I agree with this, too. It feels like we got a lot right by accident. The inconsistency between city and state governments meant the places with a lot of population density, interstate and international travel, and crowded indoor spaces mostly did the right thing by shutting at least those things down, whereas the rural areas that probably didn't need to do much didn't. Even the one thing that seemed like the Dallas equivalent to LA's exception for film production being considered "essential," allowing all construction to continue so our local real estate boom could continue unabated, turned out to be fine once we learned outdoor gatherings were fairly low risk.
This bothers me about basically *every* discussion of covid safety measures. Despite there being large variation in what the measures are from place to place, and despite these measures being completely unfamiliar from the Before Times, we somehow adopted loose terminology ("lockdown" or "quarantine") for everything from "please wear a mask for the 30 seconds between stepping into a restaurant and going to a table" and "the apartment doors will be welded shut and food will be delivered by government workers".
Unfortunately, it also seems that calls for/against mitigation measures have often gravity-welled somewhere around either "nothing is necessary" on one side to "welded doors with doordash portals" on the other - doesn't seem to be just in the way we talk about existing measures (not referring to you, Scott or Wendy, more in the public parlance...)
When debating the severity of measures and their ramifications, and trying to strike that balance in a more specific and granular way...I think there'd have to be a way to do a holistic analysis of effects of both specific and cumulative measures next time, or at least make a show of doing so, to avoid or decrease some of the political tension before it starts as well as just generally optimizing policy.
For example, once we get through this, I'd be interested to see an analysis of long-term deaths correlated with or caused by economic downturns a la March 2020 (despair, starvation from disruptions in food chain, etc.) vs. estimated disease death count. Maybe this analysis could help us hone things - I'm not sure what one does about deaths of despair, or how avoidable they are, but maybe we could prioritize/secure the food system to be robust against the specific issues that are leaving swaths of the third world at risk of starvation.
For example, this analysis seem to clearly indicate that, at least in the short term, COVID well outnumbers deaths of despair here in the U.S. but I had to look pretty hard just to find one analysis - which is now about 3 or 4 months out of date.
i am inclined to agree with speculations i heard that depressed people are likely to blame their sorrows on the pandemic (and to think they'll wain once it is over). i personally expect a spike after delockdownment coming from people noticing their problems not suddenly dissapearing.
The one other point is that we have to compare not *actual* covid deaths to the deaths of despair, but the *difference* between the actual and counterfactual covid deaths that was produced by the policy, which is just one more level of difficulty. (But if we make a very rough guess and estimate that the number of prevented deaths was the same order of magnitude as the number of actual deaths, that could help.)
People are terrible at this, largely because they do just what you said, which is assume that measures are only weakly helpful in the direction that opposes their instinct. You will find many people who are 100% convinced that if no public measures at all had been taken, the deaths from COVID would've been maybe 10-20% more --- at which point they'll say, so what? Not worth it.
But what if the actual counterfactual is that without any countermeasures at all the deaths from COVID would've been 10x or 20x higher (e.g. in the neighborhood of 5-10 million)? If people knew for a fact this was the case, it would totally change the argument. But unfortunately, it is essentially impossible to even construct an informed guess about that during the pandemic, and so we fall back on reasoned argument. If reasoned argument could establish the truth or falsity of complex questions of viral transmission and disease, the Romas would've conquered the plague in AD 200. And unfortunately most people are very willing to believe that their own WAG is no more than 20-30% off, tops. *Nobody* will believe he's wrong by two orders of magnitude.
I am sure the pros will be combing through the records of all nations and individual states that tried various things, and the outcomes, for decades to come, and making arguments about what can be shown empirically to have worked or not worked. But the nature of record-keeping is such that there is a very good chance we will never have any really solid answer, at least as far as this particular pandemic goes.
An examination of over 80 Covid-19 studies reveals that many relied on assumptions that were false, and which tended to over-estimate the benefits and under-estimate the costs of lockdown. As a result, most of the early cost/benefit studies arrived at conclusions that were refuted later by data, and which rendered theircost/benefit findings incorrect. Research done over the past six months has shown that lockdowns have had, at best, a marginal effect on the number of Covid-19 deaths. Generally speaking, the ineffectiveness of lockdown stems from volun-tary changes in behavior. Lockdown jurisdictions were not able to prevent non-compliance, and non-lockdown jurisdictions benefited from voluntary changes inbehavior that mimicked lockdowns. The limited effectiveness of lockdowns ex-plains why, after one year, the unconditional cumulative deaths per million, andthe pattern of daily deaths per million, is not negatively correlated with the stringency of lockdown across countries. Using a cost/benefit method proposed by Professor Bryan Caplan, and using two extreme assumptions of lockdown effec-tiveness, the cost/benefit ratio of lockdowns in Canada, in terms of life-years saved, is between 3.6–282. That is, it is possible that lockdown will go down asone of the greatest peacetime policy failures in Canada’s history.
[I'm not sure I agree with the paper, in particular with its part III, but I really enjoyed parts I and II where it shows the various ways scientists/economists/politicians can shoot themeselves in the foot when trying to reason about and model lockdown effects]
Just to complicate things further, the overall prevalence of mandatory lockdowns would signal to people whether they should voluntarily lock down. In that way, mandatory lockdowns could have a broad health benefit even if mobility is uncorrelated with mandatory lockdowns across countries.
Douglas Allen is an economist, not a public health expert or epidemiologist. I would value his analysis of the economic costs of lockdowns some and his analysis of pandemic public health literature close to not at all.
> Unfortunately, it also seems that calls for/against mitigation measures have often gravity-welled somewhere around either "nothing is necessary" on one side to "welded doors with doordash portals" on the other
As someone who has been very conflicted on this specific topic, I think there's an expectation (which I honestly don't know if it's warranted or not) that strong blanket rules with no nuance are needed because any level of nuance will be taken as "whatever the minimum acceptable restriction in any situation is, is the maximum acceptable restriction in any situation".
So, the fear is that saying "wear masks while indoors, but masks are unnecessary outdoors" will translate to a sizable portion of the population hearing "masks are unnecessary, regardless of if you're indoors or outdoors". This seems uncontroversial, to me.
The more controversial take is this: When instead given guidance of "wear masks all the time, and when you're indoors, wear 2 or 3 of them!" this will translate to a *larger* portion of the population hearing "masks are necessary when you're inside or outside" and will result in a greater proportion of the population actually wearing masks indoors, where it matters. The sacrifice to get there, of course, was needing to wear masks outdoors, where maybe it didn't matter.
I've felt this most acutely with the response to restaurant openings - in my area, we had restrictions on indoor dining for the majority of 2020, but essentially no restrictions on outdoor dining. One of the results of this was multiple restaurants building free-standing structures on either the sidewalk or, in places where they had them, parking lots, and then (because many of these were operating during the fall and winter) sealing them up and putting space heaters inside (because dining outdoors in the winter isn't very comfortable) and just pretending like "a sealed tent with space heaters" is actually outdoors as though this was a pointless regulation that had no actual purpose, rather than an intentional risk mitigation strategy.
That's a really interesting point. My first instinct would be that telling people that masks are needed outside/ you need more that one mask/other takes, would cause people to mistrust and disregard the rules overall if they pick up on the exaggeration...but I could totally see that happening as well.
If you tell people Noble Lies to get them to do what you think they ought to, they will *eventually* figure it out. Then the next time you think they ought to be behaving differently, you'll need literal men with guns to make it happen. Our society gets one "massive lockdowns with voluntary compliance" bullet per generation or two, and it only works for a few months. I do not think it was wise for us to have used that bullet in March 2020.
Makes sense if you think about it. What actually happened for a lot of people is that they got to work from home/got some extra time off. Fewer people were bugging them at work. Kids who were in school were given a weird sort of pseudo-vacation/remote learning thing.
If your stress comes from work or from school, your mental health would likely improve significantly.
The people who struggled with the lockdown were probably outnumbered by the people who saw less stress due to less stress from work/school.
Moreover, more people are better off after the pandemic than worse off economically - about 2:1.
But, like, 'necessary' in this case is kind of weird, right? The (largely liberal) NE/CA/cities that attempted to lock down way harder than the (largely conservative) TX/FL/etc seem to have similar outcomes in the US?
Unless you're comparing the US as a whole to say Australia, which was basically the only success story.
You're right that I should look into this more. Right now my model is something like more cases -> more lockdowns -> fewer cases, which means that the correlation kind of washes out. In support of that, see the graph on the top of page 3 here, and also the scatterplots at https://www.nytimes.com/interactive/2020/11/18/us/covid-state-restrictions.html (which are a point estimate from November 2020). In order to be really confident in this I would need to see these kinds of estimates up until the present, which I can't find. I'll register this as my suspicion for now and hopefully remember to confirm with more data after the pandemic is over.
This is honestly kind of sad, but I actually don't trust the NYT as a source anymore because they've become so agenda-driven (and I say that as a former long-time subscriber).
I don't trust them either, but it's hard to mess up a scatterplot. I trust them to lie in subtle, hard-to-detect ways, rather than literally putting scatterplot points in the wrong place. I eagerly await learning I am still too gullible.
Oh, true, I don't think there's any strict data falsification* going on, but I think they pick and choose what to report, e.g., that article is on cases (and to be fair, hospitalizations) but not overall death rate, which I think is a better metric - one can very much disagree with that though!
*Though I will say that my tipping point was when I read about the 'largely peaceful' BLM riots last June after watching - literally, watching, at 8p at night - the drugstore and liquor store across the street from me (in Flatiron!) get broken into the night before.
I think it was a matter of timing. That particular scatterplot was made after the northeast was finished getting hammered (at least Round 1 of hammering) and before California got hammered. Remember they're looking at "current" (Nov 1) hospitalizations, not total.
I can't believe that urban/rural split isn't relevant here. I live in Indiana. While I live in the city and we locked down and locked down early, I can't help but feel that such measures would have been draconian in rural counties with 2-3 digit max number of cases (certainly were perceived that way out in the hinterlands from my friends and family).
At a strict level, no. At a definitional level, you can certainly have metrics that either operate on a per-capita basis and/or a density basis (per-capita per-square mile), depending on what the models used are able to predict is most effective.
While I'm moderately annoyed that the state didn't actually stick with it, Illinois at least early in the process had specific metrics in place to be able to treat urban (ie Chicago) areas qualitatively different from rural areas (ie by adjusting the restrictions based upon pre-defined metrics per area, rather than evaluating for something like the state/country as a whole).
If the data ever becomes available I'd really like to see what fraction of variance can be explained by pre-existing Vitamin D levels, age demographics and population density.
I feel like "more cases -> fewer cases" seems to work even without "more lockdowns", the "more lockdowns" just correlates very heavily with "more cases". Not that "more lockdowns" can't (also) have a causal effect, it's just really hard to estimate what it is because the counterfactuals everybody produces are nonsense.
In this case, "everybody" includes rationalists who are no better at predicting the specifics of the pandemic than anyone else (see e.g. all the fuss about variants earlier this year).
Indeed. Similarly, the argument that lockdowns were solely/primarily responsible for downturns in cases has long been plagued by the issue that declines in cases often started before (or far too quickly after) lockdowns were imposed. "Wait two weeks" goes both ways.
Pretty much all of Asia is laughing at that last line of yours.
It's certainly possible that what the government did in (largely liberal) NE/CA/cities didn't make a big difference versus what the government did in (largely conservative) TX/FL/etc. But that doesn't mean that lockdowns, wearing masks, etc. wouldn't have worked. They clearly did work in quite a few countries, including, yes, Australia.
But nowhere in the US did a real lockdown, certainly not for long enough to actually push cases down small enough to mostly wipe it out. Our population just didn't have even near the patience required.
I don't think it's patience as much as the overall attitude/tolerance for big government to impose its will...which, yeah, Asian countries 'won' in this case but I think on the balance overall (standard of living, economic mobility, etc) tilts in the direction of the US and similar countries. But that's an entire other can of worms and I'm obviously super-biased.
Fair enough. And I agree that, in general, there is a tradeoff, and I am very much on the side of individual choice etc.
But it's important to acknowledge that hundreds of thousands of people in the US died because of COVID, and they didn't have to. Was it "necessary" to lock down, well the world didn't end, so I suppose that wasn't "necessary". But if you actually care about your fellow man, as the chorus of epidemiologists sing above, more of us should have made other individual choices. And our government(s) should have done more to help us. And I'm going to continue to be pretty mad about that.
Also fair! FWIW, I'm generally one of the selfish ones who cares most about the statistical health/well-being of my fellow man rather than the specific health of individuals (which I know isn't a popular opinion) so that's coloring my thoughts as well.
> And our government(s) should have done more to help us.
This is, I think, a big part of why lockdown took the shape it did in the US, and why it was so resented: we foisted the cost of lockdown on the people themselves.
Class was a big issue here: the professional/managerial class, which I'm part of, could mostly work from home, and so the cost of lockdown was substantially less for us.
Yes, our government should have done more to help us -- by getting out of the way! By allowing challenge trials and allowing people to waive full testing and take vaccine months earlier, saving hundreds of thousands of lives; by allowing private companies to do testing early on rather than monopolizing; by allowing importation of good masks; and many other things. I wrote on many of the bad moves by the CDC a few days ago: https://medium.com/conjecture-magazine/failure-at-the-cdc-8cbb772182de
"But it's important to acknowledge that hundreds of thousands of people in the US died because of COVID, and they didn't have to."
Lots of people agree on that. The disagreement is about what should have been done to prevent it.
I saw something, I think from the CDC, boasting of the fact that we got vaccines in less than a year. As the poem says, we got the first vaccine in about two days. The remaining nine months or so were spent on the testing needed to get the FDA and similar agencies elsewhere to approve a vaccine.
With challenge testing and not insisting on a rigorous, controlled experiment approach, that would have been more like three months — one month to check that the vaccine itself was not seriously dangerous, which doesn't require challenge trials, another two months with challenge trials of a thousand or so healthy young people, resulting in between zero and one deaths (IFR about 1/10,000), after which the only limit is how fast they can produce the vaccine and offer it to anyone who wants it and is willing to have his health status thereafter kept track of.
That would pretty clearly have saved several hundred thousand lives.
china isnt doing well but the more liberal asian countries (Japan, Korea, Singapore) are doing a lot better. these same countries also prove asian culture is not contradicory to a high standard of living.
Heavens, no, China did much better *economically* than the US by controlling the virus very early and quickly, and then re-opening their economy. China GDP growth in 2020 was +2.3%, which while crappy beats the heck out of the US (-3.6%) or the eurozone (-7.4%), who basically rode this hideous roller-coaster of partially-effective counter measures followed not quite long enough, followed by somewhat overoptimistic let-ups, rinse, repeat, which achieved the magical combination of *neither* controlling the virus nor allowing for normal economic activity for the entire year.
My company has offices in the US and China. US offices have been closed for 13 months. Those in China closed in January and re-opened in April of 2020 and have been functioning normally ever since.
Asia locked down before it was endemic. By the time the elites and public health establishment in the US woke up we had several million cases (backfilling from 30k deaths by mid April to 2-3 million cases by mid March).
Australia is the one counterpoint - they had a 2.5 month lock down (seemingly helped by seasonality) that got a very serious outbreak back to contained status.
The question is not do NPIs "work" in the sense of do they in the short term lower cases - quite clearly they do (imagine an extreme where soldiers patrol the streets keeping everyone in their houses - there won't be much COVID!). The question is:
A. Is all you are doing just pushing the pain back (relatively straightforward in the absence of a vaccine)
B. Are hospitals in danger of being overloaded
C. Is the cost in liberty/inconvenience/economic damage/societal damage (eg school closures) worth the deaths saved (in light of A and B above)
Reasonable parties may differ on the answers to any of those, but I think they have to be evaluated in light of the fact that COVID19 had already spread significantly before the public health establishment advocated much of anything.
I would assume the alternative to utilitarianism is deontology, ie you can't transgress some rights no matter how many lives it saves. And the problem with that is that government is already following the official deontology , the set of rights that are de facto recognised. You might feel that the government can't order you to wear a mask, however many lives it would save, but that would be your own personal ideology.
yeah perhaps the better thing would be "did something dramatic before it got endemic, up to and including border closures, centralized quarantines, widespread mask wearing, and contact tracing with teeth". Obviously not all of those apply to all countries but hopefully you get the drift.
It's numbers are even worse if you look at excess deaths, although it's impossible to tell how many are uncounted lockdown-induced covid deaths or lockdown-induced other deaths.
It was still winter, and barely past the coldest point, when the lockdown & other rules definitively turned the tide. The inflection point was the first week of August; the mean temperature for that week was marginally lower than the mean for July, and the mean for the whole month of August (by the end of which we were clearly on the way to zero) was just 1 degree warmer.
I mean seasonality doesn't necessarily literally mean the seasons - we don't even really know why the flu season behaves the way it does (and lots of other things about the flu). Just that the virus appears to move in cycles that defy explanations that rely on only human behavior.
Australia has been an exception in lots of ways, so definitely possible it was all just human action! Just wouldn't rule out disease cycles.
so substack is telling me this is in reply to me but I'm having trouble telling what comment it is in reply to - so I'll just riff a bit haha.
So it's true we got vaccines quickly, which is a point in favor of an Australia-type lockdown route (or better yet, early travel restrictions to put us on a better path). I've seen some discussion around 100-day vaccines in the future.
But at the time of lockdowns last year:
A. it was already endemic
B. Vaccines would be available by the end of the year were being called crazy!
Which is where I think if you believe A & B it gets harder to advocate for lockdowns - you're basically hoping you can be Australia instead of every other country that had an endemic outbreak and never got it back under control.
(one of the coauthors - D.A. Henderson - lead the smallpox eradication effort). They discuss a pandemic influenza with a CFR similar to COVID and a vaccine available within 6 months - and still do not advocate for widespread NPIs!
I'm neither arguing in favour of nor against masks, but you have to admit that citing whether overall cases in a country have increased or decreased is an insufficient proof of their efficacy either way.
I haven't worn a mask once in the last year, and I only remember one lockdown for like 2 days before it was called off after it turned out we didn't have some crazy variant that spread via fomite.
What worked, I assume, was the border controls. None of the other measures like masks and lockdowns and whatever else other countries are doing were necessary because there are no cases to begin with. International arrivals got quarantined in hotels before they were allowed into the community, that's essentially all that was needed.
Some states fucked up these hotel quarantine measures at times, which is where some of the brief lockdowns come from. Victoria I think was the only state to have lockdowns for an extended period of time rather than the brief one or two lasting less than a week after a hotel quarantine scare. But for most of the country, for the vast majority of the time, there are no masks, no lockdowns, no cases. Everything has been pre-pandemic normal.
Too rigfht. Always amazes me that people wave away the case of China, which could have been like India is now, with some airy wishcasting that the Chinese just lie about everything so they probably really did have 100 million cases and 1 million deaths and are just...really good at hiding the carnage. These people fit right in with the faked Apollo landing people, you cannot reason with them at any level.
But clearly anyone who makes a good faith assertion that public health measures could not have controlled the virus needs to explain China, where the virus *was* controlled, and limited to a small population and a tiny per capita mortality, in a very quick (if exceedingly brutal) amount of time.
>Pretty much all of Asia is laughing at that last line of yours.
As others have noted, "Asia" did real lockdowns in time to plausibly matter. Asia also seems to have done real contact tracing, which nobody in the West really managed. And Asia bought up pretty much the world's entire supply of Real Masks at the outset, leaving the rest of us with symbolic fake masks.
Asia was, if not physically immunized, societally and institutionally immunized by the SARS non-pandemic a decade or so earlier. Their local versions of the CDC were prepared to act promptly and effectively; ours weren't.
Maybe Asia should not count their chickens before they are hatched...At this point, it's difficult to say a restart is out of question: most of the circulating variants are quite different that the original wuhan strain, it seems the chinese vaccines are quite weak compared to the main western ones (and the russian one), and, in europe at least, countries/regions that got of easy during the first waves suffering more during the latest ones looks like a common pattern. common but far from systematic...Asian countries managed to stomp the first wave early, sometimes so early it did not happen at all...but some now have difficulties. Thailand for example had almost no cases early on, controlled the few case among returning workers and was cruising along happily since may 2020. But since last month, it seems they lost control. Japan seems similar to thailand, keep control early on for not-so-clear reasons, but seems more problematic this time. India also seems much higher than during the first wave, where western governments predicted a catastrophe than never really materialized...at the time, while now it looks like it's happening. The rest of SE asia is a mystery, no covid news at all. Very few cases during first wave, then I think there were a significant but not catastrophic waves in Malaisia and Indonesia, but lately I didn't hear anything, good or bad, in the region. For Myanmar it's expected, They have a far more pressing issue than covid, but the rest of SE asia? No news. So what will happen in China? not sure? Will we know about it? even less sure. I do not really trust covid news even locally, my covid impact is that I have lost my last few remaining bits of confidence in mainstream news, so reporting about China is basically fiction.
San Francisco, Portland, and Seattle have had systematically better results than Los Angeles, Houston, and Miami. I believe DC has also been on the better side, while New York and Boston have been hard to understand.
Looking at the state level obscures a lot of this.
Yeah, I'm not claiming it's easy to understand at the city level, or that anything immediately clear emerges! Just that we see more phenomena that are completely obscured when you lump "California" or "Texas" as one place, while St Louis gets divided between Missouri and Illinois.
Geographically, baselines are going to be all over the place. But if you look at lockdowns *that are actually obyed* , in a certain area, across time , they make a difference.
As an inhabitant of a city in Texas, let me assure you that the major Texas metroplexes did lock down, just largely without state-level support outside of the initial three months. They're not remotely locked down any more, though.
What do you mean by "lock down"? It's an incredibly vague term. Enforced masks (in what circumstances?)? Limited building occupancy? Forced certain businesses to close? Quarantined people testing positive? Etc.
I mean restaurants, barber shops, other small businesses closed. Grocery stores instituted quotas of how many people could go inside at one time. Anything else considered "essential" became curbside pickup. All office workers who weren't doing classified work had to work from home. Roads were empty. Downtown Dallas looked like a ghost town for nearly a year. At one point, my employer had to give me a note saying I was part of the Defense Industrial Base and need to be on-site for my job duties in case the police stopped me to ask why I was out driving.
I was never in San Francisco but can't imagine the restrictions there were ever any more severe.
Seems vaguely notable that every state that is worse than Texas has a Democrat governor. And Texas has done better than other big states like California and New York for many months.
In looking through Texas, my own area of DFW was not hit that hard, not nearly as bad as most other Texas metros. Unrelated to strength of the lockdown, I suspect that is because our biggest industries, defense, IT, financial services, outdoor construction, simply weren't hit hard or even flourishes, whereas oil and gas and shipping totally tanked and other Texas metros are heavily dependent on those industries.
A lot of states here moved quite a bit, most notably Hawaii. It's easy for people engaged in motivated reasoning to say "oh, they have a Democrat for governor," but it seems likely to me the reason for that is Hawaii's economy is almost entirely dependent on tourism and imports. Those tanked globally and have nothing to do with local policy. There is nothing any governor could have done to prevent that. It's also notable that California was hit so hard considering the very well-publicized and rather hypocritical exemptions given to the film and television industries to consider them "essential" and allow them to stay open. But most major productions stopped filming anyway, even though the government didn't force them to. Even then, the metro-level numbers show that farming regions were hardest hit. I don't see any speculation as to why, but apparently the research service of the USDA shows the values of crops and livestock over the first six months of 2020 was revised downward $29 billion compared to start of year estimates, which tanked farm incomes for the year: https://www.ers.usda.gov/webdocs/publications/100177/ap-086.pdf?v=9134.7
So you can do what is called a shift-share exercise, which is where you take a state's (e.g. Hawaii) industry composition, and say "suppose for each industry, the decline had been as big as the decline in that specific industry everywhere else in the country. How big would their total decline in employment have been?" And this allows you to compare what they actually experienced versus how the "average" place would have done given their industry composition. FWIW, Hawaii does have an industry composition that makes it particularly exposed. But it did much worse than the benchmark controlling for this.
Didn't Hawaii actually shut down, as in close its borders and not allow non-Hawaiian's to travel there? Trying to look up what the restrictions from the past year, it looks like it may have just been a required two-week quarantine and negative Covid test, but the unavailability of tests may have made that a de facto travel ban early on.
It seems to have had the desired effect in that they apparently have the lowest death rate of any US state: https://www.statista.com/statistics/1109011/coronavirus-covid19-death-rates-us-by-state/. But definitely not a great economic strategy for a state that so heavily relies on visitors for income. I guess we'll see what impact that 15% unemployment rate has. That is brutally high.
Texas has a middling deaths per capita level among US states. The high restriction states have fared no better. Many broken eggs. No omellete.
This is a pretty concerning level of disregard for human rights even disregarding the empirical lack of "necessary". The evidence is already in. Texas didn't do it, and it had no measurable consequence.
No; Texas's per capita COVID deaths are slightly below national average, while 4 out of the 4 top states in COVID deaths per capita are blue states (in 5th place is Mississippi, which has extremely Republican Whites and the highest Black percentage of any state in the country). Scott, read this description of anti-pandemic measures from 1911; they are much smarter than anything done in the U.S., and proof of a general decline in Western epidemiology during the 20th century:
In Alameda County, where Scott lives, people have been careful and it worked; there have been 1,486 Covid deaths among 1.71million people, which is about half the Texas death rate. If Texas had had Alameda County's death rate 25,000 more Texans would be alive.
Looking at state rather than county numbers will end up confusing you because the LA area of California did very badly while the Bay Area did extraordinarily well. But Scott is in the Bay, and the Bay had an extremely low death toll.
In Travis county there has been 847 deaths for 1.274million people. Austin had a local lockdown early on, but was quickly overridden by state mandates. But that's significantly better than Alameda county even.
The book is certainly interesting: there are obvious parallels between the Manchurian plague and COVID. Some paragraphs could easily have been written in the modern day. I took the book from the Internet Archive; the Google Books link didn't work for me.
There are some important differences between the Manchurian plague and COVID. For one, it was extremely deadly, killed quickly after infection (by another source, 2 - 5 days), and affected all bar the elderly and infants.
> Its specially marked feature was its unvarying fatality : 43,942 cases are recorded, and 43,942 deaths. There was no authenticated case of recovery.
> The strong and vigorous seemed as susceptible as the weak, and the infection often passed over the aged and the very young.
> Each morning we consulted together on the telephone, and on Tuesday he mentioned casually that he was not feeling up to the mark. I went straight to the station and found that he was feverish. [...] The disease ran its rapid course, and he died in little more than twenty-four hours.
> A man in a small hamlet came home ill from Moukden and died, his family attending on him and performing the usual rites. A few days later the entire household of seven died within twenty-four hours, except one infant who was found wailing beside the dead mother.
For two, it was more obvious whether or not one had it in the later stages (and I think it was only infectious in the later stages, though obviously it's difficult to tell):
> In the evening the unmistakable Plague sign appeared, the bloody spit.
> They were helped by the fact that pneumonic plague is usually easy to recognize, once it has declared itself.
Moving on to the actions taken, I'm not sure what you're referring to precisely. It seems to me most of the success from dealing with this plague came from widespread support from the populace, who went above and beyond to protect themselves because the plague was so extremely deadly, and secondly from the weather, which prevented people from travelling who may otherwise have done so.
First, they locked down the railways where they could. This seems similar to the travel bans. Not all travel was banned, but there was much less of it back then, so presumably it was easier.
Second, they moved infected people (and those in they were with) into quarantine centres (protected by the military). These appeared quite good quality, and people were paid back for their time:
> A warm kang was provided, and plenty of good food, and there was no need to work. The members of each household were encouraged to keep by themselves, and when they returned home after their ten days' holiday, they found that their houses had been well guarded, and that they received full compensation for anything burned by the police.
This seems similar to the requests to shelter in place, and schemes like UK's furlough or Germany's Kurzarbeit. I don't know why we've moved away from centralised quarantine locations (even the Chinese secured people in their homes instead of sending them elsewhere) -- perhaps the logistics are simply too hard with the much larger modern day populations.
Third, they burned the possessions of the dead and compensated the families. This isn't something we do any more, though I don't know why. Perhaps we can clean more thoroughly? Perhaps it never had that much of an effect?
Fourth, they vaccinated the doctors and gave them protective equipment. I think we did that too, before moving on to vaccinate everybody else. It's difficult to tell what exactly the vaccine involved:
> He had been inoculated against Plague, was closely masked, and, as we thought, well protected from infection
It might have been Yersin's serum, designed for the bubonic plague.
Fifth, they sent people around to each house to inspect people for plague and take them to the quarantine centres. This was made significantly easier as the effects of plague were more obvious and could be detected even by untrained people, while COVID requires a test (though I think South Korea had mandatory frequent testing). They did the same thing for businesses, which was not done anywhere I'm aware of:
> When a shop was forcibly dosed and disinfected, and twenty-nine persons removed from it to an isolation station because of the death of a thirtieth, the merchants were highly incensed. The co-operation of the general public could thus hardly be expected.
> One day a foreigner saw a sudden stir and excitement in a restaurant on a main street; a man had fallen down ill, and quickly became unconscious. The police were called, and within two hours all was complete, the premises empty, disinfected, and closed, with a cordon of soldiers round them. When the day's work was over each member of the Plague staff visited a disinfecting station, where he had a bath and left his outer garments to be disinfected.
Sixth, they used propaganda: they sent placards around telling (possibly exaggerated, possibly true?) stories on how the plague could spread and what to avoid. This seems similar to the modern day briefings / news.
Seventh, they demanded people not visit their friends for Chinese New Years. This is our lockdown. This was helped by the weather: it snowed a lot, so people couldn't travel easily.
People reacted much as we see today: with doubt, conspiracy theories, and attempts to evade the rules:
> At the beginning there was general disbelief in the necessity or usefulness of preventive measures. It was an absolute novelty to the Chinese mind to attempt to check the spread of any infection, and apathy naturally accompanied their fatalism. "This is the scourge of Heaven," said many. "All will die whose time has come, and no others. Then why take people away to isolation stations? Why burn good clothes and bedding?"
> The Japanese were credited with encouraging or even causing the epidemic in order to destroy the people and possess the land.
> It was said that everyone who went [to the isolation camps] would die, that people were sent from them to the Plague hospital who had not Plague, and that some were buried before they were dead
> Day by day men saw their neighbours fall by their side; in five days seventy died. Panic seized the remainder, the military cordon was not very strict, and a number escaped one night, carrying infection into the city.
> It was some time before opposition to Government measures altogether ceased. The most serious resistance was on the part of some merchants, who determined that their business should not be interfered with.
The plague defense was helped by the plague's extreme deadliness: people kept to the lockdown and quarantined themselves as much as possible. Visitors from outside were rejected. Goods were bought in bulk and people kept to themselves. The plague of Manchuria lasted three months; here we've locked down for over a year with more to come. Lockdown opposition isn't obvious in those it kills: if going to a party without a mask lead to the deaths of everybody at the party and their entire households, I'd expect more people to follow the rules.
"Seventh, they demanded people not visit their friends for Chinese New Years. This is our lockdown. This was helped by the weather: it snowed a lot, so people couldn't travel easily."
This might be a Chinese, Australian, or Thai lockdown, but nothing like this happened in America; thousands moved out of New York City during the 2020 stay at home orders without consequence. The more obvious symptoms of the Manchurian plague were more than compensated by the greater availability of PCR testing today (the most ambitious users of it being China, New Zealand, and Australia), which also enables governments today to test sewers for signs of COVID -something the Hong Kong government did.
"here"
-Not in the vast majority of the U.S., where the stay at home orders were over by June.
South Korea is actually fairly unambitious at testing; it has far lower tests per capita than the U.S., and certainly hasn't tried anything like China's testing of whole cities. It did have harsh penalties for breaking quarantine for COVID cases and contacts, which I don't believe the U.S., Germany, or Britain ever had.
Okay fair, I'm in the UK and I don't really know what's been going on in the US. It hasn't been constant here, so "over a year" is probably overstating it. We were locked down thoroughly from December through March. We initially shut down in April, opened up briefly with a disastrous "Eat out to help out" scheme to encourage spending, then shut down again, blaming everybody who followed the scheme for increasing infections.
The UK had fines (up to £10k if you're both caught and can't convince the officer to let you off), but in the Manchurian plague the penalty for breaking quarantine was that you and your entire household caught the plague and died, which is definitely a lot more serious.
What were the "smart anti-pandemic measures" you were referring to? Thorough, frequent testing? Centralised quarantine? Complete transport lockdown?
With all due respect to both of you, the word "necessary," like the related "need," is a way of avoiding rational thought on such issues. Lots of people didn't lock down and most of them are still alive, so it wasn't "necessary." The relevant question is whether it was desirable.
And we don't really know. The alternative of strictly quarantining the vulnerable and letting the vaccine run through the rest of the population until it burned itself out wasn't tried, and we don't know whether it would have killed more or fewer people. It would almost certainly have imposed less of other sorts of costs.
Well, it wasn't necessary for *you* because you're still here. In the same way, seatbelts aren't necessary for anyone who isn't involved in a high-speed car crash, and defibrillators in ERs are a complete waste of money for all the many millions of people every day who don't have a heart attack.
...on the other hand, if we ask the families of some of the 50,250 Texans who died of COVID in the past year, they might differ in their opinion.
A ban on driving for non-essential purposes wasn't necessary for *you* because you're still here. On the other hand, if we ask the families of some of the 40,000 Americans who died in auto accidents in the 2019, they might differ in their opinion.
I do the podcast - I'm going to do a very simple version of this now by myself but it would be so fun to do it properly :D Send me an email if you like! :) astral codex podcast at geemail.
I sure hope this is poetic license, because drinking from a random pool of water will likely kill you in all kinds of interesting non-Covid-related ways :-)
At least in the wilderness in developed countries, I think the risk tends to be somewhat exaggerated - giardia, for instance, is actually quite rare if you take reasonable precautions (drink from fast-moving clear streams, favoring higher elevation). See e.g. sci-hub.st/10.1016/s1201-9712(00)90102-4. "Drink from random pools of water" seems to work pretty well for most species, and for humans for most of history.
For what it's worth, both me (as a child) and my dog (later in life) have had giardia, and though it really sucks, it's extremely unlikely to kill you. Hell, CDC seems to estimate 280 million people a year are infected (worldwide, so not nearly all of the US).
Clearly, though, yeah, stick with running water in most places and you're probably safe even from giardia. It's a somewhat interesting instinct to see that my cat won't drink water unless she watches me pour it. She otherwise goes up to the bowl and meows at it until I replace the stagnant water with new water.
It's funny, the one time I took a drink from a mountain stream (okay, okay, second time) I got giardia. We didn't pack enough water for a long alpine hike on a sunny day, and those mountain streams look so pleasant...probably wasn't deep enough. I can confirm the giardia, while annoying, is not so bad if you're in good health.
So I read through that article and it doesn't seem very convincing. It is basically arguing that there is very little evidence in the literature for outbreaks of giardia from drinking back country water. I believe that this is probably true but a) the vast majority of people who are in a situation to drink such water know not to unless necessary and b)As others have pointed out, it's not really that serious of a disease (very unpleasent but unlikely to be life threatening) so I would be very surprised if most cases ever got reported.
They are pretty much all higher than I am comfortable with.
Yes, _some_ streams in some areas are likely ok and not all waters are equally risky. But I think the risk of getting _some_ kind of disease, from unpleasant diarrhea to debilitating illness, is very non-negligible. I think I'll keep bringing my katydid with me.
judging by the number of my friends who say the vaccine means they can lick [eyeballs/subway polls/toilet seats, depending on who i talk to], I'm guessing it's a tongue-in-cheek participation in "the vaccine makes me invincible" jokes
Also, I recently* drank from a giant** pool/river of water that formed during a terrific rainstorm. It was great. Do I need to worry about any health risks?
*less than a month ago
**about 5 feet(?) wide, 100 feet(??) long, and a few inches deep
How much time delay did the FDA actually cause? Wasn't a lot of the time necessary for figuring out a formulation of the vaccine that would deliver the mRNA in an effective way, and working out manufacturing and other logistical concerns?
Yeah, I don't buy into all the FDA hate. Sure, they could've acted faster than they did, but that's not the same as saying they shouldn't exist at all. When I take a vaccine, I want to know that it's effective (and ideally, how effective), and that it is unlikely to kill me all on its own. Without the FDA, how would I get this information ? From the New York Times ?
From pharmacies like CVS, Walgreen's, Rite-Aid, etc. They have an extremely strong incentive to get it right. The FDA has a strong incentive to not get it wrong, which are not the same things.
Agreed on the distinction between right/not-wrong, but how do pharmacies have any strong incentive to get it right? They make their money from selling drugs to people, not curing them. Obviously killing all your customers is bad for business, but there's plenty of profit to be made selling placebos, or for that matter drugs that kill only a small percentage of customers.
If that were even remotely true, the FDA would not exist in the first place, because the history of 19th century drugs would've been one of serene effectiveness, instead of the shop of horrors it actually was, and which drove the outcry that resulted in the creation of the FDA.
Is this the same FDA which allows people to eat industrial lubricating oils (PUFAs) as if they were food?
Yes, yes, it is.
I'll put my trust in the power of competition. We have much easier access to much better information about drugs now. Can you possibly agree that times change, and the need for government regulation changes, and that government bureaucrats don't want to shut down their bureau?
CVS and co demonstrably do not care about efficacy. Good news, though, they give free shipping on their homeopathic medicine if you order at least $35. Even better, they're "CLINICALLY PROVEN."
Yeah sorry. Anyone who calls polyunsaturated fats "industrial lubricating oils" has such a deficient understanding of basic chemistry that means his opinion on this subject is worthless.
Hey, did you hear the FDA allows hydrogen oxide as the first (most plentiful) ingredient in *childrens' shampoo??* An outrage!
Yes, that it was. But the general state of human nature, credulity, and tulip-bulb memetic mania hasn't changed a bit.
That said, I personally am much in favor of the FDA assuming a purely advisory role. That is, it should simply state for the record, in language accessible to anyone who paid close attention in high-school chemistry class, whether food or consumer product ingredient X, or medicine Y, is/is not safe, and for whatever reasons. To additionally restore a little budgetary rationality and enhance its support of Darwinian natural selection, it should charge a very modest consumer fee (like $100/year for a subscription) for public access to this info.
Because that would restore a little of the natural evolutionary advantage of intelligence, prudence, and due diligence.
No they should not. The potential *future* victims completely dwarf the number of present victims, and there is no respectable collective moral basis for sacrificing their welfare -- which is critically affected by whether we gather reliable data on vaccines and treatments today -- for the welfare of the much smaller number of current suffers.
We also had data relatively early on that the vaccines weren't killing people like COVID was. The time spent in trials meant more precise numbers on how effective it was (particularly as the epidemic increased the number of cases in the control group, which wouldn't have been necessary had we done human challenge trials), but we really only needed a rough idea of efficacy in order for it to good idea to prioritize distributing the vaccine:
Dual-track is worthless. There's no way of ensuring the control and experimental groups are sufficiently similar in what amounts to an open-label study with self-selection. That is especially true in something where the inherent risk is already painfully (from the point of statistics) low.
The fact that precision in efficacy was necessary should have been amply demonstrated to you by the recent J&J vaccine hold. How are we to *know* whether a low number in a million risk of severe blood clots does or does not outweight the benefit of the vaccine, if we don't know the benefit precisely enough? The *only* reason they were able to rapidly resume using that vaccine is because the clinical trial data was precise enough about the benefit. And this kind of thing is not rare, it happens *all the time* in new therapies. They didn't just pull those requirements out of their ass, you know -- they are the results of a century of painful experience.
Certainly a challenge study would've gotten the data much faster. If you want to write your Congressman and suggest enabling legislation that would fully immunize Big Pharma from liability should they produce a thalidomide-sized screwup in the process, go right ahead. Until then, this is magical thinking out of contact with the real world of legal liability in which they dwell. Vaccine developers do not crave FDA approval in the altruistic interests of their future customers, they do it so that when the inevitable $1 billion lawsuits arise because winsome photogenic 3-year-old Susie became autistic 18 hours after vaccination, or possibly 12 hours before, hard to be fully sure, and besides why does it matter? -- they have more than a fig leaf heading into the pre-trial settlement talks.
The data relevant to potential *future* victims, can be as easily gathered It looks to me like you're suggesting that we need to hold the present victims hostage, because you can't imagine any other way to get the tests and data you want except by saying "no vaccine for you until you've done your chores!"
That's correct. I don't know of any other way to get the critical data than the tedious procedures laid down by empirical science. If you do, you should publish. An instant Nobel in medicine awaits you.
No, no, you mistake me. I'm fully in favor of freedom of choice. In fact, while I think the FDA should still required these tests, and continue their effective taxation on new therapies to pay for them, I think their conclusions should be merely advisory. I am 100% in favor of fools rushing in where angels fear to treat being allowed to exercise their personal liberty, as this has a decent chance of restoring the health of the Social Security trust fund by lowering the number of fools who reach age 65.
When Pfizer announced the results of their stage 3 trial, I skimmed through the report they released. I do statistics for a living (almost purely frequentist, not Bayesian, their report used a Bayesian method but something simple enough that I had learned it in a first year PhD class). Unless they had completely faked their data (not sure the FDA is set up to catch this in any case), then yeah, any of the thousands of people with this level of training would have been able to tell you that "yeah, it's highly effective". More trustworthy than the NYT I would think.
And really, we can almost ignore the statistical training - when you have 90 cases in the placebo group, and 5 in the control, you don't need to know how to construct confidence intervals to be able to correctly infer that "huh, seems to help a lot".
The FDA then took something like 3 more weeks to read through this report and give Pfizer the green light, so some pretty low hanging fruit I would think. Could have been done in one afternoon.
Plus they absolutely did need to test the vaccines thoroughly. Lots of potential vaccines don't work, and using mRNA for vaccines is obviously not something that had ever been done before.
The FDA certainly did drag their feet more than they should have! Challenge trials could potentially have done a lot, and every extra week added to the timeline killed people! But no, we couldn't have just started giving out vaccines a year ago.
No, lots of time wasn't necessary for figuring out the formulation - IIUC they had more or less the current formulation around the time COVID cases first started being reported in the US.
I don't know how the manufacturing supply chain worked, but I do know from other medications that most of the difficulty of manufacturing drugs is getting the factories FDA-approved. It looks like Moderna's first factory was approved by the FDA in December, but I don't know enough of the backstory to know if this was good or bad.
Like I said in my comment below, I think there's a lot to be said for having vaccines manufactured in a factory that was inspected by someone more competent than, say, a New York Times reporter. Yes, this process does take time; but it's a tradeoff. What percentage of people who take a new vaccine should die from it ? The answer is not as simple as "as long as fewer people die from vaccine than from the virus, we're all good" -- witness what happened with AstraZeneca, for example.
> The answer is not as simple as "as long as fewer people die from vaccine than from the virus, we're all good" -- witness what happened with AstraZeneca, for example.
Yes, that IS the answer! Or maybe you just answer trolley problems differently and are never willing to 'act' if it kills people.
I'd think we'd always want to choose the option (including NOT 'acting' as an always available option) that kills less people, but hey, what do I know – maybe death is really good?
And "what happened with AstraZeneca" is what we're arguing AGAINST. It seems circular to use that as evidence against not doing that.
No, that's completely the wrong answer. The right standard is "fewer people die from vaccine than are saved by vaccine" - otherwise, I could give hundreds of thousands of people bottles of strychnine with vaccine written on them, and claim it was OK because coronavirus was killing even more!
With that said, if you want to minimise the sum of an increasing function x and a decreasing function y then setting x=y as you're doing will never be worse than the best outcome by a factor of 2.
And the number of people killed by the vaccine is probably in 1 or 2 figures, and the number saved is probably in... I don't actually know, but I'm going to guess 6, although high 5 wouldn't surprise me, so in practice minimise total deaths from vaccine+covid is basically an identical problem to minimising deaths from covid.
If you gave hundreds of thousands of people strychnine, they'd almost all die – that seems WAY more dangerous than COVID-19.
All-else-equal – we want to prevent as many deaths as possible.
> With that said, if you want to minimise the sum of an increasing function x and a decreasing function y then setting x=y as you're doing will never be worse than the best outcome by a factor of 2.
I don't understand how you interpreted what I wrote as being equivalent to "setting x=y". What's the "increasing function x" and "decreasing function y"? What are the domains and ranges of these functions?
My point is the we want to minimize 'deaths from COVID-19' AND 'deaths from vaccine'. But – in this world – the latter seems FAR smaller than the former, hence anything we've done to delay the latter has increased the former.
You are too focused on what's in front of the trolley to see what lies a little further down the track. Like I said, look at what happened with AstraZeneca. By all accounts, it is almost perfectly safe; however, just a few highly-publicized deaths were enough to get it effectively shut down in most places. If the probability of dying from it was even as much as 1% (as opposed to 0.001% or whatever it is now), most people would rather take their chances with the virus; what's worse, many of those people would be *correct* to do so. If you are young and healthy, COVID is not that big of a deal, so why risk blood clots ?
But the situation is even worse than that, because, absent FDA, a person couldn't even make an informed decision as per the above. Right now, the proposition is, "take this vaccine, and you will have a 90% immunity, with a 0.001% chance of death, and a 95% chance of mild side effects". Without the FDA, the proposition becomes, "take this vaccine, and it may or may not help or kill you or whatever, who knows, we are piping it directly from the experimental chemical tank into your veins". In this situation, most people would wait for the New York Times reporters to write some articles; and again, they would be arguably *correct* to do so (depending on how much you trust journalists, and personally I do not). Again, most people would skip the vaccination, and would be correct to do so.
The bloodclots were only even seen in young women. The people with the highest mortality have been elderly men. They should have had access to AZ long ago.
I think your model is that 'the public' panicked over AZ whereas mine is that 'the authorities' are the ones that panicked (supposedly) _on behalf of_ the public.
The problem is not that the FDA gathers data and indicates that things are safe and effective, the problem is that it prevents people from taking medicines that they want to take. I have no doubt whether the FDA exists or not, there will be a demand for rigorous evaluation that the market/society will provide.
The FDA certainly deserves a massive share of the blame for ruining rapid testing, which might've been able to massively cut case counts from what I've heard (though not sure why it didn't really help Europe, maybe they had similar regulations). But I would be very interested in a full accounting of how much the FDA actually held up vaccine manufacturing. I'd assume the Moderna factory was already producing vaccines at peak capacity before it was approved, right?
Shockingly wrong, actually. I had assumed that Moderna at least started to ramp up production a few months in advance of approval, but it really looks like they didn't turn the dial up to 11 until the same week they sent their final approval package to the FDA.
We were talking on the old blog about how Bill Gates was going to fund building all the factories for the vaccines, even those that didn't work, so the winners would all be ready to go.
And then . . . nothing happened. No one did it. I don't know if Bill Gates actually promised something and quit or it was just a rumor or he wasn't able to do it, but one of the easiest ways to trade dollars for the end the pandemic was just shrugged off.
Yeah, I also haven't followed what happened with the Gates funding efforts. To be clearer, I guess the question is: how much was the factory holding back on production due to uncertainty about FDA approval, vs. just facing material constraints in scaling up the production process?
The upshot is very few of these factories exist, and the companies that can have been scaling their manufacturing capability for some time now. I think the fact manufacturing / distribution still took months to scale up after approval shows that FDA requirements weren't the only bottleneck.
In hindsight, it was entirely bad. We took forever to test and approve production/use of a drug that turned out to be safe and effective. Once you get out of hindsight, it's harder.
I hold that in a world where we aren't all cowed into thinking this should take forever and that an absolutely-overkill and absolutely-overcautious testing regime aren't normal, we could have been in semi-serious production by May or so and escalated from there.
Whenever I say this the normal response I get is something like (incoming strawman) "Well, even though it Covid killed many millions, there's some small chance it might have killed some very small-in-comparison number of people in testing if it had been much more dangerous than it plausibly could have been, and those 1-1000 lives are much more important than the hundreds of thousands or millions who died otherwise", which reinforces my view. And we are totally prepped to take exactly as long next time; if we were willing to sacrifice 2 million people to save 50 there's no reason we wouldn't sacrifice 20 million.
> We took forever to test and approve production/use of a drug that turned out to be safe and effective. Once you get out of hindsight, it's harder.
You're looking at less than half of the problem. How many vaccines turned out to not be safe and effective? How many neither-safe-nor-effective vaccines would've been manufactured, marketed, and sold if they knew before hand that they didn't have to go through the FDA approval process?
You're looking at the holes in the planes that come back and deciding that it's useless to invest in armor on the places without holes.
First thing I should bring up is that you are pretending I said to abandon all testing rather than suggest that the testing isn't some sacred cow that's immune to improvements, impossible to speed up. That's a bullshit thing to do; you do it knowingly and that makes you sort of a bullshit person.
Moving on: Some non-strawman version of your metaphor is valid if a bunch of assumptions it makes are true - i.e. that the armor plating you are saying I'm abandoning is all necessary, and delaying the deployment of the plane isn't balanced against any other costs.
In the case of efficacy testing, the risk with faster testing isn't a risk of less efficacy - we get the same result (works or doesn't) with about the same certainty in a much shorter time with, say, challenge trials. But the decision was made that a risk to a small amount of people (say, 1000 deaths in some kind of worst-of-all-worlds testing disaster if I'm being really, really generous) was not ethical, while condemning hundreds of thousands of people to for-sure deaths, miserable and choking to death slowly on their own lungs, was ethical as shit and something we should definitely do.
I want to be gentle here, but I can't: You are a fine with this and it's going to happen again. This time it was 2-3 million, probably a quarter to half of which we could have saved at a really minimal risk to informed volunteers.
To flip your metaphor, this is you looking at a plane that could go intercept a nuke headed towards NYC; it's ready to go and the pilots are begging to be allowed to save those people, and everyone is aware that this would probably work, but you want and demand a 2 hour prayer to be chanted over the wings at no advantage because that's how it's always been done and you refuse to consider anything else.
I remember being exposed to the "notice when you are confused" framework back on SSC.
I'm not sure I personally am confused, per se, but as a society, we really should put some thought into how we found ourselves in a situation where the people anointed as "ethics experts" were seemingly all bar unanimously opposed to attempting to save millions of lives by letting a few thousand people voluntarily face a <1% risk of death, because we've always done it that way.
If I were designing the attributes needed to be endowed as such an "ethics expert" from the ground up, this would be a disqualifying, not a necessary characteristic.
The original claim in the verse was, "Two days to find the key
The rest, because bureaucracy."
Bugmaster claimed, "I think there's a lot to be said for having vaccines manufactured in a factory that was inspected by someone more competent than, say, a New York Times reporter. Yes, this process does take time; but it's a tradeoff."
You responded, "In hindsight, it was entirely bad."
"It" is a bit ambiguous here, but it certainly looks like you're arguing in support of the claim in the original verse. This is what I was disagreeing with.
The narrower claim that, "There are other options we could've used to rigorously test these candidate vaccines that would've been quicker without sacrificing the FDA's credibility or flooding the market with vaccines of dubious quality" seems obviously true, but, per Bugmaster, it's still a tradeoff.
On the contrary, if you got all 50 candidates, and did challenge trials immediately, you could have results on efficacy for all 50 options within a few weeks. Of course, we couldn't have that, because apparently it isn't acceptable for people to voluntarily and knowingly take a fairly small risk, even though doing so would likely save millions of lives, because to allow this would prevent medical ethicists from feeling smug (or some such - obviously this is uncharitable, but one has to wonder where self-appointed "ethicists" acquire there "i have the right to insist on killing huge numbers of people on personal whim" card). Notably, polling of the general public suggests they would have been ok with challenge trials!
Even without the trials, a policy of "each potential vaccine spit out doses and start vaccinating without delay, we'll record data to make clear what works as we go" would have saved lots of lives. In states like NJ, 10% of the nursing home population died. Did any of the 50 candidates kill 10% of the people in their respective trial? Has any vaccine ever in human history killed 10% of the people it was given to? There are trillion dollar bills on the ground here.
And it's not like incentives to prevent quack medicine are FDA or bust. If we didn't have the FDA, what would stop a company from putting arsenic in a vial and marketing it as an effective covid vaccine? Well, the criminal code, for one (fraud, for example, being illegal).
>You're looking at less than half of the problem. How many vaccines turned out to not be safe and effective?
Lots of vaccines turned out to be only moderately effective. Even a moderately effective vaccine six months ago, would have saved more lives than our perfect vaccines will now. I don't think any vaccine ever has turned out to be unsafe in a way that would be even a rounding error compared to COVID deaths.
But, more importantly, it doesn't take a year to figure out whether a vaccine is safe and effective. If the metric is "better by an order of magnitude or two than the status quo", we probably knew that by June 2020, and we certainly could have known it by June 2020.
The Pfizer phase 2/3 clinical trial only started in July 27th, so we certainly didn't know by June 2020. They had 162 cases in the placebo group when the stopped and that's certainly more than necessary even for a modest effect size. But how much faster would it have been to get just, say 60 cases? I'm guessing two months faster: you still need to recruit them, give them two doses vaccine, and then wait to see if they get COVID. COVID case numbers were growing rapidly into the fall, so I imagine they got most of the cases towards the end of the study. Even with the count they had, they just barely exclude 90% efficacy, so just barely making it past your "order of magnitude" metric. And all that is ignoring the possibility of examining any subgroups.
That said, I'm all for creating better systems so we're more prepared the next time something like this happens. Plenty of room for improvement, but the vaccine development, testing, and roll-out has been pretty reasonable by my judgement.
You don't want moderately effective vaccines. That puts strong selective pressure on COVID for escaping such immune responses and creates the perfect environment for cultivating and propagating such strains.
It's the same reason why you want people to do their full antibiotic treatment course - because if you have some strains that are marginally resistant, they will be wiped out by the full course but survive the partial course and then reproduce and if you do that several times over you can end up with antibiotic resistant bacteria because each time you selected for the strains that could resist a bit better, and with the weaker challenge there were actually survivors rather than total annihilation.
In the end, the US went with three - Moderna, Pfizer, and J&J. Sure, Moderna's vaccine was developed in two days, but maybe one of the failed 50 vaccines was developed in one day. We're not omniscient, and we tried many vaccines because we weren't sure which would work.
I mean, yes, it does quite a bit. We have at least 5 vaccines that work (Pfizer, Moderna, J&J, AstraZeneca, Novovax) plus several more that probably do but are harder for western countries to verify (Chinese + Russian + Indian vaccines). How many actually failed outright and were abandoned (rather than merely being given up on because development was going too slowly such that they would be uselessly far behind)? Plenty of others at still at various stages of development (unclear if they will go anywhere, because we have at least 5 high quality vaccines at this point).
Overall, the success to failure ratio here is quite high!
(Also, if you have 40 candidates, and each has a low chance of working, challenge trials are exactly the sort of thing that would be helpful).
Peter Hotez has been pushing the vaccine his group has been developing on the basis that it's much cheaper than the vaccines currently in use. IIRC he gave $1.50 as the cost per dose they're aiming for. Just looking quickly, it appears that COVID vaccines have the same lack of transparency in pricing that's endemic to the pharmaceutical industry generally, but that price point would be considerably below all the figures quoted here.
My understanding is that the Russian vaccine works much better than the Chinese one. Even the Chinese government doesn't have much confidence in their own vaccine.
It isn't actually survivorship bias if you have 50 candidates start, and 8 are successful and start being used while the other 42 are still going through the process. The relevant thing is successes vs failures (and it also depends on what the failures were - if they are just "abundance of caution" and "seems less effective than other candidates that are further developed" then it is questionable to count them as failures at all).
If they're still in Phase 1 or Phase 2 at this point, then they probably needed to re-do parts of them. They didn't get the initial dose levels quite right, there was some unexpected difficulty, etc. So they had to re-do the Phase with a different formulation. Same is true if their Phase 3 was delayed.
Looking at the handful of manufacturers that jumped through the hoops right the first time and decided that this means the hoops don't matter is a form of survivorship bias.
> IIUC they had more or less the current formulation around the time COVID cases first started being reported in the US.
This is true as I understand it, but how many other people claimed they had cures for COVID at the time? IIRC, there were dozens of vaccine prospects last summer. (This was one reason I was optimistic about getting a vaccine much faster than historic standards indicated was possible; there just seemed to be so many potential options!)
Point being that you need to weed out the vaccines that work from the ones that don't, and you have to show that they're safe. The rest wasn't bureaucracy. And even if you say, "Human challenge trials", well, those still take some time. So now you're quibbling about how quickly the bureaucracy went, not whether it serves a necessary function.
Is the vaccine manufacturer really going to market a vaccine they haven't tested? Do you think that would be a profitable proposition for that company?
Not Pfizer or J&J or whoever, but once you decide that we don't need to test anything and everyone's proposing "novel" or "experimental" development regimes, "vaccine manufacturer" quickly becomes "anyone with a needle and some essential oils".
My point is that the FDA had a crucial role to play in giving assurances to the public that the things it says work actually work. You can argue that they didn't balance the risks optimally in this case (I agree!) but that's different from saying that all the clinical trails were needless bureaucracy because the final Pfizer (or whatever) vaccine was very close to the original formulation that scientists had created last February.
We didn't know if any given vaccine would actually work.
If drug development was this simple, we wouldn't see most drug candidates fail out of testing.
Numerous vaccines have been tried over the years. We still don't have one that works against a number of diseases (including HIV).
Merck made two vaccine candidates against COVID - V590 and V591 - and neither panned out, as they failed to give sufficient immune response. So it's not like we knew for sure that this stuff would work beforehand.
We're having enough problems with vaccine reluctance as is. If you gave twenty million people shots that didn't work, it'd be much worse - not just for COVID but future vaccines in general.
Properly testing the vaccine and making sure it actually worked before giving it to everyone was absolutely the right thing to do.
It's also worth noting that giving a bunch of people subpar vaccines that only gave a marginal immune response would select for COVID strains that could escape weak immune responses, which could make future vaccination efforts more difficult as well and potentially make the disease more dangerous in general.
Well, essentially all of it, from one point of view. Almost all the time was the time needed to allow volunteers who got the vaccine as part of the Phase II and III trials -- mandated by the FDA before approval -- to go about their business and either (1) die of unanticipated side-effects, or (2) catch and die of COVID because the vaccine didn't work, or (3) *not* catch COVID and not die while their compatriots in the control arm are -- which is what tells us it works.
We need a lot of time because it is considered unethical in these circumstances to give half the volunteers the vaccine and then deliberately expose *all* of the volunteers to COVID and see whether fewer people in the control arm die, which would be much faster. We also can't just wait a certain amount of time, because we don't know the rate at which people are being exposed. So we have to wait until we accumulate enough cases in the control arm that we can be reasonably confident everyone (control and experimental) has been exposed.
We certainly could have had the vaccine a lot sooner if we didn't insist on knowing whether it worked ahead of time.
The FDA did a good job with the vaccines. Knowing that they are safe and effective is important. Giving people a bunch of vaccines that didn't work would have been a disaster for vaccine acceptance. Giving people a bunch of vaccines that killed a bunch of people or made a bunch of people would have likewise had a large negative impact.
While pausing the J&J vaccine was dumb, spending the time to get the vaccines right was a good thing, both for the current pandemic and for long-term vaccine acceptance.
Long time reader - first time commenter. This is the best. I love this. I don't even know the song and I love it. Thanks for the nuance and articulation.
It seems like there's a new divide growing between the "Vaccinate but change nothing yet" segment and the "I'm vaccinated now; I'm going to live my life again," segment. That may well map to altruistic vs. selfish — cynically, the latter type were only pro-lockdown to save their own skins — but there's also a trust dimension: a lot of people, and I'm one of them, think the CDC's recommendations aren't being made rationally.
Then there's the plain feeling that "I've done everything I was told I needed to do, and they moved the goalposts YET AGAIN. No. No more." Not a great argument, granted, but it may prove a major obstacle to compliance.
Yeah, the constant moving of goalposts is NUTS, and it's worsened by the (I think very corporate-informed!) refusal to ever acknowledged that what happened before was wrong or mis-guided and an insistence on the message of "oh you must have not understood properly before, you idiot, this is why we need to big brother you to death with more regulations"
I think this is backwards. 'Corporations' are a relatively recent social technology, but I'd bet at long odds that that same 'bullshit' has been going for basically ever. (I mean, have you ever read The Bible?)
But this is a particularly maddening political tactic/strategy that I wish would be widely punished :)
<rant target="them">Admit your fucking mistakes people! Everyone else would trust you _more_ if you didn't try to gaslight us all the time.</rant>
My viewpoint (and I know others with this opinion) is that I was last in line for the vaccine. At this point, if I’m fully vaccinated, everyone else I come into contact with could be too. If they decided not to be, that’s on them.
Yes, there is some risk still from COVID, but it’s much less so now. My odds of killing someone who got the vaccine are pretty low at this point, so I’ll accept those odds and live my life.
So I’m a mildly bad person, but I feel like the dichotomy you posed isn’t accurate, as my reasoning is in the middle of “I still must not hurt people” and “I’m safe now whoopee”
Anyone who has commentary on this stance, I welcome it.
My commentary: it sucks that you feel the need to say you're a mildly bad person! I'm, firmly in your camp.
Small data point: my wife is having a daughter soon, and my sister asked if she'd be allowed to be around the baby if she wasn't vaccinated - I said "no" not because I actually give a sht, but because if you need a reason to get the vaccine and I'm in the position to give you one, then I'm going to do that.
I’m a utilitarian, so unless I’m living in a box and donating all of my money to the poor, I qualify as a bad person by my own moral scheme. Ah the cursed existence
But you're a person, hence you 'deserve' utility, your behavior is constrained by your nature (now, so including the history of any nurture), and obviously guilt and shame haven't made you a _better_ person, so maybe there's some other means for you to be a 'good enough' person (or maybe – what I suspect! – you _already are_ good enough).
Utilitarianism doesn't imply that _everyone_ should be a utility maximizer – or, at least practically, it seems silly to just _assume_ that transforming people as they are into utility maximizers isn't also just another thing to run thru the 'how to maximize utility?' function we'd like to approximate.
I kind of agree and disagree with you. I think that almost everyone ends up as a bad person (or at least an insufficiently good one) in their moral systems. I mean even seriously fundamental christians generally feel like they break some of the rules occasionally.
But I don’t think that means we should feel guilty per se. I see “being a better person” as a good, just like apples or reading. It costs things to be a better person, but it’s not the only thing we are optimizing against.
I do agree that not considering yourself in your utility function is an easy trap, but I do think I’m probably past the point where in many ways it would generate more utils if I gave my stuff away.
But who lives at the edge cases of their moral code?
" I mean even seriously fundamental christians generally feel like they break some of the rules occasionally."
Just to be pedantic: as someone raised at least fundamentalist adjacent, they're more likely to feel like they break some of the rules constantly, rather than occasionally. All have sinned, and all that. As I kid I understood there were obvious sins, like lying or stealing, but also had a sense that I was probably sinning constantly anyway and was just unaware of it.
To be clear...I don't. I think I'm a good person. My moral system is screw over other people as little as possible via direct action. Everything else is permissible. Be kind is gravy (I am) but isn't a moral imperative.
I might not be a good person in other people's moral systems, but they can go eat a d*ck.
> I think that almost everyone ends up as a bad person (or at least an insufficiently good one) in their moral systems.
Oof – that's not my view or belief. I think most people are (overall) good people.
> It costs things to be a better person, but it’s not the only thing we are optimizing against.
That's a good point. I think it's true too.
> But who lives at the edge cases of their moral code?
Some people sure seem _closer_ to the 'edges' of their morality.
But I think 'maximizing' is a trap too – 'maximum' is _better_ but not mandatory. We should _aim_ at it but not beat ourselves up for falling short – unless beating ourselves up (in some way, and some amount) is actually effective at making ourselves better people.
I admit it's all tricky! I'm very sympathetic to the idea that deontological and virtue-ethic moralities are, in a very real sense, 'practical approximations' to morality-complete consequentialism (which is impossible to calculate or estimate directly).
It seems to me that a utilitarian should attach a quantitive value to themselves rather than a categorical one.
Or if you are going to use binary category of good/bad, setting the threshold at perfect doesn't strike me as useful. By your logic, a Rolex is a bad watch because it's less accurate at timekeeping than an atomic clock.
Really, even a quantitative value shouldn't matter right? For results based ethics, the past doesn't matter so much, mainly what matters is the consequences of your next choice, regardless of where you fall on the good/bad spectrum. When I said "I'm a mildly bad person" I mainly meant that the current choice (almost ignoring CDC once everyone has had the chance to be vaccinated) is sub-optimal.
So while overall many of my choices fall short of utilitarian, if I give myself an honest D&D categorization over life choices' outcomes, I'd say I'm clocking in at Lawful Neutral so far.
My girlfriend’s mom is like 70, and the mom freaked out on my girlfriend for requesting that she wear a mask - the mom has some weird trauma response to anything that looks like being controlled, so it was like “you’re withholding affection to control me!” That definitely wasn’t the motivation. But I’m now imagining that my gf and I have a kid, and tell her she has to get vaccinated, not because we first order care, but because we actually want to control her. All of that conspiracy thinking would suddenly be true..
Totally agree with you. Everyone can get it now. There’s tons of open appointments where I am. I shouldn’t have to suffer for someone else’s choice to not get vaccinated. But neither should our healthcare system. That’s the really frustrating part.
This is entirely reasonable IMO, though it does seem like some places in the US are still in a place where vaccine availability is limited and people who want shots can't yet get them. Assuming that's not the case where you live, I don't think there's anything to feel bad about in this approach.
If "vaccines have been widely available for 6 weeks" isn't the bright line for getting back to something pretty close to business as usual... what is? Covid and it's variants will presumably become endemic, not stamped out entirely. Any case-count or mortality based cutoff will be arbitrary. Variants have some uncertainty but all indications I've seen so far are that the vaccines remain very effective in preventing hospitalizations and deaths. Kids aren't eligible yet but their risk of serious illness is very low.
On the exact morning that vaccines opened up to all over-16s in my state (CA), I signed up for the earliest available appointment slot from my medical provider. Thus, I am scheduled to get my first shot in 2 weeks from now (4 weeks from when I was allowed to sign up), and I expect to have full vaccine effectiveness around the end of June.
Bay Area? In Los Angeles we have an over abundance of slots and even opened city sites early to 16 and over. I feel like the rollout of vaccines could’ve been better. There have been too many open slots here for a long time.
The one part of this that's complex is the issue of mutation. The virus can only mutate when it's actively infecting, and the more people who have active infections -- whether or not they are themselves symptomatic, or affected -- the better chance there is of something really nasty arising out of the evolutionary soup that is proof against all known vaccines and sending us back to square one.
So there is actually a bit of a collective action problem here: your vaccination cannot defend you against a mutation that can evade the vaccination, and your neighbor's refusal to get vaccinated actually boosts the odds of that happening. It can't, quite, be every man for himself because of the problem of mutation. A rational collective response would perhaps *enforce* sufficient vaccination to drive the probability of bad mutants low enough that the collective would not be likely to suffer the consequences of some independent-minded Typhoid Mary unwittingly breeding a super-variant.
I don't have a good answer to this, but I think it complicated the moral reasoning rather, since there is a way in which you *are* your brother's keeper, so to speak, in terms of the threat of mutants. It's possible for actions you take to have almost zero consequences for you but deadly consequences for others, perhaps even many others, and it's possible they have no defense at all against those consequences (other than compelling your obedience to some social directive).
The kids aren't vaccinated yet, though, and I have no idea how to assess the potential long-term impact of an infection for them. Having thankfully never had to come to grips with the subject, I was surprised to discover that 90-95% of polio infections are asymptomatic and less than 1% resulted in paralysis (I'm hoping I'm remembering those figures rightly), yet we can see in the historical record what an enormous impact polio had given the size of the denominator.
My son is starting kindergarten in the fall. If I've understood correctly, the school district has full surveillance testing in place, though I have more research to do and a conversation with the principal to schedule. If their protocols are sensible and not just cargo cult practices, I'll be willing to take that risk given that they will have the data to trace infections in near real time. Given how little we know, I'm inclined to be quite conservative with him outside of school and would appreciate others' efforts to keep the amount of virus circulating in the population as a whole low.
Buy an air purifier for the kindergarten room (unless you know for sure they have already taken care of ventilation in a way you can trust, such as having windows open).
It'll still take some time to get shots into arms. My state is only at ~1/3rd vaccinated despite being "available to everyone" for a month. On the other hand, it seems like the rate of vaccination is falling, so maybe we're more constrained by vaccine hesitancy than finding an appointment.
I'm definitely on the, "I'm vaccinated, time to party" side of that line. For context though, I was locking down about a week before anyone else was, and wore a mask AT ALL TIMES (outside my home) until I was two weeks out of my second dose. Basically, my view has always been that the point of locking down was to buy time for a vaccine. Vaccination is my schelling fence - before vaccination, I was extremely cautious, but after vaccination I've given myself permission to go back to normal. From a utilitarian approach, full vaccination means that the impact to my own happiness is no longer outweighed by the tiny amount of good I might do for others by curtailing my activities. From a categorical imperative perspective, my behavior would universalize quite well.
Not all of us anti-lockdown people are anti-vax (I'm currently suffering through Moderna side-effects), and we're happy to welcome the "I'm vaccinated now; I'm going to live my life again" segment in blowing raspberries at the CDC et al.
I've always been conflicted on lockdowns: on the one hand, they almost certainly resulted in at lease some people still being alive who would have died already, though it's not yet clear how many; on the other, the way they were implemented was capricious and amounted to yet another form of regulatory capture benefiting large businesses and hurting their small competitors.
That, and I think the burden of the lockdowns fell disproportionately on the working class; members of the PMC like myself were much more likely to be able to work from home. And don't get me started on the hypocrites who were pro-lockdown… for everyone else. Give me an honest anti-masker any day over someone who believes no price is too high for other people to pay for their safety.
Yes, there’s something especially grating about people who advocate for a certain rule-set then violate it. In a way, I’ve found this to be a good practical argument against the sort of intuitive “Parliament House” utilitarianism which some people fall into, where some individual advocates a policy for the masses, yet privately violates it themselves.
Not only is it, as you say, hypocrisy, but it also burns the commons in a sense- “the masses” won’t respond by thinking “oh this person is merely individually flawed, or indeed so intelligent that they’d figured out how to do a more nuanced version of the right thing, which We could not do”, they’ll think “these swines don’t care so the rule is just virtue signalling or an attempt to exert control”. And that’s how you get even greater non-compliance with life saving measures.
I think externally-imposed lockdowns mostly served to postpone deaths from the (roughly) Spring 2020 "first wave" to the Summer 2020 "second wave", after which all the exemptions and violations rendered the lockdowns ineffective. The people whose lives were actually saved were the people who were scrupulous and disciplined in their own self-isolation, or other precautionary measures, and I think we need a word for that other than "lockdown".
If you have a high-trust society, externally-imposed lockdowns probably shouldn't be necessary. And if you want to keep living in a high-trust society, the sort of capricious, arbitrary, and hypocritical lockdowns we got probably shouldn't be allowed.
Locking people into sterile iso-pods "for their own protection" is altruistic only under the most warped and perverse definitions of the term.
This maps to aptitude for risk management. Some Americans still have the knack, as all humans necessarily once did, of saying "this is dangerous, but it is *worth the risk*, which I have taken reasonable measures to reduce". Other Americans, first-worlders generally, insist on the thing that only recently became possible of dividing the world entirely into that which is Absolutely Safe and that which is Intolerably Dangerous, and insisting that the Intolerably Dangerous be wholly excluded from their life.
For the first group, once they and their friends have been vaccinated, the remaining risk (of death, of serious illness, or of making someone else dead or seriously ill) is relatively low, probably within the background noise of modern life. COVID + 95% effective vaccine is about as deadly as seasonal influenza + usual crappy flu shot, and justifies about as much effort.
For the second group, COVID was flipped over into the "Intolerably Dangerous" category a bit over a year ago, and the only way for it to flip back to "Absolutely Safe" is for there to be no press coverage of ongoing COVID deaths, etc, for several full news cycles. Or maybe one news cycle of unambiguously positive "vaccine works, crisis over!" coverage, but let's be realistic. Until then, they have to continue treating it as an Intolerable Danger, in the manner they have become habituated to.
I'm thirty-eight and not fit, but working on it. Yes, I am in the herd, and I'll continue to wear a mask in public indoor spaces during cold & flu season. I always used to think that the East Asian kids wearing masks on my nearby university campus were paranoid, but the bacteria & viruses really are out to get us.
Chris Hayes' MSNBC show informed me of some CDC stats last month that I still find mind-blowing. The US usually records between 150-200 pediatric deaths associated with the flu each winter. This past winter, there was 1! Positive flu specimens dropped by two orders of magnitude year-over-year in late February.
We're going to have to give some of this back to reopen the economy and I certainly don't think any level of the government should mandate masking on a long-term basis. That's a blunt instrument to reserve for a crisis; this is the sort of thing that should evolve as a social norm if it proves worthwhile. If we can keep even part of that improvement with masks & better hand hygiene, it would be criminal to drop that because we just want to go back to the way things have always been, especially as a mask is just about as close to a free lunch as you'll find in this life.
I wish we could run this experiment on a national basis this coming winter to parse how much of the reduction in flu infection came from distancing and how much came from masks & hand hygiene, accounting for flu vaccine effectiveness & takeup, but that ship sailed many months ago. Maybe we'll be able to collect some data if a norm about masking establishes itself in a few smaller university towns like mine.
For what it's worth, a decent chunk of people I know (though far from the majority) are planning to continue wearing masks both whenever sick, and possibly just whenever in large crowds moving forward.
Interestingly, at least one of the reasons had nothing to do with health - it serves as a barrier (in the same manner as large headphones) to being harassed by strangers in public.
I feel quite the opposite. Masks are a signal of tacit support for massive human rights abuses, including abuses directed at me. I consider past abuses to be a predictor for future abuses, and for that reason I will treat people who I see wearing masks in public differently.
I already gave my explanation. Past human rights abuses are a predictor for future abuses. If you were willing to go along with the regime on this, then I should be fearful of what future abuses you will support them committing against me.
What are you going to do, give me a dirty look? I suppose you must have the right to swing your arm freely and touch my mask, given that its thickness delineates a zone of a millimeter or two between your fist and my nose, but I must warn you that I flinch easily and retaliate immediately. Tit for tat is a dominant strategy in most two-person games.
It's the lockdownists that threw the first punch, but I'm not one for tit for tat. I'll just generally disregard the existence of people who wear the symbols of the regime. No direct hostility, but no help or courtesy either.
I know who Marshal Petain was. I do not take an accusation of being a collaborator lightly. Be specific. What regime would I be supporting by continuing to wear a mask in public?
I mean, you've already acknowledged that you don't care about other people. You're just trying to justify your pre-existing selfish behavior.
Honestly, if we'd locked down all international travel and people had actually gone into real quarantine for a few months, we could have completely eliminated the virus without a vaccine, at least in many countries.
That which is a barrier to being harassed by strangers in public, is also a barrier to being helped by strangers in public. To being seen as a fellow human being worthy of dignity and respect by strangers in public. And given the way humans are wired for face-to-face social interaction, if you've only got half a face you'll probably only be seen as half a person.
Masks have been one of the most polarizing issues of the past year; you and your friends propose to continue that indefinitely?
Being forced to stop doing something because someone else polarized it is a general summary of 2020-21. Regardless, I don't see anyone here suggesting forcing an end to mask wearing.
Nobody said anything about "forced". But the OP specifically called out, as one reason to keep masking up, the bit where being masked reduces the frequency of unwanted social interaction with strangers. Which, yes it does. But that includes reducing *favorable* social interaction, and it probably does so preferentially. And including passively favorable social interaction like subconsciously classifying "Edward Scizorhands" in the "basically a decent person like me" bin on the basis of e.g. having seen you smile.
If you choose to remain masked, for whatever reason, that consequence is part of what you are choosing. If you are part of a recognizable group that chooses to remain masked, that consequence will apply at the group level. But nobody (other than maybe Tucker Carlson) is going to force you.
It used to be that the ones who worked sick could claim to be tougher than everyone else and more valuable employees. But even pre-Covid I had begun to detect frustration being directed toward that employee when they had coughing fits in the room with poor air circulation. Now, I hope workplaces don’t give up the remote-work capability. Then people could be actually sick, potentially without even taking sick time if they still put in a work day remotely.
Yes – tho, sadly, I think what little evidence we have of this dynamic is that it's an obvious net-loss in productivity (because one sick person often makes others sick).
I'd say 90% bureaucracy and 10% we really, really, really don't want to inject something that could be harmful in our entire population of 300 million people, though. And to ascertain that, no matter how big the trial, a few months of observation are necessary.
To be fair, if Covid-19 were more lethal, this would have probably been faster: as long as vaccines are decent, you want to give them to people asap. But with this low-but-not-so-low lethality... it was the perfect bureaucrat storm.
This is a great song! Please write more content like this.
(Also, my brain somehow automatically generated a melody to the chorus as I read it, which I realized later to be similar to a song I wrote once but never published. Here's the melody, ROT13'd to encourage musical creativity:
I'm jealous that you can already get the covid shot; here in Europe the schedule just keeps lagging more and more. Luckily summer is just around the corner.
And now my subscription has paid for itself. Thanks, Scott! But if the fan community can put together a chorus capable of actually putting voice to this, that would be perfect.
Let's say the FDA had immediately authorized and recommended challenge trials at the start of the pandemic. How long would it have taken challenge trials to get efficacy results of similar significance to the observational RCTs that were actually performed?
"Hi there, we're looking for 60,000 volunteers for a trial of a new vaccine against the scourge that you've heard has been wiping out 2,000 people a day recently. The way this works is, we'll be giving half of you the vaccine, and half of you an injection of sterile water. But then we're going to deliberately give *all* of you COVID, so we can see what fraction of the vaccinated group die relative to those who drew the short straw.
"Since we're most interested in the results for people at high risk of death from this particular disease, we are particularly interested in volunteers who are obese, old, diabetic, and otherwise have risk factors that make it much more likely they'll snuff it once infected, i.e. we're not very interested in young healthy volunteers who correctly infer they have little risk even in the control group. But if COVID reasonably scares the piss out of you, you're who we want, and we will be sure to write a nice condolence letter to your family if you end up in the control group and we give you COVID and then you die, emphasizing the value of your contribution to science and the bulletproofness of the legal disclaimer you signed."
You would need 100-200 people for a high quality challenge trial to determine efficacy. Large scale safety testing would be done separately and does not require a challenge. If you are testing multiple vaccines, you could combine them all in one multi arm trial and further reduce the total size of the control arm (which is where most of the risk is, since vaccines mostly work). I believe the way the statistics work out is that you always "kill" *fewer* people in a challenge trial than a conventional trial for the same statistical power.
You could pay everyone in the trial $5 million and it would be a rounding error compared to the benefit.
1daysooner.org collected 38,000 "volunteers" for challenge trials. I imagine some people who talked a good game on the internet would chicken out, but millions more people would volunteer if there was actually something important to volunteer for, and way more people would happily do it for money.
Many many millions of people have signed up to fight in wars where they had orders of magnitude higher risk of death.
This is really really far from being a problem with challenge trials for something like COVID.
> i.e. we're not very interested in young healthy volunteers who correctly infer they have little risk even in the control group
All serious proposals I've seen for challenge trials suggest exclusively using young, healthy volunteers. The idea is that if the vaccine works very well in one population of humans, it's fairly likely to work at least somewhat well in all populations of humans. The goal shouldn't be "find precise scientific proof of the exact percentage by which the vaccine reduces the incidence of COVID in the target population", it should be "find out whether the vaccine works or not so we can start distributing it ASAP if it does."
I know people poked fun about there being two "report" buttons, but perhaps cutting the number to zero was an overcorrection.
I'm in awe of what it must be like where you live. Very little of this has proved necessary in Texas.
With all due respect, it was necessary, you just didn't do it.
I agree with this, too. It feels like we got a lot right by accident. The inconsistency between city and state governments meant the places with a lot of population density, interstate and international travel, and crowded indoor spaces mostly did the right thing by shutting at least those things down, whereas the rural areas that probably didn't need to do much didn't. Even the one thing that seemed like the Dallas equivalent to LA's exception for film production being considered "essential," allowing all construction to continue so our local real estate boom could continue unabated, turned out to be fine once we learned outdoor gatherings were fairly low risk.
I think you are both suffering from insufficient specificity of what "this" in "very little of this" means and this conversation is the worse for it.
This bothers me about basically *every* discussion of covid safety measures. Despite there being large variation in what the measures are from place to place, and despite these measures being completely unfamiliar from the Before Times, we somehow adopted loose terminology ("lockdown" or "quarantine") for everything from "please wear a mask for the 30 seconds between stepping into a restaurant and going to a table" and "the apartment doors will be welded shut and food will be delivered by government workers".
Unfortunately, it also seems that calls for/against mitigation measures have often gravity-welled somewhere around either "nothing is necessary" on one side to "welded doors with doordash portals" on the other - doesn't seem to be just in the way we talk about existing measures (not referring to you, Scott or Wendy, more in the public parlance...)
When debating the severity of measures and their ramifications, and trying to strike that balance in a more specific and granular way...I think there'd have to be a way to do a holistic analysis of effects of both specific and cumulative measures next time, or at least make a show of doing so, to avoid or decrease some of the political tension before it starts as well as just generally optimizing policy.
For example, once we get through this, I'd be interested to see an analysis of long-term deaths correlated with or caused by economic downturns a la March 2020 (despair, starvation from disruptions in food chain, etc.) vs. estimated disease death count. Maybe this analysis could help us hone things - I'm not sure what one does about deaths of despair, or how avoidable they are, but maybe we could prioritize/secure the food system to be robust against the specific issues that are leaving swaths of the third world at risk of starvation.
For example, this analysis seem to clearly indicate that, at least in the short term, COVID well outnumbers deaths of despair here in the U.S. but I had to look pretty hard just to find one analysis - which is now about 3 or 4 months out of date.
https://www.nber.org/papers/w28303
Famine risks:
https://www.nytimes.com/2020/09/05/world/africa/coronavirus-famine-warning-.html
There's been some recent research suggesting an unexpected absence of suicides..but more research is needed.
https://www.bmj.com/content/372/bmj.n834
i am inclined to agree with speculations i heard that depressed people are likely to blame their sorrows on the pandemic (and to think they'll wain once it is over). i personally expect a spike after delockdownment coming from people noticing their problems not suddenly dissapearing.
The one other point is that we have to compare not *actual* covid deaths to the deaths of despair, but the *difference* between the actual and counterfactual covid deaths that was produced by the policy, which is just one more level of difficulty. (But if we make a very rough guess and estimate that the number of prevented deaths was the same order of magnitude as the number of actual deaths, that could help.)
that's a really good point...
People are terrible at this, largely because they do just what you said, which is assume that measures are only weakly helpful in the direction that opposes their instinct. You will find many people who are 100% convinced that if no public measures at all had been taken, the deaths from COVID would've been maybe 10-20% more --- at which point they'll say, so what? Not worth it.
But what if the actual counterfactual is that without any countermeasures at all the deaths from COVID would've been 10x or 20x higher (e.g. in the neighborhood of 5-10 million)? If people knew for a fact this was the case, it would totally change the argument. But unfortunately, it is essentially impossible to even construct an informed guess about that during the pandemic, and so we fall back on reasoned argument. If reasoned argument could establish the truth or falsity of complex questions of viral transmission and disease, the Romas would've conquered the plague in AD 200. And unfortunately most people are very willing to believe that their own WAG is no more than 20-30% off, tops. *Nobody* will believe he's wrong by two orders of magnitude.
I am sure the pros will be combing through the records of all nations and individual states that tried various things, and the outcomes, for decades to come, and making arguments about what can be shown empirically to have worked or not worked. But the nature of record-keeping is such that there is a very good chance we will never have any really solid answer, at least as far as this particular pandemic goes.
You might enjoy http://www.sfu.ca/~allen/LockdownReport.pdf "Covid Lockdown Cost/Benefits:A Critical Assessment of the Literature" by Douglas W. Allen.
The abstract:
An examination of over 80 Covid-19 studies reveals that many relied on assumptions that were false, and which tended to over-estimate the benefits and under-estimate the costs of lockdown. As a result, most of the early cost/benefit studies arrived at conclusions that were refuted later by data, and which rendered theircost/benefit findings incorrect. Research done over the past six months has shown that lockdowns have had, at best, a marginal effect on the number of Covid-19 deaths. Generally speaking, the ineffectiveness of lockdown stems from volun-tary changes in behavior. Lockdown jurisdictions were not able to prevent non-compliance, and non-lockdown jurisdictions benefited from voluntary changes inbehavior that mimicked lockdowns. The limited effectiveness of lockdowns ex-plains why, after one year, the unconditional cumulative deaths per million, andthe pattern of daily deaths per million, is not negatively correlated with the stringency of lockdown across countries. Using a cost/benefit method proposed by Professor Bryan Caplan, and using two extreme assumptions of lockdown effec-tiveness, the cost/benefit ratio of lockdowns in Canada, in terms of life-years saved, is between 3.6–282. That is, it is possible that lockdown will go down asone of the greatest peacetime policy failures in Canada’s history.
[I'm not sure I agree with the paper, in particular with its part III, but I really enjoyed parts I and II where it shows the various ways scientists/economists/politicians can shoot themeselves in the foot when trying to reason about and model lockdown effects]
Cool, look forward to reading through this one!
Just to complicate things further, the overall prevalence of mandatory lockdowns would signal to people whether they should voluntarily lock down. In that way, mandatory lockdowns could have a broad health benefit even if mobility is uncorrelated with mandatory lockdowns across countries.
Douglas Allen is an economist, not a public health expert or epidemiologist. I would value his analysis of the economic costs of lockdowns some and his analysis of pandemic public health literature close to not at all.
> Unfortunately, it also seems that calls for/against mitigation measures have often gravity-welled somewhere around either "nothing is necessary" on one side to "welded doors with doordash portals" on the other
As someone who has been very conflicted on this specific topic, I think there's an expectation (which I honestly don't know if it's warranted or not) that strong blanket rules with no nuance are needed because any level of nuance will be taken as "whatever the minimum acceptable restriction in any situation is, is the maximum acceptable restriction in any situation".
So, the fear is that saying "wear masks while indoors, but masks are unnecessary outdoors" will translate to a sizable portion of the population hearing "masks are unnecessary, regardless of if you're indoors or outdoors". This seems uncontroversial, to me.
The more controversial take is this: When instead given guidance of "wear masks all the time, and when you're indoors, wear 2 or 3 of them!" this will translate to a *larger* portion of the population hearing "masks are necessary when you're inside or outside" and will result in a greater proportion of the population actually wearing masks indoors, where it matters. The sacrifice to get there, of course, was needing to wear masks outdoors, where maybe it didn't matter.
I've felt this most acutely with the response to restaurant openings - in my area, we had restrictions on indoor dining for the majority of 2020, but essentially no restrictions on outdoor dining. One of the results of this was multiple restaurants building free-standing structures on either the sidewalk or, in places where they had them, parking lots, and then (because many of these were operating during the fall and winter) sealing them up and putting space heaters inside (because dining outdoors in the winter isn't very comfortable) and just pretending like "a sealed tent with space heaters" is actually outdoors as though this was a pointless regulation that had no actual purpose, rather than an intentional risk mitigation strategy.
That's a really interesting point. My first instinct would be that telling people that masks are needed outside/ you need more that one mask/other takes, would cause people to mistrust and disregard the rules overall if they pick up on the exaggeration...but I could totally see that happening as well.
If you tell people Noble Lies to get them to do what you think they ought to, they will *eventually* figure it out. Then the next time you think they ought to be behaving differently, you'll need literal men with guns to make it happen. Our society gets one "massive lockdowns with voluntary compliance" bullet per generation or two, and it only works for a few months. I do not think it was wise for us to have used that bullet in March 2020.
The suicide rate went down in the US in 2020.
Makes sense if you think about it. What actually happened for a lot of people is that they got to work from home/got some extra time off. Fewer people were bugging them at work. Kids who were in school were given a weird sort of pseudo-vacation/remote learning thing.
If your stress comes from work or from school, your mental health would likely improve significantly.
The people who struggled with the lockdown were probably outnumbered by the people who saw less stress due to less stress from work/school.
Moreover, more people are better off after the pandemic than worse off economically - about 2:1.
But, like, 'necessary' in this case is kind of weird, right? The (largely liberal) NE/CA/cities that attempted to lock down way harder than the (largely conservative) TX/FL/etc seem to have similar outcomes in the US?
Unless you're comparing the US as a whole to say Australia, which was basically the only success story.
You're right that I should look into this more. Right now my model is something like more cases -> more lockdowns -> fewer cases, which means that the correlation kind of washes out. In support of that, see the graph on the top of page 3 here, and also the scatterplots at https://www.nytimes.com/interactive/2020/11/18/us/covid-state-restrictions.html (which are a point estimate from November 2020). In order to be really confident in this I would need to see these kinds of estimates up until the present, which I can't find. I'll register this as my suspicion for now and hopefully remember to confirm with more data after the pandemic is over.
Yeah, I was looking a bit more at the total outcomes in terms of deaths (honestly informed largely by this article, which is excellent: https://www.politico.com/news/magazine/2021/03/18/ron-desantis-profile-covid-florida-2024-476777)
This is honestly kind of sad, but I actually don't trust the NYT as a source anymore because they've become so agenda-driven (and I say that as a former long-time subscriber).
I don't trust them either, but it's hard to mess up a scatterplot. I trust them to lie in subtle, hard-to-detect ways, rather than literally putting scatterplot points in the wrong place. I eagerly await learning I am still too gullible.
Oh, true, I don't think there's any strict data falsification* going on, but I think they pick and choose what to report, e.g., that article is on cases (and to be fair, hospitalizations) but not overall death rate, which I think is a better metric - one can very much disagree with that though!
*Though I will say that my tipping point was when I read about the 'largely peaceful' BLM riots last June after watching - literally, watching, at 8p at night - the drugstore and liquor store across the street from me (in Flatiron!) get broken into the night before.
I think it was a matter of timing. That particular scatterplot was made after the northeast was finished getting hammered (at least Round 1 of hammering) and before California got hammered. Remember they're looking at "current" (Nov 1) hospitalizations, not total.
Dear god his wife looks like a female version of him. I'm going to have nightmares.
I can't believe that urban/rural split isn't relevant here. I live in Indiana. While I live in the city and we locked down and locked down early, I can't help but feel that such measures would have been draconian in rural counties with 2-3 digit max number of cases (certainly were perceived that way out in the hinterlands from my friends and family).
Ok, but is there a boundary between the city and the country?
At a strict level, no. At a definitional level, you can certainly have metrics that either operate on a per-capita basis and/or a density basis (per-capita per-square mile), depending on what the models used are able to predict is most effective.
I just want to respond to this with an image of a "[name] city limits" sign
While I'm moderately annoyed that the state didn't actually stick with it, Illinois at least early in the process had specific metrics in place to be able to treat urban (ie Chicago) areas qualitatively different from rural areas (ie by adjusting the restrictions based upon pre-defined metrics per area, rather than evaluating for something like the state/country as a whole).
If the data ever becomes available I'd really like to see what fraction of variance can be explained by pre-existing Vitamin D levels, age demographics and population density.
I feel like "more cases -> fewer cases" seems to work even without "more lockdowns", the "more lockdowns" just correlates very heavily with "more cases". Not that "more lockdowns" can't (also) have a causal effect, it's just really hard to estimate what it is because the counterfactuals everybody produces are nonsense.
In this case, "everybody" includes rationalists who are no better at predicting the specifics of the pandemic than anyone else (see e.g. all the fuss about variants earlier this year).
Indeed. Similarly, the argument that lockdowns were solely/primarily responsible for downturns in cases has long been plagued by the issue that declines in cases often started before (or far too quickly after) lockdowns were imposed. "Wait two weeks" goes both ways.
Pretty much all of Asia is laughing at that last line of yours.
It's certainly possible that what the government did in (largely liberal) NE/CA/cities didn't make a big difference versus what the government did in (largely conservative) TX/FL/etc. But that doesn't mean that lockdowns, wearing masks, etc. wouldn't have worked. They clearly did work in quite a few countries, including, yes, Australia.
But nowhere in the US did a real lockdown, certainly not for long enough to actually push cases down small enough to mostly wipe it out. Our population just didn't have even near the patience required.
I don't think it's patience as much as the overall attitude/tolerance for big government to impose its will...which, yeah, Asian countries 'won' in this case but I think on the balance overall (standard of living, economic mobility, etc) tilts in the direction of the US and similar countries. But that's an entire other can of worms and I'm obviously super-biased.
Fair enough. And I agree that, in general, there is a tradeoff, and I am very much on the side of individual choice etc.
But it's important to acknowledge that hundreds of thousands of people in the US died because of COVID, and they didn't have to. Was it "necessary" to lock down, well the world didn't end, so I suppose that wasn't "necessary". But if you actually care about your fellow man, as the chorus of epidemiologists sing above, more of us should have made other individual choices. And our government(s) should have done more to help us. And I'm going to continue to be pretty mad about that.
Also fair! FWIW, I'm generally one of the selfish ones who cares most about the statistical health/well-being of my fellow man rather than the specific health of individuals (which I know isn't a popular opinion) so that's coloring my thoughts as well.
> And our government(s) should have done more to help us.
This is, I think, a big part of why lockdown took the shape it did in the US, and why it was so resented: we foisted the cost of lockdown on the people themselves.
Class was a big issue here: the professional/managerial class, which I'm part of, could mostly work from home, and so the cost of lockdown was substantially less for us.
Yes, our government should have done more to help us -- by getting out of the way! By allowing challenge trials and allowing people to waive full testing and take vaccine months earlier, saving hundreds of thousands of lives; by allowing private companies to do testing early on rather than monopolizing; by allowing importation of good masks; and many other things. I wrote on many of the bad moves by the CDC a few days ago: https://medium.com/conjecture-magazine/failure-at-the-cdc-8cbb772182de
"But it's important to acknowledge that hundreds of thousands of people in the US died because of COVID, and they didn't have to."
Lots of people agree on that. The disagreement is about what should have been done to prevent it.
I saw something, I think from the CDC, boasting of the fact that we got vaccines in less than a year. As the poem says, we got the first vaccine in about two days. The remaining nine months or so were spent on the testing needed to get the FDA and similar agencies elsewhere to approve a vaccine.
With challenge testing and not insisting on a rigorous, controlled experiment approach, that would have been more like three months — one month to check that the vaccine itself was not seriously dangerous, which doesn't require challenge trials, another two months with challenge trials of a thousand or so healthy young people, resulting in between zero and one deaths (IFR about 1/10,000), after which the only limit is how fast they can produce the vaccine and offer it to anyone who wants it and is willing to have his health status thereafter kept track of.
That would pretty clearly have saved several hundred thousand lives.
> economic mobility
https://en.wikipedia.org/wiki/Global_Social_Mobility_Index
china isnt doing well but the more liberal asian countries (Japan, Korea, Singapore) are doing a lot better. these same countries also prove asian culture is not contradicory to a high standard of living.
(by better i mean good enough to beat freedomloving usa)
Heavens, no, China did much better *economically* than the US by controlling the virus very early and quickly, and then re-opening their economy. China GDP growth in 2020 was +2.3%, which while crappy beats the heck out of the US (-3.6%) or the eurozone (-7.4%), who basically rode this hideous roller-coaster of partially-effective counter measures followed not quite long enough, followed by somewhat overoptimistic let-ups, rinse, repeat, which achieved the magical combination of *neither* controlling the virus nor allowing for normal economic activity for the entire year.
My company has offices in the US and China. US offices have been closed for 13 months. Those in China closed in January and re-opened in April of 2020 and have been functioning normally ever since.
Asia locked down before it was endemic. By the time the elites and public health establishment in the US woke up we had several million cases (backfilling from 30k deaths by mid April to 2-3 million cases by mid March).
Australia is the one counterpoint - they had a 2.5 month lock down (seemingly helped by seasonality) that got a very serious outbreak back to contained status.
The question is not do NPIs "work" in the sense of do they in the short term lower cases - quite clearly they do (imagine an extreme where soldiers patrol the streets keeping everyone in their houses - there won't be much COVID!). The question is:
A. Is all you are doing just pushing the pain back (relatively straightforward in the absence of a vaccine)
B. Are hospitals in danger of being overloaded
C. Is the cost in liberty/inconvenience/economic damage/societal damage (eg school closures) worth the deaths saved (in light of A and B above)
Reasonable parties may differ on the answers to any of those, but I think they have to be evaluated in light of the fact that COVID19 had already spread significantly before the public health establishment advocated much of anything.
There's also D, that is "is the government allowed to make trade-offs between restrictions and lives saved in this way".
It really bugs me when everybody assumes that just because some trade-off is "worth it", it should be done. Too many utilitarians around here.
...what's the alternative? You can't reconcile everyone's individual preferences otherwise.
I would assume the alternative to utilitarianism is deontology, ie you can't transgress some rights no matter how many lives it saves. And the problem with that is that government is already following the official deontology , the set of rights that are de facto recognised. You might feel that the government can't order you to wear a mask, however many lives it would save, but that would be your own personal ideology.
"Asia locked down before it was endemic."
Taiwan, South Korea, Japan, and Hong Kong never had a lockdown.
yeah perhaps the better thing would be "did something dramatic before it got endemic, up to and including border closures, centralized quarantines, widespread mask wearing, and contact tracing with teeth". Obviously not all of those apply to all countries but hopefully you get the drift.
"(imagine an extreme where soldiers patrol the streets keeping everyone in their houses - there won't be much COVID!)"
Peru begs to differ. The military was deployed to patrol the streets in Lima to enforce curfews. It did not work, not even slightly. After Peru adopted Lockdownism, it's deaths per day went from negligible to massive. https://ourworldindata.org/explorers/coronavirus-data-explorer?zoomToSelection=true&time=2020-03-01..latest&pickerSort=asc&pickerMetric=location&Metric=Confirmed+deaths&Interval=7-day+rolling+average&Relative+to+Population=true&Align+outbreaks=false&country=~PER
It's numbers are even worse if you look at excess deaths, although it's impossible to tell how many are uncounted lockdown-induced covid deaths or lockdown-induced other deaths.
"seemingly helped by seasonality"
It was still winter, and barely past the coldest point, when the lockdown & other rules definitively turned the tide. The inflection point was the first week of August; the mean temperature for that week was marginally lower than the mean for July, and the mean for the whole month of August (by the end of which we were clearly on the way to zero) was just 1 degree warmer.
(I'm talking about Melbourne, for anyone wondering. The second wave happened in the state of Victoria, and almost entirely in metropolitan Melbourne.)
I mean seasonality doesn't necessarily literally mean the seasons - we don't even really know why the flu season behaves the way it does (and lots of other things about the flu). Just that the virus appears to move in cycles that defy explanations that rely on only human behavior.
Australia has been an exception in lots of ways, so definitely possible it was all just human action! Just wouldn't rule out disease cycles.
The thing is, we did develop vaccines. Very rapidly, in fact.
Many people are now fully immunized having never caught COVID.
so substack is telling me this is in reply to me but I'm having trouble telling what comment it is in reply to - so I'll just riff a bit haha.
So it's true we got vaccines quickly, which is a point in favor of an Australia-type lockdown route (or better yet, early travel restrictions to put us on a better path). I've seen some discussion around 100-day vaccines in the future.
But at the time of lockdowns last year:
A. it was already endemic
B. Vaccines would be available by the end of the year were being called crazy!
Which is where I think if you believe A & B it gets harder to advocate for lockdowns - you're basically hoping you can be Australia instead of every other country that had an endemic outbreak and never got it back under control.
It's worth mentioning a famous disease mitigation paper: http://www.upmc-biosecurity.org/website/resources/publications/2006/2006-09-15-diseasemitigationcontrolpandemicflu.html
(one of the coauthors - D.A. Henderson - lead the smallpox eradication effort). They discuss a pandemic influenza with a CFR similar to COVID and a vaccine available within 6 months - and still do not advocate for widespread NPIs!
I'm neither arguing in favour of nor against masks, but you have to admit that citing whether overall cases in a country have increased or decreased is an insufficient proof of their efficacy either way.
I haven't worn a mask once in the last year, and I only remember one lockdown for like 2 days before it was called off after it turned out we didn't have some crazy variant that spread via fomite.
What worked, I assume, was the border controls. None of the other measures like masks and lockdowns and whatever else other countries are doing were necessary because there are no cases to begin with. International arrivals got quarantined in hotels before they were allowed into the community, that's essentially all that was needed.
Some states fucked up these hotel quarantine measures at times, which is where some of the brief lockdowns come from. Victoria I think was the only state to have lockdowns for an extended period of time rather than the brief one or two lasting less than a week after a hotel quarantine scare. But for most of the country, for the vast majority of the time, there are no masks, no lockdowns, no cases. Everything has been pre-pandemic normal.
Too rigfht. Always amazes me that people wave away the case of China, which could have been like India is now, with some airy wishcasting that the Chinese just lie about everything so they probably really did have 100 million cases and 1 million deaths and are just...really good at hiding the carnage. These people fit right in with the faked Apollo landing people, you cannot reason with them at any level.
But clearly anyone who makes a good faith assertion that public health measures could not have controlled the virus needs to explain China, where the virus *was* controlled, and limited to a small population and a tiny per capita mortality, in a very quick (if exceedingly brutal) amount of time.
>Pretty much all of Asia is laughing at that last line of yours.
As others have noted, "Asia" did real lockdowns in time to plausibly matter. Asia also seems to have done real contact tracing, which nobody in the West really managed. And Asia bought up pretty much the world's entire supply of Real Masks at the outset, leaving the rest of us with symbolic fake masks.
Asia was, if not physically immunized, societally and institutionally immunized by the SARS non-pandemic a decade or so earlier. Their local versions of the CDC were prepared to act promptly and effectively; ours weren't.
Maybe Asia should not count their chickens before they are hatched...At this point, it's difficult to say a restart is out of question: most of the circulating variants are quite different that the original wuhan strain, it seems the chinese vaccines are quite weak compared to the main western ones (and the russian one), and, in europe at least, countries/regions that got of easy during the first waves suffering more during the latest ones looks like a common pattern. common but far from systematic...Asian countries managed to stomp the first wave early, sometimes so early it did not happen at all...but some now have difficulties. Thailand for example had almost no cases early on, controlled the few case among returning workers and was cruising along happily since may 2020. But since last month, it seems they lost control. Japan seems similar to thailand, keep control early on for not-so-clear reasons, but seems more problematic this time. India also seems much higher than during the first wave, where western governments predicted a catastrophe than never really materialized...at the time, while now it looks like it's happening. The rest of SE asia is a mystery, no covid news at all. Very few cases during first wave, then I think there were a significant but not catastrophic waves in Malaisia and Indonesia, but lately I didn't hear anything, good or bad, in the region. For Myanmar it's expected, They have a far more pressing issue than covid, but the rest of SE asia? No news. So what will happen in China? not sure? Will we know about it? even less sure. I do not really trust covid news even locally, my covid impact is that I have lost my last few remaining bits of confidence in mainstream news, so reporting about China is basically fiction.
San Francisco, Portland, and Seattle have had systematically better results than Los Angeles, Houston, and Miami. I believe DC has also been on the better side, while New York and Boston have been hard to understand.
Looking at the state level obscures a lot of this.
By the very virtue of your post, city level seems hard to understand as well :)
Yeah, I'm not claiming it's easy to understand at the city level, or that anything immediately clear emerges! Just that we see more phenomena that are completely obscured when you lump "California" or "Texas" as one place, while St Louis gets divided between Missouri and Illinois.
It was quite a while before I learned that East St. Louis is a city inside Illinois rather than just being the east side of St. Louis, Missouri.
Geographically, baselines are going to be all over the place. But if you look at lockdowns *that are actually obyed* , in a certain area, across time , they make a difference.
As an inhabitant of a city in Texas, let me assure you that the major Texas metroplexes did lock down, just largely without state-level support outside of the initial three months. They're not remotely locked down any more, though.
What do you mean by "lock down"? It's an incredibly vague term. Enforced masks (in what circumstances?)? Limited building occupancy? Forced certain businesses to close? Quarantined people testing positive? Etc.
I mean restaurants, barber shops, other small businesses closed. Grocery stores instituted quotas of how many people could go inside at one time. Anything else considered "essential" became curbside pickup. All office workers who weren't doing classified work had to work from home. Roads were empty. Downtown Dallas looked like a ghost town for nearly a year. At one point, my employer had to give me a note saying I was part of the Defense Industrial Base and need to be on-site for my job duties in case the police stopped me to ask why I was out driving.
I was never in San Francisco but can't imagine the restrictions there were ever any more severe.
Ofc. Texas actually has the thirteenth highest unemployment rate in the country now, just after Illinois: https://www.bls.gov/web/laus/laumstrk.htm
Seems vaguely notable that every state that is worse than Texas has a Democrat governor. And Texas has done better than other big states like California and New York for many months.
Thanks for that link. I think it's instructive to look at the metro-level stats: https://www.bls.gov/web/metro/laummtrk.htm
In looking through Texas, my own area of DFW was not hit that hard, not nearly as bad as most other Texas metros. Unrelated to strength of the lockdown, I suspect that is because our biggest industries, defense, IT, financial services, outdoor construction, simply weren't hit hard or even flourishes, whereas oil and gas and shipping totally tanked and other Texas metros are heavily dependent on those industries.
Then look at the state-level stats from before Covid (2019): https://www.bls.gov/lau/lastrk19.htm
A lot of states here moved quite a bit, most notably Hawaii. It's easy for people engaged in motivated reasoning to say "oh, they have a Democrat for governor," but it seems likely to me the reason for that is Hawaii's economy is almost entirely dependent on tourism and imports. Those tanked globally and have nothing to do with local policy. There is nothing any governor could have done to prevent that. It's also notable that California was hit so hard considering the very well-publicized and rather hypocritical exemptions given to the film and television industries to consider them "essential" and allow them to stay open. But most major productions stopped filming anyway, even though the government didn't force them to. Even then, the metro-level numbers show that farming regions were hardest hit. I don't see any speculation as to why, but apparently the research service of the USDA shows the values of crops and livestock over the first six months of 2020 was revised downward $29 billion compared to start of year estimates, which tanked farm incomes for the year: https://www.ers.usda.gov/webdocs/publications/100177/ap-086.pdf?v=9134.7
So you can do what is called a shift-share exercise, which is where you take a state's (e.g. Hawaii) industry composition, and say "suppose for each industry, the decline had been as big as the decline in that specific industry everywhere else in the country. How big would their total decline in employment have been?" And this allows you to compare what they actually experienced versus how the "average" place would have done given their industry composition. FWIW, Hawaii does have an industry composition that makes it particularly exposed. But it did much worse than the benchmark controlling for this.
Didn't Hawaii actually shut down, as in close its borders and not allow non-Hawaiian's to travel there? Trying to look up what the restrictions from the past year, it looks like it may have just been a required two-week quarantine and negative Covid test, but the unavailability of tests may have made that a de facto travel ban early on.
It seems to have had the desired effect in that they apparently have the lowest death rate of any US state: https://www.statista.com/statistics/1109011/coronavirus-covid19-death-rates-us-by-state/. But definitely not a great economic strategy for a state that so heavily relies on visitors for income. I guess we'll see what impact that 15% unemployment rate has. That is brutally high.
Texas has a middling deaths per capita level among US states. The high restriction states have fared no better. Many broken eggs. No omellete.
This is a pretty concerning level of disregard for human rights even disregarding the empirical lack of "necessary". The evidence is already in. Texas didn't do it, and it had no measurable consequence.
No; Texas's per capita COVID deaths are slightly below national average, while 4 out of the 4 top states in COVID deaths per capita are blue states (in 5th place is Mississippi, which has extremely Republican Whites and the highest Black percentage of any state in the country). Scott, read this description of anti-pandemic measures from 1911; they are much smarter than anything done in the U.S., and proof of a general decline in Western epidemiology during the 20th century:
https://www.google.com/books/edition/Thirty_Years_in_Moukden_1883_1913/wW5CAAAAIAAJ?hl=en&gbpv=1&dq=&pg=PA234&printsec=frontcover
In Alameda County, where Scott lives, people have been careful and it worked; there have been 1,486 Covid deaths among 1.71million people, which is about half the Texas death rate. If Texas had had Alameda County's death rate 25,000 more Texans would be alive.
Looking at state rather than county numbers will end up confusing you because the LA area of California did very badly while the Bay Area did extraordinarily well. But Scott is in the Bay, and the Bay had an extremely low death toll.
Well yes, because a huge portion of the Bay economy can WFH. I'm not sure why we should generalize that elsewhere or what that tells us?
"In Alameda County, where Scott lives, people have been careful and it worked"
Well, yes, I expect college-educated people to behave more carefully. "Tech bros" embraced masks before journalists.
In Travis county there has been 847 deaths for 1.274million people. Austin had a local lockdown early on, but was quickly overridden by state mandates. But that's significantly better than Alameda county even.
The book is certainly interesting: there are obvious parallels between the Manchurian plague and COVID. Some paragraphs could easily have been written in the modern day. I took the book from the Internet Archive; the Google Books link didn't work for me.
https://archive.org/details/thirtyyearsinmo00chrigoog/page/n322/mode/2up
There are some important differences between the Manchurian plague and COVID. For one, it was extremely deadly, killed quickly after infection (by another source, 2 - 5 days), and affected all bar the elderly and infants.
> Its specially marked feature was its unvarying fatality : 43,942 cases are recorded, and 43,942 deaths. There was no authenticated case of recovery.
> The strong and vigorous seemed as susceptible as the weak, and the infection often passed over the aged and the very young.
> Each morning we consulted together on the telephone, and on Tuesday he mentioned casually that he was not feeling up to the mark. I went straight to the station and found that he was feverish. [...] The disease ran its rapid course, and he died in little more than twenty-four hours.
> A man in a small hamlet came home ill from Moukden and died, his family attending on him and performing the usual rites. A few days later the entire household of seven died within twenty-four hours, except one infant who was found wailing beside the dead mother.
For two, it was more obvious whether or not one had it in the later stages (and I think it was only infectious in the later stages, though obviously it's difficult to tell):
> In the evening the unmistakable Plague sign appeared, the bloody spit.
> They were helped by the fact that pneumonic plague is usually easy to recognize, once it has declared itself.
Moving on to the actions taken, I'm not sure what you're referring to precisely. It seems to me most of the success from dealing with this plague came from widespread support from the populace, who went above and beyond to protect themselves because the plague was so extremely deadly, and secondly from the weather, which prevented people from travelling who may otherwise have done so.
First, they locked down the railways where they could. This seems similar to the travel bans. Not all travel was banned, but there was much less of it back then, so presumably it was easier.
Second, they moved infected people (and those in they were with) into quarantine centres (protected by the military). These appeared quite good quality, and people were paid back for their time:
> A warm kang was provided, and plenty of good food, and there was no need to work. The members of each household were encouraged to keep by themselves, and when they returned home after their ten days' holiday, they found that their houses had been well guarded, and that they received full compensation for anything burned by the police.
This seems similar to the requests to shelter in place, and schemes like UK's furlough or Germany's Kurzarbeit. I don't know why we've moved away from centralised quarantine locations (even the Chinese secured people in their homes instead of sending them elsewhere) -- perhaps the logistics are simply too hard with the much larger modern day populations.
Third, they burned the possessions of the dead and compensated the families. This isn't something we do any more, though I don't know why. Perhaps we can clean more thoroughly? Perhaps it never had that much of an effect?
Fourth, they vaccinated the doctors and gave them protective equipment. I think we did that too, before moving on to vaccinate everybody else. It's difficult to tell what exactly the vaccine involved:
> He had been inoculated against Plague, was closely masked, and, as we thought, well protected from infection
It might have been Yersin's serum, designed for the bubonic plague.
Fifth, they sent people around to each house to inspect people for plague and take them to the quarantine centres. This was made significantly easier as the effects of plague were more obvious and could be detected even by untrained people, while COVID requires a test (though I think South Korea had mandatory frequent testing). They did the same thing for businesses, which was not done anywhere I'm aware of:
> When a shop was forcibly dosed and disinfected, and twenty-nine persons removed from it to an isolation station because of the death of a thirtieth, the merchants were highly incensed. The co-operation of the general public could thus hardly be expected.
> One day a foreigner saw a sudden stir and excitement in a restaurant on a main street; a man had fallen down ill, and quickly became unconscious. The police were called, and within two hours all was complete, the premises empty, disinfected, and closed, with a cordon of soldiers round them. When the day's work was over each member of the Plague staff visited a disinfecting station, where he had a bath and left his outer garments to be disinfected.
Sixth, they used propaganda: they sent placards around telling (possibly exaggerated, possibly true?) stories on how the plague could spread and what to avoid. This seems similar to the modern day briefings / news.
Seventh, they demanded people not visit their friends for Chinese New Years. This is our lockdown. This was helped by the weather: it snowed a lot, so people couldn't travel easily.
People reacted much as we see today: with doubt, conspiracy theories, and attempts to evade the rules:
> At the beginning there was general disbelief in the necessity or usefulness of preventive measures. It was an absolute novelty to the Chinese mind to attempt to check the spread of any infection, and apathy naturally accompanied their fatalism. "This is the scourge of Heaven," said many. "All will die whose time has come, and no others. Then why take people away to isolation stations? Why burn good clothes and bedding?"
> The Japanese were credited with encouraging or even causing the epidemic in order to destroy the people and possess the land.
> It was said that everyone who went [to the isolation camps] would die, that people were sent from them to the Plague hospital who had not Plague, and that some were buried before they were dead
> Day by day men saw their neighbours fall by their side; in five days seventy died. Panic seized the remainder, the military cordon was not very strict, and a number escaped one night, carrying infection into the city.
> It was some time before opposition to Government measures altogether ceased. The most serious resistance was on the part of some merchants, who determined that their business should not be interfered with.
The plague defense was helped by the plague's extreme deadliness: people kept to the lockdown and quarantined themselves as much as possible. Visitors from outside were rejected. Goods were bought in bulk and people kept to themselves. The plague of Manchuria lasted three months; here we've locked down for over a year with more to come. Lockdown opposition isn't obvious in those it kills: if going to a party without a mask lead to the deaths of everybody at the party and their entire households, I'd expect more people to follow the rules.
"Seventh, they demanded people not visit their friends for Chinese New Years. This is our lockdown. This was helped by the weather: it snowed a lot, so people couldn't travel easily."
This might be a Chinese, Australian, or Thai lockdown, but nothing like this happened in America; thousands moved out of New York City during the 2020 stay at home orders without consequence. The more obvious symptoms of the Manchurian plague were more than compensated by the greater availability of PCR testing today (the most ambitious users of it being China, New Zealand, and Australia), which also enables governments today to test sewers for signs of COVID -something the Hong Kong government did.
"here"
-Not in the vast majority of the U.S., where the stay at home orders were over by June.
South Korea is actually fairly unambitious at testing; it has far lower tests per capita than the U.S., and certainly hasn't tried anything like China's testing of whole cities. It did have harsh penalties for breaking quarantine for COVID cases and contacts, which I don't believe the U.S., Germany, or Britain ever had.
Okay fair, I'm in the UK and I don't really know what's been going on in the US. It hasn't been constant here, so "over a year" is probably overstating it. We were locked down thoroughly from December through March. We initially shut down in April, opened up briefly with a disastrous "Eat out to help out" scheme to encourage spending, then shut down again, blaming everybody who followed the scheme for increasing infections.
The UK had fines (up to £10k if you're both caught and can't convince the officer to let you off), but in the Manchurian plague the penalty for breaking quarantine was that you and your entire household caught the plague and died, which is definitely a lot more serious.
What were the "smart anti-pandemic measures" you were referring to? Thorough, frequent testing? Centralised quarantine? Complete transport lockdown?
Yes; to all three.
"but in the Manchurian plague the penalty for breaking quarantine was that you and your entire household caught the plague and died"
Not necessarily, and if you caught it, you were dead anyway. If you were just a contact, you might have been unscathed.
With all due respect to both of you, the word "necessary," like the related "need," is a way of avoiding rational thought on such issues. Lots of people didn't lock down and most of them are still alive, so it wasn't "necessary." The relevant question is whether it was desirable.
And we don't really know. The alternative of strictly quarantining the vulnerable and letting the vaccine run through the rest of the population until it burned itself out wasn't tried, and we don't know whether it would have killed more or fewer people. It would almost certainly have imposed less of other sorts of costs.
Well, it wasn't necessary for *you* because you're still here. In the same way, seatbelts aren't necessary for anyone who isn't involved in a high-speed car crash, and defibrillators in ERs are a complete waste of money for all the many millions of people every day who don't have a heart attack.
...on the other hand, if we ask the families of some of the 50,250 Texans who died of COVID in the past year, they might differ in their opinion.
A ban on driving for non-essential purposes wasn't necessary for *you* because you're still here. On the other hand, if we ask the families of some of the 40,000 Americans who died in auto accidents in the 2019, they might differ in their opinion.
It would be interesting to hear this deepfaked in the style of Les Mis
Yes! Please!
Or at least a fan-made video in the style of We Love Xkcd <http://olganunes.com/projects/we-love-xkcd/>
Ups, I meant to reply to this. See my comment above :)
I would love this. Also Daniel Speyer's ACX Still Alive <https://astralcodexten.substack.com/p/still-alive#comment-1098394>.
I can contribute my voice for this project. This one would benefit from a number of different voices though. And a chorus.
I do the podcast - I'm going to do a very simple version of this now by myself but it would be so fun to do it properly :D Send me an email if you like! :) astral codex podcast at geemail.
I'll mention for common knowledge purposes that Solenoid and I corresponded and would love to join with some more people to record an actual chorus!
ENCORE!!
I love this! You had me holding in laughter so my coworkers don't notice.
> Drink from the pool. How clean the taste.
I sure hope this is poetic license, because drinking from a random pool of water will likely kill you in all kinds of interesting non-Covid-related ways :-)
At least in the wilderness in developed countries, I think the risk tends to be somewhat exaggerated - giardia, for instance, is actually quite rare if you take reasonable precautions (drink from fast-moving clear streams, favoring higher elevation). See e.g. sci-hub.st/10.1016/s1201-9712(00)90102-4. "Drink from random pools of water" seems to work pretty well for most species, and for humans for most of history.
For what it's worth, both me (as a child) and my dog (later in life) have had giardia, and though it really sucks, it's extremely unlikely to kill you. Hell, CDC seems to estimate 280 million people a year are infected (worldwide, so not nearly all of the US).
Clearly, though, yeah, stick with running water in most places and you're probably safe even from giardia. It's a somewhat interesting instinct to see that my cat won't drink water unless she watches me pour it. She otherwise goes up to the bowl and meows at it until I replace the stagnant water with new water.
It's funny, the one time I took a drink from a mountain stream (okay, okay, second time) I got giardia. We didn't pack enough water for a long alpine hike on a sunny day, and those mountain streams look so pleasant...probably wasn't deep enough. I can confirm the giardia, while annoying, is not so bad if you're in good health.
https://www.aphlblog.org/wp-content/uploads/2012/02/Oregon-Trail.jpg
Pretty much by definition a "pool" is not a "fast-moving clear stream".
So I read through that article and it doesn't seem very convincing. It is basically arguing that there is very little evidence in the literature for outbreaks of giardia from drinking back country water. I believe that this is probably true but a) the vast majority of people who are in a situation to drink such water know not to unless necessary and b)As others have pointed out, it's not really that serious of a disease (very unpleasent but unlikely to be life threatening) so I would be very surprised if most cases ever got reported.
Here is a source talking about the under-reporting of Giardia: https://sci-hub.st/https://doi.org/10.1016/j.ijpara.2018.07.003
Evidence that wild animals can have very high rates of Giardia (so it isn't really "working that well": https://aem.asm.org/content/aem/78/24/8762.full.pdf
This magazine article collates a bunch of studies of prevalence of various waterborn pathogens at various places: https://www.discovermagazine.com/planet-earth/backpackers-dont-listen-to-slate-science-does-support-stream-water-treatment
They are pretty much all higher than I am comfortable with.
Yes, _some_ streams in some areas are likely ok and not all waters are equally risky. But I think the risk of getting _some_ kind of disease, from unpleasant diarrhea to debilitating illness, is very non-negligible. I think I'll keep bringing my katydid with me.
judging by the number of my friends who say the vaccine means they can lick [eyeballs/subway polls/toilet seats, depending on who i talk to], I'm guessing it's a tongue-in-cheek participation in "the vaccine makes me invincible" jokes
Also, I recently* drank from a giant** pool/river of water that formed during a terrific rainstorm. It was great. Do I need to worry about any health risks?
*less than a month ago
**about 5 feet(?) wide, 100 feet(??) long, and a few inches deep
How much time delay did the FDA actually cause? Wasn't a lot of the time necessary for figuring out a formulation of the vaccine that would deliver the mRNA in an effective way, and working out manufacturing and other logistical concerns?
Yeah, I don't buy into all the FDA hate. Sure, they could've acted faster than they did, but that's not the same as saying they shouldn't exist at all. When I take a vaccine, I want to know that it's effective (and ideally, how effective), and that it is unlikely to kill me all on its own. Without the FDA, how would I get this information ? From the New York Times ?
From pharmacies like CVS, Walgreen's, Rite-Aid, etc. They have an extremely strong incentive to get it right. The FDA has a strong incentive to not get it wrong, which are not the same things.
Agreed on the distinction between right/not-wrong, but how do pharmacies have any strong incentive to get it right? They make their money from selling drugs to people, not curing them. Obviously killing all your customers is bad for business, but there's plenty of profit to be made selling placebos, or for that matter drugs that kill only a small percentage of customers.
If that were even remotely true, the FDA would not exist in the first place, because the history of 19th century drugs would've been one of serene effectiveness, instead of the shop of horrors it actually was, and which drove the outcry that resulted in the creation of the FDA.
Is this the same FDA which allows people to eat industrial lubricating oils (PUFAs) as if they were food?
Yes, yes, it is.
I'll put my trust in the power of competition. We have much easier access to much better information about drugs now. Can you possibly agree that times change, and the need for government regulation changes, and that government bureaucrats don't want to shut down their bureau?
CVS and co demonstrably do not care about efficacy. Good news, though, they give free shipping on their homeopathic medicine if you order at least $35. Even better, they're "CLINICALLY PROVEN."
https://www.cvs.com/shop/content/homeopathic-remedies
Bizarrely, I'm not actually sure if these are all homeopathic or just marketing themselves as homeopathic.
Yeah sorry. Anyone who calls polyunsaturated fats "industrial lubricating oils" has such a deficient understanding of basic chemistry that means his opinion on this subject is worthless.
Hey, did you hear the FDA allows hydrogen oxide as the first (most plentiful) ingredient in *childrens' shampoo??* An outrage!
You use many words to say very little.
The general state of science, medicine, and other technology in the 19th century was obviously quite different.
Yes, that it was. But the general state of human nature, credulity, and tulip-bulb memetic mania hasn't changed a bit.
That said, I personally am much in favor of the FDA assuming a purely advisory role. That is, it should simply state for the record, in language accessible to anyone who paid close attention in high-school chemistry class, whether food or consumer product ingredient X, or medicine Y, is/is not safe, and for whatever reasons. To additionally restore a little budgetary rationality and enhance its support of Darwinian natural selection, it should charge a very modest consumer fee (like $100/year for a subscription) for public access to this info.
Because that would restore a little of the natural evolutionary advantage of intelligence, prudence, and due diligence.
That's nice for you, but people who prioritize not dying of COVID-19 should have the option of not waiting so long.
No they should not. The potential *future* victims completely dwarf the number of present victims, and there is no respectable collective moral basis for sacrificing their welfare -- which is critically affected by whether we gather reliable data on vaccines and treatments today -- for the welfare of the much smaller number of current suffers.
They can dual-track and require companies to gather data: https://marginalrevolution.com/marginalrevolution/2019/12/fda-dual-track-legislation.html
We also had data relatively early on that the vaccines weren't killing people like COVID was. The time spent in trials meant more precise numbers on how effective it was (particularly as the epidemic increased the number of cases in the control group, which wouldn't have been necessary had we done human challenge trials), but we really only needed a rough idea of efficacy in order for it to good idea to prioritize distributing the vaccine:
https://marginalrevolution.com/marginalrevolution/2020/12/herd-immunity-is-herd-immunity.html
Dual-track is worthless. There's no way of ensuring the control and experimental groups are sufficiently similar in what amounts to an open-label study with self-selection. That is especially true in something where the inherent risk is already painfully (from the point of statistics) low.
The fact that precision in efficacy was necessary should have been amply demonstrated to you by the recent J&J vaccine hold. How are we to *know* whether a low number in a million risk of severe blood clots does or does not outweight the benefit of the vaccine, if we don't know the benefit precisely enough? The *only* reason they were able to rapidly resume using that vaccine is because the clinical trial data was precise enough about the benefit. And this kind of thing is not rare, it happens *all the time* in new therapies. They didn't just pull those requirements out of their ass, you know -- they are the results of a century of painful experience.
Certainly a challenge study would've gotten the data much faster. If you want to write your Congressman and suggest enabling legislation that would fully immunize Big Pharma from liability should they produce a thalidomide-sized screwup in the process, go right ahead. Until then, this is magical thinking out of contact with the real world of legal liability in which they dwell. Vaccine developers do not crave FDA approval in the altruistic interests of their future customers, they do it so that when the inevitable $1 billion lawsuits arise because winsome photogenic 3-year-old Susie became autistic 18 hours after vaccination, or possibly 12 hours before, hard to be fully sure, and besides why does it matter? -- they have more than a fig leaf heading into the pre-trial settlement talks.
The benefits of the vaccine are so large relative to the number of clots that we would know the vaccine was beneficial even without very precise data.
The data relevant to potential *future* victims, can be as easily gathered It looks to me like you're suggesting that we need to hold the present victims hostage, because you can't imagine any other way to get the tests and data you want except by saying "no vaccine for you until you've done your chores!"
That's correct. I don't know of any other way to get the critical data than the tedious procedures laid down by empirical science. If you do, you should publish. An instant Nobel in medicine awaits you.
"Hundreds of thousands must die, to prevent freedom of choice from gaining a foothold in our society!"
No, no, you mistake me. I'm fully in favor of freedom of choice. In fact, while I think the FDA should still required these tests, and continue their effective taxation on new therapies to pay for them, I think their conclusions should be merely advisory. I am 100% in favor of fools rushing in where angels fear to treat being allowed to exercise their personal liberty, as this has a decent chance of restoring the health of the Social Security trust fund by lowering the number of fools who reach age 65.
When Pfizer announced the results of their stage 3 trial, I skimmed through the report they released. I do statistics for a living (almost purely frequentist, not Bayesian, their report used a Bayesian method but something simple enough that I had learned it in a first year PhD class). Unless they had completely faked their data (not sure the FDA is set up to catch this in any case), then yeah, any of the thousands of people with this level of training would have been able to tell you that "yeah, it's highly effective". More trustworthy than the NYT I would think.
And really, we can almost ignore the statistical training - when you have 90 cases in the placebo group, and 5 in the control, you don't need to know how to construct confidence intervals to be able to correctly infer that "huh, seems to help a lot".
The FDA then took something like 3 more weeks to read through this report and give Pfizer the green light, so some pretty low hanging fruit I would think. Could have been done in one afternoon.
Plus they absolutely did need to test the vaccines thoroughly. Lots of potential vaccines don't work, and using mRNA for vaccines is obviously not something that had ever been done before.
The FDA certainly did drag their feet more than they should have! Challenge trials could potentially have done a lot, and every extra week added to the timeline killed people! But no, we couldn't have just started giving out vaccines a year ago.
"we couldn't have just started giving out vaccines a year ago"
Actually yeah, we could have. Maybe we couldn't have said "guaranteed!" but I think they could have manufactured a batch by then.
No, lots of time wasn't necessary for figuring out the formulation - IIUC they had more or less the current formulation around the time COVID cases first started being reported in the US.
I don't know how the manufacturing supply chain worked, but I do know from other medications that most of the difficulty of manufacturing drugs is getting the factories FDA-approved. It looks like Moderna's first factory was approved by the FDA in December, but I don't know enough of the backstory to know if this was good or bad.
Like I said in my comment below, I think there's a lot to be said for having vaccines manufactured in a factory that was inspected by someone more competent than, say, a New York Times reporter. Yes, this process does take time; but it's a tradeoff. What percentage of people who take a new vaccine should die from it ? The answer is not as simple as "as long as fewer people die from vaccine than from the virus, we're all good" -- witness what happened with AstraZeneca, for example.
> The answer is not as simple as "as long as fewer people die from vaccine than from the virus, we're all good" -- witness what happened with AstraZeneca, for example.
Yes, that IS the answer! Or maybe you just answer trolley problems differently and are never willing to 'act' if it kills people.
I'd think we'd always want to choose the option (including NOT 'acting' as an always available option) that kills less people, but hey, what do I know – maybe death is really good?
And "what happened with AstraZeneca" is what we're arguing AGAINST. It seems circular to use that as evidence against not doing that.
No, that's completely the wrong answer. The right standard is "fewer people die from vaccine than are saved by vaccine" - otherwise, I could give hundreds of thousands of people bottles of strychnine with vaccine written on them, and claim it was OK because coronavirus was killing even more!
With that said, if you want to minimise the sum of an increasing function x and a decreasing function y then setting x=y as you're doing will never be worse than the best outcome by a factor of 2.
And the number of people killed by the vaccine is probably in 1 or 2 figures, and the number saved is probably in... I don't actually know, but I'm going to guess 6, although high 5 wouldn't surprise me, so in practice minimise total deaths from vaccine+covid is basically an identical problem to minimising deaths from covid.
I don't understand your point.
If you gave hundreds of thousands of people strychnine, they'd almost all die – that seems WAY more dangerous than COVID-19.
All-else-equal – we want to prevent as many deaths as possible.
> With that said, if you want to minimise the sum of an increasing function x and a decreasing function y then setting x=y as you're doing will never be worse than the best outcome by a factor of 2.
I don't understand how you interpreted what I wrote as being equivalent to "setting x=y". What's the "increasing function x" and "decreasing function y"? What are the domains and ranges of these functions?
My point is the we want to minimize 'deaths from COVID-19' AND 'deaths from vaccine'. But – in this world – the latter seems FAR smaller than the former, hence anything we've done to delay the latter has increased the former.
You are too focused on what's in front of the trolley to see what lies a little further down the track. Like I said, look at what happened with AstraZeneca. By all accounts, it is almost perfectly safe; however, just a few highly-publicized deaths were enough to get it effectively shut down in most places. If the probability of dying from it was even as much as 1% (as opposed to 0.001% or whatever it is now), most people would rather take their chances with the virus; what's worse, many of those people would be *correct* to do so. If you are young and healthy, COVID is not that big of a deal, so why risk blood clots ?
But the situation is even worse than that, because, absent FDA, a person couldn't even make an informed decision as per the above. Right now, the proposition is, "take this vaccine, and you will have a 90% immunity, with a 0.001% chance of death, and a 95% chance of mild side effects". Without the FDA, the proposition becomes, "take this vaccine, and it may or may not help or kill you or whatever, who knows, we are piping it directly from the experimental chemical tank into your veins". In this situation, most people would wait for the New York Times reporters to write some articles; and again, they would be arguably *correct* to do so (depending on how much you trust journalists, and personally I do not). Again, most people would skip the vaccination, and would be correct to do so.
The problem is with authorities who reacted irrationally to the AZ results. In the UK they embraced AZ, and it worked out very well for them. They actually have one of the lowest rates of vaccine hesitancy: https://marginalrevolution.com/marginalrevolution/2021/03/vaccine-roundup.html
The bloodclots were only even seen in young women. The people with the highest mortality have been elderly men. They should have had access to AZ long ago.
More on AZ and trying to act to appease the most irrationally afraid of vaccines and whether that could actually have a net benefit:
https://marginalrevolution.com/marginalrevolution/2021/03/the-eu-vaccine-bungling.html
I think your model is that 'the public' panicked over AZ whereas mine is that 'the authorities' are the ones that panicked (supposedly) _on behalf of_ the public.
"Without the FDA"
The problem is not that the FDA gathers data and indicates that things are safe and effective, the problem is that it prevents people from taking medicines that they want to take. I have no doubt whether the FDA exists or not, there will be a demand for rigorous evaluation that the market/society will provide.
Jewish kosher organizations are very much like 'private FDAs'.
The FDA certainly deserves a massive share of the blame for ruining rapid testing, which might've been able to massively cut case counts from what I've heard (though not sure why it didn't really help Europe, maybe they had similar regulations). But I would be very interested in a full accounting of how much the FDA actually held up vaccine manufacturing. I'd assume the Moderna factory was already producing vaccines at peak capacity before it was approved, right?
My God, I forgot about that. Such a disgrace.
>I'd assume the Moderna factory was already producing vaccines at peak capacity before it was approved, right?
Wrong. http://eqvanalytics.com/moderna-vaccine-tracker.html
Shockingly wrong, actually. I had assumed that Moderna at least started to ramp up production a few months in advance of approval, but it really looks like they didn't turn the dial up to 11 until the same week they sent their final approval package to the FDA.
We were talking on the old blog about how Bill Gates was going to fund building all the factories for the vaccines, even those that didn't work, so the winners would all be ready to go.
And then . . . nothing happened. No one did it. I don't know if Bill Gates actually promised something and quit or it was just a rumor or he wasn't able to do it, but one of the easiest ways to trade dollars for the end the pandemic was just shrugged off.
Yeah, I also haven't followed what happened with the Gates funding efforts. To be clearer, I guess the question is: how much was the factory holding back on production due to uncertainty about FDA approval, vs. just facing material constraints in scaling up the production process?
Here's a Derek Lowe blog post on the subject I found interesting: https://blogs.sciencemag.org/pipeline/archives/2021/02/02/myths-of-vaccine-manufacturing
The upshot is very few of these factories exist, and the companies that can have been scaling their manufacturing capability for some time now. I think the fact manufacturing / distribution still took months to scale up after approval shows that FDA requirements weren't the only bottleneck.
Not the ONLY bottleneck, but a large and probably the primary bottleneck
In hindsight, it was entirely bad. We took forever to test and approve production/use of a drug that turned out to be safe and effective. Once you get out of hindsight, it's harder.
I hold that in a world where we aren't all cowed into thinking this should take forever and that an absolutely-overkill and absolutely-overcautious testing regime aren't normal, we could have been in semi-serious production by May or so and escalated from there.
Whenever I say this the normal response I get is something like (incoming strawman) "Well, even though it Covid killed many millions, there's some small chance it might have killed some very small-in-comparison number of people in testing if it had been much more dangerous than it plausibly could have been, and those 1-1000 lives are much more important than the hundreds of thousands or millions who died otherwise", which reinforces my view. And we are totally prepped to take exactly as long next time; if we were willing to sacrifice 2 million people to save 50 there's no reason we wouldn't sacrifice 20 million.
> We took forever to test and approve production/use of a drug that turned out to be safe and effective. Once you get out of hindsight, it's harder.
You're looking at less than half of the problem. How many vaccines turned out to not be safe and effective? How many neither-safe-nor-effective vaccines would've been manufactured, marketed, and sold if they knew before hand that they didn't have to go through the FDA approval process?
You're looking at the holes in the planes that come back and deciding that it's useless to invest in armor on the places without holes.
First thing I should bring up is that you are pretending I said to abandon all testing rather than suggest that the testing isn't some sacred cow that's immune to improvements, impossible to speed up. That's a bullshit thing to do; you do it knowingly and that makes you sort of a bullshit person.
Moving on: Some non-strawman version of your metaphor is valid if a bunch of assumptions it makes are true - i.e. that the armor plating you are saying I'm abandoning is all necessary, and delaying the deployment of the plane isn't balanced against any other costs.
In the case of efficacy testing, the risk with faster testing isn't a risk of less efficacy - we get the same result (works or doesn't) with about the same certainty in a much shorter time with, say, challenge trials. But the decision was made that a risk to a small amount of people (say, 1000 deaths in some kind of worst-of-all-worlds testing disaster if I'm being really, really generous) was not ethical, while condemning hundreds of thousands of people to for-sure deaths, miserable and choking to death slowly on their own lungs, was ethical as shit and something we should definitely do.
I want to be gentle here, but I can't: You are a fine with this and it's going to happen again. This time it was 2-3 million, probably a quarter to half of which we could have saved at a really minimal risk to informed volunteers.
To flip your metaphor, this is you looking at a plane that could go intercept a nuke headed towards NYC; it's ready to go and the pilots are begging to be allowed to save those people, and everyone is aware that this would probably work, but you want and demand a 2 hour prayer to be chanted over the wings at no advantage because that's how it's always been done and you refuse to consider anything else.
I remember being exposed to the "notice when you are confused" framework back on SSC.
I'm not sure I personally am confused, per se, but as a society, we really should put some thought into how we found ourselves in a situation where the people anointed as "ethics experts" were seemingly all bar unanimously opposed to attempting to save millions of lives by letting a few thousand people voluntarily face a <1% risk of death, because we've always done it that way.
If I were designing the attributes needed to be endowed as such an "ethics expert" from the ground up, this would be a disqualifying, not a necessary characteristic.
The original claim in the verse was, "Two days to find the key
The rest, because bureaucracy."
Bugmaster claimed, "I think there's a lot to be said for having vaccines manufactured in a factory that was inspected by someone more competent than, say, a New York Times reporter. Yes, this process does take time; but it's a tradeoff."
You responded, "In hindsight, it was entirely bad."
"It" is a bit ambiguous here, but it certainly looks like you're arguing in support of the claim in the original verse. This is what I was disagreeing with.
The narrower claim that, "There are other options we could've used to rigorously test these candidate vaccines that would've been quicker without sacrificing the FDA's credibility or flooding the market with vaccines of dubious quality" seems obviously true, but, per Bugmaster, it's still a tradeoff.
On the contrary, if you got all 50 candidates, and did challenge trials immediately, you could have results on efficacy for all 50 options within a few weeks. Of course, we couldn't have that, because apparently it isn't acceptable for people to voluntarily and knowingly take a fairly small risk, even though doing so would likely save millions of lives, because to allow this would prevent medical ethicists from feeling smug (or some such - obviously this is uncharitable, but one has to wonder where self-appointed "ethicists" acquire there "i have the right to insist on killing huge numbers of people on personal whim" card). Notably, polling of the general public suggests they would have been ok with challenge trials!
Even without the trials, a policy of "each potential vaccine spit out doses and start vaccinating without delay, we'll record data to make clear what works as we go" would have saved lots of lives. In states like NJ, 10% of the nursing home population died. Did any of the 50 candidates kill 10% of the people in their respective trial? Has any vaccine ever in human history killed 10% of the people it was given to? There are trillion dollar bills on the ground here.
And it's not like incentives to prevent quack medicine are FDA or bust. If we didn't have the FDA, what would stop a company from putting arsenic in a vial and marketing it as an effective covid vaccine? Well, the criminal code, for one (fraud, for example, being illegal).
>You're looking at less than half of the problem. How many vaccines turned out to not be safe and effective?
Lots of vaccines turned out to be only moderately effective. Even a moderately effective vaccine six months ago, would have saved more lives than our perfect vaccines will now. I don't think any vaccine ever has turned out to be unsafe in a way that would be even a rounding error compared to COVID deaths.
But, more importantly, it doesn't take a year to figure out whether a vaccine is safe and effective. If the metric is "better by an order of magnitude or two than the status quo", we probably knew that by June 2020, and we certainly could have known it by June 2020.
If you're saying June, then you agree that we needed four months of testing, not the two days Scott claimed (presumably glibly) in the verse above.
My point is that *some* testing regime was necessary, and it sounds like you agree.
The Pfizer phase 2/3 clinical trial only started in July 27th, so we certainly didn't know by June 2020. They had 162 cases in the placebo group when the stopped and that's certainly more than necessary even for a modest effect size. But how much faster would it have been to get just, say 60 cases? I'm guessing two months faster: you still need to recruit them, give them two doses vaccine, and then wait to see if they get COVID. COVID case numbers were growing rapidly into the fall, so I imagine they got most of the cases towards the end of the study. Even with the count they had, they just barely exclude 90% efficacy, so just barely making it past your "order of magnitude" metric. And all that is ignoring the possibility of examining any subgroups.
That said, I'm all for creating better systems so we're more prepared the next time something like this happens. Plenty of room for improvement, but the vaccine development, testing, and roll-out has been pretty reasonable by my judgement.
https://www.nejm.org/doi/full/10.1056/NEJMoa2034577
You don't want moderately effective vaccines. That puts strong selective pressure on COVID for escaping such immune responses and creates the perfect environment for cultivating and propagating such strains.
It's the same reason why you want people to do their full antibiotic treatment course - because if you have some strains that are marginally resistant, they will be wiped out by the full course but survive the partial course and then reproduce and if you do that several times over you can end up with antibiotic resistant bacteria because each time you selected for the strains that could resist a bit better, and with the weaker challenge there were actually survivors rather than total annihilation.
This is very much a 2020 hindsight thing, right? (pun intended)
Look at all the vaccines we tried:
https://www.nytimes.com/interactive/2020/science/coronavirus-vaccine-tracker.html
In the end, the US went with three - Moderna, Pfizer, and J&J. Sure, Moderna's vaccine was developed in two days, but maybe one of the failed 50 vaccines was developed in one day. We're not omniscient, and we tried many vaccines because we weren't sure which would work.
AstraZeneca is quite effective elsewhere, even if the US is too stupid to use it.
Do you think that strikes at the crux of my argument?
I mean, yes, it does quite a bit. We have at least 5 vaccines that work (Pfizer, Moderna, J&J, AstraZeneca, Novovax) plus several more that probably do but are harder for western countries to verify (Chinese + Russian + Indian vaccines). How many actually failed outright and were abandoned (rather than merely being given up on because development was going too slowly such that they would be uselessly far behind)? Plenty of others at still at various stages of development (unclear if they will go anywhere, because we have at least 5 high quality vaccines at this point).
Overall, the success to failure ratio here is quite high!
(Also, if you have 40 candidates, and each has a low chance of working, challenge trials are exactly the sort of thing that would be helpful).
Peter Hotez has been pushing the vaccine his group has been developing on the basis that it's much cheaper than the vaccines currently in use. IIRC he gave $1.50 as the cost per dose they're aiming for. Just looking quickly, it appears that COVID vaccines have the same lack of transparency in pricing that's endemic to the pharmaceutical industry generally, but that price point would be considerably below all the figures quoted here.
https://www.bmj.com/content/372/bmj.n281
My understanding is that the Russian vaccine works much better than the Chinese one. Even the Chinese government doesn't have much confidence in their own vaccine.
This is exactly the point. I would figure folks reading Scott (and, well, Scott) would be familiar with survivorship bias.
It isn't actually survivorship bias if you have 50 candidates start, and 8 are successful and start being used while the other 42 are still going through the process. The relevant thing is successes vs failures (and it also depends on what the failures were - if they are just "abundance of caution" and "seems less effective than other candidates that are further developed" then it is questionable to count them as failures at all).
If they're still in Phase 1 or Phase 2 at this point, then they probably needed to re-do parts of them. They didn't get the initial dose levels quite right, there was some unexpected difficulty, etc. So they had to re-do the Phase with a different formulation. Same is true if their Phase 3 was delayed.
Looking at the handful of manufacturers that jumped through the hoops right the first time and decided that this means the hoops don't matter is a form of survivorship bias.
> IIUC they had more or less the current formulation around the time COVID cases first started being reported in the US.
This is true as I understand it, but how many other people claimed they had cures for COVID at the time? IIRC, there were dozens of vaccine prospects last summer. (This was one reason I was optimistic about getting a vaccine much faster than historic standards indicated was possible; there just seemed to be so many potential options!)
Point being that you need to weed out the vaccines that work from the ones that don't, and you have to show that they're safe. The rest wasn't bureaucracy. And even if you say, "Human challenge trials", well, those still take some time. So now you're quibbling about how quickly the bureaucracy went, not whether it serves a necessary function.
Is the vaccine manufacturer really going to market a vaccine they haven't tested? Do you think that would be a profitable proposition for that company?
Yes?
Not Pfizer or J&J or whoever, but once you decide that we don't need to test anything and everyone's proposing "novel" or "experimental" development regimes, "vaccine manufacturer" quickly becomes "anyone with a needle and some essential oils".
My point is that the FDA had a crucial role to play in giving assurances to the public that the things it says work actually work. You can argue that they didn't balance the risks optimally in this case (I agree!) but that's different from saying that all the clinical trails were needless bureaucracy because the final Pfizer (or whatever) vaccine was very close to the original formulation that scientists had created last February.
We didn't know if any given vaccine would actually work.
If drug development was this simple, we wouldn't see most drug candidates fail out of testing.
Numerous vaccines have been tried over the years. We still don't have one that works against a number of diseases (including HIV).
Merck made two vaccine candidates against COVID - V590 and V591 - and neither panned out, as they failed to give sufficient immune response. So it's not like we knew for sure that this stuff would work beforehand.
We're having enough problems with vaccine reluctance as is. If you gave twenty million people shots that didn't work, it'd be much worse - not just for COVID but future vaccines in general.
Properly testing the vaccine and making sure it actually worked before giving it to everyone was absolutely the right thing to do.
It's also worth noting that giving a bunch of people subpar vaccines that only gave a marginal immune response would select for COVID strains that could escape weak immune responses, which could make future vaccination efforts more difficult as well and potentially make the disease more dangerous in general.
Well, essentially all of it, from one point of view. Almost all the time was the time needed to allow volunteers who got the vaccine as part of the Phase II and III trials -- mandated by the FDA before approval -- to go about their business and either (1) die of unanticipated side-effects, or (2) catch and die of COVID because the vaccine didn't work, or (3) *not* catch COVID and not die while their compatriots in the control arm are -- which is what tells us it works.
We need a lot of time because it is considered unethical in these circumstances to give half the volunteers the vaccine and then deliberately expose *all* of the volunteers to COVID and see whether fewer people in the control arm die, which would be much faster. We also can't just wait a certain amount of time, because we don't know the rate at which people are being exposed. So we have to wait until we accumulate enough cases in the control arm that we can be reasonably confident everyone (control and experimental) has been exposed.
We certainly could have had the vaccine a lot sooner if we didn't insist on knowing whether it worked ahead of time.
The FDA did a good job with the vaccines. Knowing that they are safe and effective is important. Giving people a bunch of vaccines that didn't work would have been a disaster for vaccine acceptance. Giving people a bunch of vaccines that killed a bunch of people or made a bunch of people would have likewise had a large negative impact.
While pausing the J&J vaccine was dumb, spending the time to get the vaccines right was a good thing, both for the current pandemic and for long-term vaccine acceptance.
Long time reader - first time commenter. This is the best. I love this. I don't even know the song and I love it. Thanks for the nuance and articulation.
It seems like there's a new divide growing between the "Vaccinate but change nothing yet" segment and the "I'm vaccinated now; I'm going to live my life again," segment. That may well map to altruistic vs. selfish — cynically, the latter type were only pro-lockdown to save their own skins — but there's also a trust dimension: a lot of people, and I'm one of them, think the CDC's recommendations aren't being made rationally.
Then there's the plain feeling that "I've done everything I was told I needed to do, and they moved the goalposts YET AGAIN. No. No more." Not a great argument, granted, but it may prove a major obstacle to compliance.
Yeah, the constant moving of goalposts is NUTS, and it's worsened by the (I think very corporate-informed!) refusal to ever acknowledged that what happened before was wrong or mis-guided and an insistence on the message of "oh you must have not understood properly before, you idiot, this is why we need to big brother you to death with more regulations"
> the (I think very corporate-informed!) refusal
I think this is backwards. 'Corporations' are a relatively recent social technology, but I'd bet at long odds that that same 'bullshit' has been going for basically ever. (I mean, have you ever read The Bible?)
But this is a particularly maddening political tactic/strategy that I wish would be widely punished :)
<rant target="them">Admit your fucking mistakes people! Everyone else would trust you _more_ if you didn't try to gaslight us all the time.</rant>
My viewpoint (and I know others with this opinion) is that I was last in line for the vaccine. At this point, if I’m fully vaccinated, everyone else I come into contact with could be too. If they decided not to be, that’s on them.
Yes, there is some risk still from COVID, but it’s much less so now. My odds of killing someone who got the vaccine are pretty low at this point, so I’ll accept those odds and live my life.
So I’m a mildly bad person, but I feel like the dichotomy you posed isn’t accurate, as my reasoning is in the middle of “I still must not hurt people” and “I’m safe now whoopee”
Anyone who has commentary on this stance, I welcome it.
That makes total sense. I managed to get the vaccine earlier than most due to my job, but I feel similarly now that it's been available to everyone.
My commentary: it sucks that you feel the need to say you're a mildly bad person! I'm, firmly in your camp.
Small data point: my wife is having a daughter soon, and my sister asked if she'd be allowed to be around the baby if she wasn't vaccinated - I said "no" not because I actually give a sht, but because if you need a reason to get the vaccine and I'm in the position to give you one, then I'm going to do that.
I’m a utilitarian, so unless I’m living in a box and donating all of my money to the poor, I qualify as a bad person by my own moral scheme. Ah the cursed existence
But you're a person, hence you 'deserve' utility, your behavior is constrained by your nature (now, so including the history of any nurture), and obviously guilt and shame haven't made you a _better_ person, so maybe there's some other means for you to be a 'good enough' person (or maybe – what I suspect! – you _already are_ good enough).
Utilitarianism doesn't imply that _everyone_ should be a utility maximizer – or, at least practically, it seems silly to just _assume_ that transforming people as they are into utility maximizers isn't also just another thing to run thru the 'how to maximize utility?' function we'd like to approximate.
I kind of agree and disagree with you. I think that almost everyone ends up as a bad person (or at least an insufficiently good one) in their moral systems. I mean even seriously fundamental christians generally feel like they break some of the rules occasionally.
But I don’t think that means we should feel guilty per se. I see “being a better person” as a good, just like apples or reading. It costs things to be a better person, but it’s not the only thing we are optimizing against.
I do agree that not considering yourself in your utility function is an easy trap, but I do think I’m probably past the point where in many ways it would generate more utils if I gave my stuff away.
But who lives at the edge cases of their moral code?
" I mean even seriously fundamental christians generally feel like they break some of the rules occasionally."
Just to be pedantic: as someone raised at least fundamentalist adjacent, they're more likely to feel like they break some of the rules constantly, rather than occasionally. All have sinned, and all that. As I kid I understood there were obvious sins, like lying or stealing, but also had a sense that I was probably sinning constantly anyway and was just unaware of it.
To be clear...I don't. I think I'm a good person. My moral system is screw over other people as little as possible via direct action. Everything else is permissible. Be kind is gravy (I am) but isn't a moral imperative.
I might not be a good person in other people's moral systems, but they can go eat a d*ck.
> I think that almost everyone ends up as a bad person (or at least an insufficiently good one) in their moral systems.
Oof – that's not my view or belief. I think most people are (overall) good people.
> It costs things to be a better person, but it’s not the only thing we are optimizing against.
That's a good point. I think it's true too.
> But who lives at the edge cases of their moral code?
Some people sure seem _closer_ to the 'edges' of their morality.
But I think 'maximizing' is a trap too – 'maximum' is _better_ but not mandatory. We should _aim_ at it but not beat ourselves up for falling short – unless beating ourselves up (in some way, and some amount) is actually effective at making ourselves better people.
I admit it's all tricky! I'm very sympathetic to the idea that deontological and virtue-ethic moralities are, in a very real sense, 'practical approximations' to morality-complete consequentialism (which is impossible to calculate or estimate directly).
It seems to me that a utilitarian should attach a quantitive value to themselves rather than a categorical one.
Or if you are going to use binary category of good/bad, setting the threshold at perfect doesn't strike me as useful. By your logic, a Rolex is a bad watch because it's less accurate at timekeeping than an atomic clock.
Really, even a quantitative value shouldn't matter right? For results based ethics, the past doesn't matter so much, mainly what matters is the consequences of your next choice, regardless of where you fall on the good/bad spectrum. When I said "I'm a mildly bad person" I mainly meant that the current choice (almost ignoring CDC once everyone has had the chance to be vaccinated) is sub-optimal.
So while overall many of my choices fall short of utilitarian, if I give myself an honest D&D categorization over life choices' outcomes, I'd say I'm clocking in at Lawful Neutral so far.
That’s hilarious to me.
My girlfriend’s mom is like 70, and the mom freaked out on my girlfriend for requesting that she wear a mask - the mom has some weird trauma response to anything that looks like being controlled, so it was like “you’re withholding affection to control me!” That definitely wasn’t the motivation. But I’m now imagining that my gf and I have a kid, and tell her she has to get vaccinated, not because we first order care, but because we actually want to control her. All of that conspiracy thinking would suddenly be true..
Totally agree with you. Everyone can get it now. There’s tons of open appointments where I am. I shouldn’t have to suffer for someone else’s choice to not get vaccinated. But neither should our healthcare system. That’s the really frustrating part.
This is entirely reasonable IMO, though it does seem like some places in the US are still in a place where vaccine availability is limited and people who want shots can't yet get them. Assuming that's not the case where you live, I don't think there's anything to feel bad about in this approach.
If "vaccines have been widely available for 6 weeks" isn't the bright line for getting back to something pretty close to business as usual... what is? Covid and it's variants will presumably become endemic, not stamped out entirely. Any case-count or mortality based cutoff will be arbitrary. Variants have some uncertainty but all indications I've seen so far are that the vaccines remain very effective in preventing hospitalizations and deaths. Kids aren't eligible yet but their risk of serious illness is very low.
On the exact morning that vaccines opened up to all over-16s in my state (CA), I signed up for the earliest available appointment slot from my medical provider. Thus, I am scheduled to get my first shot in 2 weeks from now (4 weeks from when I was allowed to sign up), and I expect to have full vaccine effectiveness around the end of June.
Bay Area? In Los Angeles we have an over abundance of slots and even opened city sites early to 16 and over. I feel like the rollout of vaccines could’ve been better. There have been too many open slots here for a long time.
The one part of this that's complex is the issue of mutation. The virus can only mutate when it's actively infecting, and the more people who have active infections -- whether or not they are themselves symptomatic, or affected -- the better chance there is of something really nasty arising out of the evolutionary soup that is proof against all known vaccines and sending us back to square one.
So there is actually a bit of a collective action problem here: your vaccination cannot defend you against a mutation that can evade the vaccination, and your neighbor's refusal to get vaccinated actually boosts the odds of that happening. It can't, quite, be every man for himself because of the problem of mutation. A rational collective response would perhaps *enforce* sufficient vaccination to drive the probability of bad mutants low enough that the collective would not be likely to suffer the consequences of some independent-minded Typhoid Mary unwittingly breeding a super-variant.
I don't have a good answer to this, but I think it complicated the moral reasoning rather, since there is a way in which you *are* your brother's keeper, so to speak, in terms of the threat of mutants. It's possible for actions you take to have almost zero consequences for you but deadly consequences for others, perhaps even many others, and it's possible they have no defense at all against those consequences (other than compelling your obedience to some social directive).
The kids aren't vaccinated yet, though, and I have no idea how to assess the potential long-term impact of an infection for them. Having thankfully never had to come to grips with the subject, I was surprised to discover that 90-95% of polio infections are asymptomatic and less than 1% resulted in paralysis (I'm hoping I'm remembering those figures rightly), yet we can see in the historical record what an enormous impact polio had given the size of the denominator.
My son is starting kindergarten in the fall. If I've understood correctly, the school district has full surveillance testing in place, though I have more research to do and a conversation with the principal to schedule. If their protocols are sensible and not just cargo cult practices, I'll be willing to take that risk given that they will have the data to trace infections in near real time. Given how little we know, I'm inclined to be quite conservative with him outside of school and would appreciate others' efforts to keep the amount of virus circulating in the population as a whole low.
Buy an air purifier for the kindergarten room (unless you know for sure they have already taken care of ventilation in a way you can trust, such as having windows open).
It'll still take some time to get shots into arms. My state is only at ~1/3rd vaccinated despite being "available to everyone" for a month. On the other hand, it seems like the rate of vaccination is falling, so maybe we're more constrained by vaccine hesitancy than finding an appointment.
Moving the goalposts is an own goal.
I'm definitely on the, "I'm vaccinated, time to party" side of that line. For context though, I was locking down about a week before anyone else was, and wore a mask AT ALL TIMES (outside my home) until I was two weeks out of my second dose. Basically, my view has always been that the point of locking down was to buy time for a vaccine. Vaccination is my schelling fence - before vaccination, I was extremely cautious, but after vaccination I've given myself permission to go back to normal. From a utilitarian approach, full vaccination means that the impact to my own happiness is no longer outweighed by the tiny amount of good I might do for others by curtailing my activities. From a categorical imperative perspective, my behavior would universalize quite well.
Not all of us anti-lockdown people are anti-vax (I'm currently suffering through Moderna side-effects), and we're happy to welcome the "I'm vaccinated now; I'm going to live my life again" segment in blowing raspberries at the CDC et al.
I've always been conflicted on lockdowns: on the one hand, they almost certainly resulted in at lease some people still being alive who would have died already, though it's not yet clear how many; on the other, the way they were implemented was capricious and amounted to yet another form of regulatory capture benefiting large businesses and hurting their small competitors.
That, and I think the burden of the lockdowns fell disproportionately on the working class; members of the PMC like myself were much more likely to be able to work from home. And don't get me started on the hypocrites who were pro-lockdown… for everyone else. Give me an honest anti-masker any day over someone who believes no price is too high for other people to pay for their safety.
Yes, there’s something especially grating about people who advocate for a certain rule-set then violate it. In a way, I’ve found this to be a good practical argument against the sort of intuitive “Parliament House” utilitarianism which some people fall into, where some individual advocates a policy for the masses, yet privately violates it themselves.
Not only is it, as you say, hypocrisy, but it also burns the commons in a sense- “the masses” won’t respond by thinking “oh this person is merely individually flawed, or indeed so intelligent that they’d figured out how to do a more nuanced version of the right thing, which We could not do”, they’ll think “these swines don’t care so the rule is just virtue signalling or an attempt to exert control”. And that’s how you get even greater non-compliance with life saving measures.
I think externally-imposed lockdowns mostly served to postpone deaths from the (roughly) Spring 2020 "first wave" to the Summer 2020 "second wave", after which all the exemptions and violations rendered the lockdowns ineffective. The people whose lives were actually saved were the people who were scrupulous and disciplined in their own self-isolation, or other precautionary measures, and I think we need a word for that other than "lockdown".
If you have a high-trust society, externally-imposed lockdowns probably shouldn't be necessary. And if you want to keep living in a high-trust society, the sort of capricious, arbitrary, and hypocritical lockdowns we got probably shouldn't be allowed.
"I've done everything I was told I needed to do, and they moved the goalposts YET AGAIN. No. No more."
This is how I feel.
Locking people into sterile iso-pods "for their own protection" is altruistic only under the most warped and perverse definitions of the term.
This maps to aptitude for risk management. Some Americans still have the knack, as all humans necessarily once did, of saying "this is dangerous, but it is *worth the risk*, which I have taken reasonable measures to reduce". Other Americans, first-worlders generally, insist on the thing that only recently became possible of dividing the world entirely into that which is Absolutely Safe and that which is Intolerably Dangerous, and insisting that the Intolerably Dangerous be wholly excluded from their life.
For the first group, once they and their friends have been vaccinated, the remaining risk (of death, of serious illness, or of making someone else dead or seriously ill) is relatively low, probably within the background noise of modern life. COVID + 95% effective vaccine is about as deadly as seasonal influenza + usual crappy flu shot, and justifies about as much effort.
For the second group, COVID was flipped over into the "Intolerably Dangerous" category a bit over a year ago, and the only way for it to flip back to "Absolutely Safe" is for there to be no press coverage of ongoing COVID deaths, etc, for several full news cycles. Or maybe one news cycle of unambiguously positive "vaccine works, crisis over!" coverage, but let's be realistic. Until then, they have to continue treating it as an Intolerable Danger, in the manner they have become habituated to.
Citizen Four is just the same lyric. Thank you.
CITIZEN FOUR:
When I get free, you won't see me
You've been so cruel
CHORUS:
Lock down, lock down
We're sorry, that's the rule
Oh, I LIKE that it's the same lyric. That got me the biggest laugh by far!
Love it.
Swelling and redness.
Am I ill? Reacting? Sign
The card. Now - go home
“Side effects haiku”
I liked it.
"ME:
I'm thirty-six and fit!
NURSE:
You're in the herd"
I'm thirty-eight and not fit, but working on it. Yes, I am in the herd, and I'll continue to wear a mask in public indoor spaces during cold & flu season. I always used to think that the East Asian kids wearing masks on my nearby university campus were paranoid, but the bacteria & viruses really are out to get us.
Chris Hayes' MSNBC show informed me of some CDC stats last month that I still find mind-blowing. The US usually records between 150-200 pediatric deaths associated with the flu each winter. This past winter, there was 1! Positive flu specimens dropped by two orders of magnitude year-over-year in late February.
https://www.youtube.com/watch?v=R5g_p_TFoiw
We're going to have to give some of this back to reopen the economy and I certainly don't think any level of the government should mandate masking on a long-term basis. That's a blunt instrument to reserve for a crisis; this is the sort of thing that should evolve as a social norm if it proves worthwhile. If we can keep even part of that improvement with masks & better hand hygiene, it would be criminal to drop that because we just want to go back to the way things have always been, especially as a mask is just about as close to a free lunch as you'll find in this life.
I wish we could run this experiment on a national basis this coming winter to parse how much of the reduction in flu infection came from distancing and how much came from masks & hand hygiene, accounting for flu vaccine effectiveness & takeup, but that ship sailed many months ago. Maybe we'll be able to collect some data if a norm about masking establishes itself in a few smaller university towns like mine.
For what it's worth, a decent chunk of people I know (though far from the majority) are planning to continue wearing masks both whenever sick, and possibly just whenever in large crowds moving forward.
Interestingly, at least one of the reasons had nothing to do with health - it serves as a barrier (in the same manner as large headphones) to being harassed by strangers in public.
I feel quite the opposite. Masks are a signal of tacit support for massive human rights abuses, including abuses directed at me. I consider past abuses to be a predictor for future abuses, and for that reason I will treat people who I see wearing masks in public differently.
Er... like what? Why?
I already gave my explanation. Past human rights abuses are a predictor for future abuses. If you were willing to go along with the regime on this, then I should be fearful of what future abuses you will support them committing against me.
Sorry, I meant: what human rights abuses are you referring to? It's not clear.
What are you going to do, give me a dirty look? I suppose you must have the right to swing your arm freely and touch my mask, given that its thickness delineates a zone of a millimeter or two between your fist and my nose, but I must warn you that I flinch easily and retaliate immediately. Tit for tat is a dominant strategy in most two-person games.
It's the lockdownists that threw the first punch, but I'm not one for tit for tat. I'll just generally disregard the existence of people who wear the symbols of the regime. No direct hostility, but no help or courtesy either.
And how am I going to distinguish you from literally everyone else in a retail environment?
I know who Marshal Petain was. I do not take an accusation of being a collaborator lightly. Be specific. What regime would I be supporting by continuing to wear a mask in public?
Lockdownist regimes the world over, which includes the one in the UK.
I mean, you've already acknowledged that you don't care about other people. You're just trying to justify your pre-existing selfish behavior.
Honestly, if we'd locked down all international travel and people had actually gone into real quarantine for a few months, we could have completely eliminated the virus without a vaccine, at least in many countries.
That which is a barrier to being harassed by strangers in public, is also a barrier to being helped by strangers in public. To being seen as a fellow human being worthy of dignity and respect by strangers in public. And given the way humans are wired for face-to-face social interaction, if you've only got half a face you'll probably only be seen as half a person.
Masks have been one of the most polarizing issues of the past year; you and your friends propose to continue that indefinitely?
I don't think I should be forced to stop doing something because someone else polarizes it.
Being forced to stop doing something because someone else polarized it is a general summary of 2020-21. Regardless, I don't see anyone here suggesting forcing an end to mask wearing.
Nobody said anything about "forced". But the OP specifically called out, as one reason to keep masking up, the bit where being masked reduces the frequency of unwanted social interaction with strangers. Which, yes it does. But that includes reducing *favorable* social interaction, and it probably does so preferentially. And including passively favorable social interaction like subconsciously classifying "Edward Scizorhands" in the "basically a decent person like me" bin on the basis of e.g. having seen you smile.
If you choose to remain masked, for whatever reason, that consequence is part of what you are choosing. If you are part of a recognizable group that chooses to remain masked, that consequence will apply at the group level. But nobody (other than maybe Tucker Carlson) is going to force you.
People going into work, or kids going to school, when they're _obviously_ ill, has long IRKED me.
Just stay home!
And, yes, some people can't – that's Sad. But lots of people _could_ and just don't, and it's long seemed obviously stupid to me that they don't.
It used to be that the ones who worked sick could claim to be tougher than everyone else and more valuable employees. But even pre-Covid I had begun to detect frustration being directed toward that employee when they had coughing fits in the room with poor air circulation. Now, I hope workplaces don’t give up the remote-work capability. Then people could be actually sick, potentially without even taking sick time if they still put in a work day remotely.
Yes – tho, sadly, I think what little evidence we have of this dynamic is that it's an obvious net-loss in productivity (because one sick person often makes others sick).
TBH, we should encourage masks be worn on public transportation - trains, buses, airplanes, ect.
I'm planning on wearing masks on planes at least from here on out.
This is amazing! I forwarded it to a couple of friends who deserve to know how awesome your writing can be.
Laughing, but also crying
I'd say 90% bureaucracy and 10% we really, really, really don't want to inject something that could be harmful in our entire population of 300 million people, though. And to ascertain that, no matter how big the trial, a few months of observation are necessary.
To be fair, if Covid-19 were more lethal, this would have probably been faster: as long as vaccines are decent, you want to give them to people asap. But with this low-but-not-so-low lethality... it was the perfect bureaucrat storm.
fantastic
This is a great song! Please write more content like this.
(Also, my brain somehow automatically generated a melody to the chorus as I read it, which I realized later to be similar to a song I wrote once but never published. Here's the melody, ROT13'd to encourage musical creativity:
Sfunec-O-Sfunec-Sfunec
Sfunec-R-Q-Q-R-Sfunec
Sfunec-O-Sfunec-Sfunec
Sfunec-Sfunec-R-Q-Pfunec-O)
Also, for anyone who wants to hear the intended melody (which makes these verses make a lot more sense), skip in the video to 1:17.
Hilarious!
Even when Scott is shitposting, it's done with class.
When I read the email minus the embed, first heard this to the “Lovely Ladies” tune.
I'm jealous that you can already get the covid shot; here in Europe the schedule just keeps lagging more and more. Luckily summer is just around the corner.
And now my subscription has paid for itself. Thanks, Scott! But if the fan community can put together a chorus capable of actually putting voice to this, that would be perfect.
I was going to suggest something inspired by Marat/Sade.
Let's say the FDA had immediately authorized and recommended challenge trials at the start of the pandemic. How long would it have taken challenge trials to get efficacy results of similar significance to the observational RCTs that were actually performed?
By definition you have sufficient statistical size.
We would have noticed around day 14 that the experimental arm was pulling away from the control arm, and just gotten better from there.
So on the order of 2-4 weeks probably?
Depends. Let's imagine the ad copy:
"Hi there, we're looking for 60,000 volunteers for a trial of a new vaccine against the scourge that you've heard has been wiping out 2,000 people a day recently. The way this works is, we'll be giving half of you the vaccine, and half of you an injection of sterile water. But then we're going to deliberately give *all* of you COVID, so we can see what fraction of the vaccinated group die relative to those who drew the short straw.
"Since we're most interested in the results for people at high risk of death from this particular disease, we are particularly interested in volunteers who are obese, old, diabetic, and otherwise have risk factors that make it much more likely they'll snuff it once infected, i.e. we're not very interested in young healthy volunteers who correctly infer they have little risk even in the control group. But if COVID reasonably scares the piss out of you, you're who we want, and we will be sure to write a nice condolence letter to your family if you end up in the control group and we give you COVID and then you die, emphasizing the value of your contribution to science and the bulletproofness of the legal disclaimer you signed."
How long until we get 60,000 selfless volunteers?
> we are particularly interested in volunteers who are obese, old, diabetic, and
Why are you deliberately making it difficult?
You would need 100-200 people for a high quality challenge trial to determine efficacy. Large scale safety testing would be done separately and does not require a challenge. If you are testing multiple vaccines, you could combine them all in one multi arm trial and further reduce the total size of the control arm (which is where most of the risk is, since vaccines mostly work). I believe the way the statistics work out is that you always "kill" *fewer* people in a challenge trial than a conventional trial for the same statistical power.
You could pay everyone in the trial $5 million and it would be a rounding error compared to the benefit.
1daysooner.org collected 38,000 "volunteers" for challenge trials. I imagine some people who talked a good game on the internet would chicken out, but millions more people would volunteer if there was actually something important to volunteer for, and way more people would happily do it for money.
Many many millions of people have signed up to fight in wars where they had orders of magnitude higher risk of death.
This is really really far from being a problem with challenge trials for something like COVID.
> i.e. we're not very interested in young healthy volunteers who correctly infer they have little risk even in the control group
All serious proposals I've seen for challenge trials suggest exclusively using young, healthy volunteers. The idea is that if the vaccine works very well in one population of humans, it's fairly likely to work at least somewhat well in all populations of humans. The goal shouldn't be "find precise scientific proof of the exact percentage by which the vaccine reduces the incidence of COVID in the target population", it should be "find out whether the vaccine works or not so we can start distributing it ASAP if it does."
Related song: https://www.youtube.com/watch?v=SB7K8v7TsGk
(Another user posted this same song and it was blocked by YouTube who claimed a TOS violation, so enjoy it while you still can)
I wrote a sequel: http://livingwithinreason.com/2021/05/26/covid-confrontation/