A few days ago we had this post on Stanford people and their issues with covid. Basically, they were afraid–perhaps with good reason–that governments would use the public health emergency to implement restrictions on freedom and maybe some version of socialized medicine, and because of that they felt that various public health officials were overstating the risk, and so they reacted in the other way and made effort to minimize the risk, and in retrospect this led them to say some things that were reasonable (recommendations that lockdowns be limited and precautions focused on oldsters and other high-risk groups) and some things that were unreasonable (estimates of total covid deaths in the U.S. of 500 or 5000, and various versions of vaccine denial).
Overall it’s not clear to me if the Stanford people had a positive or negative net effect. It’s easy to point to some of their more ridiculous positions on risks–in short, it was absurd to downplay the risk of the disease while hyping the risks of the vaccine–; on the other hand, there was a lot of behavioral action which, in retrospect, was overreaction, and so I think it was good that some prominent people stuck their necks out and pushed against that.
A complication is that most people working in public health (doctors, nurses, epidemiologists, government officials, etc.) are politically liberal, and the Stanford group gives off a more politically conservative vibe. I can see how public health people mistrusted the Stanfords on the grounds that they (the Stanfords’) opposed some anti-covid policies as much on political as epidemiological grounds . . . still, people can hold a view for political reasons and it can remain a reasonable view. It’s fair enough to be skeptical about policy recommendations founded in political activism, but ultimately you have to look at the recommendations in themselves.
The other complication is that the national government in 2020 was run by Republicans. Had it been a Democratic administration, it would’ve been simpler for liberals to support whatever the government was doing and conservatives to oppose. As it was, both sides were kinda contorted, and indeed the government was partly at war with itself.
Anyway, that’s all background. A minor figure in the Stanford drama was Hoover Institution-affiliated law professor Richard Epstein, who distinguished himself by saying two different stupid things about covid–one was the aforementioned forecast of 500 or 5000 deaths, the other was a logically incoherent argument against a vaccine mandate.
And now we come to today’s story, which is that, while searching for something completely unrelated, I came across this post from a decade ago, “Politics and the English language, 2014 edition,” where I wrestled with about Epstein’s cliche-clotted language and his false or at least highly debatable claims. The guy wrote like a parody of a bullshitting lawyer.
The point is, he already had a track record of this sort of thing, long before covid ever happened. And he’d achieved academic and worldly success with this shtick! It’s no surprise that he kept doing it. Too bad for the other Stanfords, though, to have to be associated with this sort of crap. Recall that Stanford heath economist Jay Bhattacharya lamented, “In the end, Stanford’s leadership undermined public and scientific confidence in the results of the Santa Clara study. Given this history, members of the public could be forgiven if they wonder whether any Stanford research can be trusted.” Having Epstein in the mix didn’t help.
P.S. Don’t worry, Epstein still has his Hoover affiliation. Fair enough. Across the bay, the University of California is still employing the sleep guy and the torture guy, even if its statistics department did finally get rid of all three of their known sexual harassers.
Anyone who predicted 500-5000 total deaths in the US from COVID should be shunned from ever being taken in any way seriously on anything. That’s not just wrong, that’s willfully wrong with a side of encouraging deaths.
According to our world in data, we blew through 5000 confirmed deaths by the first week of April 2020 or so. We were at 20k excess deaths (estimated after the fact) by March 30.
So far excess deaths since onset of COVID are at about 1.5M and continue to rise.
Wait, we’re still doing this covid thing. I thought we were talking about lies, damn lies, and statistics again which was mentioned recently in the comment section. I’d much rather talk about that, but perhaps that’s also applicable to this covid thing in many ways.
Possible perhaps. Who knows. If there is something like “excess deaths” (whatever that means and however that’s computed), is there also something like “excess deaths” in the vaccinated? Is that being tracked as well, and how is that computed?
Or are there also computations that cover “excess vaccinations”, like people who vaccinated 8 times and still got covid might be part of the “excess vaccinated” in some way? Could be interesting to compute, to see where the optimal amount of vaccinations might lie to still think and communicate to your friends that you’re a good person and following the latest scientific advise from experts but not go unnecessarily overboard. You know, just in case something like this happens again. People would know better to, let’s say, only get 4 vaccinations and still get covid instead of getting 8 vaccinations and still get covid.
Could you please post with a more unique name? It would help those of us that prefer to scroll past your comments and make it less confusing for others who also like to post anonymously. Thanks.
Ow snap!
In all seriousness, it’s a forum where people can post anonymously. Just like you are doing. That you might not agree with what’s in my posts is not something that should matter in this light I reason.
Quote from above: “Could you please post with a more unique name? It would help those of us that prefer to scroll past your comments and make it less confusing for others who also like to post anonymously. Thanks.”
To add to my previous reply: if you would like to read something else than my comments here, and read a more unique name, you can perhaps check out my manuscript on SSRN titled “Why psychopathy might be present and prosperous in present-day psychological science”.
It contains several references, one of which mentions how overstating the impact of their own research and teaching accomplishments, while belittling the efforts of junior faculty members might be a behavior of someone in higher education with psychopathic characteristics. The more you know!
My guess is that you personally might not necessarily find that manuscript useful or interesting, but that it may be useful or interesting for others. For instance, it contains a short list of features of the prototypical psychopathic psychological scientist which may be useful for people to be(-come) more aware of.
Why do you consider looking at excess deaths problematic? Seems a sensible approach to me. You take the five-year average prior to 2020—say, 2014 to 2019—and compare that baseline to the death rates during the COVID-19 years.
For reference, here’s the death rate in the US from 1950 to 2023:
https://www.statista.com/statistics/189670/death-rates-for-all-causes-in-the-us-since-1950/
Interestingly, the death rate was much higher in the 1950s through the 1970s than during the COVID-19 pandemic. Even in the 1990s, it was still higher. What explains that? Was it due to higher crime rates, different healthcare standards, or something else?
“Why do you consider looking at excess deaths problematic?”
I may be incorrect, it’s been a while since I tried to understand this all, but I think I remember excess deaths being computed in a way that left me with many questions. It was based on some projection or something like that. And there was something about covid deaths counted, and how, that left me with many questions at the time. It all resulted in me concluding that the numbers might be presented in a certain way that might not be optimal.
I think the all cause mortality was what I was looking for at the time, because I thought it was the most “simple” way to start to begin to understand what was going on concerning deaths. I wonder whether this all cause mortality numbers might be influenced by the vaccination in the sense that the mass vaccination created an event, just like covid, that may have influenced all cause mortality numbers…
all cause mortality minus the 5 year average for all cause mortality on the same day of the year in the previous 5 years before onset of COVID is my understanding of how “excess mortality” was computed. Or something close to that (maybe a smoothing spline or whatever)
A LOT of people got a vaccine between Jan 2021-Jun 2021 (~50% of the US were fully vaccinated by Jun 2021) and that whole time excess mortality was plummeting. It started up again in July 2021 during another COVID wave, and that wave was much more concentrated among unvaccinated people if I remember correctly (at least here in CA where that information was available at one time).
All this data is available at OurWorldInData.
Anonymous –
There are many people who have computed excess deaths over the pandemic period, differentiated by vaccine status. In particular, here’s one:
https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/deathsbyvaccinationstatusengland
There are many others, easily found. They uniformly show that the vaccines prevented many deaths and excess deaths. And then hospitalization and severe illness.
Quote from above: “There are many people who have computed excess deaths over the pandemic period, differentiated by vaccine status.”
I read that it’s “age standardized”, then I looked that term up, then I thought about what I expressed earlier about possible sub-optimal presentation of data, then I wondered why there always seems to be some sort of calculation (or what’s the term) in these kinds of numbers (just like with “excess deaths”), then I remembered why I stopped reading about this stuff years ago.
Thanks for the reply though, it may be interesting for others.
Calculation of excess deaths is not a mystery. Demographic models predict a certain number of age-gender specific deaths in a period, assuming stable age-gender mortality rates. Excess mortality is when the observed data is higher than the prediction there are excess deaths. Of course we expect that observed and predicted will not be identical. It’s model based and like everything with data there are different ways of doing it. Here’s an overview https://www.swissre.com/institute/research/topics-and-risk-dialogues/health-and-longevity/covid-19-pandemic-synonymous-excess-mortality.html.
However, this is just an indirect way to try to retrospectively estimate changes in mortality and that insurers and others use to model projections.
In terms of “excess deaths” due to vaccines, you could do something similar, e.g. if we kept the all cause death rates for each age-gender group from a pre vaccine period and model it forward we could see if the number of deaths was lower, higher or unchanged after introduction. I’m sure you have seen the mortality rates before and after the introduction of the measles vaccine.
However, in terms of vaccines, I think it’s much more important and meaningful to monitor at the individual level as is done in the monitoring of adverse events in the clinical trials prior to approval and the continuous collection of vaccine adverse event data long after clinical trials are completed and a vaccine is approved. (This actually happens much more extensively for vaccines, but pharma companies have to report adverse events for all approved drubs.) And, of course, clinical trials continue for new formulations such as the combination vaccines etc. A lot of that data is publicly available; you can read the FDA minutes.
That said, there is a lot of post approval research that is done on questions like what you asked (how many/how often to optimally reduce the impact of infection and/or the likelihood of infection).
This was when Cuomo was sending covid patients into nursing homes, right? These positive tests (which have 50% chance of at least one positive after a month of daily testing) also had numerous dangerous ramifications for these vulnerable people. Then dying for any reason after a recent positive test was classified as a covid death.
The total all-cause mortality (and “excess” is simple enough to be ok) is likely accurate, besides that the stats are all junk.
Then the vaccines didn’t stop you from getting covid due to limited/no mucosal immunity, which lasts only a few months anyway. This was obvious from just reading about flu vaccine trials back to the 1970s, then confirmed with the preclinical animal tests. Yet, we needed “vaccine passports” anyway.
And finally comparisons between vaccinated and not mortality are confounded due to all the terminally ill people not getting vaccinated, yielding a 3-10x higher baseline mortality in the unvaccinated. Yet, its claimed millions of lives were saved anyway.
There should be some shunning going on for sure.
Thank you for your reply!
“Then the vaccines didn’t stop you from getting covid …” True enough; I was vaxxed and boosted, and got Covid nevertheless. However, it was like a not very bad cold, and since I was 82 at the time, I’m damn grateful for the vaccine.
as you say, excess deaths is probably a good number. It’s based on all recorded deaths, which is definitely a good number in the US, and the difference in death rates between actual and an average of 5 years before covid. Which is also a raw average of the same previously collected very good data.
So, for all the claim you’re making about confirmed deaths being any death after a positive test, and implying therefore that would be an overestimate of the true “caused by covid” deaths, in fact, excess deaths exceeded confirmed deaths right from the beginning and confirmed deaths was always likely an undercount as predicted by most epidemiologists.
https://ourworldindata.org/explorers/covid?Metric=Excess+mortality+%28estimates%29&Interval=Cumulative&Relative+to+population=false&country=~USA
But you’re not wrong to discuss data quality. excess deaths could only be computed after the fact, once CDC death certificates had been collected, collated, and computerized etc. Before the fact there were questions, but those questions should have been in the range of 10-20% errors, that we actually should expect, and not in terms of factors of 1000 as in these people predicting 500-5000 deaths.
The thing is, 500-5000 deaths for the whole pandemic was just patently ridiculous on its face, and lots of people said so. Around April 14 we had 2000 confirmed covid deaths **per day** and literally anyone with any epidemiology background who wasn’t part of some political apparatus (ie. individuals not spokespeople) accurately predicted waves of thousands of deaths per day in the US.
https://ourworldindata.org/explorers/covid?time=earliest..2024-09-13&Metric=Excess+mortality+%28estimates%29&Interval=7-day+rolling+average&Relative+to+population=false&country=~USA
You can ridicule responses like removing the basketball rims or fining people walking their dog in an empty park, or whatever, but one thing you really can’t do is point to basic epidemiology predictions from nonpolitical epidemiologists and laugh at them as patently absurd like 500-5000 deaths was.
You could listen to epidemiological scientists and be off by maybe a factor of 2, or you could listen to these idiots and be off by a factor of 500 to 1000.
Indeed, even in the massive clinical trials that were done, the outcome measured was not infection but being diagnosed when the participant went for treatment after feeling sick. They did not ever look at or attempt to measure infection itself, but symptomatic infection.
Anoneuoid, you said “Then dying for any reason after a recent positive test was classified as a covid death.” This was never true. We went through this several times a few years ago. I don’t know why you insist on repeating this easily disprovable falsehood.
Phil: Yes, this is annoying.
Anon: If you’re going to keep doing this, please stop commenting here. If you want to keep repeating easily disprovable falsehoods, go to twitter, 4chan, etc., where you can troll to your heart’s content.
I didn’t realize it was a easily disprovable falsehood. Sorry.
Andrew – I’ll limit this to one comment.
Anoneuoid –
It’s amazing how you’ll go to the same nonsense over and over, despite having no actual evidence to support your claims. Meanwhile…
<i(Besides the general "Healthy Vaccinee Effect", there were two effects specific to the covid vaccination and which come ON TOP of the general one
One is a "healthy vaccinee", the other an "unhealthy vaccinee" one
I don't think anyone has flagged this before so here we go
https://x.com/Jean__Fisch/status/1831996596407480551?t=rczliOy4VM0BX8tgx9ongQ&s=19
Maybe if you did some actual research instead of just cherry picking theoretical phenomena and pretending that they are established fact.
I’ve got another longer post held due to links, but in March 2020 the Kings College group predicted 2.2M US deaths from COVID if nothing was done. Stuff of course WAS done, but in the end we have excess deaths in the range of 1.5M. So essentially the Kings college predictions were right within a factor of 2 right from the start.
https://www.thelancet.com/journals/lanmic/article/PIIS2666-5247(21)00029-X/fulltext
predicting 500-5000 deaths from COVID was like predicting the Hiroshima bomb would kill 10-100 people
From the paper: “Initial projections built worst-case scenarios that would never happen as a means of spurring leadership into action.”
Wait, what?! Did they say the quiet part loud? What does this mean?
That’s the authors interpretation. I’d say something more accurate was “Initial projections were based on no mitigation efforts, something that everyone acknowledged was unrealistic since some mitigation efforts would certainly be used… and yet those projections were ultimately within a factor of 2 of the correct total, so they were almost certainly underestimates for the actual outcome for the computed mitigation free scenario”
As I mentioned in a previous reply thread, I had the opportunity to have a one-on-one conversation with Epstein back in the 90s. It was clear back then that he had drunk the Straussian Kool-Aide: making stuff up is not only OK but even a badge of honor in the epic battle over subjecting the social order to the rule of virtue. One way of putting it, I guess, is that Strauss’ big contribution was promulgating a conception of virtue that embraces dishonesty.
Unfortunately, this wing of conservatism has conquered the right wing political space in the US, after conquering right wing media. And now we have people like JD Vance. From what I understand, whatever the problems with the other Standfordites’ work, it wasn’t cavalierly dishonest, just motivated, as you say.
Peter:
Also I think that behaving like a hack dulls people’s edge. In Epstein’s case there’s the confounding factor of his advanced age, but we’ve seen this too with younger scholars who became political hacks. Niall Ferguson was an interesting and creative historian who then ended up doing stupid campus politics and homophobic performance art; John Yoo was presumably at some point a competent legal scholar who decided to become an embarrassingly misinformed mangler of American history; etc. I guess there’s always a temptation to move down the difficulty level–here I am, writing blog comments instead of proving theorems about the limiting properties of stochastic processes!–but the hack thing seems to change the decision making: once you’re a hack and you’ve decided not to tell the truth (perhaps from laziness or perhaps from some “Straussian” motivation), there’s less of an incentive to know the truth. If you’re just gonna write what you’re gonna write, then why go to the trouble of actually learning the facts, right?
Regarding the other Stanfordites: I think it was dishonest, not just motivated, of Bhattacharya to falsely claim that I had “incorrectly thought [he and his colleagues] had not accounted for the possibility of false positives,” as well as to claim that I “later recanted that harsh criticism.”
To write fast, you have to train yourself not to ask “could I be wrong? what might I be missing? what are the best arguments against this?” And if you mostly engage with ignorant people who didn’t think before they posted or spoke, you forget how to engage with strong arguments from thoughtful people.
Sean:
I think this has happened to Paul Krugman and Nate Silver. I don’t think they’re hacks, but what I do think is that they are both afflicted with many stupid critics, and this has gotten them into the habit of treating all criticism as idiotic, which affects both the tone and the thoughtfulness of their responses.
Andrew: Hmm, are you saying “priming” is a real strong psychological effect? ;-)
priming = “afflicted with many stupid critics, and this has gotten them into the habit of treating all criticism as idiotic”
Seth:
Effects of environment on behavior are real, and we see them all the time. The problem with the crap “priming” research is not that it’s studying the effects of environment on behavior, it’s that it’s studying the effects of irrelevant inputs. Just for example, it makes sense that recent economic events will have large and consistent effects on voting; it does not make so much sense to expect this of shark attacks or college football games. It’s not impossible for such effects to exist; it’s just that I wouldn’t expect to find such large and consistent effects, and indeed such claims don’t seem to be well-supported by the evidence. But the lack of evidence for effects of silly inputs should not be taken to imply real effects elsewhere.
Speaking of Stanford and misinformation…as well as a poor understanding of statistics…
Our likely next Surgeon General, Casey Means, is on record stating during a Senate roundtable discussion that “for each additional serving of ultra-processed food we eat, early mortality increases by 18%.”
While I am not a statistics expert, that statement appears to be innumerate even if made off the cuff and not as a part of expert testimony before Congress. Studies, such as a 2025 analysis in the American Journal of Preventive Medicine, suggest that each 10% increase in UPF contribution to total energy intake is associated with a 2.7% increase in all-cause mortality risk, not 18% per serving.
Even if we interpret her statement charitably—perhaps assuming she meant “for each additional serving per day, every day, over an individual’s entire lifetime”—it’s still an obvious exaggeration and omits critical considerations, such as confounding factors like socioeconomic status, lifestyle, and the diverse nature of UPFs themselves.
In case anyone thought we’d bottomed out with Kennedy, Makary, Bhattacharya, and Oz… Well, nope.
You know she’s a serious commentator on health because she gives an exact number, not some vague phrase like “a lot”. I wonder if she knows a percentage increase is compounding, so that after 10 servings (for example) the number of early deaths goes up more than five-fold.
Huh, this appears to be the source:
https://bmjgroup.com/new-evidence-links-ultra-processed-foods-with-a-range-of-health-risks/
In the second study, researchers based in Spain evaluated possible associations between ultra-processed food intake and risk of death from any cause (“all cause mortality”).
Their findings are based on 19,899 Spanish university graduates (7,786 men; 12,113 women) with an average age of 38 years who completed a 136-item dietary questionnaire as part of the Seguimiento Universidad de Navarra (SUN) study.
Again, foods were grouped according to degree of processing and deaths were measured over an average of 10 years.
Results showed that higher consumption of ultra-processed foods (more than 4 servings per day) was associated with a 62% increased risk of all cause mortality compared with lower consumption (less than 2 servings per day). For each additional daily serving of ultra-processed food, mortality risk relatively increased by 18% (a dose-response effect).
Seth –
Thanks for the link. It makes her statement less egregious, and there is adjustment for certain potentially confounding factors in the studies.
Still, the finding of 18% relative risk increase per serving (per day, which she left out) is for those consuming >4 servings of UPFs per day. I’d say that’s a pretty important detail. Plus, surely it’s important to differentiate among various UPFs (e.g., fortified cereals versus Twinkies)
The sad irony here is that her whole schtick is about increasing the public trust in our medical institutions.
Oops. Actually, rereading I got that wrong and the 18% increase per serving per day IS across the whole cohort (and not just those eating >4 UPFs per day).
Apologies to Casey.
Quote from above: “Oops. Actually, (…)”
+1 for the correction
and…..
Congratulations!
You are hereby nominated (pending further verification of statements and details) for the 2026 Altering Initial Statements or Titles After Thought or Study (AISTATS) award.
The 2025 Altering Initial Statements or Titles After Thought or Study (AISTATS) award has been handed out very recently, see the blog post on here dated may 5th.