’s so weird to me that people see this as a case of leadership being stupid - it’s rather a case of transparency causing all the problems. Leadership did as well as in any crisis in all of human history, but because we see there were costs as well as benefits, people have decided that we don’t want leadership any more.
Thank you, these are the words I wanted.
There is a basic immaturity to me in the view that doesn't accept:
1. the profound imperfection of human institutions across all history and places;
2. that uncertainty is a basic feature of existence that all the lovely leadership in the world cannot manage it away.
There is so much room for criticism, but the value of the criticism is so that we respond better then next time calamity hits.
Raging because the data is imperfect and complex or because the guidance in an emergency was contradictory or that some measures were counterproductive is to rage at reality as it is.
So many people worked so hard to take care of so many other people during this mess. It was traumatic and people are still recovering. From trauma, we can grow or we can freeze. We really do have some choice about that.
I get that this pandemic and the handling of it -- like 9/11 did for some other people or Watergate and Vietnam before them, and so on -- led a lot of people to lose faith in all kinds of things. And that losing faith (along with the willingness to hang in there with nuance) is a normal kind of grieving process.
But persistence in the belief that the world *should* have done more for me than it has is ultimately a childish view that needs to be outgrown in order to live the life we do have.
I think this is also the case (at least as Americans) where we really suffered from having 50 states as laboratories of democracy because people can point to ineffectual leadership on both sides to find reasons for their side rather than looking at the cases where the leadership pursued level head decisioning and made reasonable tradeoffs.
13.5 million life years, as I posted yesterday.
Interesting. I wonder how many life year-equivalents were lost due to lockdowns? 300+ million people locking down for a few weeks would surpass 13.5 pretty easily, but we would likely need to do some kind of adjustment for actual death verses the significant loss of [enjoyment, freedom, career advancement, etc.].
This is so strange. At literally NO point during Covid was I 'locked down,' and I lived in a liberal college town. The way people talk as though everyone was under house arrest is literally not true. It's so weird to me.
So in my state, entire industries were literally told they could not operate. Restaurants were either take out or closed. Other states forced the closure or churches and other public accommodations. Lots of people lost jobs over this, including a close family member of mine. People got arrested for going to the beach.
There were no people local to me that were getting arrested for going outside, but it was possible. My boss wrote me a letter that I kept in my vehicle saying I was an essential employee in case I did get pulled over. If the police were more supportive of the lockdowns they could (and often should) have been arresting people for violating the orders.
Maybe your location didn't involve such orders, or maybe the people around you just ignored them all.
I don’t see any example of lockdown in what you say - just paranoia that someone might be treated as though they were locked down. Is a snow day lockdown? What if you get to play outside in warm sunny weather during your snow day?
if the government issues an insane, massively tyrannical order, but then enforces it only 1% of the time, that's still extremely bad
not as bad as if they enforced it 100%, but extremely bad
Okay, so China literally locked people in their homes, and the West didn't do that. I still don't have a problem with the term "lockdown" (mainly because it's what the leaders who implemented these programs and the journalists who wrote about them called them) to describe the significant restrictions on movement, business, recreation, and normal life that were mandated on ordinary people in a totalitarian fashion.
"I was still allowed outside of my home" is not a reasonable response to what anyone complaining about the lockdowns (outside of China) is worried about.
This is the sort of thing that GPT-o3 with search is usually great at. It gives an estimate of about 9 million QALY lost over 2020 to 2023. Note that QALY lost due to long Covid are explicitly omitted, since the evidence on that isn't great and the numbers are pretty up in the air. The sources seem reasonable, although I haven't checked them individually.
https://chatgpt.com/share/682f26a7-47c8-8010-9b3e-b4864e9cfb17
If the deaths were just pulled forward in time a few years, we would see *less* excess deaths in the years that follow.
I think the argument can be stated as '1.2M Covid deaths did not lose nearly as much QUALYs as 1.2M totally random deaths would have, which is why it doesn't feel as bad as other events with similar body counts.'
I think that's true, but also it wasn't *all* old and sick people, lots of QUALYs were still lost here and there - even if we only lost the QUALYs we'd expect for 600,000 'random' deaths, that's still a lot, more than how we talk about it.
Also, of course, lots of QUALYs were lost to non-death... after a bout of Covid, I lost my ability to taste sweet, and it's never come back. Others have much more debilitating long-term effects.
My grandmother died a little over a year ago at the age of 99. The last few years of her life were certainly of low quality on several different dimensions, for a host of reasons I won't go into here.
She didn't die from or with COVID. But if she had died from or with COVID during the first year of the pandemic, she, I, and the rest of our family would have been deprived of the the opportunity for her to meet my son. And I know, for a fact, from conversations with her, that moments like that with her great grandchildren made her few remaining "low qualify of life years" worth it to her and her family.
This is what makes quality adjusted life years so fraught. I understand the need to make trade-offs within any system, and in healthcare that sometimes means prioritizing one life over another. But I also think it's pretty gross when people seem to write off elderly people's last few years of life as being "low quality". That's a value judgement that 1) you can't make in the aggregate and 2) you can't determine just by considering medical conditions.
The "quality adjusted" part is less important than the "life years" part, IMO. (And it's unfortunate that the whole conceptual space is now taken up by "QALYs", which imposes this inherently-subjective and extremely difficult-to-reckon concept of "quality adjustment" on what is otherwise a simple matter of arithmetic on actuarial tables.)
It's very reasonable not to discount the final years of an octogenarian's life on quality grounds, but the fact remains that if someone in their 80s dies of COVID, it's a loss of 5-10 expected years, depending. This is clearly less significant than the death of a young person who could live another 50, even without any dubious quality adjustment at all.
What accounts for the global discrepancy between officially 7 million (US makes up one sixth) to unofficially 20-30 million (US makes up something closer to its age-adjusted share of the global population). Is the gap mostly China or a lot of countries? Do death rates seem close in countries perceived to have reliable data? Should we expect undershoot of expected mortality going forward?
You could look at the excess death rates here: https://ourworldindata.org/grapher/excess-deaths-cumulative-per-100k-economist and compare them to the official statistics. I'd imagine India makes a big part of the gap. Russia too, maybe.
> Should we expect undershoot of expected mortality going forward?
I was also wondering this. I think the answer should be "yes", right?
That's a good chart there. Russia was such a strange case, they were able to develop a perfectly good vaccine but almost nobody agreed to take it, so combined with an elderly population they had the worst excess mortality in the planet. I wonder if they've had any kind of post-COVID reckoning.
It truly is insane. Russian government managed to make one good thing in years, and yet that was the thing which provoked widespread "resistance" among all social groups. Russians would happily accept war, poverty, corruption, they would ask no questions about police violence or shutdown of free speech. But vaccines? That's where they draw the line.
I know right? Like 70% approve of Putin but 10% trusted him enough to take his vaccine. It never made sense to me.
I don't think 70% approval rating is a meaningful figure. It can unpack to something like "Do you prefer Putin to the criminal 90s or some other form of political instability?" or "Do you prefer Putin to the clowns he didn't remove from public politics in one way or the other?"
OTOH, Russia has a long history of "The tsar is good, it's the boyars who screw us up". And it wasn't Putin who personally developed the vaccines.
So I don't know why Russia had high levels of covid/vaccine skepticism but this is not a contradiction.
China is another telling example. They were able to weld shut doors to lock everyone down, but weren't willing to force old people to get vaccinated.
It literally _must _ be yes, mathematically. Everyone is eventually going to die. So if some people die early, that is fewer people available to die later. No matter what the cause for an increase in deaths, there must be a corresponding decrease in deaths at some point in the future to balance it out.
India (3x the population of the US) had less than 1/2 the total reported Covid deaths of the US (~550,000 vs 1.2 million), so maybe 1/10th the per-capita death rate.
India is poorer, but has fewer overweight and obese people.
Reporting is probably poorer.
I don't think anyone knows if the number is correct, or off by 10x.
You can look at per-country data here:
https://www.worldometers.info/coronavirus/#countries
The USA is between Latvia and Greece for covid death *rate*.
Canada and Denmark have less than 1/2 the US rate.
This is just confirmed cases, right? I think the US did a lot more testing than other countries, especially compared to India.
This is *deaths* attributed to Covid. I'm confident about the death part :-)
But he US might well assign more deaths to Covid because testing was much more extensive and India might just see a lot more deaths but not score them as Covid.
One would need to look at overall Indian (and Latvian and Greek and ...) deaths over time. I'm sure the data is available. I don't know where to get it.
No undershoot of expected mortality since Covid.
https://ourworldindata.org/grapher/excess-mortality-p-scores-average-baseline?country=~USA
My guess is that it exists but too small a signal to measure vs noise. If 1.2 million undershoot is spread out over the next 20 years, it could be just a 2% decrease, and it bounces around about that much and will certainly be affected by other health trends and medical progress.
Given that everyone alive during Covid will die eventually, then yeah, any given person who died during Covid means some year is eventually going to have one less death than it would have otherwise.
"My guess is that it exists but too small a signal to measure vs noise. If 1.2 million undershoot is spread out over the next 20 years, it could be just a 2% decrease"
With the bulk of the deaths clustered in the 65+ age demographic (and that clustered in the older of *that* demographic) it should not spread out over 20 years. Lots of people who would have died within, say, 5 years of 2020-21 should have died in those two years. I'd expect that to show up soon-ish rather than spread out over 20 years.
Maybe an undershoot would be more visible if looked at in age-bins (so no obvious undershoot for the 25 year olds, but an obvious undershoot for the 85+ crowd)?
I think you are underestimating their life expectancy.
The life expectancy for a 65 year old male is 17 more years. The life expectancy for a 65 female is 20 additional years. Those are averages, so 50% will live longer.
Even the average 80 year old has 10 more years.
Counterintuitively for some, the longer you stay alive, the higher your predicted lifespan will be.
The average female newborn will live to 79. The average female 79 year old will live to 89.
https://www.ssa.gov/oact/STATS/table4c6.html
I remember COVID skeptics being asked to bet if mortality would subsequently undershoot due to getting rid of the "dry tinder", but I don't recall anyone taking the bet.
The issue is the terms of the bet. Let's say 1.2 million dry tinder might have otherwise had a mean life expectancy of five more years in the no-pandemic counterfactual, so, to simplify the math, assume 120k undershoot of an average 3 million deaths, which is -4%. The difference between 2019 (before covid) and the average of 23 and estimate for 24 (which are pretty close and so plausibly "post-covid") according to CDC is around 230k or +8%. Are expectations of future mortality numbers good enough to know precisely whether that gap otherwise should have been 50% larger? I doubt it. And if the excess death undershoot is spread out even more and unevenly while other things are changing too, it just doesn't seem the kind of dispute easy to resolve with bets with terms that are both well-defined and epistemically reasonable.
All that being said, I think something to look for in the future may be claims of progress in mortality, longevity, etc. - where small percent changes are the norm - that could be largely mere artifacts of long slow "mortality dividends" of dry-tinder undershoot.
There are a lot of countries that just don't have the infrastructure to begin to put together a count because of unavailability of tests, or just simple nonexistence of a public health system to collect data. e.g. when a man dies of covid a village in South Africa he doesn't go to the hospital and there is no coroner to record the death.
Also there is the issue of age. Over 65 makes up 90% of covid deaths. The US has about twice the percentage of people over 65 as the world population, so our death rate is just going to be higher no matter what.
Then there is China who had such a draconian lockdown that they were able to actually halt the spread of the virus until after the vaccine and the much weaker but faster spreading omicron strain came out. So, even if they reported true statistics, which I doubt, their death rate would have been massively lower than the US. And China is 1/6 of the world population, so that's going to skew the numbers a lot.
It’s probably mostly low to middle income countries that don’t have great statistics. China likely had very little undercount until late 2023 when they ended their zero covid policy - though they probably have an undercount in the month or two after that.
Total death numbers are likely fairly accurate in most countries, though there are some (like Somalia and Venezuela) where I expect we have few if any meaningful demographic statistics.
It’s probably mostly low to middle income countries that don’t have great statistics. China likely had very little undercount until late 2023 when they ended their zero covid policy - though they probably have an undercount in the month or two after that.
Total death numbers are likely fairly accurate in most countries, though there are some (like Somalia and Venezuela) where I expect we have few if any meaningful demographic statistics.
It amazes me how the rightward shift of the tech community has motivated so many previously intelligent people to accept conspiratorial ideas (like Covid denialism and antivax) they previously would have dismissed as being for low-IQ nutters. Nothing has actually changed about how plausible these ideas are, only how normalized they’ve become on the online right.
It's crazy how anti-vax jumped from the left to the right so rapidly. Before COVID it was Marin county that had third world measles vaccine rates and West Virginia with the highest in the country. But I guess that's just the tips of the horseshoe.
The antivax community has always been an exercise in horseshoe theory. It's always been granola essential oil hippies and religious homeschoolers sharing health tips.
It feels to me like the New Age movement has finally found its natural home on the right. The antivax movement’s roots in crunchy hippy counterculture once kept it mostly restricted to the cultural left. But as the hippies have grown old and cranky and normie conservatives have grown increasingly conspiratorial and less respectful of traditional authority, the distance between these two groups has grown narrower and narrower. Covid was just the catalyst that finally fused them together.
It’s like sport teams with a player moving from one team to another. The loyalty is to the team - not to the player.
It’s really not surprising at all when you remember how incredibly dishonest the official communication about COVID was.
Maybe just because I'm in another country, but I don't remember dishonesty.
I remember some mistakes that were corrected - and, in one case, the public corrections never got through (that the whole "wash your hands/clean surfaces" thing where the later correction that it was airborne never got through to the public). And it was certainly the case that if you were following experts online then they were ahead of the official communications. But that's true of just about anything; official advice always takes time to catch up with scientific knowledge, if for no other reason than that it requires consensus.
But we got things like "vaccines prevent spread", quickly followed by "no they don't; they reduce it" and then "but not by that much" as each variant made the vaccine less effective than the previous one. Sure, they were still saying "vaccines prevent spread" for three or four weeks after it had become obvious that they didn't, but that's not dishonesty, that's the process where scientific research becomes official communications being very slow.
The main example I recall of calculated dishonesty is an ironic one for people who're generally COVID denialist. The public messaging at the beginning was that masks were not effective, because they were worried about an undersupply for medical workers and didn't want regular people buying them up. They reversed course once the supply was secure. There were some cases of pragmatic (if poorly considered) dishonesty like this, but it's all consistent with public officials trying to cope with a serious pandemic where the public often responds in perverse and unhelpful ways.
If they were so good at thinking ahead and dealing with the unhelpful masses, they should have predicted that lying to them in such an obvious way would cause then to trust then even less, making their good-intentioned interventions less effective in the future.
This is the entire point of having simple heuristics like “just tell the truth” instead of galaxy-braining every specific situation: because you’re probably not smart enough to do it properly anyway.
I never said they were good at it, and I do think this cost them valuable trust, but given ordinary human behavior, I think a substantial contingent of people were going to distrust them anyway.
Was this actually a thing somewhere, or is it just something people say?
If it was actually a thing, how much of a thing? Official country-wide government policy, a local low-ranking official making an unofficial tweet, or somewhere in-between?
My bet is some random media worker mis-spoke, or something like that.
I'll have to wait until I have more time to delve into this more, but as I recall the messaging on this point at the time was concerted and coordinated, but t not very long-lasting.
It was a tweet from the US Surgeon General at the time (Jerome Adams): https://www.npr.org/sections/health-shots/2022/04/28/1095295980/jerome-adams-masking
It's particularly ironic as Twitter started removing antimask tweets later in the pandemic.
Okay, so the linked article seems to suggest the tweet occurred at a point where the US system thought covid wasn't properly airborne ("The WHO and the CDC were saying the same thing that I was saying, so the science supported me.") Or perhaps attempting to make a more nuanced argument about how the public masking wouldn't change much at that point. (e.g. if the incidence in the public at large is very low, then each individual mask won't stop many instances of covid - because the user probably wouldn't have been exposed /anyway/).
So it looks to me like this doesn't support the interpretation that this was officially sanctioned, calculated dishonesty.
"Was this actually a thing somewhere, or is it just something people say?"
The US Surgeon General posted on Twitter in late March 2020:
"... my office have consistently recommended against the general public wearing masks as there is scant or conflicting evidence they benefit individual wearers in a meaningful way..."
https://x.com/surgeon_general/status/1244020292365815809
Six weeks later you were a bad person if you didn't wear a mask. Even if the mask was cloth or knitted.
In 2016 (so before Covid ...) the medical establishment was writing articles such as "The surgical mask is a bad fit for risk reduction":
https://pmc.ncbi.nlm.nih.gov/articles/PMC4868614/
Things pivoted HARD in a very short period of time.
I'm not an expert at the twitter interface. Your link goes to tweet one of three. Am I right in thinking the other two parts are the images below?
If not, you ought to cite them too, for completeness.
If the images are parts 2 & 3 - then the wider picture seems to be that they were asking people not to wear masks if they were in a low risk situation while supply was also limited. It's not really a nefarious pivot or underhand to change advice when the situation also changes.
I agree the text in isolation was misleading, but that's a problem with the format, and the problem looks to me like a failure to realise how it would be interpreted rather than some sort of lie.
This is anecdotal, of course, but I don't remember any official messaging at all about masks not working. But I distinctly remember messaging that we should hold off buying them to ensure medical professionals had access to supplies early on, which of course implies that they do work.
It was specifically that masks didn't work for limiting transmission when worn by normal untrained people in casual settings, so the general populace shouldn't buy them because 1) medical staff needs them more 2) it won't do anything for you anyways.
This was, indeed, a pretty odd combination of ideas to hold in one's head. Since when do you need special training to cover your face? But that's what was said for about a month before it was recommended for everyone. I do think it damaged trust somewhat and I wish the government had just instituted formal rationing on masks early on to reserve supply for hospitals.
EDIT: To clarify, public health officials were not deliberately lying to mislead the public and secure their own mask supplies. There was a pre-existing belief in medical science that masking did not slow the spread of pandemic respiratory illness, influenced by an over-reliance on RCTs for evidence and a backlash away from 1800s miasma theory: https://pmc.ncbi.nlm.nih.gov/articles/PMC8504883/
Of course, at the very beginning of the pandemic when nobody had COVID except in NY and a few other places, it was basically true that wearing masks provided no benefit in (most) "casual settings." There was nothing to protect against for the majority of people. I don't know, but can imagine, it being said off the cuff as something valid today without any thought about the guidance being very contextual.
//Since when do you need special training to cover your face?//
The answer to this doesn't seem to be what you think.
At least, not for everyone.
Have you not seen people covering their mouth and not their nose?
Even after being told what to do incessantly?
Fauci specifically said that the masks you can buy at the drug store don't work and that there's "absolutely no reason whatsoever to wear a mask."
https://www.usatoday.com/story/news/health/2020/02/17/nih-disease-official-anthony-fauci-risk-of-coronavirus-in-u-s-is-minuscule-skip-mask-and-wash-hands/4787209002/
Fauci himself later claimed that the *motivation* for telling that lie was to prevent a run on masks that were needed by healthcare workers.
https://slate.com/technology/2021/07/noble-lies-covid-fauci-cdc-masks.html
It never seemed to occur to Fauci that when people find out that you have been lying to them then they stop believing what you tell them.
Fauci is wrong more often than many people like to believe. He was also wrong about how transmissible AIDS was.
I stand corrected, though I think you're overstating his initial claims. He didn't say drug store masks don't work, he said they "[don't] really do much to protect you", which implies they do do something to protect you. And you removed the two qualifiers in your quotation, where he said "Now, in the United States, there is absolutely no reason whatsoever to wear a mask." This implies that there might be a reason to wear a mask if circumstances change or in another location.
I completely agree with your final sentence though. With rare exception public officials should always tell the truth, even if they think doing so will result in some negative consequences. The long term erosion of trust has severe consequences, some of which we're now seeing.
I can confirm that this initial messaging on masks not being helpful for the public absolutely did occur. I think the concept they were trying to communicate was more subtle: “Masks are unlikely to help the public because their exposure is less concentrated than what healthcare workers are exposed to.” But it came off as more black and white. And in retrospect was just wrong (“aerosol generating procedures” being much more infective than coughing fits remains a nonsensical idea that healthcare is somehow still holding on to) on a number of levels and was absolutely an unforced error created by panic. They were terrified we would run out of PPE.
There was a lot of dishonesty. I'm not talking about things that could be charitably interpreted as mistakes or bad calls, like when the NYC health commissioner said early on that there was nothing to worry about and people should continue to take the subway and congregate for the holidays, I'm talking about flat-out lies. I'm talking about things like the head of the CDC going before a Senate committee and claiming that if you get vaccinated you can't spread COVID.
When was that said? The vaccines weren’t expected to be neutralizing, but the Pfizer and Moderna vaccines were so effective against the original wild-strain SARS-CoV-2 that they were neutralizing. The other vaccines (e.g. AstraZeneca, Janssen) were not neutralizing, but they weren’t initially available in the US (and only Janssen was ever approved). I think that “if you get vaccinated, you can’t spread COVID” was therefore true until the Beta variant.
The non-neutralizing nature of the mRNA vaccines with the Beta variant wasn’t confirmed until 10 April 2021 when data from Israel first showed that people with a Pfizer vaccination were able to spread Beta asymptomatically. And Beta struggled to spread widely outside of South Africa and Israel; Alpha mostly outcompeted it, so it was reasonable to refer to “you” (ie the typical American) as not being able to spread COVID if vaccinated until the Delta wave in late June/early July.
It wasn't true. It was never true. She made the claim that "vaccinated people do not carry the virus, don't get sick, and that is not just in the clinical trials but it's also in real world data," in March 2021 based on a CDC study that showed nothing of the sort. What it did show was that vaccinated people had a 90% lower risk of infection two weeks after their second dose, which is distinctly different from a 100% lower risk of infection, and distinctly different from sterilizing immunity.
https://people.com/health/vaccinated-people-do-not-appear-carry-spread-covid-19/
She made her claim, the one I quoted above, to the Senate committee in May of 2021. There was no evidence for sterilizing immunity. To make a claim honestly, you need evidence for the claim, not just a lack of evidence that the claim is false.
https://www.cnn.com/2021/05/21/politics/walensky-comments-cdc-guidance-fact-check
There was so much blatant dishonesty, that it astounds me when people act like there wasn’t, or “can’t recall” any. Why did they NOT lie about.
- origin
- masks
- mortality
- transmission effectiveness
- vaccine safety
- prevention of all kinds
- herd immunity
- slow the spread
- lockdowns
- protesting government response is “white supremacy”
- contradictory response to protests based on what the protest was about (transmission is not a problem if you’re at a George Floyd protest, but you’re a “white suprematist super spreader” if you’re protesting government tyranny).
- children must get vaccinated with a novel mRNA transaction therapy or they’ll DIE. In reality, children were not at risk.
The list goes on and on.
Who is your “they” here? Because I didn’t see the scientists lying, I saw the politicians lying about what the scientists said (mostly, because the politicians were too fucking stupid to understand what the scientists said and were trying to “simplify” for non-scientists and ended up taking away so many qualifications that they weren’t telling the truth).
Origin is still the same answer as ever “natural causes more likely, can’t completely rule out lab leak, can completely rule out intentional spread”
Masks: cloth masks went from “no evidence they work” (which was true; no-one had ever studied it) to “better than nothing” (also true) to “still not very good and we’ve actually got some supplies of N95s now, so why not use them instead” (also true).
Mortality: went from “appears to be really high, but we’re only seeing the worst cases” to “1% or so” - and it appears to be a population-wide 0.3%, and there are a lot of people who either never got COVID or only got it after vaccination which definitely reduced the risk of death.
Transmission: well, yeah, the vaccine effectiveness was way higher on wild-type in the original trials in July 2020 than on Delta in July 2021. So of course the story changed. That’s not a lie, that’s the facts changing.
vaccine safety: how has the story changed? There are some risks that are either rare or minor, the same as just about every other vaccine.
Prevention: now, there is a real scandal here, which is the refusal of the scientific establishment to accept that COVID is an airborne disease and the total determination to stick to the idea that it only spread in droplets. That was a scientific fuck-up. But it wasn’t a lie; it wasn’t dishonest. They were just wrong and stubborn and wouldn’t listen to the people who had the evidence.
Herd immunity: is real, but requires you to get the reproducibility number under 1; vaccines alone weren't enough after Delta. Delta changed a lot of things and was worse than expected; so any statement made before July 2021 wasn’t a lie, just wrong.
Slow the spread: Yeah, this was a stupid idea - you need to keep the R number under 1 until you reach eradication. Riding the wave up and down was correctly assessed as a risk by an Imperial College scientific report early on and yet just about every country did it. So can we blame the politicians and not the scientists, please?
Lockdowns: These worked when implemented properly. Sadly, the countries that did lock down properly generally got didn’t push vaccination aggressively enough. So Australia, Japan, Thailand, etc all had a massive wave in 2022 (or China in 2023). The only country that did get everyone vaccinated before unlocking was New Zealand, which had one of the lowest deathrates anywhere (the lowest of those with trustworthy statistics).
Floyd protests: yes this was bloody silly, though the version I recall was that the Floyd protesters were OK because they wearing masks and the anti-NPI protestors weren’t OK because they weren’t - but I admit I’m in a different country and that just never seemed like an especially big deal.
“They,” just to start off, would be Dr. Fauci (NIAID), Dr. Robert Redfield (CDC), Dr. Tedros Adhanom Ghebreyesus (WHO), Dr. Deborah Birx (White House Coronavirus Response Coordinator in 2020), Dr. Rochelle Walensky, (appointed CDC Director on January 20, 2021).
Just to name a few (the heads of major scientific institutions and groups managing public health).
All lied. All deliberately mislead. All were repeatedly, demonstrably wrong.
Those are the political leaders of scientists, they were not actively working scientists. Not one of them published a single peer-reviewed paper on COVID.
"I didn’t see the scientists lying"
Then you weren't watching. The scientsts you refer to were telling each other things in their own personal emails which they were directly denying in public.
The people who don't remember it are absolutely sold on the idea of not believing it, and have (subconsciously) taken every effort to forget anything that isn't The Narrative. The rationalizing will abound as it does with our dear host.
As a foreigner you probably didn't see it. I kept a list.
• Covid will not be a deadly pandemic
• Closing the borders is racist
• Masks don’t work
• Lockdown two weeks to flatten the curve
• Trump didn’t close the borders fast enough
• Lockdown three months to flatten the curve
• Masks work
• Lockdown until Covid is gone
• Covid CANNOT be solved at the state level
• It was never about Herd Immunity
• Saying the virus came from a lab is racist
• One mask doesn’t work but two masks definitely work
• Lockdown because symptomatic spread will kill old people
• Asymptomatic cases can still have long term damage, ground glass lesions, and such
• Lockdown until a cure
• Every hospitalization with Covid must count as a Covid hospitalization regardless of the reason for admission.
• Don’t trust any vaccine announced under the Trump administration
• The vaccine announced the day after votes were counted is safe and effective
• Vaccinated people cannot transmit Covid-19
• We must mask children because science says so.
• The vaccine has no side effects
• We will not mandate the vaccine
• We must vaccinate as many people as possible to get Herd Immunity
• You can return to your life after getting the vaccine
• The virus probably came from a lab
• Only N95 masks work
• The NIH doesn’t fund Gain of Function research
• The vaccine was never about stopping transmission
• Vaccine side effects mean it’s working
• The vaccine has no effect on women’s menstrual cycles so stop being hysterical
• The Wuhan NIH research doesn’t count as Gain of Function research because some gain of function researchers decided those gains of function didn’t meet the legal definition
• Asymptomatic transmission among vaccinated people is fine, will not kill old people, and don’t worry about long term damage, ground glass lesions, and such
• You can return to your life after getting booster shots every six months for the rest of your life
• We will take your life from you if you don’t get the vaccine
• We will use the FBI to investigate parents who complain about school vaccine mandates
• The vaccine turns symptomatic illness asymptomatic, but don’t worry about the increase in asymptomatic spread
• Go ahead and get infected with the virus after your vaccine to improve your immunity.
• Peter Daszak, the guy who organized the “lab leak is a conspiracy theory” letter in the Lancet, proposed to build Covid-19 for DARPA in 2018.
• Vaccinated people can spread Covid-19 just as easily as non-vaccinated people can.
• The spike in heart attacks and vascular conditions isn’t from the vaccine, it’s from the lockdowns.
• Unvaccinated children will be removed from school.
• Covid MUST be solved at the state level.
• The vaccine actually does affect menstrual cycles but you should get it anyway.
• Hospitalizations with Covid don’t count as Covid hospitalizations if Covid wasn’t the reason for admission.
• The case for masking children has collapsed
• Leading scientists knew it came from a lab but hid their opinions to promote “international harmony.”
• Even though the case for masking children has collapsed we’re going to do it anyway.
• Even though mask mandates are illogical, all that matters is that they align with the “community’s goals.”
• Mask mandates didn’t make much of a difference anyway.
• Fauci: “Vaccines don’t protect overly well” from Covid infection.
• The vaccine was bad because it was Trump’s fault.
• A vital part of hurricane preparedness is to get vaccinated.
• When the FDA told people not to take ivermectin that was totally just an informal recommendation.
• Department of Energy: “The virus came from the lab.”
• CIA: “The virus came from the lab.”
Shame you didn’t keep the list of where you saw those so you know who to stop trusting.
Oh I know. :)
Some of the things in your list are bullshit, some of the things in your list are wild exaggerations, some of the things in your list are “some low-level state official may have said something like this.” Every single thing in your list needs a citation.
This is a good thread. Don’t try to junk it.
The main thing that I would describe as dishonesty was the very early anti masking instructions. The reasons for this one are obvious although from my point of view it would have been better to just tell the truth and seize the supplies.
In the US there was a massive campaign of censorship where the government put pressure on social media companies to censor anyone who went against the official narrative. They even threatened scientists careers to force them to sign on to a public declaration that contradicted what they actually believed https://en.wikipedia.org/wiki/Lancet_letter_(COVID-19)
On top of that the public health bureaucracies issued decrees that were clearly motivated by politics and not public health e.g. at a time when they had banned all large outdoor gatherings and closed public parks and beaches, they issued a declaration that BLM protests were not a risk to spread covid, and were in fact good for public health.
Outright lies were relatively rare. Undue certainty, arguments from authority, and demonization of even reasonable skepticism was extremely common.
Official communication about COVID was often dishonest, but the current administration's official commissions are deeply and transparently dishonest, and it doesn't motivate people on the right to conclude "we should disregard or reverse everything the administration says." This clearly isn't a behavior people apply consistently across the board.
Right. If the politicians du jour is grossly dishonest, half the population will have absolutely no trust in their leadership, while the other half will be solidly on team "well, they're *our* lying liars and it's not really lying if it's what they have to do to beat *your* lying liars". This is catastrophically bad for society, no matter which half is in which position.
Maybe try electing politicians who aren't grossly dishonest lying liars at all?
I think this is the crux of it. If the science had been squeaky clean from Day 1, we wouldn't in the place we now are. Once the fundamental trust in professional objectivity is lost, a doorway to lunacy opens up that will not be easy to close again.
Of course, cutting edge science is never squeaky-clean. But many statements contradicted established science for public policy purposes.
It exposed how fragile the knowledge base for epidemiology actually is, that is for sure. But you can’t just not make decisions in an environment where you have all the facts. I think it could certainly be executed better in the future but in practice it is very hard to both convey nuance and uncertainty to the public while also trying to push the policy you think needs to happen in the moment. How would you do it?
I agree this is a genuinely hard problem. I would even argue it is impossible to have a useful conversation online, because it requires so much good faith on the part of the interlocutors.
I think sometimes about all the “lies” that FDR told in the run up to WW2. Do we wish he didn’t do that?
I think you're on to something here. :)
I think there needs to be more appreciation of just how great an achievement objective science was. It took us maybe a million years to get there. You throw that out the window because of a media driven panic? That's not good. People in power need to recognise the importance of the average citizen fundamentally trusting science and act accordingly. But, post the arrival of social media, many areas of gov became convinced that narrative control was all that mattered. That is fundamentally untrue.
"it is very hard to both convey nuance and uncertainty to the public while also trying to push the policy you think needs to happen in the moment. How would you do it?"
Short answer: with humility and honesty.
Longer answer: I think you start off with a bad assumption, namely that one needs to push a policy. How would I do it?
Provide the best available information while being honest about your confidence in said information.
When your opinion changes, explain why.
Only advocate for or institute policies backed by force of law if you have a very high degree of confidence, otherwise focus on providing information and advice.
Trust people.
Grant them autonomy.
Own your mistakes.
If you egregiously mess up, or consistently demonstrate poor judgement, resign from your position of authority.
If you're worried that things will fall apart without you, recognize that you're not a very good leader, because good leaders build robust systems that can operate without any one specific person. Start making corrections now.
Don't try to expand your sphere of authority anytime you see someone doing something objectionable.
Follow your own rules and guidelines.
Be selfless.
This all seems like a good aspirational list.
The thing is, in a crisis, where there is high threat and high uncertainty, when people are overwhelmed and exhausted, we humans are going to do all these things badly.
Which is why we need our systems not to depend so much on the individual character of leaders.
In the 1980s, I taught at UCBerkeley about the lack of pandemic preparedness despite the high likelihood, near certainty, that there would be one in the coming decades. There were books and articles and loads of research done on this back then. The work ahead was to improve infrastructure for preparedness, not to rely on the good character of individuals. A lot of that work never happened and so when this pandemic came were were not nearly as prepared as we could have been.
One of the reasons for that is that as a society and as a human race, we are not yet that good at long-term planning, or at making short-term sacrifices in service of long-term well-being. This human frailty exists at the individual level all the way up to big organizations.
I disagree. Populism and general anti-science and anti-elitist thinking spread via social media had been undermining trust in experts long before Covid. People were actively looking for any excuse to disbelieve the authorities from the very beginning, especially given the obvious political and social ramifications of something as disruptive as a pandemic. There was nothing new or unusual about how experts handled an event like this other than that it happened during the age of social media.
Well, I think we're fundamentally agreeing. The advent of social media and the unprecedented level of connectivity that it afforded, created an environment where previously repressed and marginalised conspiracy material could now achieve take off. Yet, the Big Gov response of attempted narrative control simply did not work.
It worked throughout most of the 20th century, but in the 21st century we raised our standards and got more honesty, and as a result the attempted narrative control no longer worked. We got more accurate and reliable messaging, and as a result people started to understand that it was not perfectly accurate and perfectly reliable (even though it was better than all the things they trusted without complaint in the past).
What you are asking for is the opposite of “squeaky clean”. What we got during Covid was much cleaner scientific information than at any other point in history, because scientists were talking directly to the public. But it turns out that seeing scientists speak made people realize that scientists aren’t an alien race of godlike intelligences who are perfect on everything, and they decided that if they’re not like this, then we might as well not listen to them at all.
We lost the fundamental trust in professional objectivity because we allowed transparency and honesty.
And in a snit, we took our footballs home, where we now sit on the internet and rage. It's rather dispiriting.
Yes, it's true that the previous centuries had imbued scientists with godlike authority. And then Covid pulled the curtain back.
But they weren’t all being transparent and honest. Too few were honest about the level of genuine uncertainty, and too many were quick to declare the current best guess as “expert consensus” and label anyone who questioned it a science denier.
And then they started using their newly elevated platforms to push obvious politics rather than just communicating scientific fact.
It probably would be better if public pronouncements were even more honest about uncertainty. There are a lot of such ways I have endorsed. But I think it’s also true that experts during the COVID pandemic did at least as good by this measure as experts at nearly any time in history before then.
My contention is that it is precisely the visibility of uncertainty and disagreement that was a large factor in the loss of trust.
Can you think of a time when there was better honesty?
It’s funny how people on the right who complain about being lied to by the mainstream media/establishment, they never bring up the countless lies and misinformation coming from Trump in 2020.
Of course, the difference is that the left's lies didn't even end up furthering their interests in the end. Just pure incompetence. How sad is that?
Speaking as a right wing guy who complains about being lied to by the healthcare establishment during covid...
What do you want me to say about Trump lying? He does it all the time, he's very dishonest, and thats bad. People don't typically cite him as a reliable source so it doesn't come up much, but it is bad
I think that the objection is less specific. I suspect that the RW-ers I know, would say the same. They don't trust what DJT says. However, they parrot the general talking points of the administration (e.g. Tariffs are both paid by foreign countries while simultaneously boosting domestic production, DOGE is eliminating loads of corruption and saving $100s of Billions). At least to me, this seems to better articulate the [edit: +apparent] hypocrisy being alluded to.
The double standards are insane. The president, who people should be looking towards in times of national crisis, lies through his teeth about how big of a problem Covid is and peddles snake oil treatments. And all you have to say is, idk it’s bad.
But, the medical establishment that has been 99% correct on everything deserves all the vitriol because there was conflicting information about mask usages in the early parts of Covid and maybe Covid made out of a bio lab (which still isn’t 100% confirmed).
> who people should be looking towards in times of national crisis
Why do you think this? No, they shouldn't. Didn't people learn this after George "Weapons of Mass Destruction" Bush's invasion of Iraq?
> conflicting information
You're making it sound like an honest mistake instead of a deliberate lie intended to manipulate the public.
My understanding is that during the early days of Covid, the health department (under Trump btw) discouraged masks because there was a shortage of masks and they wanted to save them for medical workers/high risk people. Eventually, it was corrected. Deliberate or not, lies made with good intentions happen everywhere. If other parts of academia were as politicized as epidemiology had became, you would literally trust no one.
>People don't typically cite him as a reliable source so it doesn't come up much
This statement seems to come out of a reality where the 2020 "Trump actually won the vote" conspiracy narrative didn't happen.
Is there a time in the past where you think the healthcare was more honest than it was in 2020 or 2021?
My claim is that they were at least as honest during the pandemic as they have ever been, and it is partially because of the honesty of admitting mistakes that people now think they should never be trusted.
What about.
But why believe people who were even more dishonest? During the Delta wave three antivax conservative radio hosts died of COVID in a single month. These people made a very lucrative income from partisanship.
https://www.nbcnews.com/politics/politics-news/trio-conservative-radio-hosts-died-covid-will-their-deaths-change-n1278258
The first piece of good news in this thread. I'll reciprocate by posting even better news:
https://pubmed.ncbi.nlm.nih.gov/37486680/
It is surprising to me that so many of them are so stupid. I also was distrustful of various official communications, but by the power of reasoning was able to come to reasonable conclusions such as "taking the vaccine is good", "covid is not that dangerous to me personally, but low-cost measures that inhibit transmission are not a major imposition", and "mask recommendations are not part of a new world order scheme to extract your adrenochrome". It didn't take me that much mental effort to navigate between the conspiracy theories and rigid adherence to the CDC gospel.
Who was in charge of the executive when that dishonest communication was happening?
This is such an ironclad argument. That's why the frequent dishonesty deployed by internet atheists gave me a prior that young earth creationism is probably correct
I remember many examples of dishonesty. The efficacy of masks (or not), risks to children, the need for the most draconian lockdown measures (e.g. closing outdoor spaces), double standards for which activities were allowable and which weren't (e.g. social justice protests)... I was cognizant of the dishonesty in real-time and it still makes me angry.
But I still don't traffic in conspiracy theories. I believe myself to be a savvy enough consumer of a variety of sources of information that I can differentiate between what's likely to be true and what's likely to be false. Of course I have my biases and blind spots, but I think, or at least hope, that they aren't too egregious, and that I'm self-aware enough to assign confidence levels and come up with a realistic world-view accordingly.
I understand why people have come to distrust elite individuals and institutions. I understand why certain conspiracy theories have emotional appeal. But I expect reasonably intelligent people to stay true to a core set of principles, navigate their distrust and their emotions, and not be pulled in by dishonest, biased anti-elites, simply because they stand in opposition to the dishonest, biased elites.
The problem is that the level of dishonesty and failure that you are observing here is likely comparable or lower to the levels of dishonesty and failure that have existed among medical authorities at any other time in history - you shouldn’t stop trusting someone just because they became more honest about the fact that they are only 80-90% reliable.
YES. first thing I thought when I saw the title. rationalist groups especially
have become self-parody
Interesting, I haven't particularly noticed that. Can you point to some examples of rationalist groups being crazily conspiracy-theory-ish? Of course, the rationalist community has always supported all kinds of wild thought experiments much weirder than these, I'm wondering about cases that are more than that.
I think part of the issue is that Scott's blog exists at the intersection of a few different communities: the rationalist community, the centrist classical liberal community, the anti-woke community... So I think a lot of the commenters here with the worst epistemic hygiene may not really be "rationalists", or at least, might be more centrally from some other community.
Not to say that there's no crossover with "core rationalism", and I don't want to do a no-true-rationalist thing here, but I think you see more of this stuff in Scott's comments than on LW, for example.
https://slatestarcodex.com/2019/07/04/some-clarifications-on-rationalist-blogging/
This was both six years ago, and before the "mainstreaming" via both the NYT article and move to substack. So most likely the rationalist ratio is likely far, far lower than 13% now. This at least matches my anecdotal experience, where people who I typically see making dumb mistakes that would be caught by skimming the sequences never end up identifying as rationalist. Same with using this as identifying known rationalists (from their shared handle on LW)
Yeah this has always been a property of Scott's blog, though maybe more pronounced in recent years?
I've definitely encountered more IRL people who know of ASX then I ever did (a grand total of 0, I'm pretty sure) in the SSC days, which suggests a broadening of audience.
Ah, yeah, I certainly don't think of Scott's commentariat as being very representative of rationalist discourse or the rationalist community these days. I assumed Swallow meant something different, but maybe not.
I think part of the issue is that Scott's blog exists at the intersection of a few different communities: the rationalist community, the centrist classical liberal community, the anti-woke community... So I think a lot of the commenters here with the worst epistemic hygiene may not really be "rationalists", or at least, might be more centrally from some other community.
Not to say that there's no crossover with "core rationalism", and I don't want to do a no-true-rationalist thing here, but I think you see more of this stuff in Scott's comments than on LW, for example.
One of these days I'll come out as a part-time obscurantist...
This post.
'Here are largely-undisputed all-cause mortality statistics from CDC and census.gov' is a level of crazy conspiracy-theory-ness I'm pretty comfortable with :)
The comment thread that made Scott feel like it was necessary to explain why he believes there were a milllion Covid deaths in the Us.
Oh c'mon, we all lived through a hodge podge of policies and random claims from scientists and officials. It may have not made us all conspiracy theorists, but if were paying any attention it sure changed how you perceived "experts."
And this made you reflexively believe people who lie incessantly about everything, instead of the people who occasionally lie?
It seems much more likely it's about tribal and emotional affinity and hatred of scientists for being part of blue tribe.
It really shouldn’t have! This is how experts function, and function well! People have unreasonable expectations, which past expert communities lived up to by hiding their failures.
If you do trivia competitions, you’ll understand that informed guesses that are nowhere near perfect can still be the best option, and lead to you reliably winning over the competition.
Informed guesses are often necessary. Falsely claiming your informed guesses are actually indisputable scientific fact never is.
Flat out *making shit up* because the facts you know make your desire to virtue signal about a murdered fentanyl addict inconvenient is a whole other level of credibility-nuking.
I think it's valuable to separate scientists from spokespeople from politicians from policy wonks from corporate interests in thinking about this.
Science is, as it has ever been, messy and imperfect. And yet an incredible human endeavor and still improving decade over decade.
Politicians (and their spokespeople) and corporate interests have never been very truthful because the incentives don't run that way.
If you lived through Watergate, Vietnam, the two gulf wars (with for sure WMDs!), a couple of finance-sector-engineered economic crises, the lies of tobacco and pesticide companies, the extent of child abuse and coverup in the Catholic church (and elsewhere), and all the stuff in-between, or some fraction of those, the hodgepodge is not remotely surprising.
I think the pandemic hit a bunch of people with more limited life experience and many of those people are now understandably gobsmacked that the world doesn't work better than they thought it did.
You should consider, here, whether the previous dismissal of these ideas as "for low-IQ nutters" wasn't also influenced by political considerations. It's not a given that this was a shift that made people net less rational.
The growing institution of Alternative media has herded the cattle with Candace Owens snd Catturdism, and therefore it’s politically advantageous to pander to the cattle if you’re a tech bro trying to change the world or get rich.
I wonder if people coming from countries with a hyper-inflationary past have easier time dealing with big numbers like this. I bet to somebody from Turkey or Argentina 1.2 million deaths are more believable (in fact maybe too normal as in desensitized?) than somebody from say Brunei or Switzerland.
Contrast "I only actually know one person who died of Covid" with the breathless "expert" and media hyping at the time that we should expect to see hordes of dying people staggering the streets because emergency rooms and entire hospitals would be flooded, then the hordes' corpses to be stacked like cord wood awaiting a mass bulldozer burial. Then, poof, after the beaches and playgrounds were reopened, it seemed to have been more like a bad flu season (not to mention actual flu disappearing for some mysterious reason). The cognitive dissonance is massive.
If you worked in a hospital that first winter of 2020-2021 it was certainly not a "bad flu season". It was a nightmare.
For real. This discussion is honestly just darkly humorous as someone who worked the Covid floors from March 2020 through July 2021. Those of us who were there remember.
Anyone in healthcare would be agape at this stuff. You know, the people who were actually taking care of all the Covid patients. Who also got sick, who also died, burned out, left the field despite years of training and debt for that training. Because it was such a horror. And now to see people dismiss all of that because they didn't personally stand in the hospitals and see it with their own eyes. Like Paris doesn't exist if you haven't been there yourself.
I was a healthcare provider during Covid and it changed me and my career permanently. Me along with many many others I know directly.
Hospitals aren't equipped to deal with a large public spike above baseline use of medical services. Recall how in the beginning of the pandemic, supplies like toilet paper, and rubbing alcohol became nigh inaccessible. The supply chain wasn't equipped to quickly respond to a significant change in demand, even though these are cheap and abundant goods.
Not to mention the supply chain in general being wrecked by lockdowns.
If an industry takes 20% of our GDP I would expect edge cases to be worked out and available. Handling unforeseen spikes is something expected of the tech industry, but somehow medical system needs all our money and can't be bothered to think on their feet and handle a slightly above normal spike.
Just-in-time logistics has become normal.
It takes a lot more time and resources to scale up the operations of a hospital dealing with highly infectious patients than it does to scale up the production of toilet paper.
If you are trying to build a hospital with exactly the same equipment in the same configuration sure. In an emergency I'd expect other solutions (deputizing the public like EMTs and those with some medical training) and nurses to expand the number of workers and making beds out of things not hospital beds. Especially if it's taking 20% of our GDP.
20% of our GDP is what we're spending under business-as-usual scenarios.
If you don't have proper facilities and appropriately observed safety protocols, sticking a bunch of sick people together to care for them makes outcomes worse rather than better (because they're infecting other patients and workers.) Even under ordinary circumstances, hospital-induced comorbidities are a significant risk. There simply aren't enough regular people with adequate training to safely run hospital-like facilities who can be deputized in order to dramatically increase our capacity to provide medical care under pandemic conditions.
> deputizing the public like EMTs and those with some medical training
Why would you think that would work? Is EMT training just all for show and any Joe blow off the street can do it at the drop of the hat?
The reason medicine and tech don’t behave in the same way is that the human capital of medicine has a way longer lead time and the regulatory framework is far stricter. EMTs, nurses, phlebotomists, RTs, PTs, OTs, SLPs, NPs, PAs, MD/DOs can’t just be conjured out of the ether. You either need to dramatically over expand your resource pool for edge cases, or accept that trade offs will need to be accepted when the system shock occurs.
This is why you need a federal government in a capitalistic system. Capitalism ensures that if 1 medical provider voluntarily kept excess production available, then he would be put out of business by his competitors. This massively parallel computing machine of industrial competition is what is so brilliant about capitalism but it does mean that you need intelligent governance to deal with the shortcomings.
Having 20% excess capacity sitting around means building a lot of extra hospital wings to sit around empty and hiring a lot of extra doctors and nurses to do nothing. Nobody is willing to pay for that during business as normal, so we have what we have.
I mean, we did see a million deaths. I don't know if this qualifies as a "horde" of corpses, but you could build a pretty sweet skull pyramid out of it.
This was my whole point in the last post - people act as if this was "hyped" or "overestimated", whereas in fact it was the largest mass casualty event in American history and if anyone had given the true number at the beginning of the pandemic, people would have been even more scared than they were. It's just that people are able to ignore a million deaths because that's 1/300th of the population and dead people don't write thinkpieces about how angry they are to have died.
A lot of hospitals did get overwhelmed. You didn't hear about it because you weren't in those hospitals and, I predict, don't know any of the doctors or nurses who were on the front line.
I once worked in an actually flooded-by-hordes emergency room - not during COVID, but during a blizzard in Ireland that the locals were completely unprepared for (they almost never get snow). Every bed occupied, gurneys lining the hallways, some patients on the floor, long line of people waiting to get it, stretching out the door. What percent of people in Ireland do you think were in that emergency room, or know someone who was? Would the average person on the street have thought "Yeah, I am personally connected to the overcrowdedness of this emergency room?" How exactly do you see this working?
You say this, but 1-4% mortality for the entire population was floated early on. Everyone I know expected way higher than a million death and for it to be spread across demographics equally. I remember thinking 3 to 12 million would be a reasonable estimate when they were still using NY and Italy as the estimation.
1% is an entirely reasonable estimate if the pandemic had ripped through the population without any countermeasures early on.
Not for people under 65 (and even less so the younger you go). Which started to become apparent when this got published: https://www.nytimes.com/interactive/2020/obituaries/people-died-coronavirus-obituaries.html
Are people over 65 somehow not a part of the population then?
And? Some people happen to be over 65, and still value their lives greatly. Even more people have relatives above the age of 65, and would greatly prefer if they didn't die a painful death years ahead of their time.
Yes, 1% was about right if you caught the original virus early on
Yeah, and then we shut down substantial parts of the economy for a year, folks wore masks, and we developed a effective vaccine at record speed.
It's not even clear what case you're trying to make here.
Scott said:
> and if anyone had given the true number at the beginning of the pandemic, people would have been even more scared than they were.
People were making estimates wildly worse than 1.2 million at the beginning of the pandemic.
They were also making estimates wildly better than 1.2 million at the beginning of the pandemic.
It's almost like the fact of uncertainty itself is hard to grasp as a feature of the world.
"Effective vaccine" that was not a vaccine. Did not stop people from getting Covid. Did not stop people from transmitting the virus. Did not build up any immunity keeping "vaccinated " people from testing positive for Covid over and over. The only thing left in defense of the "vaccine" was the weak, "Well, you would have been sicker without the vaccine" . . . and I would like to know how that would be proved.
You are aware that people who've had the MMR vaccine can still get Measles, right? And people with the flu vaccine can still get the flu? There is a range of efficacy in vaccines, like with all other things.
It’s only a vaccine if it is made from live vaccinia virus to stop smallpox. Otherwise it’s just “sparkling immune system training”.
The vaccines, generally speaking, reduced the risk of dying of Covid by factor of ten or so, which is far better than nothing. For the early variants, they probably would have prevented the spread well enough to eradicate the disease if everyone had become vaccinated, but the virus evolved too fast, became to virulent and too good at evading the immunity conferred by the vaccination to achieve that. As far as I recall, that was a bit of a surprise even for the experts. Still, without the vaccines, things would probably have turned out way worse in terms of casualties.
Not trying to be political (really!), but the predictions (like you mention) and the actual results reminded me a lot of the AIDS epidemic in the early 1980s.
The AIDS deaths started rising sharply, folks started projecting the curve and warning that everyone was at risk, things DID get worse, but ... there was also clustering and lots (maybe most?) people didn't personally know anyone who got AIDS or maybe knew only one person.
Other areas (e.g. San Francisco) had lots of people who knew 10s or 100s of people who died of AIDS.
But the exponential stopped exponenting as AIDS burned through the super at-risk population. And the general public might be forgiven for wondering why if things were so bad they weren't noticing anything particularly unusual. And wondering why they were supposed to panic?
Not quite the same here, but the parallels are similar enough I think. Folks in the medical community really noticed this in their own lives. Folks who worked a lot with the elderly probably noticed this in their own lives. For a lot of the rest of us we needed the news to remind us how bad things were because we weren't seeing it.
AIDS has R0 less than 1 via heterosexual intercourse in places outside Africa. So it wasn't going to break out of some relatively small subpopulations. COVID was not like that, it was a respiratory virus that hit everyone who breathes.
Right. AIDS wasn't going to break out of some relatively small subpopulations.
But in 1985 we were getting articles saying things such as this:
"By early this year, most Americans had become aware of AIDS, conscious of a trickle of news about a disease that was threatening homosexuals and drug addicts. AIDS, the experts said, was spreading rapidly. The number of cases was increasing geometrically, doubling every ten months, and the threat to heterosexuals appeared to be growing."
The doubling didn't go on forever and the heterosexual population wasn't totally unaffected (e.g. Arthur Ashe, Isaac Asimov) but that wasn't the message ordinary Americans were getting in 1985. Without news coverage, most Americans in 1985 wouldn't have known about AIDS. And would not have though to be worried about it. Folks in SF didn't need the news.
Surely one has to control a bit for the effects of people changing their behavior in response to those warnings. https://vimeo.com/43187299
I forget if Covid eventually surpassed HIV in deaths. But it seems to me that there is remarkable similarity in the entire process, but compressed into 2 years instead of 40. (Government denial, then government acceptance, then Anthony Fauci changing policy to allow better responses when demanded by patient advocates, then better treatments, and eventually a preventative that means we don’t have to worry too much.)
I agree large numbers were floated. Small numbers were also floated. I think if God came down and said "the true number will be 1.2 million, even with pretty extreme countermeasures like locking down the whole country for months", most people would have interpreted that as "the large number people were right", albeit not the absolute largest.
I never saw anyone say 4%. I hang out in groups who’ve been thinking about this for a long time, who range from totally mainstream to batshit crazy and their consensus at the start was 1%, moved lower very fast, as low at 0.1%, then rebounded to roughly whar actually happened. I would say we knew the likely scale by May or June 2020
Fairly early predictions of deaths sans countermeasures turned out to be accurate. You can project the mortality among those who *did* become infected and compare that against who didn't in 2020. It's harder to estimate what the loss of a functioning health care system would've done to mortality, but even that's something you can ballpark. The projected scenarios that seemed gloomy were on the money as far as that goes.
There were countermeasures - people aren't just going to behave normally as corpses pile up - but it's just innately hard to to predict what political will is going to look like, particularly so in the United States. Losing over a million people is staggering, and a significantly worse overall outcome than some global peers such as Japan.
I think a lot of this is missing concentration. New York City did, in fact, have piles of corpses in the street. Everyone I know from New York City knows lots of folks who died of Covid.
I know a few.
My family members who live in rural North Carolina know zero.
> My family members who live in rural North Carolina know zero.
It’s probably also an issue of New Yorkers being much more willing to say that someone they know died of Covid, whereas the rural south which got absolutely ravaged by the delta wave had folks who would have a bunch of egg on their face if they admitted that pawpaw died of that virus that up until yesterday they had been loudly proclaiming to be a hoax.
Actually living in the south and knowing a lot of Covid deniers... there's *some* of this, but not as much as you'd think. Most folks who have someone close to them die of Covid change their tunes pretty quickly. Often they're vocal advocates for the protections they argued against.
Generally people are people, not caricatures. You can show them all the facts and figures you want and they won't budge an inch, but when they have real life experiences with the things they adapt.
I lived and worked in healthcare in the rural south through Covid. At least in east TN there were plenty of people dying from small places in the mountains where their absence would be noted given how small the community was. Some church outbreaks had multiple congregants die from a pretty small population. It was surreal driving home from the hospital past the freaking freezer truck full of bodies and seeing TGI Fridays packed to the gills with old gomers just half a mile down the street.
People are people, and people the world around are extremely invested in saving face.
I got curious about the skull pyramid thing - assuming each skull is around 17.5cm wide and lies on its side, and that they pack 1, 4, 9, 16, 25 etc, you would get a pyramid roughly 25-30m tall, and probably quite a bit wider (because skulls are sort of irregularly shaped). Which is impressive but not, like, skyscraper-sized.
For comparison, the famous Christ the Redeemer statue that looms over Rio de Janeiro is 30 meters tall.
But the pyramid of Giza is something like 150m high and 200m wide.
For visual impact, I think you could instead just clad a giant pyramid in skulls, or make a bunch of smaller pyramids and put them all together in a large flat plain.
I did the numbers.
For your skull-clad pyramid, you put the skulls upright next to each other. We again use the 17.5cm width, but now each step is 25cm high. The number of skulls per layer goes 1, 4, 8, 12, 16, 20 etc. This one is a monster: 194m tall and 135m wide. I'd suggest making it look more like a Mayan pyramid, so that you can add a bit of width and make it a nice round 200x200x200m. Estimating cost is a challenge, ad building costs are so variable and depend on so many factors (land cost, labour costs, transportation costs, material costs etc). If built analogously to a skyscraper, then the cost per might be anything from $450 million to $1.8 billion (assuming a floor height of 4m and cost-per-floor at Dubai vs New York prices). Just as a sanity check, the bulk cost of the concrete alone should be something like $160 million (at around $134 per m3 and assuming you fill the whole pyramid up with it), so I suspect the higher price is more accurate. If you don't want to use real skulls, then I found ceramic skulls on sale for around $8. Assuming a bulk discount ($4 per skull), this makes them by far the cheapest part of the whole exercise ($4.8 million). The folk from construction physics can probably math this out a lot better than I can.
My final suggestion, since we've made a flat-topped pyramid, is to have a chamber of important covid-related figures' skulls in the heart of the pyramid, and put the inevitable gift shop and restaurant on the roof.
In re: the discussion above about the true rationalist commenters vs the riff raff at this-a-here blog, this comment is first ballot Hall of Fame True Rationalist material.
This comment alone makes me glad for this whole mess of a comments section.
Would like to express my agreement with this sentiment. Cry6Aa's comment was a delight to find (and will be getting relentlessly linked to my circle of friends).
To your point: the doctors and nurses I know were despondent and exhausted during this time.
Their hospitals were over capacity, they were working constant overtime and people were dying left and right.
The feeling among their crowd were they were making this enormous sacrifice every day and nobody seemed to care.
And statistically both groups of people are entirely justified in their choices. As you note, medical personnel in hospitals were totally overwhelmed, because the healthcare system of every country on earth runs on fairly tight capacity margins, so a pandemic that increases the baseline rate of ER visits by (totally making up numbers, but approximately correct) 50% is going to totally swamp the system. It was like a never-ending mass casualty event.
But for the average person, who is younger than 65 with few to no comorbidities the approximate risk of dying from Covid was zero. Compared to the many other things that tend to kill people younger than 65, it didn't even make the top 10 causes (approximately 260k covid deaths in the under 65 cohort, and 80% of that is the 50-64 cohort). So you have this massively disruptive event which altered or destroyed the careers, families, and social circles of a large number of people who were at very low risk in the first place, and you can rather correctly say we locked down to save all of the grandmas at the expense of their granchildren.
Its only natural people would feel resentment in either group. But neither of them is wrong, and the entire discussion around covid is has to take one side or the other.
I know this is not the view you're arguing for, but this we/them frame in the face of a huge shared calamity speaks to the real dissolution of a functioning society to me.
Besides the young people resenting the old people, there are millions of people with chronic illness in any given year, millions undergoing cancer treatment with wrecked immune systems, bunches of newborn vulnerable babies, bunches of pregnant women who are also medically vulnerable. Bunches of people without great access to healthcare or who can't economically afford to miss work. *And then there are all the people who love them.*
And it makes me worry that these guys who are so resentful of the price they paid during a national calamity don't have anyone to love. Because if you love someones in your life, chances are one or more of those someones are vulnerable in some kind of way, and you would feel eager in the face of such a threat to have things to do that would make the ones you love be just a little less likely to get really sick or die.
The tradeoffs in some situations were absolutely terrible, no doubt. But all of life requires tradeoffs. And exceptional times require some really big ones. None of us is promised smooth sailing for any amount of time, ever.
Fine, life requires tradeoffs. The tradeoffs made during covid were terrible, maximizing suffering and minimizing flourishing. The people responsible for those decisions should not be taken seriously.
Like many commentors here, I do not have any direct acquaintances who died of Covid. Hopefully unlike many here, I know multiple people who chose to take their own lives during lockdown, and I am convinced within a reasonable margin of absolute certainty that the isolation is what killed them. So in the face of your medically vulnerable person, I offer the emotionally vulnerable person, who was forced by threat of government force to deal with their issues alone.
You frame the issue as if its a bunch of old white conservatives who were anti-lockdown. My experience is rather the opposite, and while I have sympathy for the medically vulnerable, I have none for the decision that society must revolve around their needs at the expense of everyone else.
That's a really profound kind of experience to have people you know commit suicide and it's really awful that that was part of your Covid time experience, that direct trauma of losing people that way during a time that was already horrible.
I'm sure you're right that at least part of the reason for the suicides we saw increasing during that time were due to social isolation. For the healthcare workers who committed suicide, it seems like those were maybe fueled by the moral injury and overwhelm of being a healthcare worker at that time.
I'm a psychologist who works with individuals and families and so I had a front seat view to the small, medium, and enormous mental health impacts of peak Covid time. In addition to the people who suffered from terrible isolation (young and old), I think the tradeoffs were also especially horrible for children and for parents of children.
I also had a prolonged non-Covid physical health threat that emerged for me during Covid (cancer) and so I had a front row seat to the impacts on hospital-based workers beyond those I know from my private practice. And I saw how various doctors and nurses and other staff carried the weight of tradeoffs, everyone carrying a slightly different mix of burdens and trauma. I also worked with quite a few patients who had severe social anxiety who experienced a couple of years of real relief from it for the first time in their lives as a result of the permission to stay home and the opening of working from home. So while Covid and all that came with it was a horror for mental health broadly speaking, there are also other stories in there as well.
I think the only difference between you and me on this topic maybe is that I took from my experience of that time the sense that most everyone is doing the best they can all the time. That humans and human systems have forever been very flawed. That there's a real limit on our capacity to make good decisions in an environment of high risk and high uncertainty and where the conditions are rapidly changing. I see those limits and tradeoffs as built into the human condition and into the reality of the world we live in.
And then I witnessed up close so so many people daily doing acts of generous and heroic kindness -- including to me personally hundreds of times -- that despite the losses I experienced and witnessed, I also feel a profound gratitude to be here, to be alive in this world, as messed up as it is. I feel a lot of grace is due to all the people who worked really hard to take care of others in a situation none of us has ever been in and for which there was no adequate road map (including because we failed to do the planning for this that had been mapped out decades ago). Long-term planning being another weakness of humans.
I also think criticism is important and it's key that we all in various ways reckon with the losses, the tradeoffs, and the costs of the different choices made in different times and places through the pandemic. I see the value of this not in assigning blame but in preparing us to do it better next time. To me that means wrestling with the nuance and complexity rather than drawing quick judgments and conclusions. Bringing curiosity, in other words.
There's a view I think that everyone who was in any way on the front lines of responding to the pandemic *should have known better* about a whole range of things, *should have communicated better* and should have overall *been better people*. And of course there were in there here and there some truly venal and opportunistic people as well as the usual mix of people of mediocre competency. To me, this is where humanity is as a whole at this moment. We would like everyone to be better and having those wishes for humanity is good. But we live in this world where people are how they are -- an incredible mixed bag. Glory and horror all wrapped up together. I have no confidence that if you'd dropped me into any other frontline role than the one I had that I would have done better than those people in those positions. I had my hands absolutely full with the position I had and was bending under the weight of it every day until it kinda broke me.
I guess I want to say there is relief in accepting this. It doesn't mean don't ever be angry or give up and don't do anything to make things better. But accepting all this as where we are as humans means that one's actions can move better from a place of calm and compassion and the choices we make from that place tend to be better ones than the choices made from rage. And also it feels a whole lot better.
I think (I don't know if I'm right) that a lot of the political discord we've seen since Covid (though it obviously was well along already) was fueled by the lack of official acknowledgement of the need for grieving, reckoning, criticism, and taking responsibility. I could write a huge proposal sitting right here with ways I think institutions and governments needed and still need to create official space for people to grieve and recover as well as to reckon and criticize.
Grief unprocessed can lead to both rage and depression (as well as a whole host of other health problems) and my country at least (US) has done a very poor job of making space for grief. Not because people suck so much as that historically as a culture and an economy, we have not prioritized emotional health and connection. Instead we have prioritized economic productivity, accumulation, and status. While at the same time, we didn't do the work over the past decades to set up our institutions to respond well to this specific kind of calamity, which then exposed a lot of people to the particular foibles and weaknesses of individual humans even more than would ordinarily happen in a crisis.
I think you are entitled to whatever feelings you have about what went on during Covid, how all of it impacted you and the people you cared about. If it leads you mainly to anger and blame, I guess I want to say have hope, there is a place on the other side of that that feels better. And it runs through you validating everything you feel and felt about the pandemic and not from minimizing it (which I don't hear you doing).
Anyway, those are some of my thoughts.
Well here's the thing. If there were a new shocking ear worm parasite that was 100% fatal but only infected people the day before they were going to die of something else, that ear worm parasite could kill millions of people in a year and cause the largest fatality event in US history, and it still wouldn't really matter because those people were going to die anyway.
Now extrapolate.
Excess deaths that year would be close to 8,000 because that is the number of deaths in one day. If the parasite moved all deaths forward by two days then excess deaths would be close to 16,000. In order to get excess deaths up to 1 million you have to move deaths forward about 100 days.
I think by that time, you’re talking real sacrifice - people lose several months of time with literally everyone they love.
It’s not as bad as a war that kills tens of millions of people decades early and thus leads the average person to lose several years of life, but I think we get a sense of the magnitude.
Ooo good point.
I think it's reasonable to assume that a very large number of people who died from Covid had their deaths moved up one to two years, and a few moved up around five.
Emergency rooms and hospitals were flooded. I know a PA who pivoted to work directly with COVID patients pretty much from the start and it was an absolute nightmare for them.
It's not cognitive dissonance, it's survivorship bias. It was "only" 1.2 million deaths because various measures were taken. It's like saying the Y2K bug was overblown because nothing really happened come Jan 1st, 2000 - but that was because legions of programmers had been employed to actually fix the bug in affected systems. That is always the curse of "preventing bad things from happening" as opposed to "making good things happen" - if you prevent something bad from happening, it can look like you've done nothing at all.
Some one up above, about the harm of government-issued NPIs, pointed out that there would still be a lot of ad hoc NPIs done in the baseline.
That applies in the other direction, too. Even for a disease that genuinely kills 1% of a population acting like normal, people will react and restrict contact in a way that reduces the impact, even with no government-led interventions or vaccines.
Right. If no official measures had been in place, it probably would have been 1.5 million deaths. No official measures except vaccine development, 1.25 million deaths.
The observed death rate isn't that far off what you'd get if you postulate the disease running unchecked until the population reached herd immunity, and even in the absence of official countermeasures you'd have lots of people taking private ones. That 1.2 million people died *anyway*, means there isn't really much room for the official measures to have done much good.
And then there's the world where we imagine no official measures but also no official impediment to vaccine deployment and distribution, and I'm pretty sure the death toll is well under a million in that one.
I'm not sure how you are deriving these numbers (personal intuition?), but this isn't what you see in epidemiological publications at all.
Also, herd immunity wasn't ever achieved. We don't have herd immunity right now. Things are much better than they were from 2020-2022, but COVID is still circulating, acting like quite bad influenza waves crashing against us. The disease burden is substantial. Turns out our immune system's limitations + viral evolution is able to frustrate herd immunity for this disease. That hope was decisively dashed in late 2021 when it became clear that developing an immune response through vaccination / exposure was insufficient to produce lasting immunity in the general population.
Even assuming these numbers are true, without getting into details like your confidence in them and whether they would have been available at the time when they would have mattered: Would you have decided "Eh, only 300k additional deaths, that's not worth a lockdown"? Especially if it was a matter of public record and you had to expect someone eventually writing nasty headlines about your decision?
I'm fairly confident in those estimates, for reasons that I've discussed here too many times to want to repeat. But as to your question, oh hell yes. First off, if you read my statement carefully, that's 250,000 lives saved (US only) by vaccines, and 50,000 by mandatory lockdowns, masking, social distancing and all the rest. If anyone here is assuming that because I opposed lockdowns I'm also an antivaxxer, then please just shut up and go away.
And second, even setting aside concerns about pesky little things like civil liberties, and the economic cost of the lockdowns, and the effects on the mental and social health of American adults, if you ask me to trade fifty *million* children going through a year and a half of half-baked zoomschooling, against fifty thousand mostly septuagenarians dying five years earlier than they otherwise would have, then there's no contest.
Even if we assume 99% of the kids came through just fine, and the 1% that don't suffer only a 1% reduction in QALYs per year from the educational shortfall, the lockdowns fail the consequentialist math. And I'm not a consequentialist.
You haven't answered my question whether any of these numbers (1.2M, 1.25M, 1.5M) would have been available in early 2020 at the same confidence. Hindsight is always easy.
Oh, sorry, yes I missed that. These numbers would definitely have been available at the time, if someone had wanted them. I mostly took these numbers from my own analysis in 2020. and that of people like Zvi Mowshowitz, and the subsequent observation that the actual infection and death tolls were tracking pretty closely with the "we fuck it all up and our interventions are useless" predicts right up until the vaccine rollout. At which point there was a substantial drop in mortality, though less so with infections.
The thing that should *not* have been available with confidence in early 2020, was the belief that lockdowns, mask mandates, and "social distancing" were going to dramatically reduce the death toll. Those were unprecedented interventions, with no proper scientific studies to back them up and based on multiple dubious assumptions.
As best-guess "maybe this will work; let's put it out there as a recommendation" measures, sure, worth a try for all we knew in 2020. But an intervention that is intrusive, mandatory, and controversial, is one that really needs to be backed up by the sort of science we can be confident in.
The distortion of reality where people pretend that hospitals weren’t flooded is ridiculous. They were. It was insane. I’ve never seen anything like it before or since.
Also It’s pretty clear why the actual flu disappeared — enough people were wearing masks that it couldn’t spread. Note that flu is spread via droplet transmissions so simple masks are highly effective against it
I think a lot of the comments section is missing the concentration thing. New York City had mass graves. A random town in Iowa may have seen zero cases.
I don't know if it's clear any respiratory disease spreads via droplets rather than aerosols now. COVID is just more transmissible than the flu (measles being super-transmissible is why we're starting to see outbreaks now that vaccination has dropped a little).
Hospitals were either flooded or empty depending on where you were.
NY hospitals got demolished because Wuhan is where all the garments are made, and the places that got bombed are tied in with the fashion industry. Wuhan, Milan, London, NY. It spread after Fashion Week. One or several Chinese fashion industry execs were Patient Zero.
Statistically, nationwide, hospitals were not overwhelmed. In fact most were under-occupied because they were saving beds, or at least that was the argument they made when they claimed taxpayer bailouts. There were local anecdotes in certain places, New York most prominently. But even there, they brought in the hospital ships and built out extra capacity and nobody used it.
Meanwhile, against the anecdotes of people claiming hospital flooding, there were the three hospital visits I made during the Covid period, twice with my mother, once with a child. The places were ghost towns. I've never seen the ER so quiet and with so few patients. Like three people in a space that normally seated 200.
I'm not claiming those observations necessarily generalize. But neither do the anecdotes of flooding.
Hospital overload is like traffic congestion. 90% of the day, highways are empty. But at the time that everyone wants to drive, they get crowded. Hospitals were under capacity for most of the pandemic because there were plans to try to save capacity - but when a big wave hit anywhere, they got flooded. (See New York in March 2020, Houston in June 2020, etc)
Even in New York, they never used really overflow capacity. The Javits Center and hospital ships never saw more than a handful of patients.
COVID turned out to be significantly more contagious than influenza. The globes' attempts to mitigate COVID spread actually appears to have wiped out one of the major circulating influenza lineages, which has simplified the vaccination formulation going forward.
The excess deaths chart above does NOT look like "a bad flu season", unless you mean something like Spanish Flu. Actual flu disappeared because measures that failed to stop COVID reduced R0 enough for them, enough to kill off some old flu strains https://www.livescience.com/flu-virus-types-extinct-covid-19.html
During the Delta wave hospitals in my area -- an area with low vaccine uptake -- got slammed and had to bring in hundreds of medical personnel from out of state.
During this same period my local state representative was posting antivax crap on Facebook.
Why do you demand perfection from experts but ignore the political partisans who got people killed?
Delta hit red areas hard because the red areas dodged the earlier waves and had lower herd immunity than the northeast did. NYC fared very well during Delta because they all had immunity from earlier infections.
This needs to be analyzed regionally and without political bias.
My county had a vaccination rate lower than 30%. Even today it's only 42%. A higher vaccination rate would have made a difference.
I higher vaccination rate among people under age 35 would not have made a difference. A higher vaccination rate of people over age 65 would probably have made a tremendous difference. And neither rate matters much now that the vaccine is less than 40% effective at stopping spread anyway.
Okay man. I just think a lot of old people were listening to people like these guys on the radio.
https://www.nbcnews.com/politics/politics-news/trio-conservative-radio-hosts-died-covid-will-their-deaths-change-n1278258
Or maybe they were listening to Brian Seitz, my local state rep who was anti vax.
https://www.riverfronttimes.com/news/in-missouri-legislature-anti-vax-calls-come-from-inside-the-house-36049972
(After that made news I called him an complained because multiple tenants from my property in Branson were in the hospital with COVID. One of them was under 40 and overweight but not morbidly obese. Fortunately they all survived.)
Or maybe they were listening to these people right as the Delta wave was surging. This is literally at my local hospital.
https://www.news-leader.com/story/news/politics/2021/08/16/anti-vaccine-protest-held-outside-springfield-mercy-hospital-missouri-covid-19-misinformation/8130578002/
Just a few weeks earlier they were turning away patients because they were overwhelmed. Fortunately they brought in out of state personnel by August.
https://apnews.com/article/mo-state-wire-michael-brown-springfield-business-coronavirus-pandemic-fea74a1cc80d3bd632aa29d417f2aa87
I can retype "a higher vaccination rate of people over age 65 would probably have made a tremendous difference" if you like. Otherwise I'm not sure what your point is.
Remember that the “immunity” New York had during delta wasn’t just from higher vaccine uptake and earlier immune system encounters with the virus - it was also from behavioral changes triggered by living in a place that was hard hit. The brain is the biggest organ in the immune system, just like it’s the most sensitive erogenous zone.
Unless NYC shut the subway down during Delta then NYC has infinite vector potential no matter what other "behavioral changes" they tried to implement. No amount of separation is going to tamp down being stuck in an enclosed poorly ventilated space, given what we know about how the virus is transmitted by stale air.
You don’t have to shut down the subway to reduce spread in the subway - as long as fewer people are using the subway it will contribute less spread. (Presumably quadratically, because spread is proportional both to the number of infected people present and the number of uninfected people present).
I suppose a reduction in subway ridership may have had some impact, but the fact that basically the entire city all got Covid at the same time six months earlier probably had a dramatically larger effect. And that effect wasn't present in the red areas that had their first wave during Delta. These are areas, keep in mind, without anything remotely approximating a subway.
Did you pay attention to the refrigerated trucks and patients spilling out of New York City hospitals into tents in parks in winter 2020?
Well, there's also the hospital ship that got sent to New York to handle the expected overflow and then turned out not to be needed.
Possibly because of mitigation steps, the rise and fall of the wave and early, limited acquired immunity by the time the ship showed up.
Maybe 🤷♂️
My point is that if we're talking about people's perceptions, the news stories that would tend to show that conditions were really bad were often balanced by news stories that showex that they were not nearly as bad as anticipated.
Yes, there were errors of oversupply and errors of undersupply. It’s important to count them both, and weight them by how much harm they did.
...those scenarios were what would happen *if we didn't take precautions* to slow the spread.
Then we *did* that, in a massive way, and things turned out better.
This is classic 'Why do I need this umbrella, I'm not even getting wet' thinking.
> we should expect to see hordes of dying people staggering the streets because emergency rooms and entire hospitals would be flooded, then the hordes' corpses to be stacked like cord wood awaiting a mass bulldozer burial
This more or less DID happen in New York City, so I don't think it was unreasonable for people at the time with imperfect information to think it may happen in other places. I was turned away from the emergency room and told not to come back unless I thought I was going to die.
> actual flu disappearing for some mysterious reason
This is such a perfect encapsulation of covid denier idiocy. "It's a mystery that with everyone staying home and wearing masks and trying to prevent the transmission of a disease that is more virulent than the flu, there were fewer flu infections this year! The NWO doesn't want you to think about this!" I don't know, there's one explanation that jumps out as pretty plausible to me!
In my city, the hospital where my stepmom works had a network of tents set up outside the building to handle the overcrowding, and the (Republican) governor sent us several freezer vans to hold corpses until they could be properly autopsied because the morgue couldn't hold them all (and because the medical examiner was actually not just declaring every death a covid death if there was a positive covid test, but was ensuring every corpse had a proper investigation into cause of death).
It makes sense that COVID killed a million or so in the US. That's a 0.3% mortality rate, or 3x the flu. It also made plenty of sense to have a strict lockdown as an initial response to the virus. The question is why did we keep it shut down after we figured out that it was going to be an endemic virus and all of those people would die of it eventually anyway. And most damningly, why did we keep it shut down for another year after the vaccine came out? All the conspiracy talk is just a distraction from discussing the very real authoritarian nightmare that was inflicted on us by the administrative state. If you had asked people in 2019 if we should shut down the economy for two years and print trillions of dollars causing massive inflation in response to a virus only 3x worse than the flu, they would have thought you were insane.
It was more than 3x as deadly. About 500,000 died in the US that first year. The worst flu season of the 21st century only had 80,000 deaths in the US and many had more like 20-30,000. So we're talking 6x-15x depending on the flu season. And that's with all the measures we took to prevent transmission.
The "3x the flu" number is an average ranging from the beginning of the pandemic (more lethal virus, no vaccines) to the end (much less lethal virus, lots of vaccinated people).
Comparing it to actual flu waves, you also have to take into account that Covid was, and became even more so, much more infectious than the flu, such that sooner or later everyone got it, whereas each flu wave only hits a few % of the population.
But, yes, initally it was much more deadly than the flu (10x as lethal as the nastiest flu strains IIRC).
If you average it over a long enough time, you can just say that it had exactly the same mortality as flu (it converges to 1x as t=>oo). This reduces my sympathy in justifying the 3x number, significantly.
You're using a very weird definition of mortality. My point is not that more people die the longer you wait. It's that the probability of dying of the disease decreased with time (because the virus mutated into strains with lower mortality, and vaccines were rolled out), but for the initial measures, it's justified to base them on the mortality for the variants that circulated then in the absence of vaccines, and at that time "10× more lethal and twice as infectious as a the nastiest strains of flu since the Spanish Flu" was a reasonable estimate.
My comment was not intended as a rebuke. You said "The "3x the flu" number is an average ranging from the beginning of the pandemic...to the end...." and "[but initially] 10x more lethal." My point was just that averaging over a period of time (3x from beginning to end) eventually converges to 1x since you are averaging a finite peak with an exponential decay back to baseline. In other words, you could pick any multiplier between 10x and 1x and there exists a time period for which this is the correct average. The ability to chose any number raises the question of why one might pick "3x." Again, my comment was at least half tongue in cheek.
The virus was "less lethal" to people who'd been infected previously, which was an increasing proportion of the population over time.
That's part of it, but not nearly all. The Omikron variants had a significantly lower fatality rate, even among unvaccinated people, and by that time most people had not had the infection yet.
The difference was Covid-19 was NOVEL, not that it was that much more deadly. The difference was that unlike the flu, basically everyone caught it, all in the same year.
Flu x 3 is not a great way to look at it. If you were over 65, it was 7*[Flu], and if you were under 35 it was 0.5*[Flu] or lower.
That's because more people got it. The actual mortality rate of covid is 3-4x of the flu https://www.ahcancal.org/News-and-Communications/Blog/Pages/Flu-or-COVID-19---Which-is-Worse.aspx
Also, I'm pretty sure that the flu mortality statistics are overstated because there are so many cases of the flu that never get reported.
Note that just saying "it turned out fine so the precautions were unnecessary" is not valid reasoning. What you *can* argue is something like that countries that didn't do a lockdown had similar death rates, which might or might not be true. But my point is that the argument needs to be a little more complex than what you said.
Note that I mentioned in my comment that it was reasonable as an initial reaction. Erring on the side of caution is a reasonable approach to an unknown disease. But by June 2020, those initial unknowns were pretty much all known. Also, it didn't "turn out fine." It was essentially the worst outcome possible. Containment was a complete failure and 100% of the population got the disease.
I think some people didn't get it until vaccines were available.
Some of us never tested positive and have no reason other Bayesian priors to believe that we got it :-)
"What you *can* argue is something like that countries that didn't do a lockdown had similar death rates, which might or might not be true."
A reasonable pair for comparison (and *I* remember saying this in 2020/21 so this isn't hindsight bias, though I didn't write anything and you don't have to believe me ...) is Sweden and Denmark. Similar populations (though not identical) and Denmark locked down much harder than Sweden.
Reported Covid deaths per 1M:
Sweden: 2,682
Denmark: 1,511
https://www.worldometers.info/coronavirus/
I have seen analysis that claim that Sweden did *better* than Denmark, but that analysis requires sophisticated math and I am skeptical.
Disappointing, because philosophically I much preferred the Swedish approach of providing guidance to the population and expecting them to behave reasonably. It would be nice if that ALSO resulted in fewer deaths. But I don't think that it did.
The UK is a popular country to compare against Sweden by those who opposed the lockdowns and the UK experienced 3,389 reported covid deaths per million, but that comp tended to arise afterwards and I think Denmark is better.
if you look at excess deaths as opposed to reported deaths I believe Sweden comes out pretty dang similar to other Nordics other the whole pandemic period.
https://academic.oup.com/eurpub/article/34/4/737/7675929
I feel like you are cherry picking here. Sweden is right at the median for death rate of countries in Europe in your source. And Denmark is near the bottom. Unless you have some actual hard criteria here, like age structure of population, for using Denmark in particular I would say it's not really valid.
Buuut, you didn't keep it shut down. At what point did you have to stop getting permission from the local police to leave your house? That's when lockdown ended.
Needing permission from the police to leave your house is a lockdown. If you can just leave your house whenever you like, then you’re not locked down, so there’s no lock down.
We (England) had a lockdown from 26 March 2020 to 28 May 2020 and then again from 5 November 2020 to 8 March 2021.
Absolutely, but if you never had a lockdown, then it's unreasonable to complain about having had a lockdown.
No where in US was ever on mandatory police enforced lock down.
Yeah, I don't understand where this meme comes from. We had school closures, restrictions on indoor commerce (usually fine w/ a mask), and were encouraged to get vaccines. I know other parts of the world had stricter interventions and a bunch of lefties were enormous hypocrites about SJ protests, but even in NYC at the height of it, I don't think people were locked in their houses.
So the US never had a lockdown? So why are all these people complaining about a US lockdown?
You will justifiably find this highly annoying, but in the US, people often use "lockdowns" to refer to all the NPIs deployed here, collectively. This is true even though only a few of them could (arguably) qualify, like school closures.
You forgot the millions who were displaced, and the hydrants of thousands of businesses and jobs deleted due to the non-existent “lockdowns.”
Businesses lost their licenses if they opened.
That’s a lockdown.
I was in Italy at the time, and we were only permitted to leave our apartment to go to the grocery store for several months, and had to carry around papers certifying that we were going somewhere that was necessary for several more months after that. There wasn’t anywhere in the US that had this style of restriction, and it became common to use “lockdown” to mean “needing to wear a mask at the grocery store” in the US.
Because they are weak, soft people, who respond to any request that they act like responsible adults as a grave imposition.
Tens of thousands of people lost their businesses. Tens of thousands lost their careers. Hundreds of thousands had to relocate their families.
What the hell are you talking about?
Is it a “request” to shut down your business while strip clubs and Costco remain open?
Is it a “request” to be targeted by police if you want to walk around without a mask?
Is it a “request” to be blocked from accessing facilities if you don’t comply with nonsensical policy they just invited 3 seconds prior?
Is it a “request” to have to quit your job (as opposed to being fired for not getting an experimental mRNA transfection therapy) to watch your kids during school hours since the school won’t open for them, but your still get the joyous opportunity to pay for the closed schools as though they’re open?
I've been trying to figure out a concise answer to that question for 4 years....
In California, it was illegal for anyone to leave their home for "nonessential" reasons from roughly March 2020 through September 2020. And if you did have an "essential" reason to leave, there was a very limited set of places to go.
No, they didn't literally have a policeman standing outside everyone's door to check. If you were willing to lie about what was "essential", you would almost certainly get away with it. I'd still consider that a lockdown.
0.3% is also consistent with the Diamond Princess, as close to a laboratory environment as possible, where the mortality rate was around 0.38% (~14 deaths, ~3700 people on board).
Yes it is amazing how well that initial death rate estimate has held up!
> "All the conspiracy talk is just a distraction from discussing the very real authoritarian nightmare that was inflicted on us by the administrative state."
I don't think it's a distraction. I think that given that ~1/5th of the ACX comment section seems to believe this false thing, it is a separate but also-serious problem, and it's worth talking about it. We can't have a sensible discussion about how bad the real facts are until we agree on them!
People who disagree are going to be the loudest by far. If it's 1/5th of the comments, it's probably actually 1/25th of the readers.
> The question is why did we keep it shut down after we figured out that it was going to be an endemic virus and all of those people would die of it eventually anyway.
> If you had asked people in 2019 if we should shut down the economy for two years
We didn't do that! Restaurants reopened before the end of 2020, despite it being obviously impossible to eat in a restaurant without spreading a respiratory disease. Real "shut down everything in-person" restrictions only lasted six weeks before economic reality and cabin fever forced people to reopen while pretending that viral particles couldn't move more than 6 feet indoors.
Scott should implement a rule: when discussing COVID lockdown everyone should have to preface each comment with the jurisdiction in which they lived so we can see what actual measures they lived under.
I will start this grand tradition and admit that I lived in Florida and mostly self-isolated the first three months. I remember how ghostly downtown was when it really hit people that this was happening. I also remember most people shrugging and moving on with their lives within around four months with maybe half masking consistently. Given Florida's place in history as the flagship of low-restriction state governance, my experience may not resemble that of blue staters.
We never once shut down the economy. As far as I can tell, we shut down 5% of the economy (GDP decreased by 2.2%, while in other recent years it had been growing by about 2.4% per year, so at most 5% of the economy was shut down.)
If you want to correct someone’s exaggeration, try not to exaggerate in the opposite direction.
It was a lot more than 5%. They just made up for it by printing trillions of dollars and handing it out. This, of course, caused the inflation mess. If Biden hadn't been such an idiot and lifted all the economy at the beginning of 2021 we probably would have had a much better recovery and less inflation. We probably wouldn't be dealing with Trump today.
What have excess deaths looked like in the years since Covid, 2023 and 2024?
If I weren’t so lazy, I would look this up. I can imagine a scenario in which this number is low or negative, with fewer deaths in these years than expected. And if that were the case, I would think of Covid as having accelerated the deaths of the difference. The people meant to die in 2023 and 2024 died in 2020 or 2021 instead.
And that might explain why it just doesn’t feel like a 1.2 million deaths tragedy.
You can indeed look this up. Your hypothetical is incorrect
https://ourworldindata.org/grapher/excess-mortality-p-scores-average-baseline?country=~USA
This is the worst graph you could use to prove your point because it compares different cohorts of different sizes.
Death is strongly correlated with age (duh). If everyone dies at age 75 then you would expect a difference proportional to the difference in birth rate between 1940–1944 and 1945–1950, and those are really different periods because of World War 2 and the following baby boom. In reality it's more smoothed out of course but this still seems too big of a factor to ignore.
So from the graph alone I really cannot tell if the ~10% increase we are seeing in 2024 compared to 2019 is above or below expectation based on changes in demographics between those years.
And at least eyeballing it, the data from early 2025 seems higher compared with 2023 and 2024 which at least suggests a dip in those years.
If you look at the comment I’m responding to, this graph exactly addresses the question there. You can create scenarios to explain what the graph shows, but the comment asked if excess were down after Covid and the answer is clearly no.
The question was whether there were fewer deaths in 2023 and 2024 “than expected” (which would indicate a pull-forward effect of mostly older patients dying).
”Than expected” doesn't mean “than 5 years ago” if there are reasons to expect the death rate to be higher than 5 years ago regardless of COVID (i.e., the baby boom).
So no, it doesn't answer the question, at least not convincingly.
And you do need strong evidence if you want to deny there is a pull-forward effect, since we know that COVID deaths disproportionally affected the elderly (and secondarily the obese, who generally have lower life expectancy as well), which nobody denies, so purely statistically you would expect the death rate post-COVID to be lower “than expected”, everything else being equal.
It seems to me the "pull-forward" theory proponents should actually present some actual evidence their theory is accurate other that just saying it makes sense so it must be true. If you have a better graph I would love to look at it. Given the oldest boomers are around 80 years old, and their life expectancy is also around 80 (although really more given the life expectancy for anyone at age 75 is around 12 years), a 10% increase in deaths seems highly, highly unlikely without some other cause.
I've given you a perfectly reasonable explanation why the pull-forward theory is likely, you just chose to ignore it. It relies only on facts that I assume you already accept to be true: that COVID disproportionally killed people with shorter life expectancies, namely: the elderly, the obese, the infirm.
Then it logically follows that the survivors have higher than average life expectancy, resulting in a reduced death rate in the following years, if everything went back to normal.
Of course there is an “if” there. You could argue that the introduction of COVID has permanently increased the human death rate which offsets the expected short-term decline (which would mean life expectancy at birth is lower post-COVID). This is plausibly true at least to *some* extent, but if you want to make a specific claim about how strong that effect is the ball is back in your court. Also this is not actually an argument against the pull-forward effect per se; it's just an explanation why the death rate doesn't go below average levels *despite* the pull-forward effect.
Note that the effect on the death rate of the pull-forward effect is likely to be small because deaths don't very neatly occur at the end of life expectancy.
What, exactly, are you arguing against? The pull-forward theory is that people who died of covid would otherwise have died later, AKA, they would not have been immortal. Really the only alternative to this is immortality.
And why are people so excited about the pull-forward theory, anyway? There seems to be an implication that it makes things ok again, and yet, dying sooner is worse than dying later (assuming quality of life above 0). All sorts of humanity's most celebrated achievements - agricultural revolution, eradication of smallpox, antibiotics - only delayed death, rather than preventing it entirely.
I started writing this comment to Calvin, but then threw it out when I realized it didn't contradict him, but I guess I'll resurrect it now.
==
Let's say we have 10M elderly and we expect 5% to die each year, so 500K deaths. Also 500K people age into elderly, nice and even.
Then a virus comes along and kills an extra 10% in year N, so we're down to 9M next year. However, out of these 1M deaths, it was strongly correlated to the worst health, so 250K of them were going to die in year N+1 and 250K would die in year N+2 and 250K would die in the year N+3. The remaining 250K were from the otherwise healthy elderly.
Now, in year N+1, we would normally expect there to be a population of 10M and 500K deaths, but 250K of those deaths happened in year N. So we have a population of 9M and 250K deaths.
That's a death rate of 2.78%.
So we would really see a dip.
I'm a bit confused if you are agreeing or disagreeing with me.
What you describe is essentially what I'm arguing: that since COVID killed people with relatively low life expectancy, we should logically see a reduction in death rate post-COVID, barring other effects.
It sounds like you're arguing that if the pull-forward effect is significant, then we should see a stronger dip than we've seen, and that argues against a pull-forward effect. But I'm not sure we have accurate enough data to tell whether there is a dip, as I've explained above.
Note that you used exaggerated numbers to get the desired effect: a 200% increase in deaths during COVID while more realistic estimates are 20-25% (though over roughly a 2-year period). Additionally, you assume ¾ of victims would have a life expectancy of only 2 years; I think it's plausible the average was higher but still far below average overall (say, 5 or 10 years) which would make the dip shallower and wider.
I understand you were using the example to explain the principle, and didn't claim your numbers were accurate, but we do need to use accurate numbers to estimate the size of the dip.
Even with your numbers, the dip is significantly smaller than a peak: a 200% increase followed by 44% decrease over the next three years. If it turns out that e.g. a 25% peak results only in a 5% dip in the following years, that wouldn't necessarily be visible in a low-resolution graph that includes lots of other phenomena too.
If anything, it's up on trend. Total deaths in USA: 2.71 million in 2015, 2.84 in 2019 (+0.13), 2.98 in 2023 (+0.15).
https://ourworldindata.org/grapher/number-of-deaths-per-year?time=earliest..2024&country=~USA
It's actually on or slightly below the trend started from 2010: https://i.imgur.com/0N0TmQR.png
By the way you have to account for the fact that the US has an uneven population pyramid, with many Baby Boomers who are going to die in the coming decades, which is why I suspect the numbers are trending up in the first place.
I'll grant you that there is no obvious decline to compensate for the COVID peak, but the graph is also conveniently missing the data from 2024, which is really important: does 2024 continue the sharp decline from 2021–2023 or continue the upward trend from 2010? That seems a really important question we cannot answer with this data.
And as I said, there is significant demographic change as well, so ideally I'd like to see this data controlled by age.
I was curious about that as well, and so, as a midwit, i did what every midwit should: I asked an AI (Grok):
Here was my question:
this link shows deaths due to covid. https://www.census.gov/library/stories/2022/03/united-states-deaths-spiked-as-covid-19-continued.html however, the data ends at 2021. Im wondering if deaths dropped significantly after the pandemic due to the deaths being concentrated amongs those who would have died soon anyway. Is there any evidence of that theory?
Ill copy its answer in a reply.
The idea you're referring to is often called the "pull-forward effect" or "mortality displacement," where a spike in deaths during an event like a pandemic reduces subsequent mortality because many of those who died were already near the end of their life expectancy. Let’s examine the available evidence to see if there’s support for this theory regarding COVID-19 deaths after 2021, using data and trends from reliable sources.
### Background from the Provided Link
The Census Bureau article you referenced confirms a significant spike in U.S. deaths during 2020 and 2021 due to COVID-19. It notes a 19% increase in deaths from 2019 to 2020 (from 2,854,838 to 3,390,029), the largest in 100 years, with deaths remaining elevated in 2021 (19.7% higher than 2019). The data highlights that older adults, particularly those aged 85 and older, accounted for a disproportionate share of deaths, suggesting that many fatalities were among those with pre-existing health vulnerabilities. However, the article’s data stops at 2021, so we need to look at more recent sources to assess post-2021 trends and the pull-forward effect.[](https://www.census.gov/library/stories/2022/03/united-states-deaths-spiked-as-covid-19-continued.html)
### Evidence on Post-2021 Mortality Trends
1. **Decline in Overall and COVID-19 Deaths Post-2021**:
- **2022 Data**: According to the CDC’s provisional mortality data, the total number of U.S. deaths in 2022 was approximately 3,273,705, with an age-adjusted death rate of 832.8 per 100,000, a 5.3% decrease from 879.7 in 2021. COVID-19-associated deaths dropped significantly to 244,986 (7.5% of total deaths) in 2022 from 462,193 in 2021, a 47% reduction. This indicates a notable decline in both overall and COVID-19-related mortality after the pandemic’s peak.[](https://www.cdc.gov/mmwr/volumes/72/wr/mm7218a3.htm)
- **2023 Data**: The decline continued into 2023, with provisional data reporting 3,090,582 total deaths and an age-adjusted death rate of 750.4 per 100,000, a 6.1% drop from 2022. COVID-19 deaths fell to 49,928 (1.6% of total deaths), a 68.9% decrease from 2022. This suggests a sustained reduction in mortality as the pandemic’s direct impact waned.[](https://www.cdc.gov/mmwr/volumes/73/wr/mm7331a1.htm)
2. **Natural Increase Rebounding**:
- A Census Bureau report indicates that natural increase (births minus deaths) began recovering post-2021. In 2021, natural increase was only 144,013 due to high mortality and a 3.5% drop in births. By 2022, it rose to 236,155, and by 2023, it reached 531,432, driven by fewer deaths and increased births. The number of states with more deaths than births (natural decrease) also dropped from 25 in 2022 to 19 in 2023, suggesting a return toward pre-pandemic mortality patterns. This rebound could be consistent with the pull-forward effect, as fewer deaths in later years might reflect the prior loss of vulnerable populations.[](https://www.census.gov/library/stories/2023/12/state-population-estimates.html)
3. **Excess Mortality Trends**:
- Excess mortality, defined as deaths above expected levels based on pre-pandemic trends, provides insight into whether deaths dropped below normal after the pandemic. A study published in 2024 found that excess deaths declined from 655,735 in the first pandemic year (March 2020–February 2021) to 586,505 in the second (March 2021–February 2022). However, for younger individuals (under 65), excess deaths increased in the second year, particularly for non-COVID causes like alcohol, drugs, and homicides, suggesting that the pull-forward effect may be more pronounced among older populations.[](https://pmc.ncbi.nlm.nih.gov/articles/PMC10988222/)
- Another analysis noted that excess mortality was highly concentrated among older age groups in 2020, with those aged 85 and older accounting for 34% of excess deaths despite being only 3% of the population aged 25 and older. This supports the idea that COVID-19 disproportionately affected those closer to natural life expectancy, potentially reducing subsequent mortality.[](https://www.census.gov/library/stories/2021/02/indirect-impact-of-covid-19-results-in-higher-pandemic-death-toll.html)
4. **Pull-Forward Effect Discussion**:
- A post on X by @Voltaire765 references the pull-forward effect, citing Alex Berenson’s Substack, which argues that after the initial COVID-19 death spikes, mortality should drop below normal as those who died early (primarily elderly and sick) were no longer at risk. The post suggests that five years post-COVID, mortality data in the U.S. and Britain might reflect this. While this aligns with the theory, X posts are not conclusive evidence, and Berenson’s claims require scrutiny against primary data.
- A study from Boston University suggests that in some regions, like New England and mid-Atlantic counties, lower-than-expected deaths in 2020–2021 could be partly due to the pull-forward effect, as vulnerable individuals died earlier from COVID-19, reducing subsequent mortality from other causes. However, this was not universal, and undercounting of COVID-19 deaths in some communities complicates the picture.[](https://www.bu.edu/sph/news/articles/2023/covid-19-deaths-in-the-us-continue-to-be-undercounted-research-shows-despite-claims-of-overcounts/)
### Critical Analysis of the Pull-Forward Effect
- **Support for the Theory**:
- The significant drop in COVID-19 deaths from 2021 to 2023, alongside a decrease in overall death rates, aligns with the idea that the pandemic accelerated mortality among vulnerable groups, particularly the elderly. The high proportion of deaths among those aged 85+ (e.g., 307,169 COVID-19 deaths from January 2020 to June 2023) supports the notion that many who died were already at higher risk due to age or comorbidities.[](https://www.statista.com/statistics/1191568/reported-deaths-from-covid-by-age-us/)
- The rebound in natural increase and fewer states experiencing natural decrease by 2023 suggest that mortality is stabilizing, potentially because the pool of high-risk individuals was reduced during the pandemic’s peak.[](https://www.census.gov/library/stories/2023/12/state-population-estimates.html)
- **Counterpoints and Limitations**:
- **Non-COVID Excess Deaths**: While COVID-19 deaths plummeted, excess deaths from other causes (e.g., drug overdoses, homicides) increased, particularly among younger populations, which offsets the expected drop in overall mortality. This suggests that the pull-forward effect may not fully explain post-2021 trends, as societal disruptions from the pandemic contributed to other mortality risks.[](https://pmc.ncbi.nlm.nih.gov/articles/PMC10988222/)
- **Data Completeness**: Provisional data for 2022 and 2023 are incomplete due to reporting lags (1–8 weeks), which could affect the accuracy of death counts, especially for recent periods. This makes it challenging to definitively confirm a below-normal mortality drop.[](https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm)[](https://www.cdc.gov/mmwr/volumes/72/wr/mm7218a3.htm)
- **Undercounting**: Research indicates that COVID-19 deaths were undercounted, particularly in minority and low-income communities, which could mask the true extent of mortality displacement. If excess deaths were higher than reported, the subsequent drop might be less pronounced than expected.[](https://www.bu.edu/sph/news/articles/2023/covid-19-deaths-in-the-us-continue-to-be-undercounted-research-shows-despite-claims-of-overcounts/)
- **Vaccination and Immunity**: The sharp decline in COVID-19 deaths after 2021 is also attributed to widespread vaccination and prior infections, which reduced mortality risk. This confounds the pull-forward effect, as lower deaths may reflect improved immunity rather than solely the loss of vulnerable individuals.
### Conclusion
There is evidence supporting the pull-forward effect post-2021, particularly among older populations. The significant decline in overall and COVID-19 death rates from 2021 to 2023, combined with a rebound in natural increase, suggests that the high mortality during 2020–2021 may have reduced the number of vulnerable individuals, leading to fewer deaths in subsequent years. However, this effect is not universal across all groups, as younger populations saw increased non-COVID mortality, and factors like vaccination, undercounting, and data lags complicate the analysis. While the theory holds some weight, especially for the elderly, it’s not the sole driver of post-pandemic mortality trends.
For further details, you can explore the CDC’s National Vital Statistics System (https://www.cdc.gov/nchs/nvss/index.htm) for updated mortality data or the Census Bureau’s population estimates (https://www.census.gov) for natural change trends. If you’d like me to generate a chart comparing death rates or excess mortality across years, let me know!
Frankly, if people believing that this didn't happen at all is what it takes to keep things moving forward, so be it. This country can't afford to become more sclerotic than it is now. Lives come and go, and a few million are easily replacable when you have hundreds more to work with. It's time to move on.
I think this is a bad way to do this, and the fact people think it didn't happen is support for that argument - people don't believe this because the experts and public health officials who support it are liars who lie all the time. They don't...uh, lie in the way people think they lie, but they definitely massage the facts to fit their narrative.
And the result has been a loss of credibility and usefulness.
Giving false reasons to do good things turns out not to be reliable. It gives you goals other than getting good things done, and like a zany sitcom plot, those goals will eventually take up your whole attention, and your original goal will be tragically or comically abandoned.
(Also helpful to your goal, which I support, would be not speaking like a comic book supervillain. Lives matter. Of course they do. You can acknowledge that society can't be infinitely safe without denying that.)
The trust is lost regardless. They shouldn't have made statements that could be used against them, regardless of whether they were true or not. And, failing that, they should have thrown the people making the offending statements under the bus. But they didn't, and here we are. Still, there's some silver linings to this outcome, and it would be foolish not to take advantage of them. There's no need to make the situation worse by making mountains out of molehills.
Counterpoint: Truth matters.
Very true. And surprised Scott Alexander would throw this up considering the arguments on whoever's side hasn't and will unlikely ever change.
I appreciate Scott doing this follow-up because I will be more skeptical of his commenters in the future. It's been sometimes challenging to assess the credibility of comments on topics that are novel to me (unlike this one), and this post gives me insight into how many complete morons are among his followers.
I am pulled in two directions: it is extremely frustrating sharing a comments section with a large number of people who have no interest in truthseeking whatsoever, yet who aggressively act like they are the smartest people in the world. On the other hand, it's a huge public service for Scott to be so patient and thorough at exposing so many people with, uh, rationalist skills still at an early stage of development, to what actual reasoning looks like.
I am very much with you in this.
It’s unfortunate that the response has been to destroy all of the institutions that are engaged in progress, because you’re mad that they made less progress in 2020 than you wanted! It’s time to move on and stop blocking vaccines, scientific research, new social technologies, etc!
I really don’t think that it matters what evidence you present about this, because Covid skeptics have a remarkable ability to ignore any facts that contradict their beliefs. Anyone getting their beliefs from people like RFK and Alex Berenson, who are really shameless in the way they twist facts, are not people who can be persuaded. In general, conservatives have decided reality is whatever they want it to be to a remarkable degree
I think this is true of some people but not others. All of these people were convinced because someone argued them into their position, and if people are constantly arguing for these ideas but refuse to argue against them because "nobody can be convinced", they win by default.
More realistically, I don't think I will change the mind of any hardcore conspiracy theorist. But I think some conspiracy theorists will refine their conspiracy theories to be more plausible (while learning some useful things about good thinking skills), and people who are not yet either explicitly conspiracy theorists nor anti-conspiracy-theorists will hone their thinking skills and be less likely to believe conspiracy theories in the future.
Maybe this is true of the people you know, but the conservatives I know (which are many; I was raised very conservative) have very insane views on Covid, most of which are not remotely based on fact and many of which are mutually exclusive (ie Covid can’t be both a leaked Chinese bioweapon AND designed as a plot to enrich Anthony Fauci).
If it changes the mind of conversatives he knows, but not the minds of conservatives you know, that's a victory. Because a world where some people change their minds and some don't is of course better than a world where nobody did.
I see somebody has never been to the Historic Downtown Anthony Fauci Bio-Thunderdome, which doubles as his mountaintop fortress. Shaped like a skull, naturally.
> All of these people were convinced because someone argued them into their position
Or maybe it's because they were never willing to believe the alternative in the first place. After all, what good does it do to believe such an unavoidable, senseless tragedy occured? It's not as if you have any workable plan to prevent this from occurring again either. Better to believe it never happened, or better yet, put the blame for it on the people you want gone. Both seem more useful than making yourself miserable for no good reason.
But it was avoidable. International travel could have been suspended to prevent the virus from entering, then with a lower baseline and non-endemic virus in other countries we could more easily isolate & contact-trace the few who brought it in.
But none of that happened, did it? In fact, no country was able to accomplish that, suggesting that this wasn't a localized issue at all, and that the root of the problem is beyond the scope of national politics. Whatever would have been necessary to prevent its spread, there would have been too many incentives standing in the way to get even close to accomplishing that. And given that this wasn't even the first global pandemic... what makes you think next time would be any different? If humanity still exists by then, and they have the same hindsight and knowledge of infectious diseases that we did right before Covid, they will make the exact same mistakes again. Why wouldn't they?
So yes, in practice, it was unavoidable. And the things that would need to change in order to prevent this would not require the consent of the rabble anyways, so I don't see any point in wasting time arguing with them.
Australia and New Zealand both accomplished that!
Many pandemic preparedness plans have been written. If you truly care about preventing tragedies, you may wish to read one of them and perhaps convince people to implement it.
>All of these people were convinced because someone argued them into their position
Most people don't work like that. They simply believe whatever their social circle believes.
True, but that kicks the can down to why their social circle decided to believe it.
Mostly because that's what elites they look up to believe (or pretend to). So it's those elites that largely need to be convinced, or utterly discredited and overthrown. Both approaches prove to be pretty difficult in practice...
I think most True Believers were "convinced" of whatever conspiracy theory because
1. It flatters their egos to think that only THEY can see the truth, and that everyone else is just a sheep
2. It validates their rage at having to do something they consider inconvenient or scary. If it's a conspiracy from the NWO, the mask or vaccine requirement is not just an annoyance meant to save lives (which would not be socially acceptable to complain about), but a form of tyrannical oppression that MUST be resisted
3. It is a tool they can use to partisan ends. First Dems did this (Kamala's insistence during the primary that she wouldn't take a vaccine rushed out by Trump), then the valence flipped (afaict because the actual acting president changed, even though the vaccine really was made as quickly as possible under Trump's aegis)
I don't think these people were convinced rationally, they just accepted something they already wanted to believe. I still think it is good to make sure that true arguments for correct conclusions are ubiquitous, because otherwise low-information people end up believing conspiracy bullshit by default.
Fireflies flash in specific ways to find mates, birds pick their own song out of a squawking forest.
Somewhere out there there's a budding rationalist who just hasn't deeply weighed out the alternative views on <your topic here>. If you speak thoughtfully, it can cut through the noise and resonate where it matters, even if that's like 1% of people.
I believe this because I've often had 'conversion' experiences after reading arguments similar to this one. Though more often each argument moves me just a few degrees.
This one didn't radically flip my view, but reinforced it slightly, and better prepped me to help acquaintances and in-laws weigh key evidence in a calm and respectful way.
My most unsubstantiated conviction is that the talkorigins archive wasn't a failure at all and we should have doubled down on it and done extensive argument mapping for all controversial topics.
(Maybe AI has already quietly solved for this, "@grok is this true? @grok what are the flaws with this argument?")
Credit to just writing this post and it existing as a reference in the future.
You should testify in front of Congress to reach the nation - before you die from the covid vaccine that is.
https://youtube.com/watch?v=_JyzTAmPXcw
https://youtube.com/watch?v=Wb0kOsKYjXA
I was raised christian conservative, then went to university as a left-leaning atheist, and am now a researcher (and still atheist). Unfortunately, I would say your last sentence applies cleanly across the entire political spectrum. Science is rife with it.
The problem is and was very simple; Covid was a time with very little confidence, yet the scientific establishment pretended high certainty about a long list of things that later turned out to be provably false (masks don't work, then suddenly they are obligatory; Trump is pushing the vaccine development dangerously fast and one month after Biden's election it's suddenly 100% safe; vaccines fully stop the spread so we can open up again oops actually they only protect the person taking it; Lockdowns are sad but necessary actually countries without lockdowns did just as fine; Lab leak is conspiracy theory well actually most intelligence agencies consider it the most likely reason, etc.). And social science already had been doing the same for social justice, so it was straightforward to extend that mistrust just one category further.
When people lose trust in the institutions, they start looking elsewhere. These other places however have little to no quality control, so it's easy for grifters to get a foothold. It's stupid, it's completely avoidable, but that's what happens if you burn your long-term credibility in the name of the short-term greater good(which actually turns out to be wrong later anyway).
I agree somewhat but also disagree. I agree a lot of our institutions and "establishments" failed very badly during covid (and continue to fail in many other areas), however, the widespread embrace of conspiracies and obvious misinformation started way before any of the events you mentioned. For example, I remember many conservatives I knew promoting the "Plandemic" movie almost as soon as the lockdowns started. I also remember at least one acquaintance trying to argue with me that the covid vaccines were dangerous before they were even invented.
Also, I remember some of those events differently than you do. I know some Democrats said Trump was rushing the vaccine approvals, but that wasn't the general consensus on the left. Similarly, after the vaccines rolled out I saw a number of articles warning that if covid mutated the vaccines would be much less effective, which is what happened.
Agreed: it's a big problem that authorities believe they can't admit uncertainty, frequently playing armchair psychologist when that isn't actually their area of expertise.
I don't think the scientific establishment "pretended high certainty" at all - I kept hearing scientists make weak qualified statements and then non-scientific officials taking all the qualifications away.
Agreed.
But it's probably unfair to expect the general public to know what scientists are saying, if they can't trust the people whose job is to transmit scientific knowledge to them. It's possible for scientists to behave well, while the "scientific establishment" behaves terribly.
Perhaps we should call it the "science-bureaucracy complex" to make it sound more sinister, and to emphasize that it's not just scientists running the show. It works for "military-industrial complex", after all.
I like that. Because I really don’t want the shit of the US science bureaucracy to spill over onto actual scientists and actual science.
The "scientific establishment" is more than just "scientists". If the scientists speak perfectly hedged, generally incomprehensible academese to each other and then communicate the results to the outside world through "non-scientific officials", then those non-scientific officials are part of the establishment. And what *they* say, is what the establishment says in the only forum that matters.
If it's just scientists talking to other scientists (or science-adjacent nerds), then there's no reason for anyone else to ever fund scientists to do anything; they can buy lab equipment with the tax-free tips they earn as Starbucks baristas or whatever. If they want to be relevant (and funded), the scientists need to take responsibility for how their work is transmitted. Because they will, properly, be held responsible for such.
The problem is that those non-scientific officials are generally the bosses of the scientists, so it’s hard to hold the scientists responsible for their bosses’ fuck-ups.
Also, this isn’t a failure of “the scientific establishment”. Professor Chris Whitty (UK Chief Medical Officer) was very clear that the vaccines would prevent transmission for the current version of COVID, but that couldn’t be guaranteed for future ones (which is, perhaps, why I’m less annoyed than Americans).
It’s a specifically-American failure of your “scientific establishment”. That is, it’s a failure of Americans who were managing scientists. Not of politicians, not of scientists, but of the middle layer in just one country. Other countries didn’t have the same failures. But “science” and “the scientific establishment” are global phenomena, not national ones.
The "bosses of the scientists" are, while they are acting in that capacity, part of the "scientific establishment". There is an edifice that consists of scientists, bosses of scientists, and mouthpieces of bosses of scientists. Some of whom also have other duties, but in *this* context they are part of an edifice that is spewing bullshit masquerading as science. Words that cause many people to believe false things about important subjects.
That is a huge failure of the scientific establishment, even if one scientist named Chris was very carefully pedantically correct in the words he spoke to other scientists within the establishment. Even if *every* scientist was carefully pedantically correct in all of their intra-scientist discussions.
We all don't fit into simple boxes. Many of us were rational and skeptical, and yes, based on factual evidence early on, such that outdoor transmission was extremely unlikely. This from PubMed in May 2020. None of the media and few experts clarified this.
That isn't true. Many experts and many media people did clarify the facts around outdoor transmission. It was not hard to find that stuff.
A LOT of the skeptics claim that the media never reported facts that they learned about from the media. I will agree though that while the facts were available, the most easily available information from the "establishment" was often incorrect or nonsensical.
Also a lot of what people seem to believe was official was just vaguely authoritative social media posts from people at the more scared and gullible end of the spectrum. There’s one I saw everywhere “explaining” why masks were to help others and not yourself which was not only wrong (that’s not really true) but also from an Australia anti-flu campaign
I mean, TONs of media and experts clarified this. I know this because I listened to them and behaved accordingly.
But the media ecosystem is sufficiently fragmented and attention spans are what they are that it was easy to go through the pandemic watching popular media without having that information internalized.
This is the problem of our time. It's in no one's financial interest to do accurate, factual mass-communication, so no one does it, and we are all worse as a result.
With regard to the underestimation of COVID, allow me to point out what I think is the largest elephant in the room of public discourse, in politics, mainstream media, and incredibly even in the scientific community outside of long COVID specialists: the ongoing risks of contracting long COVID even for vaccinated people (only partially protected against it) and children (who can suffer permanent consequences like diabetes too). To grasp the extent of the underestimation, I think it is sufficient to read the following recent review article from Nature Medicine: https://www.nature.com/articles/s41591-024-03173-6
(from its abstract: “Long COVID represents the constellation of post-acute and long-term health effects caused by SARS-CoV-2 infection; it is a complex, multisystem disorder that can affect nearly every organ system and can be severely disabling. The cumulative global incidence of long COVID is around 400 million individuals, which is estimated to have an annual economic impact of approximately $1 trillion—equivalent to about 1% of the global economy.”)
or even the article for the general public by one of the authors of the review, Dr. Ziyad Al-Aly ( https://scholar.google.com/citations?user=DtuRVcUAAAAJ&hl=en )
https://theconversation.com/long-covid-puzzle-pieces-are-falling-into-place-the-picture-is-unsettling-233759
("A new study that my colleagues and I published in the New England Journal of Medicine on July 17, 2024, shows that the risk of long COVID declined over the course of the pandemic. In 2020, when the ancestral strain of SARS-CoV-2 was dominant and vaccines were not available, about 10.4% of adults who got COVID-19 developed long COVID. By early 2022, when the omicron family of variants predominated, that rate [of long COVID for infected people] declined to 7.7% among unvaccinated adults and 3.5% of vaccinated adults. In other words, unvaccinated people were more than twice as likely to develop long COVID.
While researchers like me do not yet have concrete numbers for the current rate in mid-2024 due to the time it takes for long COVID cases to be reflected in the data, the flow of new patients into long COVID clinics has been on par with 2022").
I think that two figures, the first from Nature Reviews Microbiology https://www.nature.com/articles/s41579-022-00846-2/figures/1 and the second about children from Scientific Reports https://www.nature.com/articles/s41598-022-13495-5/figures/2 are illuminating too.
These represent just a small sample of the vast scientific literature on long COVID which paints a troubling picture of the problem.
Examples of recent papers on long COVID in children from top journals are the following:
https://jamanetwork.com/journals/jama/fullarticle/2822770
https://publications.aap.org/pediatrics/article/153/3/e2023062570/196606/Postacute-Sequelae-of-SARS-CoV-2-in-Children?autologincheck=redirected
It is not coincidental, given their knowledge of the syndrome, that the already mentioned Dr. Al-Aly and another author of the Nature Medicine review article, Dr. Akiko Iwasaki ( https://medicine.yale.edu/profile/akiko-iwasaki/ ) still wear a mask virtually everywhere.
https://youtube.com/watch?v=M4Nzhpeu2Y8&pp=ygUMeml5YWQgYWwtYWx5
https://youtube.com/watch?v=rMt6ZV-hHSE&t=3681s
For people looking for concrete numbers, I like to point to this paper: [1]. In their sample from mid-2022, about 0.23% (from 1.67% * 14.12%) have apparent chronic fatigue syndrome due to COVID infection. If we assume that, on average, they'd had half a COVID case per person, that means that each COVID infection caused about a 0.5% risk of chronic fatigue syndrome. Divide that in half because of vaccines, and we can expect about a 0.25% risk of chronic fatigue syndrome every time you catch COVID, with a conservative assumption that your risk of catching chronic fatigue syndrome is independent each time. Given that people with chronic fatigue syndrome say, on surveys, that chronic fatigue syndrome is about half as bad as death, and that chronic fatigue syndrome is generally incurable, and that the average person gets COVID about once a year in my native Canada, that's really scary. (Remember that, since chronic fatigue syndrome is incurable, that number will keep going up over time. My guess is that the steady state is somewhere in the single-digit percent of the population having chronic fatigue syndrome, which will be a small but real drag on GDP growth.) It's why I personally wear a mask almost everywhere indoors: given my remaining life expectancy of about 60 years, that means I lose more than two weeks of quality-adjusted life expectancy each time I catch COVID. (Perhaps I should rate my risk even higher, since chronic fatigue syndrome is more common among young people.) I also get my COVID booster each fall, since it mRNA boosters in the fall reduce the risk of COVID during the winter season by more than half.
It's not clear to me if I should assume that the risk of chronic fatigue syndrome is independent between infections. It's also consistent with the data about chronic fatigue syndrome to believe that some people are born susceptible to chronic fatigue syndrome, and most of those people will go on to develop chronic fatigue syndrome at some point. But given that we don't know, I have to assign a pretty high probability to the risk of chronic fatigue syndrome being independent each time you get infected.
As far as I can tell, the risk of chronic fatigue syndrome is, for under-50s in otherwise good health, more serious than everything other risk from COVID combined (including the other long COVID stuff).
(The authors hesitate to say it's definitely chronic fatigue syndrome, presumably because nobody knows what chronic fatigue syndrome is or how it works, but I don't see any reason to think that COVID-induced chronic-fatigue-syndrome-like-illness is different from chronic fatigue syndrome as we know it.)
[1] https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0309810&type=printable
Scott wrote a post three years ago (now behind the pay wall) detailing the statistics that lead to his conclusion that long covid is a psychosomatic illness. His reasoning seemed reasonable, but I've had ME/CFS for 20 years with weirdly specific features that are common among people with the syndrome and/or long covid, so I have to dismiss his conclusions.
It's worth saying that "long COVID" is so poorly defined as to be essentially a useless category - generally including any symptoms that linger longer than a few weeks. I assume this is so that press releases about research can say "we studied a thing that affects 15% of people who get COVID, we're really important, please give us media interviews so we can get grant money that our institution will take a large chunk of to fund more press releases" or whatever - perhaps I'm a bit jaded. But almost all "long COVID" stuff goes away relatively quickly - presumably including whatever Scott was talking about then. A small (but still surprisingly large) portion of "long COVID", though, is actually quite serious stuff like chronic fatigue syndrome or consequences of having a heart attack or stroke (COVID can cause blood clots). I think a separate category - like "serious chronic COVID complications" or something - would be useful.
(Also, to anyone reading this who can direct grant money: please spend money researching chronic fatigue syndrome; it's a huge cause of disability and mortality but is very poorly understood.)
Scott's an ME/CFS skeptic, too. I don't really blame him. It's prevalent.
Man, I'm not sure if this is a case of "I'm a completely insane person that spent 17 hours a day on early covid twitter and knew how to counter this stuff by heart by April 2020" or "Political polarization and its consequences have been a disaster for the ACX comments section" but I'm surprised your commenters could be split on this. The evidence against the "with covid but not of covid" stuff is overwhelming just by excess deaths even before you get into estimates of IFR and the infected fraction.
I was hoping this would get to the (still utterly unpersuasive to reduce the magnitude of the tragedy, but certainly potentially making the comparison to say, the civil war unfair*) 'years of life left of covid victims', which given the old or pre-existing disease over-representation of deaths, probably makes the number seem less bad than it seems at first glance, even though still more years of life lost than events of substantial historical/cultural impact. But it seems that better argument was absent.
I feel bad singling out anyone, but the last comment in particular made me groan. I know several people who died or covid, including early on a former high school teacher than was only on her 50s and substantially scared away my parents, of similar age, that this could happen to them too. Would that convince any covid downplayer than the number was exactly 1.2 million? Anecdotes are simply worthless on statistical analysis at the national, several millions of people level
*tho with the lower life expectancy back then, even after child mortality, maybe it's closer than it may first seem
> 50s is nearly elderly,
Good grief, this is quite the rounding up. I know that to young people everyone over thirty seems ancient, but a 50-year-old woman can expect to live over thirty more years, on average.
> doesn’t justify taking away the irreplaceable joy of the college years from all of my younger friends.
That's irrelevant to the point, which is a factual discussion of whether 1+ million people died of COVID, not a political discussion over whether specific COVID policy choices were worth the tradeoffs.
There are three different things I would say to answer this.
The first is to return your preamble. I'm sorry you had to lose the joy of your college years due to the unfortunate coincidence of them being during the pandemic. If it's any consolation, experience makes me find the 'irreplaceable joy' a bit overstated: being a bit older, I personally enjoyed the social life soon after college (specially since I had my own money) more enjoyable, but part of this was people I first met in college and to the extent you lost that chance, it is truly regrettable.
The second is that you're arguing a different, and wholly conceded by Scott in his previous blogpost, point. I quote:
"Usually I’m the one arguing that we have to do cost-benefit analysis, that it’s impractical and incoherent to value every life at infinity billion dollars. And indeed, most lockdown-type measures look marginal on a purely economic analysis, and utterly fail one that includes hedonic costs. Rejecting some safety measures even though they saved lives was probably the right call."
The hedonic costs you mention are indeed what makes this not close. I would add that people are not automatons blindly following the law: people naturally relaxed, met irl and partied a bit when case counts went down, then stayed home or tried to do stuff online as a new wave of deaths came. Analyses that take this into account are the ones that make lockdowns not worth it on a purely cost-benefit perspective. This doesn't change the main question of these two blogposts, which is about the huge number of deaths and the fact they are legit.
The third is that '50s is nearly elderly' is something I'm skeptical most would agree with outside this context. In Spain, a 54 year old woman that dies from covid was expecting 31-32 years of extra life, including several years of retirement and knowing the grandchildren. It is basically equivalent to a 25 year old male dying in the civil war in 1860 (who would expect just a bit more, around 35-36 years of life if he survived rather than died in the American Civil War)
I submit that people in their 50s are very much not done growing and producing things, and if you did not want to imply otherwise you should not have said as much. If you want to argue over the tradeoff between disruption of two years of college and shortening the lives of the elderly, I suggest you review the estimated overall life-year cost and put forth your conversion factor: https://www.cidrap.umn.edu/covid-19/study-covid-pandemic-stole-nearly-17-million-years-life-adults-18-european-countries
"Math can't speak here"? What blog do you think you're reading right now that you can just brush off statistics regarding human flourishing?
It's quite easy. Take an actual measure of how many years of human life were lost to COVID, how many years of college were disrupted, and compare how much value people get out of being alive versus how magical college years are and how much worse COVID made them. I should hope there is some level of tradeoff here that would give you pause.
Whether slightly, mostly, or fully trolling, what I want to get across is that the literal reading of it is probably more inaccurate than you imagine. Scott has received plenty of mockery from people not reading literally the next paragraph as the screenshots going around for conflating deaths of old people from covid to deaths in the American civil war, and I just wanted to press the point that qualify of life and life expectancy has changed so much in the last 140 years than a 54 year old covid victim and a 25 year old civil war victim are much closer than people realize, so that basic intuition does not hold
yes, I agree it's different. namely, I would rather be a Spanish 54 year old woman in 2020 over an American 25 year old man in 1860, and nothing about 'life stage' or 'personal trajectory' compensates for the massive difference in quality of life standards, but I decided not to mention this since it's further opposition to your point.
College is mostly a waste. Young people can still enjoy themselves while getting more practical education on the job.
For how long? Most non-college young people are in service industry jobs like restaurant and store operation, which reopened almost entirely after a few months. Aside from that, very few people were ever prohibited from hanginb out in the local park or visiting friend's house. True "lockdown" only lasted about six weeks.
What is a "lockdown" by your standard, and where was it extended past the initial six weeks?
The kids flip burgers no matter what, the presence or absence of masks seems of limited impact. Getting laid up for two weeks from talking to your customers sure sucks though.
If college is the best time of your life then you are living life miserably
I was always under the impression that the lockdown was imposed not primarily to save lives but to prevent hospitals and hospital staff from collapsing under the weight of the gravely ill.
I'm extremely sympathetic to your complaints about lockdowns.
But my kids will be teenagers when I'm in my 50s. If I died in my 5th decade, would you tell my kids that the "irreplaceable joy" of your college years is more important than having their father around for some of the most formative years of their lives, to say nothing of their next few decades as adults (which I hope to be alive for)? What if I were already a widower and they'd be losing the only parent they have left?
Our lives are not entirely our own. When we die the loss is felt by family and friends. Surely they deserve some consideration, no?
I remember seeing Robin Hanson & Greg Cochran bet COVID skeptics on Twitter https://www.overcomingbias.com/p/my-11-bets-at-10-1-odds-on-10m-covid-deaths-by-2022html https://falkenblog.blogspot.com/2020/11/1000-covid-bet-with-robin-hanson.html The skeptics lost every time, but never seemed to learn anything from their track record of failure vs their opponents accurate predictions https://slatestarcodex.com/2020/04/14/a-failure-but-not-of-prediction/
I still get a bit mad everytime I remember some of these, particularly one article titled 'the bearer of good coronavirus news' focusing on Ioannidis and saying total US deaths would get to 40,000 or something. The tone was so much 'why do people resist good news?!?!?!' instead of modeling someone that truly believes the claims to be false for technical and empiricial reasons with high confidence, and thus sees people letting the guard down about the coming tragedy/disaster, very similar to such debates when it comes to climate change, AI risk, etc. I wish all of these were nothingburguers, but I was/am simply not convinced!
I realize I should probably meditate and get over it or something.
I was a big fan of Ioannidis's early work, but he really jumped the shark in the pandemic.
I should really reread his stuff on most published findings being false. That’s great stuff, but now I’m worried that he meant it to say that research is bad, rather than saying that research is good as a practice even though individual claims shouldn’t be believed.
I think it's the opposite. He made his career by being against the consensus and doubtless made some enemies that way. As I think we've seen, that takes a tole on folks and can make them become reflexively contrarian across the board. I view his COVID-era thinking in that light. Doesn't hurt that there was plenty of shoddy analysis being thrown around at the time. Back in May of 2021, every "data scientist" with a blog had a COVID tracker app and a SIR model projecting exactly how many of us would die.
I actually am a skeptic of risks from climate change & AI. And I favor public bets for them as well! https://www.econlib.org/archives/2017/01/my_end-of-the-w.html
funny you link that. I think Caplan's climate change bet is the first big one he's on an almost guaranteed track to lose already.
He thinks so as well.
It's strange to see the comments attract so many deniers given that Scott has been quite consistent on the dangers of pandemic disease in general and COVID in specific. I remember when the EA/rationalist community was a few months ahead of the curve in calling that COVID would be a big deal, or at least that the odds of it being a big deal were severely underestimated. I suspect the "skeptics" here are a significantly different crowd, but I don't know what would attract them other than the fact that Scott is unusually reluctant to ban anyone who believes themselves to be arguing in good faith.
Still, I think it's good that Scott bothers to write such detailed proofs of obvious things. It's good to have strong arguments spelled out in detail to point to when "that's obviously dumb" fails.
Indeed, if only you listened to him.
oh absolutely, to be clear I agree. it's more of a 'despair that this is necessary' more than 'Scott was wrong to post this'. while plainly wrong, I don't think the 'don't give creationists a platform by debating them' stuff applies for this sort of position
Yeah, I wouldn't argue with these guys on CNN with no moderator support, but that's not what Scott's doing. He's writing TalkOrigins for COVID, and that's something we really need.
I'm in the same boat. I think it's selection effects w.r.t. the comments section. Just a very different crowd here now than in 2016 - 2020.
>The evidence against the "with covid but not of covid" stuff is overwhelming just by excess deaths
The evidence for lab leak versus wet market should get you *at best* to a 50/50 shrug, and yet Scott spent way more time on that one.
Political polarization is certainly one reason this comment section is lower quality than back in the SSC Golden Age (ie, 2014-15), but not the only one, and not only for the reason you seem to imply.
I'm confused about both your first sentence and what you think I'm implying.
Most obviously for the former, if something is a controversial 50/50 shrug, *obviously* you will spend more time presenting the presumably strong case each side can make that results in such a close call? as opposed to simple things, which require less time? as seen in the fact this article here is short on wordcount for Scott's standards
edit: just for clarity (in case you don't check this in e-mail and do see the edit: I'm *wondering* if it's political polarization, the other option being that I'm the insane one. it's an unsure statement)
Short version: it was very unclear snark about the way Scott gives attention to covid. More clearly, I think he is stuck on wanting to write about COVID (not unreasonable) but also chooses the least interesting ways to write about it. Is this really what's worth his time?
Longer but quickly typed, inaccuracies likely abound:
I think the Rootclaim thing was a waste of time and Scott should've stuck at 50-50.
>For what it’s worth, I was close to 50-50 before the debate, and now I’m 90-10 in favor of zoonosis.
I also think that GoF research is risky whether or not covid was a result of GoF, and arguments in favor of zoonosis are sometimes (not necessarily by scott, but others) used to act like GoF isn't risky.
That's the kind of thing I think someone of his talents could spend more time on. There are all sorts of policy failures and social failures around COVID, and he's now spent *how much time* writing about one not very important thing (zoonosis vs gof is broadly irrelevant unless we're considering to destroy Wuhan specifically) and one not particularly interesting conspiracy that *also* will have no impact how to react to future pandemics.
I mean, his blog, his talents, his engagement farming! I just think he could do more, and gets stuck in these outrage-inducing but not really impactful cul de sacs instead.
I mean, I think I either disagree with you, or you advance possibilities you don't believe in but worry others will believe, but that do not apply to me. For example, I think gain of function research is risky and must be stopped and will believe that exactly as much if tomorrow they find the exact animal with one base pair difference to wild type covid on its lungs. To the extent some people use arguments in favor of zoonosis for that, they're my enemies.
I also do not think it's about destroying Wuhan specifically. Back in the pre-pandemic days, people complained about serial passage in ferrets to make a strain of avian influenza airborne, and that was within the US, UK and Netherlands if I recall correctly.
As for the reasons for Scott to spend his time, I'm afraid I'm less concerned to let him write whatever he wants. Perhaps this was just strategic considerations as he saw people around him over correct too much towards lab leak, perhaps he found it interesting and decided if he was going to sit down six hours of it he might as well take notes and make it into a blogpost. With the amount of writing output he can do I'm not overly concerned.
>I'm afraid I'm less concerned to let him write whatever he wants.
Well, yes, as I said
>I mean, his blog, his talents, his engagement farming!
Of course he can write what he wants, and he will. I am making a suggestion that he could have more impact on other topics.
And suggesting it should be unsurprising that the comment section is contrarian and split when the posts ultimately aren't... "substantive" isn't the right word. Pushing forward? Broaching unaddressed topics? Missing the forest for the trees?
Unless you're writing government lab policy, lab leak versus zoonosis is kind of a nerd snipe. Deaths with COVID versus deaths of COVID is sort of interesting, but it's also Scott picking a fight with a subset of his audience/some guys he finds irritating on twitter, rather than trying to communicate something novel.
Not every post can be a banger, not every reader has to love every post. But I think it's fair to observe a marked shift in the sort of topics of his during the pandemic posts, and these later ones.
I... have nothing to contribute to this debate. I just want to show up and be a person online who agrees with and supports things that are obviously true. In this case, I mean, the things that Scott is saying. Let no-one think that all online people believe in craziness. We believe the normal stuff, we just don't comment that often.
I have a friend whose both parents died of covid (both started having difficulty breathing, went to a hospital, died there). So the idea that covid is not real (and my friend probably just suffocated his parents using a pillow, and the nurses in the hospital finished them) is just absurd to me.
Of course, this is internet, so I expect people to argue for absurd things. And in absence of a downvote button, people can compensate for the absurdity of their opinion by saying it more loudly and more often, because that's how you win online debates.
I can’t believe you are taking people who question the mortality stats seriously. Not only was there unprecedented excess mortality, the daily mortality counts always peaked two or three weeks after clinical cases peaked.
Even in the comments section here, no one skeptical of the death counts is offering a causal model for the relationship you described. It's pretty telling...
>so if 1.2 million died of COVID, that’s about 1/300
Another way of making this more intuitive: "do you know 300 people? If not, it's not surprising that you personally don't know anyone who died from COVID"
It's actually 208 people! I'm not smart enough to automatically remember the math myself, but I'm smart enough to know this is one of those situations where you ask a mathematician before giving any specific figure (in my case I asked o3, and double-checked with Claude).
This comment made me check the math; it's correct. When I read the article I assumed you computed it yourself, and I wish you had said you got it from AIs in the article. I know they're getting good, and they did get it right — but I'd still like to know where the information I read is coming from. One reason why: as the companies do more reinforcement learning on usage-time/subscription-revenue/etc, the AIs will start trying to manipulate us more; they've been getting better at that too.
For the record, here's the math: Let p be the probability of death of covid. Either know >=1 people who died of covid, xor you know 0. It's easier to find the probability of the latter. The probability *surviving* covid is (1-p), so assuming independence the probability of all n people you know surviving covid is (1-p)^n — the more people you know, the less and less likely it is that everyone you know survived. We want to know when this reaches 50%; ie, how big does (n) have to be for the probability to fall all the way to (1-p)^n = 0.5. Solving, we get n = log_(1-p)(0.5). In our case, p=1/300, and with a calculator we get log_(299/300)(0.5) ≈ 207.6.
So if you take 207 people at random, there is a slightly higher than 50% chance that all of them survived. If you take 208 people at random, there is a slightly less than 50% that all of them survived.
Still greater than Dunbar’s number!
Even 208 is just 50-50, right?
The response will be that they know far more vaccine injured people, including people who died from myocarditis, and that at some points the VAERS numbers far outmatched the covid fatality rate because some antivax activist(s) spammed their followers to submit false reports to the database.
I will neither agree nor disagree, but could it not be the case that what COVID did if just front-load(*) deaths that would have occurred anyway due to putting stress on people in already frail and fragile condition? We do know that people with diabetes and obesity were more likely to be affected. I'd love to see how that first figure you presented continues. Does it slow down after that? For example, if you extend the line from pre-2020 trend through to 2025, does it fit? Was the bump transitory or a permanent shift upwards?
(*) A bit like incentives for people to have children; they don't work, they only make those who were going to have them anyway do it earlier. There's little to no permanent effect.
You can easily look this up yourself with a two minute search. The statistics don’t support that theory. Excess deaths are still higher than before COVID, even after 2022.
https://ourworldindata.org/grapher/excess-mortality-p-scores-average-baseline?country=~USA
Its not really correct to say that excess deaths are higher because of COVID however. Excess deaths being up after 2022 makes sense since the median US age is ~2.5 years older in 2023 than it is in 2020(https://en.wikipedia.org/wiki/List_of_countries_by_median_age) .At present the US is aging about 1/5 of a year per year and Europe is aging about 1/3 a year per year. If you take the statistics for France and Germany (about half a US in population) You see the same elevated excess deaths (https://ourworldindata.org/grapher/excess-mortality-p-scores-average-baseline?country=FRA~DEU~USA).
Since life expediency at birth is increasing at a lower rate than median age, excess deaths in a year on year comparison would need to be normalized by that before we could rule out this sort of front-loading. Otherwise its kind of pointless.
Now you could say that the people collecting that data are taking that into account but a quick search of https://pubmed.ncbi.nlm.nih.gov/34190045/ (the source for the excess deaths from your link) doesn't indicate that considered median age in their calculations or really much at all. Meanwhile the source for the European excess deaths (https://ec.europa.eu/eurostat/cache/metadata/en/demo_mexrt_esms.htm) is explicitly just a rolling average where the population (and therefore expected deaths) of a decade ago is assumed to be similar to the population of today.
Digging into the data more, you could point to the fact that you see such burn effects in the French/German data every winter (spike than fall), but that is potentially a bit of a reporting artifact, and there is no such effect ANYWHERE in the US results which raises some eyebrows. Could be the US medical system is failing, could be that people in the US are dying of a lot of things such that deaths due to cold and flu are being drowned out by other periodic sources of noise. Maybe its fentanyl or it could be that the books are cooked and the 'I think its all lies' people are correct. Who really can say? It would require a lot more digging than a 2 minute search to figure it out.
Taking a step back, I want to also bang the loss of trust drum here. Between the continued metastasization of the replication crisis*, and the politicization of every single truth finding institution its generally really hard to make ground level claims, and I see a lot of commentators (and everywhere else really, but this is a walled garden I care about) here just drifting closer to "I alone hold the truth of the world, and everyone else is too intellectually lazy to figure it out". I think actually making the claim that 1.2 million Americans died due to COVID(and would still be mostly alive in 2025 otherwise) on a solid enough level that you could reach across divides and establish it as a ground truth would be a painful adversarial collaboration, and probably come out as a long ass report. Since it seems like that's not going to happen, all this sort of article yields is more stuck priors and more evidence that the other side is dishonest.
* This isn't even getting into the weeds of the academic coding crisis, a separate but related problem where most academics can't code, most research is coding, and bugs are literally everywhere and being used to make decisions.
This is the long-ass report. Scott has written even longer-ass reports on Ivermectin and lab leak vs wet market origins. At this point, it's on you to explain why you don't think a clear chart of recorded COVID deaths track precisely with observed excess mortality. Aging populations alone wouldn't have done that.
It's all well and good to say that more transparency and open communication could have better preserved trust in the medical establishment on a broad level, but individuals have a responsibility to justify their own claims. All this "the CDC gave inconsistent early guidance on masks therefore we can never know for sure, no I won't present a superior body of evidence I'm just saying we don't know" only demonstrates a poor grasp of logic and epistemic cowardice. The skeptics are invited to make their case or go home.
Well I didn't mention mask mandates, or the CDC in my comment, so thanks for making my point about prior sticking so clearly :P.
To be more serious, its not like mask guidance was the only thing that happened over the past couple of years that degraded trust. There were things like the Ferguson Imperial Model and its subsequent code issues, vaccine safety concerns about Operation Warp Speed, and their social inversion post election. There were issues with protests vs social distancing, and issues with vaccine rollouts getting muddied by politics instead of purely by efficacy, and that's without getting into the mud of "were lockdowns effective" or "was it a lab leak or wet market origin".
Regarding the discussion here, I'm not sure what you mean by the skeptics are invited to make their case or go home. Martin Blais asked a question about if and how much is it possible that the 1.2 million people would have died anyway by 2025 if not for COVID. It's a fair question given COVID's known comorbidities with other serve health issues like age. Calvin Blick responded that excess deaths do not show the sort of effect that one would expect if COVID had a front loaded death rate and thus would have a lower excess death for the next couple of years until it normalized.
At which point I pointed out that national aging is not something considered in excess death statistics. Given the speed at which the west is currently aging, this is a potential massive cofounder for looking at things like deaths from COVID as a front loading issue (75% of the 1.2 million dead were older than 65 https://www.statista.com/statistics/1254488/us-share-of-total-covid-deaths-by-age-group/, but the number of new 65 year olds is also spiking up by year https://www.stlouisfed.org/on-the-economy/2019/may/how-many-people-will-be-retiring-in-the-years-to-come). I then pointed out the there is no seasonal effect for deaths in the excess statistics unlike France and Germany and note that that is weird and needed further analysis to figure out if we could rule out COVID front loading. Then I bemoaned the breakdown in trust and the fact that there isn't a clear neutral source that we can turn to and determine, "Hey why aren't old people dying in the winter more than the summer right in in America? That's kind of weird right?".
If I'm playing the role of skeptic here, I've made my case on this point. I disputed Calvin's point and brought evidence to back me up both that excess deaths are elevated because of population aging and that the US is statistically weird. Seriously, the US has been sitting constantly at ~8% excess mortality for 5 years now! It doesn't have a natural season cycle of death like France or Germany. Additionally both France and Germany look like what you'd expect with a front loading effect (drop down to 0 or negative % deaths on a trend line that's slowly sliding up https://ec.europa.eu/eurostat/statistics-explained/index.php?title=File:Mortality_statistics_Fig01_v2.png) If you subtract 8% from the figure the US does have a similar dry tinder cycle.
The ball is in the court of the ... (non-skeptics? skeptic-skeptics? ) now. They can pitch back something like "If there was really a dry tinder effect it would have been so obvious that it would have dwarfed aging effects beyond even what a trendline would expect, because xxxx)" or "Even if the people who died of COVID would have died of something else in their near future, it would have been lower and they would have had many years" or "some other argument that my biases are blinding me from". That is how civil truthfinding is meant to work. Not just drive by "You brought no proof", or "You must bring a full gish gallop of evidence for every inch of this debate or I'm going to dismiss you".
Finally, and this is entirely in the voice and energy of Crocodile Dundee. This is not a long-ass report. Both articles written here are 2500 words together. Scott's H5n1 more than you wanted to know sits at 4500 words. https://www.astralcodexten.com/p/h5n1-much-more-than-you-wanted-to
The old SSC adversarial conversations that I directly called in my post sit at around 8000 words (https://slatestarcodex.com/2019/12/12/acc-does-calorie-restriction-slow-aging/) to 9500 words (https://slatestarcodex.com/2019/12/11/acc-is-eating-meat-a-net-harm/).
That extra 3-4x word count gives a lot of room to dig into assumptions like "What about the distribution of deaths by age? What does that fact that ~50% (https://www.statista.com/statistics/1254488/us-share-of-total-covid-deaths-by-age-group/) of Covid deaths occur in a group that is older than the 2018-2019 US life expectancy(78.8) mean for the categorization of them as COVID deaths? What about the fact that 27% of the deaths are significantly older than the life expectancy?". I mean when we are talking about 1.2 million, that's potentially 600,000 people who were statistically expected to die.
I guess you can fall back and say 600,000 is still a tragedy, and that's true it is but that must means we have to go back to quibbling about population aging and health issues.
You have gone to the effort of making a case, I'll give you credit for that. I do appreciate it. I don't know much epidemiology or all the relevant math, but I'll give it my fair shake.
To dig into the specifics, it's true that the US has been aging, but it hasn't been aging that quickly. (https://www.statista.com/statistics/241494/median-age-of-the-us-population/). In 2017, the country was 38. In 2022, it was 39. That's fast enough to be concerning for the long-term fate of social security, but it doesn't look like we should be seeing a %5 increase in overall mortality on the slowest day of 2022. This suggests to me that we're seeing broad-scale impacts of a virus known to be more lethal than most and to inflict longer-lasting residual symptoms on survivors than is typical. COVID is past the point where it can be treated as a crisis, but it's still a respiratory disease with a seasonal cycle that's stronger than the flu, and the flu already killed ~20,000 a year.
I'm not really sure what you're talking about with the US not having a seasonal death cycle? It definitely looks like excess mortality is higher in the winter where the flu and COVID are known to flourish.
I'm not familiar with the idea of a "dry tinder effect", but I gather that you're trying to argue that the longer-term mortality rate is unaffected because as many as half of all COVID victims would be dead by now anyways? This seems contradicted by the fact that excess mortality didn't dip after COVID. If those people were going to die anyways, then their earlier deaths would have "borrowed" against later deaths since they couldn't die a year later as they would have in the counterfactual.
> "What about the distribution of deaths by age? What does that fact that ~50% of Covid deaths occur in a group that is older than the 2018-2019 US life expectancy(78.8) mean for the categorization of them as COVID deaths? What about the fact that 27% of the deaths are significantly older than the life expectancy?"
I don't dispute any of this, but I don't think it means much of anything at all. Sure, just saying "1.2 million deaths" masks the fact that most of them were old, but everyone knows that. If somebody was going to die next year of heart failure but instead they die this year of COVID, they go down as a COVID death because that's what killed them. I don't think it's very complicated. Sure it impacts the overall life-years lost to disease, but that sum is still pretty high (https://www.cidrap.umn.edu/covid-19/study-covid-pandemic-stole-nearly-17-million-years-life-adults-18-european-countries).
Mostly, I find this whole line of questioning kind of confusing. Like, you could do this for basically every disease. Most diseases in developed countries mostly kill old people. That's why fighting diseases is our main method of increasing human lifespan and healthspan. You never see somebody show up to the heart disease fundraiser arguing that everyone is wasting their time because most who die of heart disease would have just died of cancer in a few years. The fact is, COVID deaths were counted more or less exactly the same way as every other death of disease, and the number dead is about 1.2 million. I don't know what you're looking to prove by digging through actuarial statistics.
Everyone will die anyway. And yes, trivially people dying now means they won’t die later.
My cousin's husband died from COVID after being discharged from the hospital. Just couldn't get enough air at home, even with eventual supplemental O2. And yet the state of TX-or perhaps merely the city of Dallas--would not list COVID as his cause of death. Chris, my cousin, had to fight for weeks to try to get this changed. Not sure of the result, but I know the original decision.
So we know some places labeled cod as COVID if you died of something while incidentally having COVID. And we know of places that suppressed the actual cod via COVID figures. Looks like a wash, statistical noise.
1.2 Million.
> It infected hundreds of millions of weak old people of exactly the sort who die from viruses like this all the time.
I think one way it differed from expected and previous coronaviruses is effects monotonic with age - it pretty much left infants alone
Isn't the age-specific mortality pretty similar to eg seasonal flu?
Not for children/infants. https://pmc.ncbi.nlm.nih.gov/articles/PMC10760487/table/T1/ has a good table.
Comparing annual death rates from COVID vs. Flu in Japan. COVID killed about 1/3 as many 0-9 year olds, but 33% more 20-29 year olds, and >2x as many 80+ year olds.
The flu is more dangerous for children than for teenagers/young adults, while COVID is safer.
No. Covid is weaker than flu for younger people and x2 or worse than flu for everyone over the age of 65. Graph:
https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fbucketeer-e05bbc84-baa3-437e-9518-adb32be77984.s3.amazonaws.com%2Fpublic%2Fimages%2F1b4b3bc1-5a9e-4932-bfc5-71c77731d3c6_624x457.png
From here:
https://hwfo.substack.com/p/the-covid-19-red-ink-blue-ink-problem
I can share the data if you'd like to tinker with it.
To me, the reason everyone is still busy with COVID is because they're's no sense of closure. The perceived authorities have failed at primary tasks incumbent upon them.
There is much to be gained from *preventing* closure. Look how it's worked out for RFK Jr. Vaccine denial was his leverage into power.
The problem is that the perceived authorities were given the task of transparency and admission of imperfection, which was not a task they took on before. This is a task that will inevitably lead to perceived failure.
Tangential, but as a formerly-suicidal person I think the reason we (sometimes) cope with disaster better than expected has something to do with the fact that I often felt most hopeless when my life was going well. After all, if there's nothing in particular to blame the bad feelings on, and nothing you can take action on in an attempt to feel better, you're left with the conclusion the problem is "I am inherently broken somehow and will never feel good no matter how well things go." On the flipside, I broke my femur a while back and found it remarkably clarifying. Yeah, it absolutely sucked, but there was a clear external reason for the suck and a process to go through to make the suck go away (surgery, physical therapy, etc). And emotionally that's noticeably more pleasant than interminable existential angst.
(That said, I was not doing well during the pandemic).
My experience of covid-related debate has been that those who are in the 'denialist' camp are also bad at handling statistics. They don't have competency to handle the concepts needed, and will misunderstand things and draw the wrong conclusion.
Obviously there's a broad range, from people who are a bit vaccine hesitant because of what they've heard second-hand, through to the conspiracy theorists. But unfortunately that extreme has a large advantage in uninformed discussion, because they're not beholden to facts, and can wilfully misinterpret any set of data.
I think you get better at statistics by reading lots of good statistical arguments on topics of interest to you.
Some people can, but some people can't. There's the phrase "pearls before swine", and of course a literal swine can't learn statistics no matter how much you try to teach it.
More specifically, you get better at understanding and applying statistics by reading lots of good statistical arguments of topics to interest to you, and working through the math on the stats that are strongly counterintuitive to you.
A wicked fast update on a specific strongly held prior after reading some specific statistical metric (and then the math) is the most effective teacher at the nature of non-intuitive realities in the statistical domain.
* gotta love the technical term for “wicked fast update,” the first derivative of acceleration: jerk
Absolutely, but you need to appreciate that the denialists aren't doing that.
They're eyeballing numbers and misinterpreting them. Then accusing you of lying.
Many of them aren't even looking at numbers at all.
The vast majority of pro-COVID (you know what I mean) are also bad at handling statistics, and so will be ineffective, perhaps entirely counterproductive, at correcting the denialists.
On vaccine hesitancy, there's also an issue that there for a while (maybe still?) being opposed to *mandates* got you called an inhumane braindead anti-vaxxer. Words are supposed to mean things and "antivax" became an absurdly large tent during COVID when it didn't need to be.
“People only started getting vaccinated in December 2020, but there was the same amount of excess all-cause mortality before that time.”
Begs the question about how effective the vaccines were in preventing COVID deaths.
Very effective but not perfectly so - you just need to compare death rates among unvaccinated to vaccinated (obviously correcting for the necessary factors).
This feels (depressingly) like a conversation from 2022.
Scott's right. We need a reckoning!
Here's a chart: https://ourworldindata.org/grapher/united-states-rates-of-covid-19-deaths-by-vaccination-status . I think what vaccines giveth, new variants taketh away (mostly, but not entirely, from the unvaccinated).
Thanks for talking about this and bringing clarity.
Why does the graph only start in Oct 2021? It's like finding a super-interesting book but the first half is gone.
Not a new question https://www.natesilver.net/p/fine-ill-run-a-regression-analysis
This is one of the things I find so frustrating about this camp. We hear all this innuendo and vague gesturing towards supposed statistical anomalies, etc. "Hmm, that really begs the question, eh??" And then if you spend one minute googling this question you immediately find ten different lines of evidence that the vaccine was extremely effective and it probably saved hundreds of thousands of lives.
One othe point: to the extent that the Covid responses varied by state, those results are consistent with Covid being the problem. States that had higher levels of vaccination had lower mortality, states that were more open and less shut down had higher death rates.
Two states with high levels of elderly, Maine and Florida, had very different mortality (it was much safer to live in Maine), while Florida is the state that opened up more quickly and vaccinated less of its population.
Fair enough, but to deny that 1.2 million people died of Covid is to deny that there was such a choice.
Macedo and Lee, in their book In Covid’s Wake, find no association between non-pharmaceutical interventions and mortality in the pre-vaccine period. Some state/country comparisons show the inverse - high restrictions and high mortality pre-vaccine, and the studies they cite accounted for all manner of demographic and other factors. (Ebook version allows you to pull up footnotes, which link directly to the cited research).
High restrictions and high mortality pre-vaccine is an argument for vaccine effiectivess and deaths from Covid.
Yes, the difference between states only emerged after the vaccine became available.
I've never been able to take such studies seriously, because they ignore that places that are more susceptible to COVID (such as by being more urban) are going to use more strict NPIs. In other words, the reason there's no association is because NPIs are caused by high death rates, and are maintained at a level where they lead to a similar death rate compared to the rest of the country.
I’m a bit torn on this.
Statistically, the public health evidence against NPIs both pre-COVID and post-COVID is very strong. After the pandemic runs its course they seem to make no difference to deaths or case loads and of course they do economic damage. If you compare Sweden, or even Florida (pre vaccine) with California, their outcomes were about the same even though California had much stronger NPIs.
But when you look at specific smaller areas it’s not so clear. The Bay Area did better than California as a whole and it’s hard not to see that as being because NPI compliance was better. New Zealand and Australia had strict travel restrictions and even though their quarantine failed they had very low deaths and case loads even in the end.
I appreciate the effort that you put into proving things from first principles anyway but this is just so bizarre holy hell. As if flat-earthers suddenly decided to heavily comment on a rationalist blog
👆👆👆👆👆😬
(They've been here the whole time.)
I don’t comment here much, but the overall level of comment on this thread is far, far better than most discussions of Covid on social media (or, i suspect, most other places). Which isn’t to say it’s ideal, or even good, but much better than the usual.
You mention: "What about the alternative claim - that it wasn’t COVID that caused the extra deaths, but various treatments - ventilators, remdesevir, vaccines? We know it wasn’t a specific single one of those treatments, because the treatments were only used during certain subsets of the pandemic, but the excess mortality was a constant function of COVID cases."
I suspect you're saying total US excess mortality divided by total counted US COVID cases on a per-week basis looks flat.
Can you link to strong data regarding that? I'm really skeptical.
I assume that data is far too noisy to "prove" that statement. And, I think that part of the noise (COVID case count itself), is not just suffering from random noise, but time-dependent noise (I don't know when we'd be over our undercounting, but I'm sure the amount that was over or undercounted was strongly varying over time).
Restated, it'd be really interesting to state and test two hypotheses:
Hypothesis 1 (wildly held by the mainstream?): Mortality rate dropped with Omicron.
Hypothesis 2 (held by skeptics?): Mortality rate increased with ventilator use, dropped with Omicron, and increased with vaccination.
I suspect we can't confirm either.
Instead, what I think we do, is simply bolster priors. For "establishment supporters," the "relative" flatness of a mortality rate will bolster thinking that COVID was deadly, and vaccines were not. For skeptics, the "relative" flatness of a mortality rate will be insufficient given the noise, and the null hypothesis that fatality didn't increase with vaccination won't be confirmed, and that'll bolster their priors.
Right?
I'm making the much weaker and less-trouble-finding-statistics claim that about the same number of people died of COVID in 2020 as in 2021, so if all the deaths were from vaccines, the 2020 deaths would be unexplained. I agree that you would need more sophisticated statistics to detect a 10% or something effect on death rates.
I don't see the numbers clearly enough to even really believe that 2020 vs 2021 death rates are constant (maybe, maybe not - I just don't see them). I suspect that case rate reporting may be wildly off over time (example guess: 4x underreporting in one year, 2x in another). If so, then the excess mortality rate year over year has a 50% error. There's a time gap between when people test vs. report death, etc. It's just a morass.
Regardless: if your claim is something akin to -- we can definitively confirm that less than 100% of excess deaths are due solely to vaccines -- that's so weak it begs why to write the article and leaves both sides just doubling down on their priors with any whiffs of data provided?
Pro-establishment: "Indeed, excess deaths were large, and we can definitively rule out that less than 100% of excess deaths were due to vaccine. We suspect that 0% were due to vaccine, but can't really prove that."
Pro-skeptic: "Indeed, excess deaths were large, but at least as many of them were due to vaccine as due to virus. We can prove that excess deaths must be >0% due to vaccine (how could weird heart problems not be correlated with at least some deaths?). We suspect that >50% of were due to vaccine, but can't really prove that."
Your best posts tend to show data/logic that makes one or the other side recheck their priors. Here, I think what's being demonstrated (at least to me, convincingly) there are excess deaths. I think only a small fraction believed there weren't excess deaths (and your data is useful with respect to that, thanks - and thanks, as always, for good faith posts). However, I think a more important fraction think that the fraction attributable to care (vaccine, treatments, lockdowns, etc.) may be comparable or greater than the fraction attributable to the virus itself. Both sides are just doubling down on their priors after that, right?
We don't just know that vaccines didn't cause extra deaths, we can see that they lowered them. https://www.natesilver.net/p/fine-ill-run-a-regression-analysis https://ourworldindata.org/grapher/united-states-rates-of-covid-19-deaths-by-vaccination-status
Thx. That sort of data might have helped Scott's point here, but it's not going to be particularly convincing to the "other side." Regression analyses in observational work off of foundationally limited data are obviously interesting, but face profound limits.
My recollection is that most data pools treat vaccination status oddly. I think such data sets bin an individual as unvaccinated until two weeks after their second dose. So if a person got 2 shots, separated by 2 weeks, any occurring to that individual in the period from their first shot up until 27 days are registered as to an unvaccinated person. Let's assume for the moment that the vaccine itself did cause some deaths - if any of them happened in a time period of up to 27 days after the first shot, it's a really interesting problem.
First, in most analyses (including I suspect Silver's) that person's death is not added to the vaccinated bin (where it belongs). Second, it is added to the unvaccinated bin (where it doesn't belong). The observational stats need to be far richer to sway that side here.
The second link might help address that by separating out status, but I assume still has the two week delay issue (i.e. it has a pair of 13 day problems, rather than a single 27 day problem) that I described above.
Further, both make an implicit assumption that the vaccinated and unvaccinated groups are otherwise demographically equivalent. Perhaps they are, perhaps they aren't- I don't know. Maybe vaccines are always given to sick people, so excess deaths would be terrible there (and dramatically understate vaccine utility).
I understand why a pro-vaccine camp would look at these data and revel in their apparent conclusiveness.
I also understand why the skeptics camp seem them as garbage-in-garbage-out data.
My point - all these somewhat-foundationally-limited data sets aren't moving the needle. It takes something new and possibly sophisticated to move priors that are this deeply established at this point.
> Further, both make an implicit assumption that the vaccinated and unvaccinated groups are otherwise demographically equivalent.
No, Silver makes the point that there weren't big differences in mortality between states based on partisanship until the vaccines became available. If pre-existing differences between states were driving the difference in deaths, it should have shown up earlier.
It still relies on the assumption that vaccine decisions cut cleanly across state/party lines rather than some other confounder-- which they clearly don't; even the least vaccine-friendly states still got up to like 60% uptake as I recall.
> 500,000 - 700,000 excess deaths in each of 2020 and 2021, which adds up to most of the 1.2 million (although I think the full number might include some residual deaths during 2023)
You probably mean 2022 (or 2022 and later).
On this parenthetical in footnote 2: "also, suicides didn’t rise during the strictest part of lockdown, and may even have paradoxically fallen. I agree this is mysterious, but it’s only one of the many mysteries of suicides, which often fall during bad weather or social disasters - see my article here. Drug overdoses rose, but by a tiny fraction of the COVID death toll."
This is just speculation, but given that suicide seems like it would happen when people feel disconnected from others. Perhaps the lockdowns, 'social disasters' (guessing something like terrorist attack), or bad weather (hurricane, tornado, etc) are events that, although scary or 'depressing' also make us feel more connected with other people, more part of a community, even temporarily. That may be enough to depress suicide rates to some degree. Again, just wild speculation on my part.
There is some evidence for this - suicide rates are highest in the happiest countries. If misery is common and temporary people just kind of deal. COVID lockdowns gave a "maybe it'll get better if I tough it out" vibe.
All the accounts I've seen from suicidal people is that it's not the misery, it's the isolation and the sense that there is no potential end to the misery. "Only two years of lockdown" indicates participation in a collective effort with a clear end goal.
OWID has a data explorer for excess deaths btw. https://ourworldindata.org/explorers/covid?Metric=Excess+mortality+%28estimates%29&Interval=Cumulative&Relative+to+population=false
I think that you're on the right side of this one Scott, but I want to ask, because it always is overlooked in this discussion for some reason, what share of the 1.2 million excess deaths should we attribute to death's caused by the lockdowns? E.g. weren't deaths from overdoses and suicide higher than they otherwise would have been? And has anyone tried to quantify deaths due to cancer or heart disease that was caught too late because of missed appointments?
Edit: shame on me, I should have read to the footnotes!
I'm sure it was not a very large share of the excess deaths, but I think about this every time people say something like " well if there were excess deaths above baseline, it means people died of covid". I think lockdowns definitely caused some excess deaths too!
No, he specifically pointed out that suicides were lower. Cancer screening in the US is also ineffective. https://www.overcomingbias.com/p/beware-cancer-screenshtml
You are correct, this article was a reminder that I should read all the way down the footnotes before I comment!
Most countries report much more excess deaths than deaths from covid in longer term (2020-2024). How do you explain that?
Most countries have worse medical systems, which fail to detect many COVID deaths.
No, they are actually better in some ways.
Ok, less gold-plated.
Ok, less gold-plated.
I would also trust data from Sweden more than from the US. Less panic means more precise reporting.
The US was totally tribal – Trum vs. the rest. I talked to some doctors from the US and somehow mentioned one thing and the doctor rejected that by saying that Trump had supported that. I wondered why is that a problem? “Even a blind chicken happen to get grain.” Or even Trump can sometimes be right by accident. He rejected that notion and maintained that everything should be done to fight him. They even delayed vaccine because of Trump. Terrible tribalism that permeates the whole discourse in the US.
I think that the US failed to get proper statistics and other countries got this right. Yes, a lot of people died from covid but also a lot of people died because of lockdowns. And it is not surprising because that was expected.
What other countries statistics showed that people died from lockdown rather than COVID?
Trump was right to do Operation Warp Speed and get the vaccine out as fast as possible. But he can't take credit for it because so many of his fans hate the vaccine, and he wound up appointing RFK Jr despite denouncing him earlier.
See, you immediately started to talk about Trump's fans instead of addressing the issue.
Why do I care about Trump's fans? What does it bring to the discussion?
And you immediately didn't answer which statistics show deaths from lockdowns over COVID. I will remind you that you are commenting under an article that very clearly shows COVID deaths tracking excess deaths overall and both lines moving in a way that does not seem to track the severity of pandemic restrictions, which were highest during the first six weeks of response starting in March.
I am saying that most statistics show a lot of excess deaths more than from covid.
If I see one graph that contradicts this knowledge, I am right to be sceptical. Obviously the full story may be even more complicated.
You brought up Trump and the vaccine, I responded to that.
Our World In Data has the US death numbers through 2023: https://ourworldindata.org/grapher/number-of-deaths-per-year?time=earliest..2024&country=~USA
Death rate is back on trend in 2023, and it's easy to see that the total excess deaths over trend are ~1.2 million. 2.84 in 2019, 3.38 in 2020 (+0.54), 3.49 in 2021 (+0.65 over 2019, +0.51 over 2023), 3.19 in 2022 (+0.21 over 2023), and 2.98 in 2023.
I think you're right because I attended the funeral of my wife's great grandmother in the municipal cemetery of Sofia, Bulgaria and I saw the new fields the cemetery had to expand into. Hundreds of new graves with temporary wooden crosses that all said 2020.
Read the whole comment. It doesn't matter whose funeral I attended.
This is interesting, because expansion of cemeteries visible from space is also used as a measure of Russian military losses in the current war.
You gotta stop paying attention to your commenters. They just haven't got a clue. (Yes, including me.)
Very insightful information. The # of excess deaths during this period certainly show a real and honest trend that makes me less skeptical of the large # of deaths ascribed to COVID-19.
Where I still have a lingering bad taste in my mouth is the general COVID response hand the blame being thrust on the virus itself, and not the growing number of comorbidities that we as a global society have let metastasize in the general public that greatly exasperated the devastating effect that the virus had on public health.
When we speak about forest fires we have grown intelligent enough to recognize that it is not the nature of the initial spark that is the most determinant factor in how widespread the damage can be, but rather the long term accumulation of underlying pre-conditions that led up to the moment of ignition (i.e. the growing abundance of dry, dead tinder from poor forest maintenance; the expansion of high voltage transmission lines in these area; and increasingly dry/hot weather from localized climate change).
When we speak about COVID-19 I feel we really never addressed these underlying health issues. There was no big response to getting people in better shape to prevent the growing number of cases and deaths, rather we had everyone drastically slow down their lives to "bend the curve" which if anything had a counter-beneficial effect in actually making these comorbidities worse as people become less active.
I believe this is what has caused such a vague understanding of the total death count. Was it 1.2 million healthy people who all got hit with the worse global virus in a century? Or 1.2 million unhealthy people who caught a slightly more infectious and deadly virus? Or somewhere in the middle. And if its the third option, how do we move forward in preventing such devastating results from the inevitable next pandemic?
Most of the 1.2 million were unhealthy because they were quite old, and old age is a “comorbidity” that not much can be done about. Yes, our health care system’s priorities are wildly skewed when it comes to promoting overall health and well being. Yes, there are a lot of unhealthy people. But Covid mortality is so closely tied to age that it doesn’t make much sense to try to sort deaths into “healthy” vs “unhealthy” buckets
It really was caused by the virus, and the main way we're addressing chronic major health issues now isn't due to COVID at all but instead semaglutides coincidentally becoming available a few years after. After the "dry tinder" is exhausted in a forest fire you shouldn't see a fire there again, but we aren't seeing excess mortality undershoot in subsequent years the way it overshot previously.
Maybe we should try clearing out some of the old people so that we don’t have so many vulnerable next time a big disease comes about.
/s
I also sometimes people say the medical establishment has been ignoring obesity - I sometimes wonder if these people have ever been to a doctor or read anything from the medical establishment, because basically all of it tells people to lose weight.
I wonder if we can calibrate the worldwide data with the data from the US. 1.2 million is about 0.3% of the American population. If a similar percentage of people died around the world from COVID, that would be 28 million worldwide, assuming America dealt with the pandemic no better or worse than average.
Given the known age profile of the disease and of the US, you'd expect US deaths to be higher than average. And we don't have good data from countries with poor health-care systems where many people will have just died without getting any sort of COVID diagnosis.
But if we look at Europe (EU+UK+Switzerland+Norway+Iceland) then 0.3% is pretty much normal here (some other European countries report low and untrustworthy numbers, e.g. Belarus reports 0.075%)
https://www.statista.com/statistics/1111779/coronavirus-death-rate-europe-by-country/
Rich Asian countries (ie those with healthcare systems capable of producing quality statistics) tended to have more effective lockdowns and lower death rates as a result - some had huge surges when they finally unlocked, others managed to vaccinate widely and then unlock carefully (New Zealand, famously, had a 0.05% death rate, which is pretty much a best-case scenario).
This issue is a good illustration of how people hate hypocrisy 1000x more than dishonesty. The official public health community made some bad calls to say what they thought would save the most lives in the long run rather than expressing the full uncertainty and changing weight of evidence as more facts rolled in, as well documented here and in other places. And yet they probably were still among the top places to get reliable information! People recoiled from a source that claims to tell the truth and tries very hard to do so but got its priorities twisted and leapt into an alternative information world of grifters and liars who will never admit anything.
I largely agree with this comment and anyone who reads my comment as doing otherwise is wrong.
But this is also an excellent argument for, uh, just telling the truth. Lying didn't work, and permanently crippled our ability to react to crises.
In the years leading up to COVID arguments about "voting against their interests" and people being too stupid to know what's good for them gained full penetration in elite circles. As a result, we leapt directly to simplifying the narrative and keeping back information. And everyone noticed.
I think you have it backwards - the medical establishment used to lie all the time, and people trusted them as a result. But by 2020, they had adopted an ideology of admitting their failures and being more transparent than they used to be, and now people hate them.
> People recoiled from a source that claims to tell the truth and tries very hard to do so but got its priorities twisted and leapt into an alternative information world of grifters and liars who will never admit anything.
False dichotomy. You can fall back to your priors, and first-hand evidence, and heavily discount everything else. Which for most people would tell you something like "COVID isn't very dangerous, unless you're old or with comorbidities, but vaccines are pretty safe in general, if not always highly efficacious. Also the authorities are going to keep overreacting and keep claiming they're having a significant positive impact regardless of the truth, for political incentive reasons."
It worked pretty well for me. Including predicting (5-6 months in advance of the shot being available to people in my age group) that the government (this was in Ontario Canada) would impose vaccine passports. I made sure to save all the paperwork because at the time I didn't have a health card, and I figured I'd later need to prove I was vaccinated to access public places or travel. Which is exactly what happened. (Okay, I was a little too paranoid: I expected their computer system that recorded who got which shots to flat out fail as IT systems often do, so I literally had video recordings of getting vaccinated as a fallback proof. That turned out to be unnecessary: Recovering my vaccination records to get the fancy QR certificate only required multiple hours on the phone).
Other people had bad priors (mostly starting from "lots of random internet people agreeing with each other is a useful signal"). That sucks but the mistake wasn't in reducing trust in authorities.
Ok, more precisely my point is that people reacted to the authorities’ behavior by going from 95% trust to 5% trust (or whatever), leaving much less reputable sources as rivals or winners for many people. Sounds like this doesn’t describe you. Me either.
I'm reminded of Bounded Distrust (https://www.astralcodexten.com/p/bounded-distrust) though I'm struggling to phrase the exact connection, and rereading Scott's Lying Sequence will not be sufficiently edifying.
Anyways. If you want to avoid ever suffering from a pandemic, you listen to the people that predicted 10 of the last 1 pandemics. If you're a normal person that doesn't want to spend your entire life in a panic or building a survivalist bunker, you listen to the people that only accepted the reality of the last pandemic three months after it was too late. The side effect of the latter is the substantial failure mode that collapses consensus reality ("very rarely lies" can be more dangerous than "constantly lies") and leads people into epistemic nihilism.
Is it possible to see dips in excess mortality caused by COVID killing vulnerable people? Or is it somehow priced into the formula?
I personally know zero people who died from Covid and one relative of mine who died from the secondary effects of the lockdown (delayed cancer treatment in April-May 2020 => plausibly an earlier death). However, I do agree that 1.2M is a completely plausible figure. I know several people who were *close* to dying from the virus and many of them are still alive 5 years later, so it clearly wasn't "just the flu" for the elderly.
However, an arguably much more interesting question is:
- Was the ACX/LW/Tech Twitter crowd right about taking additional precautions in 2020, assuming they were under the age of 65, had a BMI under 35, and had no major preexisting conditions?
- Was the same crowd right about taking additional precautions after getting their vaccines in spring 2021, such as the insistence on testing during various LW/Rationality events that persisted until summer 2023?
I've mentioned this in a few Open Threads over the years and I still feel like the story hasn't been set straight.
I agree that the crowd I was a part in took too many precautions in early 2020. I still remember all the copper tape on doorknobs!
In https://slatestarcodex.com/2020/03/02/coronavirus-links-speculation-open-thread/, I included age-specific mortality tables and hospitalization tables. In retrospect, the mortality numbers were inflated by about 10x and the hospitalization numbers by >100x, because only severe cases had been detected and so the government was underestimating the number of cases (and, since they detected most deaths, overestimating death conditional on case). The mortality rates were low enough that even the 10x inflated ones weren't so bad, and I said that "the good news is that it’s pretty unlikely to kill young people". But the inflated hospitalization rates looked awful - 15% - 20% of young people who got COVID were hospitalized! I was suspicious at the time, saying:
> "This is a weird pattern – why are so many young people getting hospitalized if almost none of them die? ... Are these an overestimate? Maybe most cases never come to the government’s attention? There’s some evidence for this ... So hopefully the 20% hospitalization rate will prove to be a worst case scenario, and the real number will be less."
...but absent 100% proof that this was the explanation, I elected to stay inside rather than get a disease that had some chance of a 20% case hospitalization rate for young people. I'm not sure how I could have done better without being able to be confident in the rate at which the government was underestimating cases, something that's beyond my epidemiological knowledge (and apparently everyone else's). Also, later in the pandemic lots of people made the opposite mistake, assumed that the government was still underestimating cases when in fact by that time it had gotten pretty good, and confidently predicted that COVID would end near-immediately. So a heuristic of just always assuming detected cases are underestimates wouldn't have worked.
You can see this same dynamic in my more recent post on bird flu: https://www.astralcodexten.com/p/h5n1-much-more-than-you-wanted-to
> Was the ACX/LW/Tech Twitter crowd right about taking additional precautions in 2020, assuming they were under the age of 65, had a BMI under 35, and had no major preexisting conditions?
For me, the main motivation to avoid covid was to avoid accidentally bringing the virus to my older relatives. If I wanted to stay in contact with them, I had to protect myself, too.
Some of the things I did were probably excessive, but that is hindsight; and frankly I found most of the precautions relatively easy. For example, at the beginning I bought a bottle of disinfectant, and whenever I bought something in a shop, I wiped most things with a paper towel soaked in the disinfectant. Probably useless, but it took me one minute of time, so no problem.
> the insistence on testing during various LW/Rationality events that persisted until summer 2023?
In Vienna, during covid we had LW meetups under the open sky. No one checked anyone's tests.
The part about covid denialism that really irritates me is that it implies a belief that actually huge numbers of people in this country can't do their jobs at all.
Like: epidemiology and statistics has much much more sophisticated methods for attributing causality than just comparing graphs and looking at if they spike together. They know what they're doing. Likewise for developing vaccines, confirming they worked, and tracking the data on proving they're working.
And every single doctor and medical worker went through living hell for most of covid because of the spike in cases of people with clearly life-threatening illnesses of the type they are deeply familiar with. There are no doubts whatsoever if you worked in a hospital or nursing home.
And my favorite example: I lived on the island of Kauai in Hawaii for three months during Covid, which managed to almost completely prevent covid from showing up by implemented a strict quarantine (you had to quarantine for two weeks and the national guard would show up and make sure you were there). And people on the island would simply not believe covid existed because they hadn't seen it! Because their government had done a *good job*.
Outside of the conspiracy-theorist's low-trust life there is a huge world of mostly competent and well-intentioned people doing their best-obviously! But all they notice are the rare examples of incompetence or deception which they continue to harp on for years. Most of which make a lot of sense in context if you bother to read about what was going on.
Like, get over it. Reality is sitting there existing *even if* you are pissed about the media or cdc or whatever trying to manipulate you a couple times. You can be mad about that without turning delusional just to prove your point. We get it, you're mad about being powerless to prevent the bad stuff from happening. But having a tantrum and refusing to listen to everyone is not fixing that at all, it's just making it worse. If you want society to handle disasters better then maybe go contribute to fixing its general incompetence in some material way, like helping make society more high-trust, or contributing to the competence of your local government or something.
I disagree that those working with data to address pressing policy issues always know what they are doing. People are familiar with the replication crisis in psychology. I have first hand experience of research in physics and second hand experience of research in economics through my wife. Even if outright fraud is uncommon, there is plenty of work that is sloppy, rests on shaky assumptions, or is presented in a misleading way. I suspect epidemiology cannot be too different.
Psychology has been comfortable with BS for a long time. Freud & Jung were both pseudoscientists in the Popperian sense, but big deals in psychiatry. Robyn Dawes wrote "House of Cards" about the irrationality of his colleagues in psychology. Paul Meehl was writing about the things that gave rise to the replication crisis back in the 60s. For all the problems with physical medicine, they did at least switch from harming to helping people once germ theory, vaccines & antibiotics became prevalent. Dead bodies are an obvious outcome which psychology doesn't really have compared to physical medicine.
Yeah, there’s even a bit in one of Feynman’s autobiographical books about behaviorist experiments on mice and all the things you actually have to do to control for outside influence. To this day no one does any of those things. I remember this every time I hear about a social psychology finding
“ always know what they are doing” is very different from “ mostly competent and well-intentioned people doing their best”. People demand the former, and have trouble accepting that the latter is better than populism.
Highlighting the incompetence of specific people is contributing to fixing the problem. For example, you seem to put the epidemiologists and the vaccine researchers on equal footing. AFAICT the latter are much more competent in that the vaccines worked and the clinical trials, while slightly overstating efficacy, got numbers in the right ballpark.
In contrast epidemiologists continuously overestimated the mortality rate despite having access to a good source of true IFR data early on (Diamond Princess), insisted on using SIR models in all the curve-prediction efforts they made despite it fitting extremely poorly.
In the UK, where lockdowns happened semi-regularly based on epidemiologist predictions, when the 4th one was called off due to a backbencher revolt (largely driven by an open admission by one modeling team head openly admitting they don't bother modelling and presenting probable scenarios that don't require the government to intervene: https://www.spectator.co.uk/article/my-twitter-conversation-with-the-chairman-of-the-sage-covid-modelling-committee/ ), the predicted deaths failed to occur.
And the other famous UK model, the non-SIR model from Imperial College London, whose predictions are what made the UK decide to act much more aggressively, turned out to both reliably and dramatically overestimate deaths ( https://thedailyeconomy.org/article/the-failure-of-imperial-college-modeling-is-far-worse-than-we-knew/ ), trying to justify that the interventions suggested by their models saved lives by using their own model again, and when the source code became available it became obvious the entire thing is heavily non deterministic (even for the same parameters, including the random seed) and riddled with bugs and ad-hoc decisions.
Just because two groups of people are superficially similar (PhD degrees, peer-reviewed articles, etc.) doesn't actually get at the substance of who's actually doing science and who's cargo-culting it. Being able to sniff out the latter and stop rewarding them should be part of the conversation if we want anything to improve long term.
I think it’s wrong to say epidemiology is just cargo cult statistics. It’s less reliable than vaccine research, but only in the same way that cosmology is less reliable than planetary science, and nutrition research is less reliable than diabetes research. Some things are harder to study than others (and particularly harder to do controlled experiments with). That means we should put less weight on specific predictions from those fields - but it doesn’t mean we should cast out the field, because they’re still doing better than anyone else at addressing those particular questions.
> still doing better than anyone else at addressing those particular questions
That's begging the question.
In my example above, JP Morgan (not epidemiologists) were able to notice the obvious fact that the Omicron variant was less lethal than Delta, and that this made a huge difference in number of deaths expected in the UK when it started spreading there and what kinds of measures were appropriate to take.
When confronted with this fact, the epidemiologist running the UK advisory committee said such low-lethality scenarios, even though likely, were not modeled because that's not what the decisionmakers were interested in.
This is utterly broken.
I mean, that's almost certainly bullshit. Epidemiology is a sophisticated and evidence-based science. A couple examples of malfeasance doesn't invalidate the field as a whole: it invalidates those people, sure. But more generously, they were in complicated uncharted territory and almost certainly trying to do their best in spite of all the politics around them. But the way people talk it's like, one epidemiologist is wrong once and now you think it's possible their covid death numbers are off by two orders of magnitude. Like.. Lol? No way.
> one epidemiologist is wrong once
The heads of two of the biggest epidemiological forecasting groups in the UK whose work drove actual policy. And in one case not just wrong, but openly saying to anyone who can read that they weren't even trying to make accurate predictions because scenarios that don't lead to lots of deaths don't have policy implications. That's the guy who was the head of SAGE, the advisory board for the government on COVID policy in the UK (until this incident and growing fatigue caused a backbencher revolt and Boris Johnson ignored the advice to go into a 4th lockdown). I'm not nutpicking here, these are pretty central people, elevated as the experts among experts.
> and now you think it's possible their covid death numbers are off by two orders of magnitude. Like.. Lol? No way.
(Assuming you're talking about retrospective numbers, not prospective numbers): That some people manage to be even more wrong is not a ringing endorsement of the experts. You don't have to like how people react to competence crises by falling back to even worse methods such as believing the whole thing's made up. But when you decide that therefore
> huge numbers of people in this country can't do their jobs at all
is false, that's a transparent appeal to consequences.
My point is that the data people use to formulate their dismissal of 'the entire expert establishment' is massively underpowered for that conclusion. They see a few failures, don't measure the successes at all, and lazily assume the whole thing is bunk.
I think the whole issue is that epidemiology (especially for COVID) is not about scientific observation and experimentation (what the layman would view as science). It is using models and statistical methods to predict the future which is prone to motivated reasoning.
Another poster claimed that SAGE couldn't show that their models prevented deaths. That may have been too hard a level. But I have a more fundamental critique. Has SAGE release their original code and data? That would give me more confidence that SAGE just erred in their predictions and it was an honest mistake.
I really don't know and my quick search didn't turn up anything.
I forgot to respond to this, by the way, but I wanted to mention that you seem to know nothing about epidemiology. It is very very much a science that is based on observation, experimentation, and sophisticated statistics. Skim... any paper... if you want examples. (Not an epidemiologist myself, but my dad and ex both were, so I've interacted with a lot of their work, enough to believe it's very serious.) Any science, of course, is going to falter in a situation that has never been seen before. But you seem to be dismissing the field out of hand, and I cannot imagine why.
I remember that incident. In front of a parliamentary committee one of the epidemiologists defended their models saying, <paraphrased> "Well, we didn't intend them to dictate public policy." And one of the MPs responded, "Well, why are we paying you then?"
The SIR models absolutely sucked at predicting COVID deaths. A few epidemiologists raised this issue and proposed solutions, but SIR models are the bread and butter of epidemiology (when it comes to grants). No one really wanted to dump their precious models.
> make society more high-trust
If the powerful institutions of the media, CDC etc. regularly try to manipulate people, increasing trust them is bad, and you're really complaining about "conspiracy theorists" seeing them for what they are.
Unfortunately, people see evidence that we should trust these things to 90% rather than 99% as meaning that we should never trust them at all.
Trust is a difficult thing to calibrate, and unless you're confidently savvy to distinguish between good and glorious harvest announcements (https://www.astralcodexten.com/p/bounded-distrust), not trusting them at all *probably* is your "best" (a word which here means prevents the failure mode of inappropriate trust in accordance with your abilities) options.
The problem is that instead of retreating to well-tested traditions, or even some trustless epistemic nihilism, people in that situation tend to resort to whatever talking head is saying acceptable things regardless of correctness. Everyone likes a talking head that tells them what they want to hear, as was amply expressed during the pandemic.
Using your post as a starting point and extrapolating. Please correct me if I have taken the extrapolation too far. Taking your claim to the next level. Every claim by an authority should be evaluated independently and the people should trust if they believe in it 90%.
What I think most people do and what I do much of the time. I think they take a more Bayesian approach to making decisions (ie previous observations do matter) and whether to trust is binary (trust or don't trust).
So, people were trusting the authorities at the beginning of COVID. But they then started loosing trust based on the actions of those authorities. I am not going to rehash it all (others have done it in this comment section). I am referring to the mask issue and also the political issue where the authorities at least gave tacit approval to some protests.
The person assigns a 90% trust score to the specific claim based on the facts they have. But based on priors, the person assigns a 50% trust score to the authority making the claim. They discount that 90% by 50% so the score is now 45% which is now in the not trust territory. They assign the binary score of not trust and conduct their life with that belief.
No, I mean acting in trustworthy ways and helping other people feel like the world is generally high trust. That doesn't mean you should trust those media companies in particular.
"implies a belief that actually huge numbers of people in this country can't do their jobs at all."
I think it more implies a belief that huge numbers of people don't actually have the jobs their job title would naively suggest.
If someone's job title is "academic research scientist", for example, we might naively expect that their job is to learn new things about some aspect of science, things that are real and true and important, and then to report them accurately. But that's not their job, because that's not what we pay them for. Their *job*, is to get research grants.
And while research grants are correlated with doing good science, the correlation is often quite weak. Some research subjects are more likely to get grants than others, even if they are less valuable to the world at large. Some *results* are more likely to lead to future grants than others, even if they are false and the others are true, And how the result is presented, matters too.
If you're a research scientist and some journalist calls you asking "Is This Thing in your area of expertise a Major Crisis?", then if you say "no" or "who knows?", it doesn't bleed so it doesn't lead so you go back to your desk full of long-shot grant proposals and wonder if you're going to be able to pay your grad students. If you say *yes*, you get your face on TV and people will recognize you as the Expert on That Thing. And That Thing will be elevated in priority, with more research grants going into the field. You've got grad students to feed. What is your real job here?
And maybe you have the intellectual integrity to not outright lie about it, but now you're in a situation where your salary kind of depends on not understanding that the profitable answer is a lie, and motivated reasoning is a powerful thing.
People in nursing homes and with chronic debilitating conditions tend to have small social circles and low visibility. My memory of the AIDS epidemic of the 1980's is of friends dying. What most people will remember of Covid will be lockdowns, masks, and social distancing. (excluding critical care docs and nurses).
I’m sorry for your loss. I can’t imagine what it must have been like to live through that decade.
There is maybe no greater argument that this (excellent) article is an exercise in futility than the existence of the article itself. It’s a reflection of over five years of mass delusion, despite consistent, excellent analysis and advisement.
OK Scott, you've convinced me, 1.2 million is probably a close number based on excess deaths.
The real question for this reckoning is: what should we have done differently?
--More? (more NPIs, more vaccines)
--Less? (Targeted NPIs and vaccines to at-risk populations, fewer or no NPIs/vaccines for everyone else)
--Nothing? (the NPIs and vaccines did the best they could)
How would one prove it?
For an extremely annoying answer, see section "Actual Evidence 1.1" in https://www.astralcodexten.com/p/lockdown-effectiveness-much-more
Thank you, I missed it the first time, and it's not annoying at all. The CoronaGame deserves a revisit.
>CoronaGame
Instead of reading the article I'm going to assume this means putting a bunch epidemiologists and public health experts into a dome for a deathmatch, and whoever survives get to set WHO policy.
(Jk the actual game looks really neat)
Man, how did I ever forget your conflict of interest for that one paper?
That graph of actual country's performance plotted against the efficient frontier people achieved in the game is really interesting, do you remember where you got it from?
I think it would be interesting to see more countries, with an extra few years of data, to get a sense of what the space of possible outcomes looked like.
From an email conversation with Jan - I don't know if that exact picture is available publicly, but he has a different version with commentary at https://boundedlyrational.substack.com/p/on-the-dilemma-between-lives-and
Awesome, much appreciated, thanks
Vaccines work better than most medicine. The anti-vaccine movement are about as wrong as it's possible to be.
My answer is "what New Zealand did", but that (shut down all international and interstate travel entirely, yes that means that NZ citizens abroad couldn't go home for over a year) wasn't even close to the Overton Window.
Part of “what New Zealand did” is “have two time zones worth of water between you and any other land”.
I think the evidence is that the NPIs were not that effective, and the vaccine was extremely effective. So my recommendation with hindsight would be fewer to no NPIs but much stricter vaccine mandates and requirements. I would also have tried to increase production of Paxlovid and increase access to early mABs.
It’s very hard to measure the effectiveness of distancing behaviors - most of the measurements were of distancing regulations. (I also think we shouldn’t have bike helmet laws even though bike helmets are probably a good idea.)
"People still don't think it happened" was my first thought reading yesterday's post.
I think all this, from ignoring covid deaths to denying covid deaths to claiming charities are bad, is better explained with the Biggest Logical Error Of Our Age: Things that prove us wrong are never true.
You might think, "That's not even a logical error." But it is. The process works like this:
1) You have priors for how the world works.
2) You inculcate a philosophy of skeptical rationalisim - there is no higher authority that defines truth. It's something you discover yourself. And you follow the "rationalist, scientific" method where you don't update without good evidence.
3) No evidence is good evidence. Experts are corrupt, information sources have agendas, anecdote can always provide a counter-example to any statistic.
4) You are never wrong, and you know this because you did a stringent analysis of the evidence that you were wrong and it's bad.
I want to be clear that I'm not saying "other people sure do this a lot." I think I am probably the world's biggest perpetrator of this fraud. I don't know how to clear it without giving up on rational skepticism, and I don't know how you determine truth without using rational skepticism. This community in particular has lots of tools for this, which is why I'm so interested in what you write. But I still think giving people these tools mostly just gives them more ways to rationalize their priors.
But if in 2019 you thought this whole thing was an exaggeration, would you rather believe that you were wrong and your wrongness killed a million people, or that you were right all along and the number is made up.
Postscript: Deaths from Covid were extremely concentrated, especially in big cities and especially on the <corrected, said west> east coast. A huge percentage of those deaths happened in New York City alone. If you're a rural republican who doesn't want to believe it happened anyway, it's very likely the pandemic never even really reached your sparse town...
Do we know why there was a such a mess in New York City? Yes, they were the first place in the US to get COVID, but shouldn't other places have had the same problem once the pandemic reached them?
I'd have to look at it more, and I might be doing exactly the thing I complain about above...
But if I recall correctly, our early preparation was really bad, and then after New York the other major cities went "oh crap it's going to be really bad" and took precautions like increasing hospital capacity and expanding staff.
(Instead of writing ten thousand qualifiers I should do the rationalist thing and just say "I only have 55% confidence in this because I haven't read up on it" lol)
One thing a lot of people have pointed to is infected people being sent into nursing homes to infect more elderly people, but I don't actually know how what impact such policies had.
First place hit by (probably?) the worst strain and rapid dense spread during the period when most "mainstream" types were still skeptical, large elderly population, and the stupid nursing home policy.
My personal conspiracy theory is that the 'rona was circulating in the US much more rapidly or earlier than people were paying attention, but I have no good explanation for why it wouldn't show up in the excess deaths sooner because of that (even if they're not labeled covid deaths). It's mostly rooted in my personal experience. *Maybe* I happened to get the worse flu of my life and lasting lung damage early March 2020 when theoretically it wasn't in my state yet, but it's a heck of a coincidence. I hadn't been to China so my negative flu test was treated as evidence it was probably flu anyways.
I wonder how much is due to intuitive distancing behavior. In march 2020, everyone still treated staying home from work when sick as optional. But once you knew in your gut that there was a pandemic, you were going to be a bit more cautious. Everyone adding that bit of caution would slow the spread and make the waves crash more gently.
As an ICU doctor who personally has seen over 200 people die directly from Covid (guess the famous CT scans were just fake!) and saw firsthand how hospitals got overwhelmed, the comments here are making me more angry than anything I think I’ve ever read.
I feel deeply betrayed and disappointed. Humanity together was at war with a common enemy, and people found any level of discomfort or inconvenience too much to bear. Easier to practice denial instead, and leave those fighting on the front lines, like you, to carry the weight alone.
I always imagined that when the rubber hit the road, people would take a stand to fight together, but it turns out most people are self dealing spineless cowards.
Thank you for fighting on the front lines, and know that some of us do understand and remember what you did for us, and were doing everything we could to fight along side you.
> at war with a common enemy,
Yes, but they mostly fought the wrong one.
And unfortunately they are still waging war on the scientific research establishment.
I know a nurse who got completely burned out, not in 2020 but in 2021 when people who refused to vaccinate flooded the same hospitals. Again. She had a sense of purpose in 2020, but when the unvaccinated waves came she just couldn't do it anymore.
So yeah, very little patience for this crap. That's why I stay away from this whole disaster of a debate, it just makes me want to punch people.
I don't think anyone here is arguing that "no one died from COVID". The debate is over "I think fewer than 1.2 million people died from COVID". Why would that make you angry?
Some people are saying the whole thing was vastly overblown. For example, "after the beaches and playgrounds were reopened, it seemed to have been more like a bad flu season"
The debate is not over “I think it was 1.1 million rather than 1.2 million” or “I think it was 1.3 million rather than 1.2 million”.
This has been a fun experiment in different factions of Scott's audience being radically uncharitable to each other and substituting the comments actually here with the most ridiculous caricature they can imagine.
I kind of hope that was his plan all along and this will feed into a future banger describing some interesting social phenomenon.
My ex was an MICU NP in NYC throughout the pandemic and, reading through these comments, I keep thinking she would be absolutely livid to see what some people are saying here.
I've never met anyone from Rhode Island, so therefore it doesn't exist. /s
And now you have! The Rhode Island quantum wave function has collapsed into an observation.
Fake News!!!
Delaware is a state I've heard that said about.
This is how I feel about Idaho. Those potatoes all come from Montana or the state of Washington. It’s an Area 51 level hoax.
See also: Bielefeld conspiracy.
The claim is not that physicians were lazy and/or dishonest when they classified deaths as "COVID deaths." The claim is that they were following protocol. On March 24, 2020, the CDC, through the National Vital Statistics System, changed the way that the underlying cause of death was to be determined for COVID, and only for COVID.
https://www.cdc.gov/nchs/data/nvss/coronavirus/Alert-2-New-ICD-code-introduced-for-COVID-19-deaths.pdf
From the bulletin:
"Will COVID-19 be the underlying cause [stated on the death certificate]?
The underlying cause depends upon what and where conditions are reported on the death certificate. However, the rules for coding and selection of the underlying cause of death [on death certificates] are expected to result in COVID19 being the underlying cause more often than not."
Further:
"Should “COVID-19” be reported on the death certificate only with a confirmed test?
COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death. Certifiers should include as much detail as possible based on their knowledge of the case, medical records, laboratory testing, etc. If the decedent had other chronic conditions such as COPD or asthma that may have also contributed, these conditions can be reported in Part II."
No other disease or medical condition has ever been treated this way. The result is that any time someone either tested positive for COVID or had respiratory symptoms, they were required to be classified as a "COVID death," regardless of whatever else may have been going on with them.
No laziness or dishonesty was required on the part of the certifying physician.
Exactly this. No lab test was necessary to say that the person dying had covid. It would be interesting to get statistics by reviewing randomly selected records to see how many were actually announced to be dying from (or with) covid.
“Assume” in panic times should be taken with a big grain of salt.
But one other thing is that during peak times hospitals became a breeding place for covid infection. If you came in with, let's say UTI, you would get covid. During peak times no one could ensure good separation of patients, isolation etc. My mother went to the hospital with UTI and came back with covid, then had to go back to the hospital again to treat covid. Scott assumes that the background rate of covid infection for those in the hospital is the same but it is clearly not possible.
Both this comment and the parent comment miss the main point of the post. Lots (millions) of extra people died over the normal rate (“excess mortality”). Why, if not from COVID?
Yes, from covid but also from lockdowns or whatever it was.
Every single data point shows this. Most countries with lockdowns have clearly diverging lines.
Opioid deaths increased greatly during covid in the US. This graph provided by Scott seems outlier. As it comes from official statistics, I have to conclude that the most likely explanation (Ocam's razor) is that the statistics are poisoned. People who died from other causes were often counted as covid victims.
I don't know the reason for those excess deaths. It is the thing that needs to be explored. But USA is very much influenced by tribal politics and both parties don't want to investigate that it becomes almost impossible.
This paper from BC [1] suggests an excess mortality of about 6.5 per 10,000 of population in 2020 among people who tested negative for COVID (and didn't subsequently test positive any time in 2020). I would expect this to be an overestimate, given that people who are tested for COVID are likely to be otherwise unwell (early in the pandemic, you _needed_, IIRC, to have some other risk factor, or be hospitalized, to get a test).
Extrapolating that to the United States, that gives 195,000 excess deaths from non-COVID causes in 2020. (BC had very few COVID cases in 2020, but I see no reason to think that non-COVID excess deaths should be proportional to COVID cases, unless some of those non-COVID excess deaths are caused by true COVID cases in some way (people dying of heartbreak after losing a partner, for example?), in which case we should count them anyway).
So excess deaths in COVID-negative people don't explain most of the excess deaths observed in the US in 2020.
[1] https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-025-21782-9
I am interested in more than one year period.
Deaths were going up even prior to any lockdowns. The lockdowns were in response to deaths. Early on we had officials downplaying things and encouraging people to go to NYC's Chinatown.
An outlier compared to what other statistics that you find more credible?
Looking how Sweden didn't have a typical peak and trough curve but more level, I find it more credible that peaks were created by lockdowns. They caused the very thing they were supposed to prevent.
Do you have graphs of excess deaths by month in Sweden?
Well, that's the question, isn't it?
I'd certainly be interested in finding that answer. But one is under no obligation to simply accept the explanation proffered by institutions that were deliberately juking the stats. Particularly when said institutions have gone out of their way to smear and stigmatize any who would question their credibility.
All-cause mortality numbers are, one assumes, still pretty hard to fake. In the U.S., anyway. China doesn't even know what its actual population is to within a few hundred million. But that, in and of itself, doesn't tell one anything about what caused those deaths.
This is just not true? We were obsessively testing everyone in the hospital. Sure, we had some cases of nosocomial spread but they were not common.
Most likely it is true.
Maybe people were obsessive and at the same time completely negligent. Testing made no difference in controlling covid spread at all. So much money was wasted.
My own mother in 2022 went to the hospital for UTI and got covid. They tested her at admittance but not at the discharge. No one cared. She came back home, tested positive and soon had to go back to the hospital again due to breathing difficulties. (Imagine people were banned from twitter for saying that it was possible for fully vaccinated to get covid.)
My father got infected from my mother, died a week later at home. No doctor even looked at his body, wrote a death certificate that doesn't mention covid. It was good because the crazy pastor would not have allowed the open casket funeral otherwise. See, all those people were unreasonable. Had strange beliefs, didn't do their jobs properly, and at the same time moralized over others.
That's why I am skeptical about these statistics because they don't align with the total picture. I don't want to end like Scott who trusted the statistics that 20% of young people get hospitalized (https://www.astralcodexten.com/p/the-evidence-that-a-million-americans/comment/119233511). He could not explain how it is possible if all other data disagreed. Why don't be brave and say – the data is most likely corrupt and should not be trusted?
See, I am not covid denier or vaccine denier. In fact, I predicted the total number of deaths quite precisely and yet I saw that children and young people are very little affected. I saw it better than Scott.
> Testing made no difference in controlling covid spread at all
What makes you say this? In the case of other illnesses, contact-tracing can be very effective. I suppose you could argue that we should have been contract-tracing and collecting statistics by clinically diagnosed cases (that is, people who a doctor diagnoses based on their symptoms rather than a test), but that seems like a weird argument (tests are cheaper than doctors' time).
I thought it was common knowledge by now.
Is there any evidence that it worked? In medicine we should assume that things don't work by default.
People even made the mobile phone app that was based on modelling how it could reduced cases. Didn't work at all.
From first principles, it seems like it should work: isolating people who might have been infected can prevent them from spreading the illness further. It's hard to get statistics on exactly how effective it was on a population level, but that doesn't mean we should assume it doesn't work; rather, we should try reasoning about it in other ways, such as reasoning from first principles. It's true that it might have been a waste of money, but it might also not have been a waste of money. We can't really be certain either way, so it makes sense to have funded it even though it wasn't clear how effective it should be.
Contact tracing works when you've only got a small number of people spreading the disease. If we had locked down the borders to prevent entry, we could have relied on it. But instead we had enough infections for it to become endemic.
That is just a speculation without any evidence whatsoever.
I’m really sorry to hear your story. My spouse also clearly caught COVID while working in the hospital and using precautions such as surgical masks and hand hygiene, so believe me I understand that these precautions were far from perfect. But I don’t really think it is clear they did not have some degree of efficacy. The pandemic really highlighted the methodological limitations in most studies regarding these interventions.
I mentioned our frequent testing to point out that we did not observe frequent nosocomial spread, either at my institution or in reports from others. We did not have many patients who were discharged then readmitted with COVID either. It happened but was not frequent. Nosocomial spread was something the medical community was highly concerned about. It was a major driver of the initial push for early intubation that turned out to be the incorrect approach. Estimates vary, but for example this study from the UK, where shared hospital rooms are more common than the US (https://www.nature.com/articles/s41586-023-06634-z) suggests a 1-2% nosocomial spread rate.
As for the efficacy of test and trace, studies vary on their estimates and certainly have a lot of systematic bias due to their observational nature, but most reviews do suggest some benefit, such as this one (https://link.springer.com/article/10.1007/s10654-023-00963-z). I am not certain of the precise magnitude of efficacy, but I don’t think anyone can clearly say it doesn’t have some degree of impact.
These interventions are just really hard to assess in the context of a changing pandemic with changing population immunity, while multiple different interventions are being used in different ways in different locales. Sure you can do some clever “natural experiments” leveraging these differences but there is always so much residual systematic bias it is hard to really know. Nothing beats good prospective experimental methodology.
I would really love to see more robust research on this. We should have adaptive platform randomized trials, leveraging the amazing power of Bayes theorem to test multiple NPIs in combination, geared up and ready to be implemented when H5N1 or the next pandemic hits. Unfortunately, there is no political will for this in the US currently but the EU is working on it which is encouraging.
By the time vaccine became available most healthcare workers had already got covid infection. It means that any attempts to limit the spread of covid in hospitals were useless.
Ok, maybe not completely useless but no one showed a graph in advance how the rate of covid infection among hospital workers were expected to rise reaching 100% in a given period. Basically it means we have no idea how effective those measures were. You are trying to quote studies but why should I trust them?
Similarly I saw a lot of studies trying to measure how effective covid vaccines are in limiting the spread. Until at some point seroprevalence studies showed that nearly 100% of population had been infected with covid. Apparently all this modelling was completely wrong and didn't have any effect whatsoever on attempts to change the course of pandemic. I was better following the statement that original covid vaccine trials did not attempt to measure its effectiveness in stopping transmission therefore this is not guaranteed or in other words, no evidence that it stops the spread. (Ironic that Scott wrote an article that experts misuse “no evidence”=“do not work”, to assert that masks actually help but at the end it turned out that masks actually do not work).
I don't want to look into more research. I think that most people trying to do that, including Scott, are clearly over their heads. The reason is that it requires a lot of time and man hours to review and evaluate those studies. Only big groups can do that, like Cochrane or NICE but not a single person.
Scott tried several times: he reviewed masks and got it wrong. He defended lockdowns and got it wrong. He even believed that 20% of young people get hospitalized despite his strong analytical skills. What hope do we have?
Instead, I take more humble position. I assemble a puzzle from trusted evidence that has high confidence. For example, death rates are trustable. whereas average life years lost by covid death is not. “A person was expected to live 10 more years” is a modelling, not something that is measured. Simple fact that the average age of death from covid was around 84 years and average life expectancy is 83 years makes it suspect. It could be if there was great variance in age but we know that risk of death was exponentially increasing by age. It means that 74 year old person was about 10 times less likely to die from covid than 84 year old. It becomes mathematical impossibility to have 10 years lost on average. Maybe you can make a model where it seems possible but why should I trust such a model if so many other models turned out to be wrong.
I would wait for Cochrane or NICE to make a better estimate but they are probably too busy with other things. But a lot of people won't wait and will believe any model they are being presented. I am ok with keeping uncertainty about this issue.
You said, "The result is that any time someone either tested positive for COVID or had respiratory symptoms, they were required to be classified as a 'COVID death'."
What you quoted, however, says no such thing (or anything remotely similar). It says, "COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death." (sensible guidance) Is the reason for your confusion that you are misreading the first part? That just says that a lot of people are dying of COVID and we expect that a lot of the death certificates will reflect that. It is not an instruction. It is an observation.
Agreed. Since the parent comment claimed that “No other disease or medical condition has ever been treated this way,” I checked the CDC handbook on certification of death. It contains no requirement that any diagnosis be supported by a laboratory test.
https://stacks.cdc.gov/view/cdc/131005
You're missing the point, and misunderstanding the way death certificates work.
"Cause of death" is not a unitary concept, but instead contemplates both parallel causation and chains of causation. It also recognizes that not all causes are equal in either their contribution to a particular death or their clinical and broader significance.
On death certificates, "cause of death" is divided into two parts, Part I and Part II. Part I is used for those conditions/symptoms that most immediately and significantly caused/contributed to death. Part II is used for conditions/symptoms that were present, and while they probably contributed to death, they weren't deciding factors. So, for instance, a 90-year old patient with stage IV lung cancer who caught a cold and died might have just "respiratory arrest" and "lung cancer" listed on their death certificate, but another physician might also include "UTRI" as well. But we all recognize that it wasn't really the cold that killed them, even though the cold did indeed contribute to their death. It was really the lung cancer that did it. So respiratory arrest and cancer go in Part I, but the cold, if it's going to be included at all, goes in Part II.
So first off, "caused or is assumed to have caused or contributed to death" is an incredibly, incredibly broad net and normally subject to some kind of discretion. If a condition is present, it's basically impossible to say it didn't contribute at least a little bit, regardless of what else is going on. Even so, an illness or condition that is otherwise trivial but just happened to tip someone over the edge might, in the certifying physician's discretion, be omitted, particularly when we're talking about frail, elderly people who are one stiff breeze from keeling over anyway. But what the CDC guidance says is that COVID should always be listed, if present, no matter the circumstances of any given patient's death. Doesn't matter how much COVID contributed. This serves to artificially increase the number of death certificates on which COVID was listed by removing certifier's discretion about whether it was serious enough to be worth recording.
Second, the CDC guidance says that COVID always goes in Part I, exaggerating the overall seriousness of COVID's contribution and making it basically impossible to distinguish between "dying with" and "dying from". It's a nondiscretionary inclusion, and it has to be put in the most serious causation section.
No other condition is treated that way.
RE: Commenters saying they didn’t know anyone who died so it wasn’t real.
I worked briefly as a security guard at a hospital around mid-late 2020. There were bodies everywhere! Every single day we had a covid patient die, and the ICU was at max capacity for months. The death toll was never unbelievable to me.
I don't dispute that such a case could be.
And yet at the same time health care professionals on average were underutilized during pandemic. How do you explain that?
Also, last year NHS had flu season and overflowing hospitals and many reported, it is worse than it was during covid? How could it be?
Yes, the death numbers are real and yet it is not the whole story.
The art is not selecting one piece of evidence and make a story. It is by collecting a puzzle.
First were videos from China where people on streets were dropping dead from covid. What those videos mean? Were they real or showing what we were told they mean? Obviously not. Some pieces are poisoned. If you believe everything, you will not be able to make a puzzle because they won't fit.
I don't understand what you are talking about in terms of "poisoned pieces".
Health care professionals were on average underutilized during the pandemic, in the same way that I-35 through Austin is on average underutilized. Sure, during rush hour it can take 45 minutes to drive a couple miles, but 80% of the day it has less than one car per lane per five seconds. We unfortunately don’t have a good way to Target capacity to the moments when it is most needed, and during the pandemic we tried to make up for that by reducing usage generally.
We actually had. By not introducing lockdowns. Somehow, those countries who didn't introduce them generally had higher but flat curve which led to more stable usage of healthcare services.
I was in the DC area during the entire pandemic. We went to and from from the house with zero interference and so did everyone else.
.
https://www.cdc.gov/mmwr/volumes/70/wr/mm7014e2.htm
6.6% of deaths were from covid, rest were "with covid"
That's because it's arguably incorrect to list a death as being simply from COVID. Rather, death certificates would say something like "acute respiratory distress syndrome, caused by COVID". That's still a death caused by COVID.
Death certificate would not read ARDS caused by covid. Certificate would read primary cause ARDS. Other conditions etc. If ARDS caused by covid should read Covid primary with etc. CDC hides the most interesting data in technical reports and footnotes. Turns out they knew the vaccine was causing myocarditis and deaths but suppressed it. Oops.
This is a California death certificate form [1]. If section 107 would read "(a) ARDS, (b) COVID", then that should be counted as a COVID death, right? Those are the death certificates the paper is talking about (the 91.5% (94.5% * 97%) of certificates that list a "plausible chain-of-event condition").
[1] https://www.typecalendar.com/wp-content/uploads/2023/05/free-death-certificate-in-california.jpg
Scott's post was directed at you, but you still need to read past the headline.
I know someone with MS! Wait, shouldn't be that excited about that...
Same.
I recently learned that MS was very convincingly linked to Epstein-Barr virus a couple years ago. I haven't heard if there were any new treatment recommendations based on that but it seems like taking drugs to keep EBV inactive might be a decent idea.
I still can't understand my fellow Americans reactions to anti-covid measures. I thought we were a country that was proud of our resilience, and strength throughout history.
We were at war with something literally killing millions- as far as we knew the greatest threat we ever faced, especially early on when information was unclear, and some reports were suggesting 10% mortality rates. We had a chance to stand together and fight, but people were cowards, and the mild discomfort of masks and distancing was too much to bear, it was easier to pretend it wasn't happening, even if the pretending killed your grandparents, and your elderly neighbor.
Then these same traitors had the gaul to invoke masculinity and say stuff like "you're not a real man if you're cowering behind a mask." I got this from people literally spitting in my face and invading my personal space, while I was at the store volunteering to buy groceries for the immunocompromised people at the retirement home next to my house. I'm young and healthy, I am wearing the mask to protect you and the guy with failed kidneys I'm buying these groceries for, not to protect me.
I wouldn't have put this in these terms but I agree on how weird it was to see people claim masking was a sign of weakness and fragility when it was an inconvenience I was taking on for the benefit of others - something I'd think of as quintessentially masculine behavior. Also weird was people talking about personal choice literally spitting on me for making a choice, personally.
The language is important- we all had a once in a lifetime opportunity to rise to the occasion, and practice leadership in our own communities, to protect the weakest among us from being killed and disabled. Some of us showed a total lack of character and strength - and then try to hide their shame with the type of language I am using, painting themselves as some type of masculine warrior fighting for freedom.
I was followed, spit on, and mocked in public for being out with my 2 year old son in a mask. I was following strict distancing and laboratory grade sterility measures in public, because I was the only person bringing food to immunocompromised elderly people that would die if I didn't do so. I bought their groceries in full lab PPE, and dropped them off at an agreed upon location so they never came near me. I will never forget the cowards that tried to actively thwart my efforts to keep my neighbors, friends, and loved ones alive and mocked and betrayed me in the process.
Because the message that 20% of young people get hospitalized was a lie.
Not a lie, early on there was no good information yet, all anyone really knew is that hospitals around the world were overwhelmed, and the bodies were piling up faster than anyone could deal with.
Information should be presented with proper confidence levels. It was not done, so essentially it was misinformation (aka lie).
Basically this is why companies are not allowed to make any medical claims about their products unless they are licenced by the FDA.
Ok, this is slightly different but the same thing – people will misunderstand statements with terrible consequences.
Oh brother.
I’ve also encountered this stuff, as have many I know.
It predates COVID and masks, though. TBH I’ve never understood people who initiate hostile confrontations with complete strangers in public for clothing choices.
> you're not a real man if you're cowering behind a mask
…yup, you’re only truly brave if you cowardly conform to my ingroup’s expectations instead of just wearing whatever the hell you want.
It’s just football hooliganism bleeding over. I don’t think there’s anything new or complex or special here.
It may have helped if more people calling for lockdowns weren’t at the same time minimizing the level of disruption and sacrifice required to have a lockdown. Lots of “oh why do you care about the *economy*, grandma killer?” at people genuinely afraid of losing their livelihoods. “Oh just stay home, use FaceTime” as if missing births, deaths, church, and everyday socialization was all zero value stuff that only a selfish idiot could dare weigh against The Virus.
Crazy conspiracies on one side and toxic, condescending virtue signaling were two sides of the polarization coin.
> Crazy conspiracies on one side and toxic, condescending virtue signaling were two sides of the polarization coin
I agree with that, and see how they drive each other as a system. People want to believe the conspiracy theories to avoid the guilt being thrown at them from the other side. And then the virtue signaling is doubled down on in response to the conspiracy theories.
The lockdowns were a good idea for a short time when there was so much uncertainty, but they went on way too long after we discovered that masks worked, and so did good ventilation and being outdoors.
My theory is that, in much of the country, we “threw away our shot” on lockdowns too early. As soon as New York got bad, everybody locked down hard right away, anticipating an immediate huge spike that in many places didn’t come. I’m in Arizona, and people were fairly compliant with the initial orders. But then the massive emergency hospitals were empty, the wave of death was more of a ripple, and people were absolutely sick of being home by Memorial Day. At which point everyone went bonkers at the first lifting of restrictions, and THEN we had a huge wave.
Covid in America was more of a rolling wave and I think we may have been better off with keeping areas open until cases started to spike. Keep medical systems from being overwhelmed, but not tried too hard beyond that point because the returns diminished.
An additional "thumb-on-the-scale" for why you didn't hear about elderly people dying was that a large number of elderly people are socially isolated -
https://pmc.ncbi.nlm.nih.gov/articles/PMC5498693/ "more than half of elderlies were in isolation mode and high risk for isolation in terms of social network"
Meaning people are less likely to hear about an extended elderly family member dying, especially with with everything else going on during the pandemic
> The yearly mortality rate for 70 year olds is 2%, so the risk-during-the-time-they-incidentally-have-COVID is 2%/(52/2) =~ 0.08%.
This is wrong reasoning. Hospitals were the place where covid infections spread the most. The rate of testing positive in hospital was probably 10 times the background level. Even if you had no covid before hospitalization, you most likely got infected in the hospital. People, even doctors were so naively unaware about those risks. They trusted masks and hand washing which did almost nothing.
Even more, no covid lab test was actually necessary to report a person to be dying from covid or at least with covid. Did the person have cold symptoms, cough and difficulty breathing? Yes (who doesn't before death?) Let's write that this was covid. Scott clearly underestimates how much incentives can play role and misreporting cause of death is a normal thing.
Do you think anyone was worried that a death is reported as covid without testing and/or sufficient evidence? If you do, then you have forgotten the panic that existed during covid pandemic. Any talks that lockdowns are unnecessary were met with abuse. Experts and lay persons were censored for saying the correct thing, banned from social networks. Even most doctors online said all kinds of absurd things.
I can easily imagine that during outbreaks panic becomes contagious. Once people start dying from covid in the hospital, the staff starts suspecting that everyone who is dying has covid and therefore I don't trust any statistics from those times.
A good start would be reading the article. Scott addresses this - his argument is that all-cause mortality is higher during that period, not that there were a lot of confirmed deaths from Covid.
He says all-cause mortality is higher. That could be covid deaths plus excess deaths from lockdowns etc.
I am just saying that the graphs showing deaths from covid are fiction. We have no way of knowing which factors contributed more.
As most countries had similar rates of deaths relative to Covid infection, this requires a global conspiracy to doctor (sorry) the causes of death to attribute hundreds of thousands of deaths cause by lockdowns (?) to Covid. Why and how? Why would the Chinese government do this, for example? Why the government of New South Wales? Tough one for Oxfam’s razor
There is no conspiracy. In fact data clearly shows a lot of excess death more that just from covid. Especially for Australia.
This was also addressed in the article.
He just waves this away but saying it could only be covid.
We used a host of lab tests to diagnose covid, not just a positive PCR. The symptoms are actually fairly distinct, and the blood work is quite distinct. One example, flu, RSV, rhinovirus and other common colds do not cause d-dimer levels to spike up. Another, ferritin levels spike up. Another, blood clots form in the lungs. Another, quite distinct patterns on lung x-rays and CT scans. The first couple of months were quite confusing since we had never seen any disease with the pattern covid showed. Now it's routine.
Yes, but this is about what was put on death certificate. My father's cause of death on death certificate has no relation to reality. I am not going to believe that this was an a rare exception. They didn't seem to care much, to be honest.
So, according to you hospitals were clusters that helped COVID to spread, but lockdowns to prevent clusters were a bad idea ? Don't you see a contradiction.
Lockdowns were not to prevent clusters at all.
Locking people at home doesn't prevent clusters ? To me it seems straightforward.
Lockdowns actually created big peaks. See data how Sweden and Belarus avoided them.
The biggest problem with pandemic is the amount of disinformation that was distributed by the authorities. I cannot blame people that still believe these lies but these lockdown policies were bunkers and extremely damaging.
Do you have a theory with predicting power that explains how locking people at home alone or with few relatives causes clusters and helps virus diffusion ?
Because my predicting theory is straightforward : virus has dramatically less opportunities to reach new hosts, so I expect that Rt crumbles after incubation delay.
And it seems to be the case https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-021-11684-x After one month of the first lockdown Rt fell from 2-3 under 1 il all cities covered by the study.
Locking people at home happened in China not in the west.
You have to be precise when talking about these things.
People had to stay at home during lockdowns, they obeyed voluntarily or out of fear being fined or sanctioned but they were not locked. And they had to leave the house to buy food or do some other urgent things. Many people had to go to work.
Rt might be reduced but we don't really understand this that well. All models about pandemic failed, turned out to be wrong. Some places had very little lockdowns and virus didn't spread in the first year. Basically we had little information how and when any lockdowns could be introduced to achieve effective suppression.
Sweden decided not to bother. They did better than any other country.
It says something when Trump was the incumbent at the outset, that Operation Warp Speed is widely regarded as a success, yet COVID policy is an albatross for the Democrats (despite it probably costing Trump the 2020 election, and lockdowns being less of a thing as time went on).
I think it's mostly uncontroversial that there were surplus deaths from Covid. What Democrats need to contend with is that lockdown policy and the like are now regarded skeptically. This has tarnished their image and doubling-down on unpopular things (see: Biden's laundry list) is not helping.
It's not enough to be right about some technicalities. Pick the right battles.
Lockdowns lasted way longer in blue states, and while OWS is regarded as successful (once Trump couldn't take full credit), vaccine mandates were hilariously unpopular and tied strictly to the Democrats.
For me I think it is useful to zoom out a bit and ask: What do we do with overwhelming evidence when it points to something we don't want to be true?
The level of evidence here is at a similar level to "Men are significantly better at most sports than women," which is a fact that I don't want to be true but none the less is.
It's painful to see the mental gymnastics that people go through to deny the sports differences. It was similarly painful to read the comments section of a rationalist blog twisting and turning into anecdote and conspiracy over the simple fact that Covid killed quite a lot of people.
Perhaps this tends to happen when a fact comes into conflict with a core moral belief. In the sports case this might be the moral principle of "equality and the evils of discrimination." In the Covid case it might be "individualism and the evils of forced conformity."
I was also surprised this readership was so misinformed about COVID deaths, disappointing, and happy that Scott addressed it
What we do is bet against the idiots and take their money until they can't afford to bet anymore, then shame anyone who continues to BS while unwilling to bet.
How do you propose betting on events in the past?
That's harder to do, so you just need to have bet years ago. Bet against idiots on more things in the future.
Somebody managed to set up a refereed COVID origins debate with a $100,000 prize and was able to collect on it. That could be done for similar arguments.
Yeah, I was asking whether he was proposing something like that. Because in that case, you'd be betting on the outcome of a debate, an entirely different thing from what is true.
I think it went great for getting the best arguments out of the lab-leak and wet market crowds. The lab-leakers' pivot from professed confidence in the format to complaints that their representative only lost from being bad at presentation settled any doubts I had about COVID origins.
What's the problem with men being better at most sports? Men are bigger and have an easier time building muscle mass. Being significantly better at physical sports should totally be expected.
The one "oops" is chess where it's possible the brain size plays some role and it would be a hard pill to swallow. But I think fewer women finding motivation (whether because they're less competitive or because chess in particular aligns with their natural or socially conditioned inclinations worse) might be a good enough explanation in itself?
I deeply appreciate all your work in keeping these discussions open, sensible, and sensitive. This sort of analysis of what happened, how did we respond, why, and how to be better oriented in the future, is important. The commenters are appreciated and invited to respond further. We all learn and grow in this manner. Thanks.
Excellent coverage of the subject but as seen by the comments there are still some people that you will never convince with data because as they bathe in the soothing sounds of right wing talk radion and Newsmax, their "feelings" about Covid both being a hoax and a real virus that was engineered by the Chinese to kill Americans will outweigh any real data that can be presented.
As I'm deep into our $1,000,000 debate on vaccine efficacy, all this is fresh in memory. So here are the quick answers, using 'trust me bro' as a source, with the full documents to be released after the debate is done.
1. Covid caused around 1.2M US deaths
2. Covid deaths were under-reported, not over-reported. 'Deaths with covid' is a myth. There are around 10% where covid was a major cause of death but wasn't reported as such.
3. There were many frail people who died of covid. 10-20% would have died within a year or so.
4. There are other causes of death that increased due to the pandemic, especially overdoses, totaling 100-200,000.
5. Vaccines saved almost 1,000,000 in the US, and caused near zero deaths.
1. believable
2. Even Scott's graphs doesn't show this. Maybe if you think that lockdowns saved some people who then died from covid. Not impossible but unlikely.
3. Most of them were frail.
4. very likely
5. Sounds right. 0.5% projected death rate -> 0.3% died from covid, 0.2% saved by vaccine. Less than it was worth. What did we spend on vaccine development and manufacturing? Probably many billions. Saving a life for, let's say $5000/QALY is within normal expenditure limits but for a vaccine that is quite a poor result.
2. we'll show a full analysis of it. but you can see it in cdc data with non-covid deaths rising with waves.
3. depends on the definition. it's ~10 years lost per death.
5. that's from charities in the third world. in the US it's ~$100,000
2. CDC data can be quite rubbish. GIGO. Remember, this organization recommended covid jabs to kids until very recently.
3. I mentioned somewhere else, 10 years are not believable at all. Apparently you have too optimistic assumptions.
5. I mean, in the US each saved life was done at the expense of $5000/QALY. Obviously you are ready to spend up to $100,000 but overall we don't spend that for every person. If vaccine costs are $5000/QALY saved that is very bad.
2. there's a difference between bad decisions and raw data manipulation. do you have any evidence of cdc ever manipulating raw data?
3. we didn't research it heavily as it is not central to our thesis. we just used the consensus from studies. would be interesting if you can substantiate your claim that is far lower.
5. note there's probably 10x more benefit from preventing non-lethal damage and related costs.
2. Raw data does not exist for that. Even in ideal conditions there is a big uncertainty about how much covid could contribute in each death.
3. I already showed that it is mathematical impossibility. To overturn it, a lot of our understanding about pandemic needs to change. Not impossible but unlikely. If you are taking studies at their face value and not being experts in reading studies, you arrive at wrong conclusions similarly to “masks work” etc.
5. Actually it only increases the cost, since all those people who survived will require a lot of healthcare. I am not saying it is a bad thing, just that it is unusually costly compared what we generally expect that vaccines are cheap interventions to save lives as compared to costly treatments when vaccines are not available.
2. that is irrelevant to what i wrote.
3. can you refer me to the mathematical proof?
On the "knowing someone who died of COVID" point, I suspect the distribution of people who died (and who know someone who died) is actually much more similar to the categories in footnote 3 than to MS. As acknowledged in footnote 4, if you have certain characteristics--you're elderly, or you worked in a nursing home or hospital, etc.--you may know loads of people who died of COVID, like a police officer knows loads of police officers or a Rhode Islander knows loads of Rhode Islanders. The corollary of course is that more people without those characteristics, like relatively young people who work in tech and read blogs, would be less likely to know someone who died of COVID than they would be to know someone with MS, or with some other hypothetical randomly-distributed 1/300 trait. So I don't think COVID deaths feeling thin on the ground should actually surprise too many people, if they really consider how their own characteristics map onto those that would make one more or less likely to know COVID victims.
I unironically really appreciate Scott doing the work to persuasively argue the truth of something very obviously true to better inform his commenters, though am kind of bummed that there were enough commenters skeptical about this that he felt it was necessary.
Well, RN here. We had lots and lots of people die during that period. Die horribly, gasping for breath. I work in a 26-bed ICU. We expect a few people to die each week. We don't expect a few people to die each day, for 2 years running. We put people on ventilators, one or two a day. We don't expect to have 20 of 26 beds with ventilated patients.
We put people on ventilators only if there is no other choice, if they would die otherwise, because they need AIR. Not because some lab test says they have covid. Not because some wacky doctor wants to try out some new treatment. It was ventilate or die, after everything else had been tried.
Why is there no age-standardized excess mortality in Sweden? Why do so many African countries show no excess mortality?
If you are not considering disruption, panic and fear as significant drivers of excess mortality, you aren't aware of the power psychosomatic illness. Consider this study from 1980:
https://www.sciencedirect.com/science/article/abs/pii/0022399980900379
"Fifty-two volunteers were given experimental colds by nasal inoculation with rhinoviruses during the course of a 10-day residential stay at the Common Cold Research Unit, Salisbury. Prior to their inoculation with virus, they were assessed on five different measures of recent life stress and they also completed the E.P.I. Clear evidence of a psychosomatic component in colds was obtained. Introverts developed significantly worse symptoms and infections than Extraverts. And life events, when they involved change in the person's general level of activity, were significantly related to magnitude of infection."
Now consider that we, essentially, locked people in their houses and told them that the worst virus in 100 years was circulating.
The studies on stress and cancer survival are ubiquitous.
It hard to overstate just how disruptive COVID mitigations were to literally every person in the developed world. If you factor in delayed/missed health care, profound economic disruption and an incredible stress/terror burden, it's not that hard to get to a 5% excess mortality.
African countries have much younger populations.
Barry Marshall won a Nobel for disproving stress as a cause of ulcers (it's a pathogen), and says there's a Nobel for disproving every other ailment alleged to be cause by stress. https://entitledtoanopinion.wordpress.com/2010/09/18/barry-marshall-says-stress-still-not-the-culprit/
This.
It is such a simple thing. Anybody who remembers that risk of death form covid was exponentially increased by age would intuitively understand.
As I remember, studies about stress and cancer were inconclusive. Probably a very little effect if at all.
However, psychosomatic illness is not stress but entirely different thing.
So prolonged stress does not make immune system work worse?
What is the evidence that it does?
I didn't research the topic but it's a claim I've heard fairly often. Just checked, ChatGPT confirms it, too. Is it the standard of discussion here to do more serious research before asking a question?
It's a common belief, but not because of evidence behind it (hence Barry Marshall's skepticism). Similarly, many people concluded that placebo effects have a big effect on health without first checking whether regression to the mean could explain those outcomes instead.
> then total all-cause mortality during the COVID pandemic would be the same as always
No, the correct comparison would be to the deaths if governments acted as they did and instituted lockdowns and other "pandemic measures" even though the virus isn't particularly harmful.
As others have noted: people were afraid to go to the hospital for a while. I know family members who opted out of seeking care during the covid era.
Also: I don't claim to have any special knowledge but shouldn't you also look at the trailing excess death numbers AFTER the pandemic? Something like (this is from memory! sorry if it's not entirely correct!) 97.5% of covid deaths were people over 40 and 93% wfere over 50. And about 80% were over 65, I think. My expectation would be that when you look at the data, you'd find that covid largely accelerated the deaths of a cohort of people who were older and already in bad health and were going to die in the next few years anyway. Now that's still a tragedy- no doubt about it. But the headline of "COVID KILLED 1.2 MILLION PEOPLE!!!!" seems a little bit misleading, to me, when you consider that the vast majority of those people probably would be dead by now, in 2025, anyway.
So I'd be interested in knowing if the high excess deaths during the covid era were followed by a lack of deaths afterward for a period of time, when all the people who would have died from other causes, but had already died, aren't... well, dying.
Excess mortality is the number of deaths above (or below) the number predicted by a model. At a minimum the model includes the demographics of the population. This means that if, for example, an extra 20% of all those over 80 die, then fewer of that cohort are available to die later. If the mortality rate for the now slightly older cohort return to the baseline then their contribution to excess mortality should return to zero, not become negative.
There is no reason to expect the integral of excess mortality over time to tend to zero, but that seems to be widely assumed in the comments here.
Has the number of excess deaths become negative the last two years? If there's a million people who would have died in the next 10 years but instead died with/of COVID during the pandemic, isn't that 100 000 fewer deaths (on average) each year? Or maybe that should be 50 000, if the average remaining life expectancy was 10 years so half would live longer than 10 years. And perhaps it's even lower since more of the deaths will be concentrated toward the end. Regardless, wouldn't we expect fewer than expected since the weaker were killed off?
Aren’t you ignoring foregone medical care/other lockdown related death? Is there a good estimate of these deaths anywhere?
For those who are reluctant to blame covid, the first thing to think of is whether there were excess deaths compared to trend or not. If there were then its really not important whether covid was the primary reason or not.
All those excessive deaths would not have happened without the pandemic and our response.
This sounds deranged. Of course it matters whether the people who died were killed by the virus or by the "experts" entrusted with immense power to combat it, if for no other reason than to decide what to do the NEXT time a novel pathogen is discovered.
What is funny is that people find all arguments to minimize COVID deaths : all men must die anyway, those who died where mostly old and/or already sick, it's not the virus but the treatment, the hospital, the mask, the ventilator, the vaccine, the lockdown, the immune response, whatever, anything but not COVID.
However the same people are less likely to contest statistics concerning cancer, heart attack or any other cause of death.
Typical double standard.
I'm really glad the quoted commenter in III didn't have the personal experience of someone they know closely dying from COVID - but I knew at least four, including my father and two grandparents. So, I acknowledge that I come to this specific topic with a fair bit of emotion attached.
Given it's a communicable disease, and that there was a stratified response in terms of precautions, vaccine adoption, etc, it would make sense to see patterns where some people know one person, if any, and others know countless others. If you're primarily participating in conscientious, wealthy, and educated circles, I think you'd expect to know fewer dead than those who aren't.
Regardless of the merits of the ongoing debates on lockdowns, school closures, etc (which I think are 100% worth having), I do think the attitude of "well, I didn't know anyone who died of it" shows a certain incuriosity and spiritual sclerosis.
I agree with the numbers strongly indicating the excess mortality is from COVID and not from other sources.
Assuming that is true, and that around 80% of COVID deaths were among the elderly (65+) and especially concentrated among those with comorbidities, we should expect to see a significant drop in excess deaths starting around 2022 when those people would have been dying otherwise. Instead what we see is that although the numbers go down from the 2020/2021 peak, it's actually still significantly higher than the 2019 baseline trend. So why did people keep dying despite many of the most likely to die in 2023-2024 having died early?
That I don't understand, and would love some thoughts about.
I don't know the answer, but there's a fundamental vulnerability to this analysis that I haven't seen addressed so far -- while "total deaths" is a pretty clearcut and reliable number, *excess deaths* is not so much; it relies on one's chosen baseline for expected deaths.
While I expect the people who calculate excess deaths are clever folks and take such things into account, some people on here seem to want to use a linear trend for YOY death rates -- which given the combination of life expectancy vs. age being non-linear with the well known spike in birth rates starting about 75 years prior to 2020 seems unreasonable even in normal times.
I'm glad I read far enough to see someone ask this, as it was going to be my comment. I'm not disputing that the excess deaths appear in Scott's graphs, but at some point they either (a) have to be reflected in a pull-forward effect and thus show some reduction later, or (b) have to be caused by something other than Covid, or (c) Covid is permanently increasing the baseline death rate by 5-10% (which it isn't, that would be hundreds of thousands of deaths per year and I don't think any side of the debate is claiming that many). At some point, one has to ask why the excess deaths are continuing when the virus has mostly burnt out. That could include things that EA / rationalist types usually care about, like the economic fallout of lockdown induced recession, stimulus induced inflation, and reduced educational opportunity.
A non-NPI thought: 10-20 million undocumented people likely arrived in these same years, some estimate even more. Could it be a modification of the denominator, that we're measuring deaths for circa 330 million when it should be closer to 350-360 million? Or to put it another way, extras added in? If the average illegal immigrant is 31 and mostly male (which Google suggests) then the death rate is about 250 per 100,000 for that demographic - but might be higher for this subset, since they're mostly from the Third World and work more dangerous jobs. If the mortality rate was, say, 400 per 100,000 and there are 25 million, that would be an extra 100,000 deaths per year. I don't know if that would quite match the increase in baseline, but maybe in conjunction with the other economic effects it could explain the increased death trend line.
While I don't think it contributes that much, covid or its damage persist in an infected organism sometimes. So it might increase mortality for much longer than it takes to clear respiratory symptoms.
I'm not interested in disputing that this many people died from COVID, but I think you should engage the distinction between what is seen and what is unseen. If you're going to say the COVID interventions were warranted because 1.2 million dead from COVID, then you need to make some accounting of the number who died because of the COVID interventions. You've got ~163,000 non-COVID excess deaths during the pandemic itself, and then there's going to be a long tail afterwards as cancers that went undetected and untreated during the pandemic make themselves known and take their toll, liver damage accumulates in people who turned into alcoholics because they didn't have anything to do but sit at home and drink, etc. You don't seem to be taking any of that into account at all.
I agree with everything Scott has written, but the original point (why aren't we talking about the 1.2M deaths?) seems...consistent with society's other choices? After all, the Spanish flu killed more people than WWI, possibly more than WWII, and hit younger people harder, yet there is an order of magnitude more discussion, media, memorials, etc. for the wars than for the flu.
The point is denying that COVID was really a health issue rather than a government conspiracy to make wear masks so we'd be more compliant or something. Evidence doesn't change minds on this.
Don't you sort of answer your own question here? 1 in 300 people having multiple sclerosis is really bad but most people don't go around thinking about it because they likely don't even have a close relative or friend that has multiple sclerosis and may not even have any acquaintance with multiple sclerosis.
1 in 300 people dying from COVID is really bad but most people don't go around thinking about it because they likely don't even have a close relative or friend that died from it and may not even have an acquaintance that died from it. Further, to the extent they know anybody that died from COVID, it is overwhelmingly likely that they were really old or had some very serious comorbidities, and often times both.
In contrast, basically everybody had their life negatively impacted by the reaction to COVID. Even if they liked some of the results (work from home; extra time with family; slower pace of life, etc), overall it was likely negative, especially financially if they did not get any PPP or ERC money.
There’s something that seems a bit contrived about the from Covid/with Covid distinction. I lost my aunt in 2020, my father in 2021, and my uncle earlier this year. They all died in the hospital, none from Covid, but cause of death obviously wasn’t a clear-cut checked box on a piece of paper. Once s*** started slipping, keeping them alive was a gauntlet of tradeoffs – stopping treatment of one thing to fight off some other thing. Especially my aunt went from seemingly quite healthy, going in for relatively low-risk surgery, to dying surprisingly fast. No one thought it was her time yet. Would they have been able to save her life, and give her another 10 years, if Covid hadn’t been part of the mix, messing up her recovery? I don’t know. But all this is just to say that excess deaths seems like a much better way of tracking this than what anyone reported on some form in a million different cases.
"This number - 1/300 - is also the prevalence of multiple sclerosis3. Do you know someone with multiple sclerosis? No? Then it’s not surprising that you also don’t know someone who died of COVID"
I have no particular reason to know that my paralegal's sister has MS. I would definitely know if my paralegal's sister just died.
Would be interesting to compare US covid death rates with that of other countries, particularly those with even better record keeping than the US to see if the US is an outlier. Then figure out why. Then figure out what interventions and restrictions actually made a difference. This reconning needs to be done. It is OK to panic initially, but then cooler heads and evidence based practices should take over collectively. The tribal hostility that became the public manifestation of the pandemic in the US was a disgrace, and people wanting to move on may want to do so to avoid facing that part of themselves.
There is a lot of data available on that, e.g. on worldometer, and analysis on excess death and other things on our world in data. USA seem to be on the high end of covid deaths (0,36%), but not really an outlier.
Thanks for the input. Compared (on Worldometer) to the countries I would expect to have reliable tracking, like Canada, Europe: Scandinavia, Netherlands, Germany, Austria, Portugal, Spain, France, most countries seem per my math to have significantly lower rates, UK, Poland and Italy (which was so hard hit) are closer but lower. Didn't on that limited search find any comparable country reporting higher rates. So definitely "on the high side", as you say, which is a less dramatic way of putting it.
I hear a lot of people say they don't know anybody who died from covid, meanwhile two people in my family died from it within a month or so of each other (these are people who i've met later in my life and don't even know i lost people to covid).
In my case, my grandmother was sick (she had tested negative once but a day later was still feeling bad), when my mom went over to her home to bring her some things to help her feel better, my grandmother had collapsed in her bathroom after using the toilet (i think my mom might have had fecal matter exposure which could be a very very high viral load, or so I have heard something about that). That was how my mom was infected as well. My grandma was dead about 10 days later, and my mom about a month and a half after that. In my mom's case, she was on a ventilator for a bit, and when she didn't improve, and then worsened, put on ECMO for about a month. After worsening, stabilizing and then improving, they took her off ECMO but after the ECMO was removed she just kind of... never came back. She was gone, so we pulled life support a couple days later.
Anyway, I didn't post to share too many details, but I wonder if this is perhaps an explanatory reason many people don't know someone who died from covid. That is, the nature of how it spreads and our lives at the time means only very close family members may have caught it from one another, and some genetic/environmental weaknesses could have led to some families being devastated while others went relatively unscathed.
> also, suicides didn’t rise during the strictest part of lockdown, and may even have paradoxically fallen.
School being closed substantially reduces youth suicide.
https://jhr.uwpress.org/content/59/S/S227
> Leveraging county variation in the timing of reopening, we find that returning to in-person instruction increased youth suicides by 12–18 percent. Analysis of Google search data suggests that bullying is a likely mechanism.
> School being closed substantially reduces youth suicide.
That's not such a big number to begin with.
https://upload.wikimedia.org/wikipedia/commons/e/e4/US_Suicide_Rate_by_Age.svg
1 in 8 suicide deaths are peopled aged 15 to 24, how many were going to school? Even if it's 10%, that's only about 1.5% drop of suicide rates explained.
Ok, so how much did overall suicides change? If it's 1.5% out of 5%, that's a pretty big portion.
https://www.cdc.gov/suicide/facts/data.html
Between 2015 and 2019 the average rate was ~13.8 per 100k per year, from 2018 to 2023 excluding 2020 it was ~14.1. 2020 was the only year with noticeably lower rates, 13.5. So we need to explain the difference between 0.3 and 0.6 out of ~14 in 3/4 of a year (from Jan to March 2020 there was nothing unusual). That's 2.9-5.7%. Indeed, 1.5% is a big chunk of that.
Thanks for checking!
>Is it possible that the CDC and National Center For Health Statistics are lying about all-cause mortality patterns? Seems unlikely, because individual states reporting separately found similar patterns, and so did the other countries that reported data. This would take a truly global conspiracy.
I *don't* think these numbers were misreported or that any of your conclusions are wrong, but just as a general note:
It could also happen if all of those institutes had similarly strong incentives to lie in the same direction.
Lots of toddlers point at the dog when a parent demands to know who broke their vase. but this isn't an international conspiracy to libel dogs, it's just lots of people reacting in the obvious way to similar incentives in similar situations.
Heavily correlated actions can look like coordination if you suspect coordination going in, or don't think about the parallel incentives producing the behaviors.
Every single personal point of data that I encountered ran counter to the public narrative.
1. My son was born in the Autumn of 2020. Due to some complications, we had to stay in the hospital for three days. It was an absolute ghost town, with virtually no cars in the parking lot and no one walking around the hospital. I was watching the news on the TV about hospital overflows while sitting in a hospital in a major metro that appeared to be completely empty.
2. I have 2 close family members that work in hospice care. They both indicated that every single person who tested positive for covid in their facilities was marked as a covid death, even ones that entered hospice with stage 4 cancer.
3. I knew two people who died of Covid: one was morbidly obese (400+ pounds), and one suffered from Alzheimer's and was in his 90's.
It's not so much that I doubt the data on excess mortality, I just think that the counts don't tell the full story. The hospital system stress never materialized, the masks were a non factor, and the deaths were generally in very sick or elderly people. If we had the data to convert "excess deaths" to "missing life years" by collecting this data on an individual level, the whole thing would look much less important.
This post does a good job of reminding me why 2020-2022 did more damage my belief in humanity's ability to think and act rationally than any other period in my life. The people who continue to obsess over conspiracies surrounding the vaccine, the number of deaths, the "lockdowns" are simply so stupid I don't know how they are able to draw breath.
I've never analyzed the covid conspiracies so deeply before, but now that you point it out, it's absolutely hilarious that two mutually exclusive conspiracies have such a high rate of correlation in the population.
1. Covid was started by a group of Chinese scientists trying to engineer a deadly virus, potentially intentionally to disrupt the world economy.
2. There's a secret plot by the governments of the world to lie about the damage of covid when in reality almost nobody died.
The probability of either one of these being true on its own is low. The probability of them both being true feels nearly impossible given that they are contradictory.
You do realize that the people who were paid extra to report COVID deaths were also paid extra to report COVID cases?
Also that if the original burst, prior to the lockdown, was COVID, this does not mean that the drastic reaction to it did not, in itself, cause excess deaths.
How did they know to make the fake covid case numbers line up with the real actual death numbers?
What on earth are you asking here?
Are you claiming that deliberately declaring deaths and cases are COVID when they are not would somehow make the numbers not match?
Yes, did you look at the graphs? Do you think there was some other factor that caused monthly death numbers to swing double-digit percentages in the exact same pattern in every Western jurisdiction on the planet over the course of two years?
Given that the data that it was COVID was falsified, why we will never know
Really? You think there could be some other unspecified thing which caused millions of excess deaths worldwide that somehow evaded the notice of every government and respectable intelectual organization on the planet?
Now then, if it escaped their notice, why would they falsify data to cover it up?
I know we're discussing this elsewhere, but, any chance you have a link for this claim? Your argument rests pretty heavily on it, so I think it's fair to show us the strength of this claim: who was paid, how much, by whom, when? All that good stuff.
Will you admit that it refutes your case if I do?
Because your reasoning skills displayed do not fill me with confidence that you are asking in good faith.
If you can find me
1. Government programs in at least 3 countries
2. With a source that provides an order of magnitude estimate for the total amount paid
3. And that total amount, distributed among a proportion of hospitals sufficient to account for at least 10% of reported COVID deaths in that country, accounts for more then 10% of the annual budget of those hospitals
I will concede that there is reason to believe that COVID death counts were incentivized upward sufficiently to make a material difference. How material, I reserve the right to judge depending on the actual countries and numbers.
In case you want to hear my reasoning: my case rests partly on international comparison, so I want enough countries to be sure you can't just be explaining the US. I want to make sure these payments can have materially affected COVID death counts, so I want to see the money going to places where large numbers of the COVID dead came from. And finally, I want to make sure the payment amounts are enough to incentivized real change in behaviour: if you find me $1M spent total, I don't think that's enough to buy a material change in number of COVID deaths on the scale of a country.
I also acknowledge that these criteria are just my first pass at trying to capture those concerns, and especially as I know nothing about hospital budgeting, I am willing to accept arguments that I should modify some of these criteria due to misunderstandings or mistakes; I just want to make sure whatever evidence you provide shows that paying for more COVID deaths happened widely enough that it can actually explain a large fraction of COVID deaths.
Shifting your goalposts again only underscores that you are not making the demand in good faith.
What goalposts? I asked you to provide evidence, I'm now telling you what I think counts as good evidence. I've explicitly said I'm willing to change these specific conditions so long as the evidence you provide is sufficient to show that this is widespread enough to materially affect reported COVID death rates in multiple countries.
If you really want to prove I'm arguing in bad faith, just post your evidence and if it's any good, surely everyone else will see so, and make up their minds about my good faith based on my response.
Denying that you are shifting the goalposts is not evidence toward your good faith. Or that you are not actually shifting the goalposts.
Go back to the first post. A medical professional in the comment thread will explicitly tell you that the payments corrupted the reporting
Why on earth are you demanding to know the total sum of COVID payments for instance? What could the lack of the sum do to change your opinion?
Seems to me he's actually trying to work something out with you. He's clear about his criteria, and offering to modify them if someone points out misunderstandings or mistakes. I think you are the person being unreasonable about goalposts. He asks for a link to your claim. Instead of just giving it -- and jeez, why on earth *not* give it, it's a url on the internet, not your wallet -- you demand that he commit to saying you are right if you hand over the url.
How would you like this?: I say I know of a website that offers excellent evidence that you are wrong. But I'm not going to tell you what it is unless you state in advance that unless there's nothing there but a dirty limerick you will agree that I am wrong. You don't get to point out ways you think my data is inaccurate or does not prove what I think it does. You just kneel down and kill my ass.
No, he's literally demanding something irrelevant and refusing to back up why he demands it even after I call him on it.
Why do you require a precondition for providing rudimentary evidentiary support of your claim? Shouldn't you be eager to provide such support?
Why? Given that the person demanding has clearly shown bad faith?
For one thing, because it's a public discussion and you can provide useful information not just to the person who asked it but to all the people who read your discussion without participating.
It's good to have goalposts marked before they're moved.
Hey, here's a URL proving you are wrong. I was going to insist that you agree in advance to admit I am right after you see it, but I am so confident you will admit it that I'm just gonna give you the url no strings attached: https://tinyurl.com/224cnwbh
Grow up.
Alas, Mary, I’m afraid I’ll just never be as highly evolved as you.
You can do a lot better than you are now even if that height is beyond your reach.
It doesn't also not *not* mean it. You'd have to bring vastly more data to actually debunk what seems the rather obvious proposition that a communicable disease going around the population at the time actually did kill people in large amounts. The skeptics who are denying the covid=deaths link generally don't bring in that data and just go "well, there might be an alternative explanation, no?"
No, we bring in the data that shows that the data was falsified on a large scale.
Offering bounties for COVID cases and promising that no one would ask questions is all any rational person needs to doubt the data. (New York went so far as to state that anyone found dead without an immediately obvious other cause of death was a COVID case.)
Why on earth would they have to do something that would obviously inflate the numbers if the numbers were sufficient for whatever purpose they had?
"No, we bring in the data that shows that the data was falsified on a large scale"
Except, well, you haven't brought the data, you've been too busy litigating whether I'm arguing in good faith.
There are a bunch of claims in this comment, care to provide a link for even just one of them?
Go to the last thread, the one that Scott posted this in response to. You will read a health professional telling you it happened.
It's a 1000+ comment thread, at least link the comment.
Other countries did Covid reporting too you know? The conclusions were, in general, similar - so the fake reporting you are talking about is unlikely to have had such a very large effect. There may have been underreporting as well. That's why it's good to compare to excess deaths. Maybe the true number is 1M, maybe 1,4M, but how much does that really matter?
In what significant ways are their governments different, so that they would act differently?
In a great many ways I would imagine. If you want to believe that there is a world wide conspiracy to inflate the deadliness of covid, while at the same time a lot of excess death happened for other reasons, I don't think there is much I can say to change your mind.
And if you think the openly admitted fraud about COVID deaths is -- what? Doctors accidentally diagnosing deaths as other causes when they were caused by COVID and accidentally putting down the right cause while trying to commit fraud?
There were, even at the time, many admissions that there was massive fraud.
Incidentally, we had the lockdowns. Lockdowns cause many problems, including mental. Doctors had a motive to declare suicides were COVID.
I mean that I am not sure that it's possible for someone to "know" than a specific acquaintance with cancer died "significantly earlier because he couldn't access the treatments he needed because of COVID."
It seems that the strain of virus at the time it erupted in NYC was particularly potent. I am glad that as it morphed, it weakened. People in their 40s died within two weeks. This, and not having effective treatment for those who had severe disease, is what freaked people in medical care, out. So you’re saying, hey, as long as my circle of loved ones and acquaintances didn’t die, it’s all fake? I am surprised that this comes from you. Did you volunteer to work in a NYC hospital in March 2020 until March 2021? Do you know for certain people could not die from this infection? How do you know? People often have many chronic diseases simultaneously. It’s the kicker that gets them.
"I like these less because they’re class- and location- stratified, so your chance of knowing them goes up or down a lot depending on your own characteristics."
COVID deaths were also very class and location stratified! Poor people in New York are a lot more likely to know someone who died of COVID than a rich person in Seattle or Minnesota or whatever.
There's an argument to be made on ventilators making COVID much more *lethal* in the early phases and that being part of why New York especially was hit so hard before treatment protocols evolved. I don't think there's any argument that would have shifted deaths by more than a high 4 digit/low 5 digit number however.
Yes some people died from disruption to the health care system. You can still attribute that to COVID since it was the disruptor.
And folks who ignored practices which could slow the infection rate, such as isolation and masks, added to the disruption. You could see at the time clearly different rates and deaths in countries with better behavior patterns. Or worse patterns.
I think I may have said this before, but there’s a huge difference in experience between people who, say, worked in hospitals or with elderly or as EMS workers in 2020 vs. people who experienced the pandemic from a distance.
I think that difference underlies a lot of the polarization.
As a professional debugger who deals in problems with tough unknown solutions often, I think a lot of people talking about the COVID response as though the people in charge had a lot of certainty about events / decisions made while the pandemic was unfolding.
In reality, the people guiding the pandemic response were dealing with messy logic of an unfolding pandemic that might cause an extreme amount of deaths, where best practices weren’t totally known at the time, and measuring effectiveness of lockdowns versus the damage to the economy / etc was hard to do.
In good tech companies, after the “end” of a bad outage / problem we come together and do a big blameless postmortem, where we talk about all the correct and incorrect actions we undertook during the outage, and try to come up with better practices for the next time something similar happens. This both acknowledges that we are acting with a lot of uncertainty (and are likely to make mistakes acting under uncertainty), and that we were trying really hard to fix the thing. And it provides a sense of closure.
America / the world needs a big post COVID postmortem, but this would be impossible to have the desired effect due to all the reasons in the comments here. I think the best thing we can do is make a big memorial to honor the hospital staff that took the brunt of this in stress / burnout / etc and add a national holiday of remembrance for the suffering of health / doctors / hospital workers.
TBH adding a national COVID Remembrance Day would help us all cope with the fact that it damaged everyone in some ways, and we are all upset about that and still coping with the aftermath.
This article is sort of boring because this isn't being framed right by either side. The honest question shouldn't be whether 1.2 million people died of Covid-19, it should be how many of those weren't going to die of something else in the next couple of years anyway.
The main reason we don't personally know anybody who died of Covid is because Covid cleared out the nursing homes.
A much more interesting article would be to correct the Covid death numbers by removing everyone over age 80 and everyone who already had a terminal disease, and see what's left.
You could even look at the lower-than-average death rates post-Covid as part of the analysis, which is an artifact in the numbers left over from a bunch of soon-to-die people all dying at once in 2020.
Indeed, excluding the surrounding years is a flaw here. Moreover, the excess mortality in other countries with different policies is relevant, with Sweden as the oft-discussed outlier.
What would really help is a study of co-morbidities. Assuming that the people dishonest enough to list “COVID” as a co-morbidity for an 86-year-old in hospice who coughed once before succumbing to stage 4 cancer were still honest enough to ALSO list cancer on his death certificate, we might learn something from seeing how the rates of other causes of death changed over this time period.
I don't think this is quite right either - it definitely killed a lot of <80s as well. A lot of them were more morbid than average, but nonetheless their deaths were real.
The question is how much could we have influenced the death tolls with the NPIs were implemented and did those pass basic tradeoff math. I think the answer is clearly "No" for everything other than vaccines, and would like to see the Democratic party have some sort of public acknowledgement that they were bad so we don't have to worry about going to remote school for several years the next time a pandemic rolls around.
The real tragedy is that if we get Measles-19 out of a secret lab and the lockdowns ARE warranted, nobody is going to do them because 2020 was a giant exercise in crying wolf.
well I think people will do things even absent Government intervention, because people are sorta pseudo-rational about some of this stuff. Like restaurant reservations collapsed in NYC before any lockdowns because people were like "maybe I'll wait a few weeks to see what happens with this whole Wuhan-flu thing"
If kids started dropping in droves from Measles-26 than I expect there would be pretty dramatic changes in behavior that would be largely justified.
BJ. You said: “The main reason we don't personally know anybody who died of Covid is because Covid cleared out the nursing homes.”
I do. My sister was 39. She lay in a coma for six weeks and died in April 2020 FROM COVID. She left a husband and an 8 year old adopted son. She had adopted him after his mother, her best friend, had died in a car accident a few years earlier.
My Uncle was 81, he was NOT in a nursing home and died from an aortic aneuyrism - he lay at home for five hours waiting for an ambulance because none were available, whilst I was on the phone with him - he was over 400 miles from me and I could not get to him, in May 2020. He had tested positive for Covid after his death. He had been a merchant seaman and I made a donation to the RNLI towards a new lifeboat in his memory.
My mother died in August 2020 from cancer that had not been spotted because she was unable to get to a GP because of Covid shutdown. She was 78. She was NOT in a care home. She had only been diagnosed in May 2020.
Please be careful of your comments on this issue. None of my family were ‘chronically sick’ prior to going into lockdown. Their deaths were not marked with funerals. I was unable to say goodbye to any of them. They were just gone. It has been five years and it still hurts that they died alone, in severe pain, without the help they should have had.
We do know the people who died. We remember them, even if the country doesn’t.
People who post in this and similar communities often presume that everyone involved in the discussion is thinking in a stochastic, Bayesian way where there are statistical distributions and outliers buried in all statements.
I apologize for not clarifying that "we" did not mean "every single person in the country," and instead meant "we who exhibit quality X which is clearly stated in the sentence, and who vastly but not universally have the same statistical experience."
There is a constant tension between lingual sensitivity and lingual brevity.
Don’t be so cruel. Turning this around on my ‘ignorance’?! Stop it. These were human beings not some philosophical experiment. You can’t even bring yourself to apologise properly, can you? I hope you never have to suffer what we have endured. You couldn’t take it, you really couldn’t.
Half an year ago, my mother died. At the point where it was confirmed that this was going to happen for sure it was obvious that she'd only have a few days to week until she was gone. I would still have been angry as shit if someone had broken in and killed her during that time.
If this were a major effect, wouldn’t we see a below-previous-baseline death rate for the years after the pandemic? Have we?
The Brooklyn neighborhood and Jewish community I live in was one of the first NYC areas to be heavily affected by COVID. We have a community website that shares obituaries when someone passes, and pre COVID, we averaged 1-2 a week. During COVID, we had dozens. My grandfather survived a few bouts of COVID, but literally every single one of his weekly study partners died. Until he passed earlier this year, all his new study partners were young men because almost all the men of his generation were dead.
We have a local volunteer EMS service that was doing dozens of hospital transportations a day, and the Jewish cemetery was doing only one car allowed in at a time burial services, with burials going from sunrise to sunset wall to wall. Multiple new sections of the cemetery were inaugurated in a six month span.
So in short, the 1.2 million number makes total sense to me.
I think the argument that "you can't trust the experts" is not equivalent to "Covid is not a thing." Pretty clearly, something happened. The issue with trust is that there was never a mature discussion about what occurred at any acceptable level of seriousness, which lead to distrust.
That's why we can't point to hard numbers on Covid deaths and, when challenged, people point to excess deaths. I agree that something clearly happened in 2021 and it's pretty obvious it was Covid but technically we're not even establishing causality here. We just see a spike in deaths, that's where the level of discourse is.
And allow me to clarify this for some of the medical professionals here, because Covid has become very partisan. Let's imagine an ideal "Jane Doe" patient who might be misclassified as a Covid death. Our imaginary Jane Doe is a 79 year old female with a BMI of 38.1, diabetes, and COPD, who passed away in September of 2020 with Covid as the listed cause of death. I think everyone can agree it's fair to say that this patient's cause of death is...less clear than we might like and also not unrepresentative of the average person who passed from Covid.
And, to the best of my knowledge, these kinds of cases were never identified or investigated, certainly not in April-June of 2020, and overwhelmingly for political reasons. Because there are, to a rounding error, no technical challenges that would prevent this information from being gathered and summarized.
Take a trivial example. Say we wanted to know what % of patients who passed away from Covid at Kaiser Permanente facilities in 2020-2021 had 2+ high risk diagnosis, like diabetes or chronic lower respiratory disease. Well, I guarantee Kaiser has a SQL database somewhere in the backend to manage all the data and I guarantee there either a PatientDeath table or a PatientEncounter table listing every patient who died and there's also got to be a PatientDiagnosis table listing every active diagnosis for every patient, or at least a historical Diagnosis table with start and end dates.
And presuming your hospital has the most minimal technical infrastructure imaginable, the query to get this data would rhyme with:
SELECT
pd.UniquePatientIdentifier
, COUNT(DISTINCT(apd.DiagnosisName)) AS NumberOfHighRiskDiagnosis
FROM PatientDeath pd
LEFT JOIN AllPatientDiagnsosis apd
ON pd.UniquePatientIdentifier=apd.UniquePatientIdentifier
AND apd.HighRiskDiagnosisFlag=1
WHERE pd.DateOfDeath BETWEEN "2020-03-01" AND "2023-01-01"
GROUP BY pd.UniquePatientIdentifier
HAVING COUNT(DISTINCT(apd.DiagnosisName))>=2
We're talking about 10 lines of code. To be fair, not 10 lines of code I would expect a casual user to come up with but even the most junior BID should be able to knock this out in a few hours. It is simply not credible that no one at Kaiser, or any other major health system or hospital, was unable to provide those numbers for technical reasons. No one is credibly that incompetent.
This information was not provided, to the best of my knowledge, to either the public or to medical experts for obviously political reasons driven at the local level by hospital administration.
And, again, if you're a medical professional, can you tell me what % of Covid patients had 2 or more diagnosis from the top 10 causes of death (excluding Covid) at your hospital from 2020-2022? If not, why do you pretend that this is an acceptable level of discourse? Why do you pretend that Covid was managed and investigated to an appropriate standard? Why do you trust reports generated by a medical administration that has never treated you terribly well?
Because that's why we're all mucking around with excess deaths and other garbage metrics; because an intentional decision was made by hospital administrators under political pressure to not provide clearly relevant data, regardless of its impact on public discourse or, more importantly, patient outcomes.
(1) https://www.healthline.com/health/leading-causes-of-death
100% Exactly my feelings on the matter. It seems like the individual level data that could clear this up must exist, so I find it incredibly frustrating that we have to talk about things like "excess deaths".
The truth is that legal and privacy requirements mean that the table you reference probably doesn't actually exist. I run into this problem all of the time in adtech: easy access to information to answer useful questions is a legal liability, so information is impossible to access (or simply not tabulated) unless it done in the context of a product that is making the company more money than the legal liability of exposing the information.
This is well stated. The difficulty in believing the official numbers is not solely a function of a priori distrust in the relevant institutions. It is that the relevant institutions deliberately set up the data collection system to make this information opaque and difficult to assess. Skeptics are naturally going to see deliberate choices by CDC et al. which actively made it harder to get reliable information as evidence that there's something nefarious going on.
Let's suppose we had all this data for every single death, would it let us come up with a better estimate than the excess death data? How would you write that SQL query?
Obviously we can't just say that anyone with any other health conditions apart from covid doesn't count as a covid death. Maybe we could do some maths with the actuarial tables for 79 year old female diabetics with COPD in 2019 vs 2020 and then decide to attribute Jane's death as 30% of a covid death?
But I suspect that if you did all this maths then you'd just wind up getting your excess death numbers back again.
1. COVID killed lots of people and was quite bad
2. Pretty much all non-vaccine policy choices had low efficacy and should not have been mandated (people would have chosen to change behavior some on their own)
3. Vaccines should have been sped up even faster
4. Vaccine should never have been mandated
People end up in this weird spot where they think either:
NPIs were unjustified, so COVID was not actually bad
Or
COVID was really bad, so NPIs should have been even harder!
Whereas the actual answer is sometimes you just have an illness pop up and there's only so much you can do. Focus on origins, vaccines, and everything else is roughly de minimis (especially on a QALY impact basis). None of this was controversial pre-2020 - the types of NPIs we rolled out had basically no scholarly basis in pandemic literature as responses to flu pandemics. Go back and read the Pandemic response papers by the guy who eliminated smallpox (DA Henderson). Suggesting closing schools for years would have been anathema.
This study demonstrates that people mostly mix with people their own age, and that older people have fewer contacts than younger people, in general. https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.0050074
Excess mortality during Covid was concentrated in older people, if your circle includes fewer of those, then you're less likely to know someone who dies of Covid. People in their 30s and 40s were likely to have already lost their contacts among their grandparents' generation, while their parents are not yet old enough to be at high risk. I'm almost 40, but my parents have older siblings; two of them (late 70s/early 80s) died during Covid. Among my social circle I probably have 20-30 people over 75, mostly family members, family friends and a smaller number of parents of friends that I'm closer to.
I would be interested in a study on how different age groups form social connections with one another, particularly outside of blood connections.
All of this Monday morning quarterbacking about COVID, its source, and arguing about how many folks died from it is absurd in the extreme.
Nature has been in the deadly viral and bacterial pathogen competition game with all forms of life on planet earth for billions of years. Mankind has only been around for maybe a few hundred thousand years, and written “history” for maybe a couple of thousand years. All religions are part of mankind’s search for meaning, understanding and finding a way to get along without killing each other. But, no religion or political belief has any linkage to nature’s pathogen game, and never has. Only mankind’s hubris allows it to think it does.
Religious zealots and politicians are mere observers and actually help the pathogens spread by denying their existence and impacts.
Pandemic viruses and bacteria are not political or religious. They don’t care about religion or politics, either way, but will exploit weaknesses wherever it finds them, to mankind’s and all life-form’s detriment, and occasionally to their benefit.
Thus far, nothing can change that tension, but vaccines can slow them down.
The Covid facts are: 1, It did exist; 2, It was deadly; 3, Millions died from it worldwide; 4, Approximately 1-1.2 millions of Americans died from it; and, 5, The various vaccines slowed it down and saved millions of lives.
Period.
So, please stop with the yabut’s, woulda shoulda coulda’s, and accounting nonsense and face the above facts.
Ditto about the Covid vaccines. They worked, otherwise millions more would have died.
Just stop with all of this political and religion driven naysaying nonsense. It’s irrelevant and deadly. That includes RFKJ and his pack of pseudo scientific “advisors.”
Pumba said it best: “put your past in your behind.”
Also, "opinions are like assholes, everybody has one,"
In a sane and sensible world, there's a sane and sensible discussion to be had about how the public health response to covid went (really the hundreds of separate public health responses aroud the world) and what we can learn for next time.
Obviously the correct thing to do in hindsight would have been to lock down earlier and harder to eradicate the virus completely. But what would the second best response have been, once we'd missed the deadline for that?
It's still a pretty open question to what extent any particular NPI was worthwhile.
>Obviously the correct thing to do in hindsight would have been to lock down earlier and harder to eradicate the virus completely.
Let's hope the next novel virus originates in Scotland, then. Or Poland.
The case is a little stronger than simply tracking reported COVID deaths as indicated by cause of death and excess deaths because the rise and fall of excess deaths also tracks with a slight lag positive COVID tests and wastewater data where available. Excess deaths were rising and falling with *infection rates.* Excessive deaths were appearing where and when COVID was. The excess death numbers are consistently modestly worse than the reported deaths, which also tracks gaps in attribution you'd expect to see.
I know or know of 5 people who died of Covid: my mom's pastor and two of his congregants, and two inlaws--mother and son. All of them dedicated anti-precaution, all of them old or unhealthy or both.
I imagine there are some people just as distant from clusters of deaths as my connection to two clusters has me closer than average.
Fwiw I also know two people vaccine-injured and suspect a few more, with at least one possible death as well--but that one could have been post-infection damage too. I'm not close enough to have any idea, just that they "died unexpectedly" ie not while actively sick.
The spike is dangerous no matter how it's introduced.
So a lot of people keep saying that many people who died were over 85, but few seemed to talk about what the number actually was. So I looked up COVID deaths by age and the answer is 307,169 https://www.statista.com/statistics/1191568/reported-deaths-from-covid-by-age-us/
But what's more interesting that no one talks about is that nearly all the COVID deaths (93%!!!) are people over age 50. Less than 100k people under 50 died according to this data! This actually explains a lot of the intuitive discrepancies we keep encountering.
Also this might mean that we could have just locked down all the people over 50 and let everyone else function as normal and it would have possibly been fine?
But also shouldn't this be a bigger deal? Am I missing something?
I mean, if people over 50 all lived in a different city from people under 50 that might have been possible, but...
In 2023 the internet suggests that 36% of Americans were 18-44, and 24.6 were 44-65. Assume that those account for all "working age" Americans, then over 50s are ~40% of working Americans; locking down 40% of workers doesn't strike me as all that much less disruptive than locking down ... whatever percent were actually locked down.
Well the percentage that was locked down was technically all of them at any time where there was a lockdown, so I don't see how having less people is less disruptive than having no people.
I just mean that a workplace where half of it's employees are on lockdown presumably functions at much less than half capacity; I'm imagining some non linear phase change type behaviour where having 5% fewer employees makes you 5% less effective, but at some point you pass through a critical value and suddenly at 50 (or whatever) percent, you're now at like 20% effectiveness.
But yeah, I don't totally discount that this idea might have worked (I think it's basically the Great Barrington Declaration); I just think "lock down only the fifty year olds" isn't logistically trivial and probably still has pretty big economic downsides.
I'm not saying it's easy or that it doesn't have down sides, just that it could have had less downsides than locking down everyone. Also it's not necessarily about work, you could have still had young people do stuff outside and go places and possibly congregate, etc.
Yeah, sure, I'm not taking a hard stand against this.
Fwiw, in my city we were allowed to go outside and congregate much of the time; I held my birthday in a public park in July 2020, was doing regular bike rides with one of my friends that whole summer; and definitely was eating in restaurants again (albeit very rarely) by August 2020 so how different your scenario is from what actually happened depends on jurisdiction.
Sure but if you're arguing that knowing that it's mostly people over 50 who died from COVID wouldn't have changed much in a practical sense then that's possible, but I disagree with that, or think it's at least it's worth considering what we could have done with this information
Some people, at least in hard-hit areas such as New York, may have died because overload of the healthcare system and/or fear of coming to healthcare facilities prevented or delayed treatment of conditions that would have been cures without delay. This includes prompt treatment of primary cancer. Also prompt thrombolytic treatment of acute cardiovascular disease.
These are not direct Covid deaths, but victims of system overload are just as dead as those who died directly from Covid. In my opinion, excess deaths constitute the most meaningful indicator of the health effects of the Covid pandemic, and will remain so unless the national invests in highly redundant capacity.
This is a master class study on how approach a divisive topic without angering people but still being firm.
Great follow up post.
In continuing from yesterday, I think I would quibble with your tally along the 5-10% range, for the reasons you listed in your article. Which is to say I’d go along with an estimate of approx 1 million died “of covid”….which is still a very large number….but maybe not 1.2 million, but which doesn’t detract from the thrust of your original post.
Great point about the 1/300 ratio and the value of personal experience of actually knowing anyone who died. I fall into the category of your example: I didn’t know anyone directly, but I heard a friend of a friend did.
My personal prior is coloured by the fact that the hospital I admit to kept covid tallies based on “with covid” and not “of covid”…which is why I question some of the official numbers (at the margins).
We used to believe government statistics. Now (and especially after the last four months and DOGE), even if they tell us the truth, we won't believe it. That's what happens when you set the public trust on fire, pour gasoline over it, add a few sticks of dynamite, fan the flames, stir the ashes, and then salt the ground after for good measure.
No one trusts official spokesmen anymore. And even less official spokeswomen.
Agree 100% here, but I think we should show some sympathy to those who don't trust government statistics. Not because they are right, but because we have given them ample reason to be distrustful. The last time we had a genuine national security crisis before Covid was 9-11, and the public reacted with almost complete trust in the government to take the proper steps to fix the situation. To say they didn't deserve our trust is the understatement of the century.
Most people have absolutely no ability to make decisions on what to believe based on the facts involved, because there aren't enough indisputably facts to make a decision. It's all vibes and trust. The implementation details of vaccines, lockdowns, masks etc. doesn't matter at all, because a substantial fraction of people didn't believe anything they were being told.
Had a friend high up in the CDC during the first outbreak of West Nile. He said that stats and reporting just became completely politically captured *at that point*. Bad information went out and the doctors just fell in line like the government dupes they have turned out to be. It's disappointing. We are on our own.
> No one trusts you anymore
Who is the "you" here? Scott isn't the Government.
The Government has lied in the past, so we should throw out normal truth-seeking behaviour and just believe the opposite of whatever the Government says in the future? This doesn't seem like it's going to get us closer to the truth. The untrustworthiness of official sources is a good reason to sharpen our truth-discernment abilities, not to throw them out completely.
I meant to be addressing the mouthpieces of the government. Sorry for the confusion. I don't necessarily believe the *opposite* but the opposite is probably closer to the truth than the narrative of the previous 50 years.
Unfortunately, this is a bit more complicated.
We know that covid incidence is highly age- and co-morbidity stratified. So, we would find the excess deaths in the old and folks with such comorbidities.
Conversely, excess deaths outside these groups (especially the younger than 50) would not be prima facie evidence of covid, but of other related causes, be they lockdown related (e.g. overdose deaths, missed "elective" procedures), treatment related (ventilators), or pharmaceutical (vaccine related).
I don't have sourcing, but I would hope someone has done this. But even allowing for these effects only in theory and in plausible numbers would suggest that the true deaths FROM covid are a lot lower than 1.2 million.
Try this one: "COVID killed a million people" is neither /wrong/, nor /semantically coherent/. Let's go:
Collapse "died with-but-maybe/probably-contributed" and "died from" COVID. Gander at the mortality rates by age (I think the risk-doubles-every-seven-years-of-age is still good?), plus the mortality by comorbidity (general ill/fragile health, lung weakness by any cause). All together, we can see COVID as a major driver of death in any cohort vulnerable to exactly what you'd expect from (especially the early, more aggressive variants of) COVID's lung-inflammation pathology on top of general flu symptoms.
To abstract, summarize the above as "a particularly potent and transmissible strain of influenza ripped through the population; the expected cohort died". This recalls that /the flu/ is a major contributor-and-cause of death, in exactly the cohort that is in general and pneumatic ill health.
This gives us a bit of Science: we predict from this that all-cause mortality would spike at viral introduction, stay high until everybody's had it, and then return to normal levels. If society were to, say, grind itself into worse health with lockdowns, ever-worsening chronic illness, and generally higher population, we'd see the raw numbers level back down a little higher than previous (don't forget population still rose at least ~3% over the span, beware raw figures).
So what do we see? Deaths by millions, starting 2019: 2.85, 3.38, 3.46, 3.28, 3.09, 3.07 (provisional 2024). This all makes sense, if COVID largely (largely! again beware the states are only meaningful epidemically) contributed-or-caused death in those who were, sorry to say but this is life, already going to bite it for some reason soon enough. Pneumonia, pre-COVID flus, or whatever other big systemic shock would have gotten them next, in the COVID-free alternate universe.
Semantic coherence breaks down if you simplify this into "COVID killed a million people", because COVID didn't "kill a million people" the way an big asteroid hitting Austin would kill a million people. Thus, if you want endless talking-past-each-other internet flamewars, keeps saying "COVID killed a million people", and you'll never be at want of gnashing teeth, without anybody in the conversation "being wrong". Everyone gets to be more-or-less-uselessly "right".
Thus the semantic incoherence of using "pandemic" for COVID: the death tolls matches our expectation from an introduction of a novel influenza, which finished off lots of people in a brief span, and then faded back into the more general tableau of holy-crap-flu-is-surprisingly-deadly statistics, which are now a smidge higher than their previously puissant death toll. This is different than an ebola or bubonic outbreak, which will rip through a population pandemically, but don't reflect a previous endemic problem, or simmer down into a chronic condition.
Semantically, "ebola killed a million people" coheres with the usual notion of "X killed Y total people", like a meteor or an airplane crash. It does not cohere with "a particularly bad flu season, lasting about two years, and then returning to near-normal". YES YOU CAN STILL BE RIGHT every time you say "COVID killed a million people", but as soon that gets used to justify pandemic/lockdown policy, you're semantically incoherent to the contexts where we might, say, go into incredible debt and immiserate hundreds of millions in order to "prevent the killing of a million people", who will go on to have healthy normal life expectancy afterwards. I figure there's still quite a lot of hot neurons on that connection, prompting the hissing and spitting every time "X killed Y people" gets applied to COVID.
You mean to tell me that there is a lack of critical thinking in the US. Or a majority of our citizens whom are totally dismissive of data driven decisions. Just trust your gut, huh? What would it take to convince the deniers of the 1.2M COVID deaths? My guess is there is no possible argument that will ever change their minds.
There were also medical establishments which intentionally underreported Covid deaths. A friend's mother, who was in her 70s, was admitted to her local hospital in Alabama with Covid. She died ~8 days later. Cause of death: pneumonia. My friend asked if her mother died of Covid and was told: no one dies from Covid. My friend didn't pursue it because she doesn't care what her mother died from.
That's just one anecdote. Were the underreporting and overreporting equal? I have no idea but the data which Scott refers to indicates that ~1.2mm Covid deaths is approximately accurate. Was the actual number 1.0mm or 1.4 mm? Maybe.
I believe covid easily caused the excess deaths mentioned. That said is there any factoring of the increase in all cause mortality caused by loneliness? As most people in lockdown were much more lonely.
Teen suicide went *down* during remote instruction! Whatever effect loneliness & isolation may have had, it was swamped by the reduction of in-person bullying.
https://www.nber.org/papers/w30795
Let's just put it directly: arguments are soldiers, and there's a lot of people who are so committed to the idea that the most important thing about Covid was that restrictions were bad that they are just going to never accept that Covid was actually deadly, since someone might use that as an argument that restrictions were good. They're going to just keep throwing up an endless amount of arguments of what *might* have happened (and that "might have happened" is always implied though generally not directly claimed to be equivalent of "did happen") instead of accepting the most obvious, Occam's-razor explanation that, indeed, it was the disease that mostly caused the excess death rates.
"Yes, Covid did kill all those people and yes, all restrictions were still immoral" is a rigorous and in many ways correct position (though at this point, this implicitly also is already close to the mainstream position - Covid still goes on and excess deaths still have larger-than-normal-flu-season spikes in winter, and no-one important is indicating that even larger death spikes would bring back the restrictions), but at this point even refusing to consider the idea that Covid did indeed kill pretty much the amount of people generally claimed to have been killed by Covid is just mendacious.
Furthermore, what burns me is seeing various and conspiracy theorists basically running victory laps in the sense of "everything the "conspiracy theorists" said was true! Everything!" whenever there's any indication of the mainstream accepting previously not-so-mainstream theories, like lab leak, when the common Covid-era conspiracy theory discourse was replete with predictions that didn't turn out to be true, like the vaccines basically being poison that will kill or sterilize hundreds of millions of people, or the restrictions and NPIs going on forever and ever as a tool of NWO slavery, or Covid being used as a reason for cancelling elections or so on.
> There are 340 million Americans, so if 1.2 million died of COVID, that’s about 1/300. This number - 1/300 - is also the prevalence of multiple sclerosis3. Do you know someone with multiple sclerosis?
I feel like you buried the lead here. Isn't the above the larger point, more important than the 1.2 million number? The per capita number is the meaningful stat, not the raw aggregate number. It just turns out that lots of things produce a death toll of 1/300, and it's not such a huge deal. Sure, if you multiply 1/300 by 340 million people, you get a big number, but that doesn't and shouldn't reflect our intuitions about seriousness.
Oh sure if some disease kills a million people then it's not a big deal. But if _I_ kill a million people then I go to prison. Where's the justice in that?
The whole point is that, given a large enough base population, *lots* of things kill a million people. Scott himself noted this, giving multiple sclerosis as one example of many. Are you freaking out about multiple sclerosis?
The question is not "Is a million a big number?" It's "How does COVID-19 rank among the various risks we're familiar with?"
Cancer yearly death rate in the US is about 0,14% - so similar to that during the worst years?
Look, I'm not going to argue that people in general are particularly good at how they prioritise various different causes of death.
Certainly far more column inches have been spent on the untimely end of George Floyd than on the 72 people killed by bees or 20 people killed by cows in the US in the same year.
But a million excess deaths is a lot however you slice it.
I actually do know someone with MS.
But yes, Covid obviously did kill a ton of people. And I'm not sure why anybody finds that hard to believe. Most of us got it; as a young and healthy man who suffered through an episode, I'm not sure I would've survived the same infection as an elderly man with other health issues.
https://fred.stlouisfed.org/graph/?g=1Ex50
Wow, that does not look good. Any statistics from other countries? I checked my own country, and there is no similar trend there (raising, but continuing trend from before covid).
edit: Wait, what is this saying - there are more people who are registered with a disability or more people with a disability are able to work (due to more remote work?), seems like it's both? Is this just for St. Louis?
So I've heard skeptics cite this paper as concerning:
The extent and impact of vaccine status miscategorisation on covid vaccine efficacy studies https://www.medrxiv.org/content/medrxiv/early/2025/01/24/2024.03.09.24304015.full.pdf
I think the context in which they offered it was in traditional academic papers looking very formally at effectiveness data.
Although I've heard skeptics cite it frequently, I've not heard any mainstream (pro-vax) folks tear into it.
My interest - I suspect that if that logic/paper were appropriate, it would apply to the conversation. If people die in a ~2 week window after vaccination (or anytime before 2nd vax, or 2 week after second vax, depending on study), those deaths would be attributed to "not-vax" group status because the vaccine hadn't been assumed to be at full power yet. That would undercount vax deaths, and overcount non-vax deaths.
For the folks who think there are attributable vax deaths, this is important because it doubly-changes the numbers (reduces vax efficacy, and increases non-vax excess mortality).
Is this paper not valid? Or is the effect assumed (by pro-vax) folks so small that it doesn't move the needle (either on vax-efficacy or on overall excess mortality) so the paper is effectively ignored by one side yet used by the other as a prime support?
I remember hearing the paper eons ago, but don't recall ever hearing a refutation.
It seems to me that few people give a shit about the numbers. Most are angrily, personally invested in their familiar set of beliefs about covid and about the stupidity and character flaws of people who disagree.
"Drug overdoses rose, but by a tiny fraction of the COVID death toll." Hrm, but do we know that? I personally had someone die during the pandemic but whether it was because of covid or poor health or a prescription drug overdose was never determined. The pandemic was very hard on people and all deaths does include those deaths that may not have had the floridity of a gun to the head.
I know people with MS and I know people who died of COVID. In 2021 I attended a dinner with the administrators of a large American university. The black women all had several relatives, including close ones, who had died of COVID. All anecdotal, but made me wonder at the time if African-American deaths were actually underreported.
The CDC has a fantastic site with a ton of raw data on deaths:
https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm
Anyone questioning excess deaths directly caused by COVID needs to explain the first chart on that page.
None of the alternative proposed causes make any damn sense with the raw data.
This is the first chart that should be in the above post (as opposed to the above "excess" death chart referenced by Scott, which naturally creates questions about assumptions) - just show the raw data.
I guess one question I have that I think is the real reason people are so mad - well, one of them - about all COVID things is the sense that not only were the measures taken to combat the virus mostly useless in the end, but they exacerbated excess deaths. That's what I'm most interested in. How many of those rolling excess death rates are attributable to deaths of despair, alcoholism, suicides, etc., in timing with lockdowns or the like? Could be misreading the graphs I'm a tired dude.
At one of the Ivies (which I won't name because it will just make this post irritating) there was a debate club that kept track not only of people's wins by vote, but of their knockouts, so to speak, and also of their being "knocked out." The debate equivalent of being knocked out was a debater's admitting, on the debate floor, that he could not refute his opponent's argument, and that his opponent had in fact convinced him of his point of view. This public admission of defeat was called 'being broken.' Members who had broken opponents and had also been broken were most admired.
I think the norms that group had were admirable, and would have fostered excellent training of both minds and egos. I wish we had similar norms here.
Is there anyone who here was strongly committed to the idea that there were far fewer than 1.2 million covid deaths, and is now at least considering the possibility that they were wrong? In fact has anyone reading and posting about either or these related posts actually tried on the idea that they were wrong about *anything at all* covid-related? If so, I sure haven't seen many signs of it.
If practically nobody is changing their views in either direction? Then we are all so fucked.
As a way of modeling the capacity to be influenced, I'll disclose a way I have changed my view a bit. Various people have pointed out here that while the total deaths is a high number, the US population is a number 300 times as big. In other words, covid killed about one person in 300, which, as somebody pointed out, is about the fraction of people MS kills. Marinating in that info, I did move in the direction of thinking that in the big picture, covid was not all that lethal. In fact I'd say it was probably on the darker side of a gray zone when it comes to decisions of whether to sacrifice a lot of money, time and human well-being in an effort to mitigate its effects. (But I also think the US could have slowed it and mitigated its effects with far less collateral harms if it had approached the problem more sensibly, and been honest with the public.)
And for those who don't know: Despite all the chaos and suffering efforts to mitigate covids harms inflicted on its citizens, the US did a rat-shit job with covid. Only 16 out of 200+ countries had higher per capita death counts than ours. I am sure that some countries reporting very low death counts simply did not have the resources to keep track of covid deaths, and some are flat-out lying. But even after you take that into account by ignoring the data from those countries, we still look terrible.
https://en.wikipedia.org/wiki/COVID-19_pandemic_death_rates_by_country
MS is responsible for 0.04% of deaths globally (2017 data from wikipedia). Do you sum them over the years? But covid is not over.
Out of 200+ countries, most are nowhere near the US's league when it comes to the number of elderly people. Not saying the US did a good job but any rankings should take demography into account.
(Tangential: when it comes to maternal mortality, the US are now well behind most of Europe. And if you look at black women specifically, they're IIRC some 2.5-3 times worse off, on par with Kyrgyzstan or Mexico. So something about the US healthcare system does make it subpar for some tasks, compared to the countries of comparable income/HDI. Edit: Much of the elevated maternal mortality could be due to obesity. ChatGPT says severe obesity increases it x2-x5.)
You won't see any minds changed in the comments. People making comments are full enough of ourselves that we think other people should actually take the time to read what we're saying. The changed minds occur in lurkers. And they do happen.
If the guys in the Yale Political Union can enjoy debates in a setting where both breaking and being broken are admired, we can. Egos don’t come any bigger than the ones those guys had.
No one I know personally died of COVID. A friend of mine lost her father, she says to COVID, but she is a scientist so she has reason to lie about it. I was not with him when he died. I never met him. So how can I even know that her father died, let alone from COVID. Or that she even has a father.
My mother lives in a retirement community since 2022. The whole time she's been there, no one there died of COVID. She's still alive.
I got COVID and I was fine. I mean I am not even sure I had COVID. You can't believe those tests. I got the vaccine and over the curse of the next three years, my hair got grey, I gained 5 lbs, I stubbed my toe, I got in a fight with my husband, my alcohol tolerance went down, I had one episode of a-fib back in February (2025), I had trouble falling asleep one night in 2024, and some days I am really tired. Also my dog sometimes ignores me.
How'd I do? Do I sound like some of the other posts?
Simple question. What is the average age of people who died from COVID?
Eyeballing it somewhere around late 70s for the US https://www.statista.com/statistics/1191568/reported-deaths-from-covid-by-age-us/
Well over 80 in Australia https://www.statista.com/statistics/1245896/australia-number-of-coronavirus-deaths-by-age-group-and-gender/
Worldwide no idea
I’ve been through two pandemics, HIV and Covid. The rhetoric was identical, the conspiracies identical, the “where are the dead people” identical, the downplaying of disease spread identical. NEJM published articles by 1987/8 that HIV would be gone by 1990.
Humans are incapable of simple reasoning about anything with changing values except in linear patterns. Human eyes can track something moving but not something changing acceleration. We have no feeling for viral growth therefore it all is “fake”, made-up or false.
Compound growth whether personal debt or viral spread is not understood.
Always expect challenges to non-linear patterns and you will never be disappointed.
On the other hand AIDS was also accompanied by overly-panicky public health messaging like https://www.youtube.com/watch?v=OJ9f378T49E
And I remember right-wing talkback hosts in the early 90s complaining about how anti-discrimination laws would prevent people from setting up AIDS-free towns and that we'd all be dying as a result when the virus mutates to become airborne or something.
Basically I think epidemiology is hard, people are stupid, public health messaging is optimised for something other than truth, and the errors aren't necessarily all on one side.
>public health messaging is optimised for something other than truth
Ding ding ding! Succinct, important, underrated.
I wouldn’t know in the 90’s. My last close friend died when I lived in LA in late 1989. I decided to move on and got a place in Paris January 2000 and gradually people I met stopped dying.
I found out two decades later that I was immune (CCR5 Delta 32, homozygous; I have other strange mutations I’ve slowly realized).
The last person I knew with HIV died two years ago.
Fortunately for most people behaviors change very abruptly then slowly assume a new dynamic when a major disease hits.
Doctors don’t really understand the behavior of new viruses. They are new. Goes with the territory.
My colleague Billy Quilty had a thoughtful piece in 2020 when the UK reached the point at which most people would, on average, know someone who’d died from COVID. The key point, of course, is that affected communities aren’t evenly distributed - some were far more affected by others (and often those without the loudest voices): https://www.theguardian.com/commentisfree/2021/jan/22/silence-died-covid-figures-grief-pandemic
Thank you for participating in this (used-to-be rational) discussion Adam. If anyone knows this data and this disease, it’s you.
Orthogonal thoughts after reading some of the comments:
I think it's important to be able to separate being personally angry at something and thinking that something was unjust. I wasn't really able to visit my mother in hospital while she was dying (liver issues, not COVID-19), but I think it was generally right not to let me go there, it was the time before we had vaccines generally available. I'm deeply sad about it and I still haven't processed her death even though it's been multiple years; but I honestly also don't really know if I would have more closure if I'd been with her more often in those final days, we were on the phone every day until the end. But I was definitely upset that I couldn't see her - it felt 'unfair', but I had a lot of flexibility in deciding where to point my anger, and I pointed it at COVID-19 and the timing of it all, not at the people who were (to my perception) trying to do their best choices with limited data.
I'm still largely happy with the way my government handled things (Germany). I am pretty sure it wasn't ideal (Scott has posted other analytical articles, such as https://www.astralcodexten.com/p/lockdown-effectiveness-much-more, and generally you shouldn't listen to me about any of this anyway, I'm not even trying to convince you that Germany did a good job, which would be secondary to my point even if I believed it exactly like that, I'm just trying to convince you that it's possible to think people made wrong decisions and also not get angry at them for it), but I think honest efforts were made.
I realise some of the commenters who are saying "lockdowns were a big awful thing" are not necessarily also saying "and therefore the people who decided to do lockdowns are personally responsible in a morally reprehensible way," the claim can be much weaker, it can be "and therefore we can tell that the people who decided to do lockdowns were working with incomplete data." But I do see that this is sometimes lumped together, and I'd like to make an argument to be less angry at the people with the incomplete data trying to make decisions under pressure.
Caveat: "What about the lying?" As someone who habitually avoids the news I honestly can't say much about this (how much it was happening, etc). I especially can't say much about what was being said in the US. I got most of my information from either my biochemist girlfriend (who lives in Australia), or through three layers of indirection and refinement online (e.g. this blog). Sorry, I realise this is a point that has a lot of people really angry, but I just can't judge it at all. The closest I can come to this is that there was a ton of discussion about whether masking works and doesn't work, which through those sources looked like "things are changing as actual hard data emerges." I was super annoyed at the changes at the time (I wanted to have solid advice), but I cannot get worked up about it in hindsight.
(Maybe you're tempted to throw tweets and news articles at me to convince me that it was really bad. I'd personally advise against it, because my general attitude is "lying is indeed bad," but we'll probably get caught up in a lot of discussion about whether something is lying, strategic misrepresentation for political gain, strategic misrepresentation for damage reduction, confused misrepresentation (i.e. fumbling an explanation the person technically understands, but is simplifying in a way they're not cognisant is eliding something important), sincere misunderstanding (i.e. scientists say one thing, but you misunderstood what they say, and now you're telling other people what you (mis)understood), or some other category I'm forgetting about right now. If you think there was lying, take the short-cut and understand that I think this is pretty damn bad - but I also think you can be deeply angry at someone for lying about one thing but still be OK with how they handled other parts of the response, so I would probably continue to be OK with how people here handled other parts of the response.)
Also, as an SRE, +1 to people saying we could really use a blameless COVID-19 postmortem to learn from all the mistakes.
> I'd like to make an argument to be less angry at the people with the incomplete data trying to make decisions under pressure.
There's an associational problem where the people *informing* the decisions (the public health agencies, one assumes) get conflated with the people *finalizing* the decisions (in the US, mostly politicians), and then many of the latter going on to be absolute hypocrites- Gavin Newsom quite famously throwing a party while the rest of California was shut down and going to the beach, alone and socially distanced, might get you arrested.
Also from the US perspective, I think Fauci enjoyed being in the spotlight way too much and this contributed to that association. Among other messaging failures that had little to do with incomplete data and everything to do with incompatibly alien value sets.
> Gavin Newsom quite famously throwing a party while the rest of California was shut down
Yikes, that's definitely an exciting data point. Thanks for sharing more info on the US side of this!
Can you just admit the comments here are increasingly just conservative hackery?
> I think this disproves the claim that it was just normal deaths being reported as COVID deaths. What about the alternative claim - that it wasn’t COVID that caused the extra deaths, but various treatments - ventilators, remdesevir, vaccines?
I think that this weak claim, and the refutation thereof, masks a much stronger claim that goes entirely unaddressed in this post: what gets politely called "deaths of despair," or more bluntly, "people killed by the lockdowns."
We saw significant spikes in suicide and self-harm, in depression, and in drug use, and that's just the big obvious effects. Then there are the more insidious ones. For example, my grandmother died during Covid, not *of* Covid, but because she couldn't get into the hospital for her cancer treatments. (The hospital was not over capacity; they just didn't want to let her in because she might catch Covid!) She's by no means the only one.
So, the US suicide rate is about 0,014% per year. That is about an order of magnitude to small to account for excess deaths - and that is the total rate, the change in suicide has to be a much smaller number.
I think you will find that an analysis of increases in drug and alcohol related deaths, and reduced medical service will similarly amount to numbers that are much to small to account for the excess deaths.
I don't know many old people, but I know no-one who died of COVID, and four people with MS, so that example really has the opposite effect for me!
I forgot how fun it is to argue about covid. takes me back. thank you.
Thank you for making this post. I wanted to reply to some of the commenters the other day but it felt like it would be blowing into a hurricane.
> They’re pretty much the same, maybe margin of error of 10% or so but not consistently in any direction.
How are you getting this? Looks to me like reported Covid deaths is consistently under excess deaths for the entire time period, just timeshifted later by a few weeks in some areas.
> Do you know someone with multiple sclerosis? No? Then it’s not surprising that you also don’t know someone who died of COVID.
This seems like a bad argument. Multiple sclerosis is not necessarily visible, and people have a tendency to not publicly discuss their illnesses. It's a lot harder to hide "death" from the people you know.
Best steelman of the "excess deaths not due to COVID itself but the response would probably be something like:
1. Most died of COVID cases were died with COVID cases.
2. Excess mortality is due to lack of access/willingness to get health care, which tracked COVID waves due to correlated-in-time rises in distancing/fear or medical capacity shortages.
Even if this somehow worked out, it would probably require that the (2) involve almost entirely acute lack of care causing death (e.g. heart attacks) rather than lack of chronic care (e.g. chemo). Otherwise there would be a big lag in the excess deaths relative to the cases.
It's just so much more parsimonious to accept that a bad respiratory virus caused a lot of deaths directly.
Fwiw, I buy the death numbers. My biggest problem with the government response is the several steps along the way where they burned their credibility in response to public madness. Not a fan of Fauci. I was decent friends with a guy who died of COVID. He didn’t get the jab. He was Nigerian and just felt like they were lying because of the mask guidance. Couldn’t talk him into it.
I think Scott has inadvertently exposed the reason why the world has such trouble mourning the Covid dead and moving on. It’s because no one is sure what actually happened during the Covid years. It really destroyed faith in mainstream institutions. I hope we get some clear answers going forward.
> If people died “with” COVID, ie of normal causes like flu, cancer, or car accidents while only incidentally having a positive COVID test, then total all-cause mortality during the COVID pandemic would be the same as always. If people died “of” COVID, then total all-cause mortality during the COVID pandemic would be higher than usual.
No, that's not true. Or at best misleading.
Basically, all-cause mortality could have gone up because of the lock-downs (eg from lack of exercise or whatever).
You could say that COVID caused the lockdowns, and thus caused the excess deaths. But that seems misleading.
I think we have to work a bit harder than just looking at excess deaths to figure out how many people died of COVID.
(I don't know whether the lockdowns caused excessive deaths. This is just an example of indirect causation of deaths that we shouldn't attribute directly to COVID. Reality might differ.)
EDIT: ok, you seem to address this in the second section.
> But is this really so mysterious? There are 340 million Americans, so if 1.2 million died of COVID, that’s about 1/300. This number - 1/300 - is also the prevalence of multiple sclerosis3. Do you know someone with multiple sclerosis? No? Then it’s not surprising that you also don’t know someone who died of COVID.4
I suspect many people who died of COVID were elderly and didn't have many friends.
So there's clear selection bias. It's the usual result that most of your friends have more friends than you do.
I think it would be more useful to have data up until the present day.
I do not question that there were many more deaths due to COVID during the period, but this doesn't give us any information about how many quality life years were lost.
As everyone has to die at some point, there will be fewer deaths than there would have been otherwise in the following years.
If there is an extremely sharp drop that would indicate most of the people who died from COVID would have died from other causes in a short space of time anyway. If deaths are just slightly below where the expected trend would have been without COVID it would indicate that many of the people that died with COVID would have gone on to live considerably longer.
It is not given.
Covid disruption could have caused deterioration of health for many people which shortened their lives and dying sooner after pandemic.
In fact, after pandemic we have elevated mortality. No one knows why. Could be for various reasons.
You're saying that it's not a given that in a counterfactual universe without covid, that the people that died of/with covid in 2020-2021 wouldn't have died at some point further into the future?
Or are you saying that somehow the interventions (quarantines, masks, school and business closures etc) somehow increased the mortality rate in the subsequent years to such an extent that it cancelled these out?
I.e. Person A died of Covid in 2021. In non covid Universe they would have died in 2022, so we'd expect to see a +1 in '21 and a -1 in '22. But person B dies in '22 from "Covid disruption", who would have otherwise gone on to live a long and healthy life, so overall we see a +1 in '21 and no change in '22?
That sounds very improbable to me. Does anyone have any actual data for all cause mortality from 22 to present?
In alternative universe Trump comes to power in 2021 and not being senile yet he turns America into communist country with devastating effects on people's life expectancy (like in North Korea).
See, this type of reasoning doesn't work.
>> Or are you saying that somehow the interventions (quarantines, masks, school and business closures etc) somehow increased the mortality rate in the subsequent years to such an extent that it cancelled these out?
Yes, I am saying that it is a very likely possibility. Why would you say it is improbable? It is strange that you are asking for this data because the data shows exactly that – greatly increase mortality after covid for unclear reasons.
Most people asking those questions have never seen or analysed anything. When I tell them exactly what are the data, they just are shocked in disbelief. I tell them no evidence that masks work and they are in so much disbelief (didn't Scott provide evidence that masks work? No, silly, Cochrane report clearly shows that evidence is of poor quality and even that shows no or little effect).
Another point – healthcare services were underutilized during covid. Yes, in some places and times certain hospitals were overwhelmed. But in general it was underutilized. It could explained increased mortality after covid because these chronic cases of non-covid were not properly seen and treated.
Supporting my statement that underutilization of healthcare during covid could lead to many more non-covid related deaths is the wonderful thread today by Crémieux:
https://x.com/cremieuxrecueil/status/1910067925870154115
In short: oversupply of doctors don't improve life expectancy but undersupply make causes people to die.
If someone had told me before covid that there will be pandemic, I could believe it. If he had said that during this pandemic, healthcare services will remain underutilized, I would call him nuts. And he had said that even after pandemic people would refuse to believe that such underutilization caused more deaths, I would think I am dealing with a crazy persons.
And here I am, in the world where all these 3 statements are true.
"In alternative universe Trump comes to power in 2021 and not being senile yet he turns America into communist country with devastating effects on people's life expectancy (like in North Korea).
See, this type of reasoning doesn't work."
>> My point was that everyone has to die at some point of something.
You can argue about numbers, but it is incontrovertible that A some people did die of COVID in 2020 and B had COVID not happened those people would have died from a different cause at some point.
> Yes, I am saying that it is a very likely possibility. Why would you say it is improbable?
>> I agree there were some tragic deaths that could have been avoided - missed cancer diagnoses in young people etc.
What I think is highly improbable is that the number of these deaths was anywhere near as high as you are suggesting.
> It is strange that you are asking for this data because the data shows exactly that
>> It's strange that I'm asking for the data? Why is that strange?
Around a quarter of all seniors in the US suffer from social isolation. There was also a clustering of deaths in retirement homes, where many people are known to receive next to no visitors. This, to me, explains some of the "but I don't know anyone who died from covid!" argument, which is already a poor one from a probabilistic perspective.
( https://www.ncbi.nlm.nih.gov/books/NBK557972/ )
I actually know three people who were confirmed Covid deaths, so I guess that puts me above the trend for a 1/300 situation.
I wonder if folks just find it hard to believe that 1.2 million people could die of this thing, and society would not just move on as normal, but a huge contingent of folks would pretend it never happened.
>but a huge contingent of folks would pretend it never happened.
Is it that surprising? Public health decided to torch institutional trust. Society spent three months on semi-house-arrest before going insane. The groups that can make things stay in a news cycle were ultimately embarrassed in hindsight, and everyone else was so tired of the whole problem they'd rather just ignore it.
Lots of meat for interesting social observations by a talented writer known for interesting social observations, but no one is taking up the charge.
I know someone from rural Missouri who lost five family members between his and his wife's family – some deaths coming after the vaccine was available. All were morbidly obese, all refused to take the vaccine for political identity reasons.
My point is that the people posting on this forum might not have less contact than average with the socio-economic groups who got hit the hardest by COVID.
We also happen to live in the Loneliness Era. People with no friends and acquaintances died, too. A lot of them.
My dad's friend was the first reported death in Utah. He lived up the street from me. My friend's husband died a month before vaccines rolled out. His daughter is friends with mine. My grandma died from COVID. Granted it was in 2022, she was vaccinated, and she was in her 90s. My in-laws lost a number of friends in their church community. My high school friends lost their dad. He was so genuine and kind.
Also, funerals were not the same then. That made it even harder. People dying in isolation in the hospital, then friends and family not being able to gather the same to grieve the loss. Let's not do this again, mmkay?
I guess I post this to say that it amazes me people don't know a single person who died from it. Perhaps this is more about how we're not as connected to community anymore?
"One commenter gives a plausible argument for doing this: '1.2m dead is just not believable. I only know 1 person who died from catching covid'."
This is literally the argument of a toddler (lacking object permanence): Roughly 700,000 people in the US die of heart disease every year. I haven't personally known someone who died of cardiovascular disease in several years. By that logic, the CVD death stats are inflated or false, which is obviously nonsensical.
The effect of increasing birth rate 1940-1950 seems like an important part of the story. US Births increased by ~400k/y from 1940-1945 and then another ~800k/y from 1945-1950. The peak inflection of the Baby Boom was 1945-1946 at ~1M/y..so that's 75 years before 2020-21.
https://x.com/JoeBrenton7/status/1926251349873693077
Also some amplifiction by the decrease in child mortality
https://www.statista.com/statistics/1041693/united-states-all-time-child-mortality-rate/
Very well presented and I think all the arguments you addressed you did so well. I would be interested in getting your take on how one could determine if the lockdowns themselves were the cause of the excess death. The argument being that the drastic and draconian lockdown measure contribute to increases of stress both physical and mental on the entire population. And doing so across a large enough cross section of the population would inevitably result in the increased mortality rate we see as it would essentially take anyone who was already in a borderline situation and "push them over the edge" so to say. This excess death is then attribute to Covid through whatever means and presto you have a non-virus explanation.
I don't know if the above explanation would be persuasive to anyone but it's a novel hypothesis I didn't see directly addressed above. I think what one could do is try and correlate the intensity of lockdowns in various states to the reported excess death one way or the other. That likely wouldn't be perfectly useful one way or the other but is a good starting point.
By my narrative, presented for reference without the expectation of being persuasive, the analysis of excess deaths compared to covid effects, which is the central pillar of the argument, is a reversal of cause and effect.
That is, excess deaths caused reported covid deaths.
The Wuhan Institute of Virology did exactly what it set out to do, which is to create a virus that would kill large amounts of old and infirm people who would naturally have died in short order anyway, and which could be used as cover for statist programs which would kill large numbers of people, and cause massive harm to the general populace.
I find that a lot of details are provable in a sense that the other side denies. Covid vaccines are deadly. Masking is deadly. Covid death counts are wildly inflated. Lockdowns and the accompanying economic damage are deadly. A proper analysis must start from these confirmed facts, and work out the remainder.
The reasons for these beliefs are out of scope, and available elsewhere. I claim that if you follow the evidence from the other direction, you reach a very different destination, thus the crux is whether physical claims and direct experiments regarding, say, mask efficacy are more trustworthy, or whether government statistics are more reliable. That is, internal bias versus external bias, aka, manipulative liars.
I do think here that, the local ingroup is confused about the outgroup. I see a lot of things that are treated here as reliable data, which to me, sound like pronouncements of a local priesthood. I can't confidently assert that my positions are entirely typical of the outgroup here referenced, but I hold that they're a lot more common than the other side expects, and treating "science" or institutional claims by states and associated credentialist entities like universities which can be easily falsified, as gospel truth, will continue to generate surprise when outgroup members, aka, conspiracy theorists, don't take such arguments seriously.
I'm late on this one, but I was an internal medicine intern at a large US academic hospital during covid. I did see many patients die of covid in the ICU during this time, and I think the characteristics of patients who died somewhat explain your phenomenon of not knowing anyone who died of covid.
1) They were mostly elderly, and skewed older even within that group. Like, picture 80 year olds not 65 year olds. I realize you and the commenter you quoted do know some elderly people, but I bet it's not very many. Most people who are not elderly themselves really only know their own elderly family members, but likely have many non-family contacts in other age ranges.
2) They were also mostly very unhealthy. I'm sure everyone has heard the association between things like obesity, diabetes, heart failure, etc and covid deaths already. But even among people with their diseases these patients were on the extreme end of doing-badly. Think a1c of 13, BMI of 45, or EF of 20%. Of the non-elderly patients I saw die they were all in this very unhealthy category. And I suspect that like the elderly, the chronically ill do not make up much of the average person's usual contacts.
3) A lot of them came from nursing homes. Should not be a surprise that there's a lot of overlap between the chronically ill, elderly, and nursing home patients. But it is another place where these people were essentially siloed off from the rest of society. They go back and forth from nursing home to hospital until eventually they don't go back to the nursing home. And unless you're in healthcare or they're your close family member you're probably not aware of them.
Every report, my personal experience and experience of others is in the same direction – that those who died covid where already quite close to death.
Also, the fact that risk of severe covid was exponentially increasing by age is in the same direction. The older or more sick someone is, the more risk he had from covid. But the same was true for risk of death from all causes.
It was as if mathematically the risk of death was suddenly doubled. Young people who has low risk, this doubling meant nothing, covid was less risky than driving a motorcycle or riding a horse.
For elderly people who had high risk of dying suddenly it was a big deal. If you had 10% chance of dying in one year, that would suddenly became 20%.
If your chance of dying within a year was 50%, it meant that you most likely died from covid.
People paid too much attention to ventilators. Great majority died without ventilators. They are not used that often at all. Even if we had no ventilators at all in hospitals, the death rate would not be any different. Even several thousand more deaths are rounding error in the million.
But we had ventilators and survival rate from using them was about 20%. It shows that we needed a better triage – using them only for those who are more likely to survive to increase it at least up to 50%. Maybe it was impossible, we had no diagnostic tools to understand who would benefit from ventilators more than others. I am just saying that creating panic that we ran out of ventilators was unnecessary.
This post missed two effects that I think are very important. First, lockdowns killed old people by keeping young people from becoming immune earlier, so old people were _more_ exposed to the disease before less deadly variants arose. Second, lockdowns will also kill young-during-Covid people later in life because lockdowns deprived those people of education and opportunities during a critical period of their lives, which will lead to crappier, shorter lives.
None of this is to say that it's not awful that 1.2 million people died, or even to question that number. I'm saying that it's facile to say that lockdowns weren't a major contributor just because the proximate cause was the virus, and we haven't seen all the damage of the lockdowns yet. Lockdowns were a stupid, deadly response to the pandemic.
One interesting thing was that Sweden had about the same amount of people dying from covid than in the UK (but much lower excess death) but also that their covid deaths did not form strong peaks, they were more equally distributed.
It shows that stricter lockdowns actually cause more peaks that tend to overload healthcare system. Lockdowns were sold with the promise to protect hospitals from overloading but they achieved the opposite effect. Hammer and dance indeed become a hammer and dance and more deaths.
Yes; the thing about lockdowns is that they *end*. Either officially, or when people stop caring. And then the disease spreads from its remaining reservoirs worse than it otherwise would have, because you've got a mostly immunologically-naive population all going out and engaging in higher-than-normal levels of social interaction to make up for lost time, all at the same time.
"Flatten the curve" was never going to work; that was I think known to be mathematically nonsensical from the start.
"Hammer and Dance", theoretically could have worked but in the way communism could theoretically work - the math is way too complex, the data is unreliable, and the system isn't responsive enough.
Locking down until either a reliable vaccine is developed, or the disease is literally extinct so there are no reservoirs from which it can rebound, takes too long - nobody in the West had the discipline or social cohesion to enforce a really strong lockdown for more than three months or so.
What *might* have done some good, is to realize that we probably could maintain a strict lockdown for two months without too much trouble, *once*, and that the best time to use that silver bullet would have been two months before we were ready to roll out the vaccine.
Perhaps the reason many people posting on blogs don't know anyone who died is that COVID mortality was highly correlated with other mortality causes which were already correlated with low socioeconomic status, diabetes, obesity, and so on. Based on my dear SO who worked on a COVID care floor in a hospital during 2020 and 2021, a typical COVID patient might be an overweight Hispanic man in his 50s, not a 39-year-old college graduate typing on a MacBook.
Let's look at the 5 year average of excess deaths after the pandemic to see if the spike just brought forward impending deaths.
> but it’s only one of the many mysteries of suicides, which often fall during bad weather or social disasters
I don't know if this is original but now I have a theory. Suicidal tendencies correlates with how far you're different (below) from the current average. If the current average is rising (through industrialization or rising technology), you're more likely to commit suicide. But if the current average suffers (through disaster or just seasonal trend), you aren't. I'm pretty sure there'll be an easy data that immediately debunks this utterly, but I'm open for discussion.
I think the most plausible contrarian argument is that most of the people who died were old and fragile, so they would have died some time later anyway. We would need some "lost life years" measurement to measure the impact adequately.
COVID resulted in mortality indirectly as well as directly. It might, for example directly kill someone on whom another person depended, indirectly causing the death of this other person, even though the latter did not himself contract COVID. Mainly, it produced a remarkable social response, with people altering their behavior to avoid the risk of dying directly from COVID. The measures taken to reduce the risk directly from COVID probably increased the risk from other possible causes of death: thus there were extra deaths over and above the deaths directly from COVID.
This social response was likely suboptimal; still, these extra indirect deaths were *caused by COVID*.