Covid 4/​15: Are We Seriously Doing This Again

Link post

Yes we are. It can happen here.


If any of them don’t have mustaches, we need to get them some clip-on ones, because while lots of people die at least they should get to enjoy the pleasures of twirling.

Yes I am fully aware that it is technically a particular rarer blood clotting disorder that is happening here and thus in that subclass it is above the base rate and that there’s an argument this might be ‘real’ in some sense and no I do not care even a little bit about any of that and no I am not going to treat this with the dignity and respect that it does not in any way even potentially deserve. There are scientific details and if you find them interesting by all means read about them but I am ignoring them because like the points They. Do. Not. Matter.

In case you were wondering how people were going to react or what this would do to public confidence, these are from less than an hour after the announcement:

I mean, they’re wrong, but I can’t fault their reasoning from where they sit, if you asked me the ‘which is more likely’ game back in 2019 I would most definitely have not have gone with ‘no really they’re doing this in a pandemic because of six cases.’

Several hours in:

Again, seems logical to conclude they were rushed if they act in a way that would only make sense if they actually did rush.

The first time around with AstraZeneca, I could sort of understand the argument for the other side of the hesitancy effect when I squinted, that this would look like the Very Serious People Take Vaccine Safety Seriously and therefore we should now expect the people to trust the FDA more, and being untrustworthy stewards who kill a bunch of people in order to fool the public into thinking we are trustworthy is a tradeoff they thought we should make, but we ran the experiment on that hypothesis, and, yeah, no.

Seriously, people are not so stupid:

Also, when you keep saying loudly that any adverse things that happen will destroy your credibility, consider your credibility preemptively destroyed already because you’re either right or wrong:

This is going to permanently supercharge the anti-vax movement, not only on Covid but also in general, and kill a huge number of people. Over six cases. Note deaths. Cases. Six.

You know how many people died?

ONE. F***ing ONE.

No, this was not ‘going to come out.’ It was going to completely correctly get ignored. All they had to do was put it in the list of side effects and note it was extremely rare.

This is the Washington Post’s attempt to chronicle what was going through their minds, it’s sympathetic but doesn’t make the decision look any less absurd if you actually think about the physical situation at all or how real people would react to it.

If I had to steelman the case being made, it’s a combination of believing that acting over-the-top paranoid about side effects makes people feel more confident rather than less confident, that a pause to inform people can meaningfully impact care for this rare type of blood clot, and thinking that until one looks at the data who knows how big the problem might be and one shouldn’t assume the math is right until you check it, so we should halt and catch fire for a day and then quickly convene a meeting to confirm that this is only going to kill one person in six million.

Even in a world in which the initial pause wasn’t crazy, there was a meeting the next day to go over the information, and the decision was made to wait 7-10 days and then meet again without making a choice about the pause (and obviously, here, if you choose not to decide you still have made a choice). They didn’t even make the ‘compromise’ decision of halting for young women (and yes, ‘people who are in the subpopulation that is often on birth control which causes orders of magnitude more blood clots than this seems like it’s a hint on what’s happening) and continuing for everyone else, since you can then swap doses between different groups and keep up your pace of vaccinations while you ‘investigate further’ whatever that means here. The failure to at least make that decision is obviously completely bonkers even if you somehow think the initial decision to halt and catch fire was reasonable, as laid out in this thread by someone who supported the first decision but at least supported the ‘compromise’ option at the meeting.

Here’s an argument that this isn’t so bad in the United States, as it will mostly only destroy faith in Johnson & Johnson, rather than faith in the mRNA vaccines as well, or all vaccines generally:

FDA Delenda Est. The FDA must be destroyed.

At a minimum, while we prepare to do that, we can at least implement Tyler’s modest proposal.

Let’s run the (other, not equal to one or six) numbers.

The Numbers


Prediction from last week: Positivity rate of 5.9% (up 0.4%) and deaths decline by 8%.

Once again Washington Post’s numbers baffle me, although this being six rather than seven days later makes them not impossible. Somehow tests fell, cases rose, and the positivity rate barely budged.

A key question whenever one gets good news on deaths is whether this is good news or whether it’s time shifted. If it’s cases shifting into the future, it means the next week looks doubly worse and on top of that you were fooled by what looked like a downward trend. Similarly, bad news can be a mirage from old cases. It now looks like the death rate decline has stalled out, which is unfortunate.

Predictions for next week: Positivity rate of 5.8% (up 0.2%) and deaths unchanged.


Feb 25-Mar 3383416695610195813071
Mar 4-Mar 1025951775371415399623
Mar 11-Mar 1714921010321714027121
Mar 18-Mar 241823957289512946969
Mar 25-Mar 311445976256412626247
Apr 1-Apr 71098867178911604914
Apr 8-Apr 1410701037162111454873

Half or more of the Midwest increase is quirky data in Missouri, but that doesn’t make any of this good news, and it’s likely deaths are going to now be stable or go slightly up, along with cases, until we get enough vaccinations to turn things around.


Feb 25-Mar 366,15158,295151,253115,426
Mar 4-Mar 1062,93557,262114,830109,916
Mar 11-Mar 1749,69659,881109,141115,893
Mar 18-Mar 2447,92172,81099,568127,421
Mar 25-Mar 3149,66993,690102,134145,933
Apr 1-Apr 752,891112,84898,390140,739
Apr 8-Apr 1460,693124,161110,995137,213

Looking at this chart, it seems clear the Midwest’s problems are real. The finale wave is out in force there, even if it’s relatively tame in other places.

Given the increase in positive tests, and the report of a continued decline in test counts, I’m willing to believe that positive rates did go up ~0.4% in the past week, which Johns Hopkins confirms (although they have lower numbers on both ends than WaPo does), so the prediction miss was mostly about doing it based on Friday’s number or some similar quirk (or a math error on their end somewhere).

Things in many places other than the United States are quite bad. In India, they surpassed 200,000 cases per day and things are rapidly getting worse, and there are many other places that have big problems. Aside from the places that successfully did full suppression, the places doing actively well are the ones with strong vaccination campaigns. Facing the new strains while not keeping up in vaccinations is a very bad place to be right now.


That small decline at the end might have something to do with the J&J suspension, or it could mostly be a random quirk. Either way, even without J&J, we still should be able to continue slowly expanding our vaccination rates until we hit a wall where we run out of people who want a shot put into their arm. There are signs this is starting to be an issue in some places, but mostly there’s still plenty of people eager to put the pandemic behind them.

As furious as I am at the J&J suspension, and as many people as it’s going to kill (most of which will be from disadvantaged groups and areas, which J&J’s one shot at room temperature made much easier to reach), it is important not to lose perspective. J&J was a small portion of our vaccine effort, and case growth is not that rapid, so it’s not going to kill hundreds of thousands of people, at least not in America. If we’re lucky, it will only kill thousands.

Vaccine Passport Hype

Washington Post reports on New York’s Excelsior pass, the first one of its kind. Conclusion was that it’s relatively easy to use and isn’t got reasonable privacy protections given the circumstances, but that unless you’re counting on ID to catch fraud it’s trivial to fake it via copying someone else’s pass. That sounds about right. It’s clear by now that there isn’t going to be a national system and that New York is the exception rather than the rule. That doesn’t render the questions moot, but it lowers their urgency and importance quite a lot.

Tyler’s position is that we should be planning full reopening, and that passports seem more likely to hinder that than help. That’s one of the key disagreements. Is the alternative to passports a full reopening, or is it more restrictions? My guess remains that not being able to check leads to more restrictions in the medium term (next few months), but there’s a point when that flips, and things that would have fully reopened without checking would, if given the opportunity to easily check vaccine status, continue to check that status for a while longer. We then have to balance these needs. My guess is that the ‘overtime’ period’ is 50% to last at least two months or so, but highly unlikely (<10%) to last for six, and that the ‘extra game time’ period when passports would help starts now and has at least three months to go most (75%) of the time, and there’s a decent chance (25%) it’s six months or more in at least many blue areas, so one can do a cost/​benefit calculation with this plus all the other objections. Here I’m counting the extra restrictions as pure downside, because even with them the net risk is likely higher than with the pass, unless we’re checking physical cards at the door, which is a different cost/​benefit tradeoff.

The other half of that is the argument from focus. If the country and discourse only have so many focus points (Imperial Focus Points!), which seems basically right, then it’s plausible that all the work on passports delays the full reopening not because of lowering the costs of not reopening fully, but by preventing the attention and blame pressure required to generate the reopenings. Doing anything at all, in this model, has high opportunity costs. I don’t think I value this as highly as Tyler but I’ve likely not been giving it its due.

Vaccines Still Work

The J&J suspension goes hand in hand with the ongoing campaign to convince the public that vaccines work, but don’t work in the sense of accomplishing anything for people. In the name of some combination of proving one’s Very Serious Person credentials, maximizing the quantity of economic harm and scaring people as much as possible, there’s a competition on how to give the impression that being out there is unsafe for the fully vaccinated.

Zeynep points us to an especially creative entry here:

What CNN is saying might be technically correct. A model where 90% of people who are vaccinated are fully safe, while 10% remain at similar risk to before vaccination, is simplified but mostly plausible. What CNN is technically saying here is that there are 100k people who are being exposed to possibly getting infected (look around, could it be you?) and Zeynep is pointing out that this is damn well written to give the ordinary person the impression that if we didn’t Do Something About It that one in ten people who fly vaccinated would get infected, so if you’re vaccinated and fly that way there’s a 10% chance you get infected, which is of course complete nonsense.

Even without this willful mislead it’s still terrible and leads to scaremongering, but this here is something special. There should be some kind of award for such things.

Also, Nate Silver is correct here, and it would be dishonest to treat ‘we don’t know if vaccinated people can transmit’ FUD spreading as anything but gaslighting.

Here’s the Business Insider article linked there, which notes:

Vaccinated people are almost certainly less infectious when they do get infected, on top of not getting infected. The reduction in risk to others is ‘we don’t know’ to the extent that it might be much safer than the 90-95% range in which it reduces risk of infection.

Anyone who tells you otherwise is either lying to you, or is believing the lies told to them by others. Those who continue to treat vaccinated people as risky to others, and avoid living life on that basis, are making a choice to not live life in order to send some sort of social message or tell themselves a story about the type of person they are, or some other not-physical-reality based motivation. Or they just aren’t that into you and it’s a convenient excuse.

That doesn’t mean risk for the vaccinated is zero, precautions that are cheap are worth taking and ‘stupid stuff’ is worth avoiding, and one should follow mask norms for social reasons cause it’s really not that big a deal, but on this ‘we don’t know if it works’ thing, seriously: Stop. Just stop.

Similarly, Zeynep has a thread here about wildly misleading headlines about effectiveness of the vaccines against variants. Studies that find the vaccines work fine are being reported as ‘vaccines don’t work as well’ in a way that has nothing to do with any practical implications. The practical implications are that they work just fine, thanks, and it’s so clear I’m not even going to go into it beyond that. Attempts like this one to scare people about this are pure gaslighting.

How about we instead do this:

And act more like this (video, which is living its best life):

Otherwise you get this:

Which is why this is basically where we are:

That’s before the whole Johnson & Johnson mess.

In Other News

Meanwhile in Germany:


I mean, at least it’s true.

More support for first doses first. As Tyler notes, it’s too late for America to benefit from this, but the rest of the world still could. The same goes for fractional dosing. From what I’ve seen, a lot of people are having a really unpleasant day or two after the second dose of Moderna, and my strong hunch is the severity is caused by using a dose that’s twice as big as it needs to be, and it would be actively better for them to get half doses plus we’d have twice as many doses to give out.

Canadians return home via taxis from Buffalo to avoid quarantine. They did indeed solve for the equilibrium.

Australia might not open its borders even after full vaccination. The hypothesis that Australia succeeded because it was using good epistemics to make decisions is not holding up well in the endgame.

Covid Tracking Project offers thoughts on data source issues. I miss them deeply.