Quite a lot of deaths got reported by quite a lot of different states this week. Case numbers did not spike, and instead actively declined. What’s going on? It is right after Christmas and New Years, which means the least reliable reporting of the year, and no one cares much anymore – Indiana and Illinois didn’t bother reporting anything on either front – so the obvious explanation is this is a lot of backlog dumping with the new year. It could however also be that XBB.1 is doing its thing and no one is bothering to test anymore. Error bars are very high for next week.
What I am not worried about, given the case numbers, is that this wave is going to get out of hand. Even if things are worse than we thought right now, it is a bounded worse, and mostly should not much alter our behaviors.
Executive Summary
- Huge spike in deaths, likely holiday reporting issues. Some chance it reflects a lot of unreported XBB.1 cases. Cases actively declined this week.
- New ‘Twitter files’ on vaccine information is a hack hit job, nothing to see there.
- The origin of the athlete death claims is dumber than you expected.
Let’s run the numbers.
The Numbers
Predictions
Predictions from Last Week: 450k cases (+7%) and 2,950 deaths (+7%)
Results: 367k cases (-13%?) and 3,666 deaths (+36%!!)
Predictions for Next Week: 385k cases (+5%) and 3,100 deaths (-15%), with a very large helping of what the heck do I know and a lot of this is reporting issues and I don’t see the value in doing a deep dive to correct for them. Huge error bars.
A number of states reported what looks like multiple weeks of backlogs of deaths, numbers that were absurdly high. Indiana and Illinois did not report at all. I decided to not adjust any of it.
Instead I am simply noting that this week’s number doubtless reflects real deaths, even more than usual following a holiday, is clearly not ‘fully real.’
Case numbers actively declined, even if you were to add back in Indiana and Illinois, which I definitely did not expect. In Florida they almost doubled even after being smoothed out.
Deaths
Cases
Physical World Modeling
New bivalent booster papers from David Ho, Barouch Lab and Dr. Paul Offit. Offit thinks only the vulnerable need to be boosted. On the flip side, Eric Topol comes around, sees bivalent booster working better than expected. None of it much changed my views – mono-valent booster would have been better, updating for the actual circulating strain would have been better, it is likely a small mistake (not medical advice!) for most people to get or not get boosted, the vulnerable benefit enough it seems clearly worthwhile for them.
StatNews on why wastewater testing isn’t getting more rapid adaptation. Boils down largely to ‘this is not what we are used to doing,’ ‘it doesn’t organize the data by the same geographic categories we are used to’ and ‘getting long term investment in things is hard.’ There are also data reliability concerns, especially with comparing inconsistent methodologies and having incomplete source samples, but those concerns seem like clear cases of this still being superior to its alternatives.
Zeynep updates us on things on 8 January.
First some very good news.
Then some news she says is bad.
It is bad that people are being hospitalized. Conditional on the number of hospitalizations, I do not see why them being older is bad news.
Being close to zero Covid at this time strikes me as a failure to engage with reality. So this is that much clearer confirmation to dismiss this particular worry.
She closes with your periodic reminder that Long Covid is a real problem that deserves attention despite many common claims about its severity and frequency being, as she puts it, incompatible with a napkin and second grade math.
A thread from Tom Wenseleers analyzing the growth rate advantage of XBB.1, which is not small. It will clearly displace all other existing variants. Tom says this has ‘predictable effects’ including a new wave, but unfortunately I do not see any prediction with respect to magnitude, which I still expect to be modest.
In Other Covid News
Biden Administration plans to renew Covid public health emergency for another 90 days, says it is preparing to transition into long term threat management and this will be the last time. We will see.
A pilot Home Test to Treat Program being launched by NIH (announcement). This seems excellent, although I worry we will spend all remaining Covid dollars on tests.
Stanford Medicine scientists pinpoint Covid-19 virus’s entry and exit ports inside our noses, highly useful information for potentially creating a nasal spray to prevent infection.
People who believe themselves to be attractive are less likely to wear masks. This is presented as a ‘novel finding’ as opposed to confirmation of a baseline assumption. Masks partly disguise your level of attractiveness. Solve for the equilibrium.
Last week I discussed Bryan Caplan’s post dividing libertarians into two camps. To reinforce what happened to the second camp, the only libertarian counterpoint thus far comes here from David Henderson, who has spent the pandemic documenting government restrictions of our freedoms and attempts to deny that those restrictions were real or pretend that the restrictions made sense. He claims that the ‘freedom from Covid’ camp also wants the ability to fight Covid. His first example of how to fight Covid is… Ivermectin. His second is the government not sabotaging the effort. There is no third.
Anti-vax advocates hound frequently-cited-here health expert Celine Gounder to tell her that her husband Grant Wahl, who died at the World Cup in Qatar right after being famously denied entry because he was wearing a pride shirt and definitely wasn’t murdered, died because he took the Covid vaccine. Often they tied it to the Obvious Nonsense athletic death statistics.
Remember
I also mention this in How to Bounded Distrust, seems worth going over the details here: The origin of the false ‘more athletes died this year than in the past 38 years’ claim is even dumber than you would think (original letter).
A superficially credible doctor writing to a credible journal. Bounded distrust now allows anyone to report the result without looking into it.
We can safely skip the examples. Two completely different-in-kind lists were compared. This is not a good faith mistake.
Road deaths increased 6.8% during the pandemic, with the increase coming during the night. The post cites a ‘lack of traffic jams’ on ‘too-wide roads.’ There was also a large increase in the number of accidents with drivers that did not have licenses, and a 31% increase in hit-and-runs.
In earlier AAA survey, 4 percent of Americans made the shocking admission that they actually drove more during the earlier months of the pandemic when most people were only leaving their homes for essentials — and that when they did, they participated in a range of deadly roadway behaviors, like speeding, texting, and intentionally running red lights, at rates that far outstripped people who drove less.
None of this seems shocking to me. If you don’t have a license, and there is a pandemic, taxis and shared rides become dangerous. You might not be able to depend on others for rides. You also might be bored out of your mind and have nothing else to do, while the roads are suddenly open and clear, and have little tolerance for additional safety talk or worries.
Of course you would run red lights and go faster if there were no other cars on the road. When Serious People say things like ‘the roads were too wide and there were too few traffic jams’ it tells you how unaligned they are with human values. Sometimes riskier behaviors are right.
Consider our family. We drove dramatically more in the first few months of the pandemic, because to avoid being stuck in the city we moved to a town. That moved us from no driving at all to some driving, and by someone very out of practice. That is going to be dangerous. It seems odd to find this shocking.
In a surprise to no one and in response to White House pressure, Facebook, in addition to removing what it believed was false misinformation, ‘focused on reducing the virality of content discouraging vaccines that does not contain actionable misinformation.’ Also:
“We’ll remove these Groups, Pages, and Accounts when they are disproportionately promoting this sensationalized content,” said the Facebook executive in an email responding to White House demands for censorship. ”More on this front as we proceed to implement.”
The Twitter Files: Vaccine Data Edition
Alex Berenson is out with a new ‘Twitter Files’ entry, also available in Tweet form.
It was, shall we say, an interesting choice by Musk to give full archive access at Twitter to someone this dedicated to opposing vaccinations. I don’t know enough to confirm Zeynep Tufecki’s characterization as ‘one of the pandemic’s top grifters.’ That’s a high bar. It does seem highly plausible given what we see here.
This was a textbook example of the rules laid out in How to Bounded Distrust.
I don’t doubt the emails in question were sent. The framing is that this was a corrupt secret mercenary censorship operation to hide the ugly truth that vaccines are useless in the name of corporate profits, while asserting the proven truth of patently false anti-vax claims.
Here is the first ‘secret’ email that is Just Awful:
Scott Gottlieb is absolutely right here that:
- Given the source was a former FDA commissioner, virality was likely.
- The statement is false and misleading. Nothing of the kind was clear.
- One practical result would be less vaccinations and more infections and deaths.
Scott’s other listed attempt to take issue with a Tweet was denied, correctly, by Twitter, and the tweet in question remains up. I can kind of see why he complained here when I squint, but it is an odd choice and clearly not a commercial one.
As Scott’s thread in reply illustrates, there is a bit of a vendetta here, along with threats to his safety. Alex is not shy about their history.
I found this to be an interesting test Tweet on here to draw the boundaries:
If you are reading this, doubtless you can see what Alex is doing here. The Pfizer study has no power when looking at overall death rates, and does not ‘show’ anything at all on that front. Covid simply isn’t that large a percentage of total deaths, and this is not how any of this works. So yes, this Tweet is at least deeply cynical and intentionally misleading, with the goal of stopping vaccinations.
Are you going to strike it down and potentially ban the account for it? Are you going to ‘suppress debate’ and make such judgments from on high?
Good question.
I’m not saying I would have banned him. But I understand.
Zeynep Tufekci is more blunt in her thread. Like her, I was wary at first, yet found Scott Gottlieb to be a fair and reliable source of good information. Some bias and disagreements, as one would expect, no one’s perfect. I am sad he is not still running the FDA.
I sincerely hope Musk figures out why giving this man access to the Twitter Files, and linking people to this entry, was a serious blunder. If he continues to make these kinds of moves, it will go badly all around.
China
Travel is back its old normal self, time for a road trip.
China suspended Korean and Japanese visa applications in retaliation for bans on Chinese travelers.
Crematoriums and hospitals reported to still be super busy, but less busy than a few weeks ago. Things will continue to be not good for a while, but decent chance they have already peaked, or at least peaked until the next strain rolls in.
Other Medical and Research News
Thread of the experience singing up the grandparents-in-law for ACA coverage. Required tricking the system to let it create a new account, denying them any subsidies because they filed their taxes separately, needing someone to prevent them from buying a worthless plan, and a perverse incentive he declines to inform them about where they would have more money if they outright reduced their income.
NHS in the UK is failing to provide adequate emergency care as response times spike to deadly levels.
In many places I am a skeptic that more healthcare results in better health. Emergency care is one of the places where this is very clearly not the case. Emergency care saves lives. Delayed emergency care kills people. My heart goes out. I don’t doubt those involved are doing their best with the resources available. The right thing to do, assuming a fixed NHS budget, would be to shift away from other less impactful healthcare, and reallocate to the emergency systems. Alas, I do not expect this.
There’s a whole big to-do about gas stoves. I moved all that content into a draft post I might put up tomorrow.
It’s official: Be extra cautious while looking for health information in the Twitterverse.
Don't Worry About the Vase 
You gave the nail a glancing blow: everyone worries about unaligned AI when our current problem is unaligned experts.
Malign experts.
“If you are reading this, doubtless you can see what Alex is doing here. The Pfizer study has no power when looking at overall death rates, and does not ‘show’ anything at all on that front. Covid simply isn’t that large a percentage of total deaths, and this is not how any of this works. So yes, this Tweet is at least deeply cynical and intentionally misleading, with the goal of stopping vaccinations.”
No, having not read the study that you did not link, I do not see anything wrong with Alex’s claim. 14-15 deaths in a (small?) time period per group doesn’t obviously stand out as “too small a sample size to get any information from the overall death rate”–it suggests a lot of people were in these groups since most people don’t die. And regardless of covid’s share of total deaths, the vaccine’s (once-in-use) share of total deaths is unknown and one of the things being investigated by the study, so looking at total deaths in the first place seems appropriate.
What does the study show in areas where it does have more power?
The traffic fatality rate jumped 3-4 months after the pandemic began. Early in the pandemic, the fatality rate was lower. In fact, the jump temporally coincides with the George Floyd incident. The study noted covered a lot of ground, yet ignored race for some reason. Steve Sailer does not neglect his beat: “In the CDC database, motor vehicle accident fatalities were up 12.8% overall in May-December 2020 compared to the average of May-December 2018-19. But black deaths were up 37.4% in 2020, followed by Hispanic deaths up 20.4%, white deaths 5.8%, and both Asians and American Indian deaths were down slightly.” That’s fairly dramatic and runs to 1,300 excess deaths just among black Americans in 2020. These elevated death rates, along with the elevated homicide rates that show the same time and race patterns, remained elevated at least into 2021. The de-policing info-op looks like a more significant causal factor than the pandemic.
A failed emergency response system spares neither the affluent nor the top bureaucrats. I wonder what kind of QALYs are lost here and what the cost of correction is. The excess deaths presented are around 5% of all English deaths. There will also be harder to quantify non-fatal harms, such as from delayed stroke intervention. I suspect that in a non-communist healthcare system people would be willing to pay for prompt EMS service. It does seem easier to eliminate a communist government than a communist agency like the NHS. It may not have turned out to be the road to serfdom, but it appears to be a permanent partial serfdom.
Berenson: He was chosen to enact the Law of Compensation. The Left had controlled all of the MSM and social media throughout the pandemic. They lied. Then, they lied some more. Next, they insulted and injured conservatives. Does one receive compensation or obtain power by truth-telling? If this were so, I would expect different reporting from the NY Times and different narratives from the politicos. Can we hold the powerless outer party to higher standards? Unfortunately, mendacity is a means to power. Berenson is a liar. He is also popular with the outer party. And he is keen to give his enemies a dose of their own medicine. It’s a nasty, deadly game. Would I choose Berenson here? No. A guy like that is not necessary. The bare facts of the censorship regime cast the ruling class in a dark light. Adding his trashy malice detracts from the effect. He’s a distraction.
I don’t think the specific censorship request Gottlieb made against Giroir was justified. Readily granting the difference in veracity, what is the operative difference between the roughly accurate claim that a prior infection is as good as a vaccination or claiming it’s “a lot” better? I think it changes very few vaxx decisions–therefore should not be censored. There was a presumption among these people in favor of censorship as the default. Gottlieb went further with his censorship push in the next example Berenson provides, related to exiling children from normal life and vaxxing them. It’s totally unjustified commissaring. Then Gottlieb, the censor, has the audacity to claim repeatedly that he favors debate. Gottlieb is a liar.
The NHS breakdown is predictable, and disastrous. The Conservative government’s austerity plan has not only gutted the health system here, but it has also created numerous problems upstream.
If you look at all metrics for NHS performance, they improved markedly when Labour was in power. The memory of a well-funded NHS is now a distant memory, and it’s being hollowed out as we speak. Labour made some big mistakes (PFI was an abomination) but at least they got results — the system improved on every front.
The tories have sat on their hands since 2010, implemented a pointless and contentious reorganisation of the NHS, underfunded everything (real terms pay for nurses is down double digit percent), failed to reform care in the community & nursing homes and now *everything* is broken. And the more broken it gets, the more people quit, which heaps even more pressure on the remaining staff.
The government encouraged us to clap for ‘the heroes’ during the pandemic, then stiffed them. Now we’re perilously close to being banned from protesting _and_ striking.
Meanwhile, there’s been no adjustment for an ageing population, either. We haven’t kept pace with health spending compared to the richer nations in Europe, but it will be sold as “The NHS is a dinosaur, and privatisation is the solution”.
Meanwhile, I had to get a private consultation for a chronic health issue because I could not get an appointment with a specialist. I was loathe to do this, because by doing this I am helping to fund a private system that offers treatment using the very same NHS staff (in the UK, most medical staff at private hospitals tend to be … NHS staff).
And guess what? Yeah, I saw someone semi-quickly and got a diagnosis, but dealing with private health companies is a complete pain in the arse. I had to make numerous phone calls to get authorisation codes, and they don’t make it easy to understand what insurance covers.
I literally had a conversation that went like this:
Health company: We need to send you for an MRI, but we haven’t had your insurance authorisation code.
Me: OK, I’ll phone them.
Insurance company: OK sir, where is the MRI scan taking place?
Me: Uh, no idea. They didn’t tell me where it’d be.
And around I go. I’ve *never* had to deal with this bullshit before, and it’s tiring.
Also, the moment you get diagnosed with something that’s beyond their ken, you get passed back to the NHS — the private lot wash their hands with you. And that baton pass is often fumbled (as it was in my case).
The same thing also happens if you’re receiving private treatment and they botch it — if you have an operation in a private hospital that goes awry, they just dump you back into NHS critical care. Very much a ‘privatise the profits, nationalise the losses’ playbook.
Things are going to get worse, and it makes me so angry because it’s night and day compared to even 5 years ago.
Sorry for the rant, but urgh. You don’t appreciate what you’ve got until it’s walking out the door.
When you discuss bivalent boosters (in this or other posts), please specify if it’s the one targeted at BA.1 or the one targeted at BA.4/BA.5. One, the other or both may be available in different countries.
Do you have an opinion about which one is better (or whether there is a significant advantage to the BA.4/BA.5 one), either speculatively (which one are the currently spreading variants closer to) or based on experimental studies?
Zvi, I’m 32, and I took two Pfizer shots ~18 months ago, with mild to moderate symptoms lasting for 2-3 days after each shot. How much should I be now concerned with the long-term elevation in mortality risk? What kind of steps (extra medical tests, blood-thinning foods and supplements?) could I take to reduce it?
I regret getting vaccinated, because I have likely avoided the infection, but getting a jab seemed reasonable at the time given the infection risk. Now I’m scared immensely about the spike protein, microclots, and long-term health risks…
https://www.youtube.com/watch?v=B3_bqcvDxvI – no need to watch the video, just please take a look at the description and the comments. I cannot imagine it’s just a matter of extra attention and selection bias…
It’s real bad, but keep a sense of scale. 16% increase of a tiny number is still a pretty tiny number. The people who suppressed evidence of harmful outcomes need to hang, but it’s not a “32 year olds should fear death in the short term” situation yet.
> Of course you would run red lights and go faster if there were no other cars on the road. When Serious People say things like ‘the roads were too wide and there were too few traffic jams’ it tells you how unaligned they are with human values. Sometimes riskier behaviors are right.
This is not as crazy as it sounds.
Actual safe speed on a road is a function of how far away you can reliably see things that require action for safety. For example, if you’re driving in the right lane and there’s a driveway from which someone might enter the street.
Perceived safe speed on a road is a function of lane width and density of overhanging trees. It’s a combination of intuition crudely pattern-matching limited access highways and pre-car instincts that are completely divorced from reality. But it’s a psychological reality that replicates in repeated studies.
If a road has higher perceived safe speed than actual safe speed, people are very likely to travel at unsafe speed. Setting a lower speed limit doesn’t help unless you enforce it reliably, which approximately no one does.
What does help is traffic jams.
The fact that a lot of US streets are only as safe as they are because of traffic jams is a massive indictment of US city planning. As if we needed another one.
Academic city planners know about these effects, and can absolutely use them to make streets safer, and often faster as well. But US city planning is primarily optimized to boost car sales, so city planners are kept out of the field.