New research suggests: “Targeting interventions – including transmission-blocking vaccines – to adults aged 20-49 is an important consideration in halting resurgent epidemics and preventing COVID-19-attributable deaths.”

In recent weeks we’ve been hearing a lot about the priority of vaccinations. Should we be vaccinating older people first? Essential workers? Just vaccinate as many as possible without worrying about who gets it?

Giving out the vaccine is partly about protecting people and partly about slowing the chains of transmission. The results of a new research paper by Mélodie Monod et al., “Age groups that sustain resurging COVID-19 epidemics in the United States,” suggest that for the latter goal it could make sense to vaccinate young adults.

Here’s what Monad et al. say:

Following initial declines, in mid 2020 a resurgence in transmission of novel coronavirus disease (COVID-19) occurred in the US and Europe. As COVID19 disease control efforts are re-intensified, understanding the age demographics driving transmission and how these affect the loosening of interventions is crucial. We analyze aggregated, age-specific mobility trends from more than 10 million individuals in the US and link these mechanistically to age-specific COVID-19 mortality data. We estimate that as of October 2020, individuals aged 20-49 are the only age groups sustaining resurgent SARS-CoV-2 transmission with reproduction numbers well above one, and that at least 65 of 100 COVID-19 infections originate from individuals aged 20-49 in the US. Targeting interventions – including transmission-blocking vaccines – to adults aged 20-49 is an important consideration in halting resurgent epidemics and preventing COVID-19-attributable deaths.

One convenient aspect of this paper is that many of the subheadings are declarative statements, so we can pull them out as a summary of the findings:

Cell-phone data suggest similar rebounds in mobility across age groups

Estimated disease dynamics closely reproduce age-specific COVID-19 attributable death counts

SARS-CoV-2 transmission is sustained primarily from age groups 20-49

The majority of COVID-19 infections originate from age groups 20-49

No substantial shifts in age-specific disease dynamics over time

School reopening has not resulted in substantial increases in COVID-19 attributable deaths

Always good to see people managing uncertainty by fitting hierarchical Bayesian models using Stan.

140 thoughts on “New research suggests: “Targeting interventions – including transmission-blocking vaccines – to adults aged 20-49 is an important consideration in halting resurgent epidemics and preventing COVID-19-attributable deaths.”

  1. Finally! I have been driving my family crazy for the past 2 months ranting about how, along with the vulnerable and “essential”, people aged 18-40 should be vaccinated first. IMO it’s blatantly obvious that this age cohort has been driving the pandemic since last fall.

    And tangentially, why in the world aren’t we delaying second doses of vaccine to get as many first doses out as possible? There is plenty of evidence pointing to this being the right thing to do.

    A recent post on this blog (by Phil, I think) made reference to models not being entirely necessary when something can be seen in the data clearly with the naked eye. We have been there for months!

    • The UK seems to be set on determining how safe these mRNA vaccines actually are.

      Supposedly ~90% of the elderly have been vaccinated at this point. They are delaying the second dose, and they’ve got some new strains circulating that showed 20-40% enhancement of infection in the presence of antibodies towards the original strain in vitro.

      The only thing to make it more likely is giving a few months for waning, but looking at the frail and delaying the second does should make up for that. Give it a few months and look at all cause mortality.

    • I would argue that this is exactly when good models are needed. To date there has been little study of vaccines’ ability to inhibit transmission of the virus. This information is essential to predict the lifesaving effect of vaccinating likely spreaders vs. high-risk individuals. We also need validation of cellphone tracking as a measure of contact. We need direct observational data on mask-wearing, distancing and other behaviors. We need data on likelihood of infection as a function of viral loading, and more, to allow the model to predict the better vaccination strategy.

    • So let’s see: the first people should be

      Vulnerable
      or
      Essential
      or
      people age 18-40

      Which is about 80% of everyone right? because the vulnerable are the ones age 65+ so you’re basically just excluding people age 40-65 but a bunch of them are essential (healthcare, grocery store, utility maintenance, govt etc etc)

      So your advice basically comes down to “let anyone get it”…. which is exactly my advice. In fact, I’ll go one further. Just auction it off and use the results to pay people who don’t get it to stay home.

      • A flower market style auction. Start at say $10,000 a dose, and drop the price by some percent every hour. Hold the price constant as long as there are more people with appointments than doses available. Use the money to pay UBI style payments. Allow employers to pay for employee shots with pre-tax money. Problem solved.

        • The. Problem. Is. Only. Solved. When. The. Doses. Are. Given. To. Vast. Numbers. ASAP. Sorry. Clever “economical” arguments a la Malcom McGladwell-macDowel or whoever the devil sells the most clever books does not best the problem we noew face. All children get and transmit Measles, Polio, etc. Rich or poor is a red-herring. Like the roads, the sewers and the police, and the fire, the vaccination must cover all — and *fast* — or else it covers no one. You know this!

        • But Ron, the rate we’re giving vaccines is ~ 1.5M/day or something similar. Let’s say we can somehow get it up to 5M/day which seems unlikely, it’s still 320*2/5 = 128 days minimum until the country is vaccinated. realistically we’re more likely to do around say 2M/day in which case it’s 320*2/2 = 320 days. That’s more or less a year. Over the next year it makes a big difference whether we do it stupidly, or we do it smartly as to how quickly we recover economic value.

          Realistically we can argue about how it’d be really great if we could vaccinate everyone by 7 days from now… but it’s pointless, because the soonest we’re going to vaccinate everyone is like June and realistically it’s more like November.

          If it were going to take 15 days, shenanigans would be meaningless, I agree. If it’s realistically going to take 200 to 300 days shenanigans could make the difference if a couple hundred billion or a Trillion in GDP which is not peanuts, and is associated with lower suicide rates, lower bankruptcy rates, and lower lots of other knock on effects.

          there’s no way around it we need to consider the effect of *order* and age alone just doesn’t cut the mustard as comprehensive information about who should go first…second…300millionth

        • Ask the army corps of engineers how it’d be done.
          Ask the Germans right now how they plan to do it.
          I do not pretend to know; but I do know that in historical and even recent historical times mountains have been moved in cases of emergency.
          How was the Polio Vaccine managed?
          Come on. Do we have to pretend that those people come from the Anglo-Saxon age and had access to recondite secrets in logistic and engineering and leadership which are beyond now the comprehension of us mere pampered fools and knaves?

        • As far as I know even if we pull in the Army and Natl Guard and volunteers with transport or logistics or healthcare experience etc… We just rapidly run into the rate of manufacture as the limiting factor. It’s not clear we can say 10x the rate of manufacture even with emergency type measures. I think 150-200 days is likely to be the fastest we could possibly go under the best circumstances. In reality I think it will take 300 to 500 to give double doses to everyone even with extra efforts from distribution.

          Reordering the order of who gets it to optimize the effect is not something we can do by central planning because the information isn’t available. However markets do this all the time. Auctioning off half the doses would be hugely more economically efficient than continuing our centrally planned priority method.

        • Daniel said,
          “As far as I know even if we pull in the Army and Natl Guard and volunteers with transport or logistics or healthcare experience etc… We just rapidly run into the rate of manufacture as the limiting factor.”

          Yes, I think rate of manufacture is the limiting factor. I don’t know if the Army or National Guard could help at all with this.

        • I take strong exception to the formula that we must turn to the ageless wisdom of the “knights of the green table” when what we need is expertise in: [1] Manufacturing; [2] Supply-chain; [3] Logistics; [4] Communications (PR).

        • I doubt we could triple or quadruple the rate of mfg of vaccine in a short enough timeframe. I mean this isn’t making cloth masks it’s precision sterile biochemical reactors and they’ve had since last March to ramp up manufacturing under an at-risk protocol. Maybe there’s a whole bunch of mfg capacity sitting idle somewhere because noone noticed they could be making super valuable vaccines that everyone wants…. But it seems about as plausible as not noticing your flooring is made of pure gold.

        • Also I fear that while public health people will assume everyone wants a vaccine, in reality only about 1/3 to 1/2 of the country will actually want one by the end. I mean it’s reported that 40% of men believe widespread election fraud… I suspect a large fraction will believe the vaccine is a plot by whoever… Then there are people who had covid asymptomatically and maybe figure it’s no big deal, and there are various other subgroups. I suspect after a rush of people trying to get it in Jan, Feb March etc… You’ll see them relax who gets it and the lines won’t fill… They’ll be relaxing more and more. By Jun they’ll be talking about how few people are showing up…

          But that still leaves tremendous opportunity to get the jab to the people who want it more quickly by letting them express how much they want it. Guys who don’t care can go to the end of the line at no cost to anyone.

        • “But that still leaves tremendous opportunity to get the jab to the people who want it more quickly by letting them express how much they want it. Guys who don’t care can go to the end of the line at no cost to anyone.”

          The “Guys who don’t care or who cannot get it” are presumably wandering around infecting the rest.

          There really is, sir, a point where we have to calibrate out ideas once again to some notion of the public-good. Would you allocate sewers on the clever basis that those who wish to or can pay more should get them first? Let the poor stifle in cholera. Problem is, the poor work in your god-damned Ralphs and Vons where you get your precious organic dill-weed and so on and the cholera doesn’t give a *crap* whether you are “willing” (or able) to pay more for your better water district or not.

          The Danes were kept out of Wessex finally when Alfred (the Great) raised a host of *all* the Englishmen. All, whether they could summon a horse and a hauberk or not.

      • I think an auction as you describe is a terrible idea, at least if the idea is to save as many life-years or lives as reasonably possible, or to return the economy to normal as quickly as possible, or any superposition of these. I think a lot of people who are at very low risk of either getting or transmitting the virus would be among the first to get it, while many who are potential super-spreaders or at least potential spreaders to vulnerable people would get it much later. I think there are also disturbing ‘economic justice’ considerations that I probably could have mentioned in the first sentence along with the other two.

        I’ll use my personal circumstances to illustrate what I’m worried about, because I think that I represent a large class of people, perhaps including many people on this blog. I am able to live comfortably and easily with very low risk of exposure to the virus. I work from home, I have the flexibility to shop when the grocery store is empty, and I have safe fitness and recreational activities out my door or in my house. I have an outdoor space where I can have friends over for lunch, weather permitting, with adequate distance, so even my social life has been able to continue to some degree. This doesn’t just describe me, it also describes many of my friends.

        Contrast this with many lower-income people. They have to go somewhere to work, where they have the potential to be exposed to the virus or to pass it on to other people. They often share living spaces with others in a similar situation, so if any of them get infected there is a high potential to pass it on to others. They may live with, or help care for, older family members or others who are vulnerable in one way or another.

        Of course there are many parameters, and everyone is on a continuum on each of them; it’s not like there are just these two groups. But if we think about these two groups, I think the former would have much higher early vaccination rates under the Lakeland Plan, and that that’s backwards from the way it should be by pretty much any metric I would want to use for evaluating the success of the program. I think even a lottery would be better than this, if there were some way to do it: just pick people at random. This can’t actually be done but we could perhaps come close.

        But we should be able to do better than random. There’s a list of high-transmission jobs, let’s vaccinate those people early, along with vulnerable people. I don’t know if it would be legal to use census data at the household level, but if it is we could vaccinate large households in small apartments.

        At any rate I’m not a fan of the auction idea. But feel free to try to persuade me!

        • > feel free to try to persuade me!

          Ok. I will!

          First let’s be clear on what the proposal is… The price starts very high, and declines through time like a flower market auction. By design it stays constant so long as the supply and the demand are matched… so there could be weeks and weeks where it’s at say $113 a dose.

          Second off, the money raised is explicitly used to pay people to stay home. So the “contrast this with many lower-income people… they have to go somewhere to work” is ideally not true. They’re getting paid say $1000/mo maybe to stay home and take care of their kids, and stay out of the hospitals.

          Third, I’m proposing that the **employer** can (but does not have to) pay the price, and that the price starts VERY high. $10,000 a dose for example.

          So, the first people to get vaccinated would be those for whom having them stay home is costing employers a bundle. If you’re say an R&D person in a biotech and you can’t get access to your expensive HPLC or mass spec equipment etc, and your tens of billions of dollars of drug discovery research is stopped and the company is paying rent and utilities and financing on expensive equipment… this is a HUGE loss to society. These kinds of jobs can’t be done from home, and need to get back to work right away. Letting these people pay $10,000 per employee to get their workforce going potentially 6 months earlier than otherwise could be tens to hundreds of billions of dollars of net GDP gain. Might as well do this efficiently, and pay the poor people to just not get sick by staying home.

          Now let’s think about someone like yourself. You can work from home, you’re at low risk, your friends too. My guess is you’d rather have $1000/mo than pay $10,000 to get vaccinated today… Because a few weeks after getting the vaccine, you’re going to lose your $1000/mo. And in fact you might rather have $1000/mo than pay $400 to get vaccinated today… You might even decide to wait all the way til say Nov getting $1000/mo extra, while working from home! So I’d expect you to actually stay home a long time.

          I think you’re thinking like “hey I could pay $400 and get back to work, but I don’t really need to, someone else should get it first”. But I doubt you’d pay $10,000 now and give up $10,000 for waiting 10 months.

          So that’s the proposal, I think it maximizes economic recovery while redistributing wealth from the wealthy to the poor, and limiting risk a lot.

        • This sounds like a proposal by a neoclassical economist. That’s not a compliment!

          If the boss at the chicken processing plant says “show up at work on Monday or you’re fired”, I think those workers are showing up. I understand why you think they wouldn’t, but I think they would.

        • Not if they’re getting paid enough to stay home. Obviously I don’t think we should just do this without thinking through the quantitative questions, but we should absolutely think through the quantitative questions.

          The GDP/capita in the US was previously $60k/yr/person or so. That’s $5000/mo/person (including babies, the elderly, and other non-workers). So it’s quite clear that we could get people to give up their meat packing jobs for $5000/mo tax free ;-) that’d be a raise for lots of people. On the other hand, it’s probably clear we can’t afford to do that without bad consequences.

          On the other hand MZM increased by $5T since start of the pandemic through govt “printing press”. So we’ve already “spent” 5e12/320e6/12 = $1300/mo per man woman and child.

          https://fred.stlouisfed.org/series/MZM

          My sense is if you gave people something like $1300/mo per man woman and child each month since the start of the pandemic provided they stay at home or are an authorized “essential worker”, we’d have a DRAMATICALLY lower toll in lives lost and “real” economic damage, at zero extra monetary cost.

          So, the key here is thinking outside the HIGHLY NONFUNCTIONAL standard box. The standard box says guys who run poultry plants have all the power and this is how it “ought” to be, and they should be able to say “come to work” and the workers should jump up and get sick and do the work.

          Burn that box to the ground.

        • Yes, I understand your argument and I agree with it…in principle. The idea that we could get sonething like this up and running in the next few weeks or even months, -in practice-, seems like loony tunes.

        • But if you have people like meatpacking plant workers all stay home, how do you keep the supply chain from collapsing?

          I mean sure there is enough surplus food in the US that by itself that shouldn’t cause starvation – but it’s not going to be *just* that one group.

          That was what really worried me back in March/April – that given how complex modern cities are, a supply chain collapse would be far, far worse than anything COVID would ever do. That was why I really thought we should just be telling people over retirement age to stay home.

          Now we didn’t have a major collapse – but I am not sure we *couldn’t* have if things had gone slightly differently.

        • > The idea that we could get sonething like this up and running in the next few weeks or even months, -in practice-, seems like loony tunes.

          In practice I think we’ll continue to do a TERRIBLE job. So I agree with you. But I don’t think it’s impossible to do this by any means. I mean, politically impossible, sure, but like in terms of practical difficulties… not hard at all.

        • Confused: I also worried a lot about supply chain. The thing that worried me the most was they were plowing food into the ground because people couldn’t afford to pay for it, and then people out of work were in multi-mile long lines at food banks. If we’d printed money and given it to workers staying home we’d have had a lot better chance of feeding people in an orderly fashion, of course food prices would shoot up, but hey they SHOULD shoot up, because hey that’s what gets people to get out there and get the food distributed.

        • I absolutely believe that in a well functioning economy without tremendous gaming of the system, chicken prices for example would have been easily as high as fancy fish is normally (ie. like $20/lb). That would have been fine. More pandemic sourdough and eating beets or whatever would have been good and saved a lot of lives.

        • “we should just be telling people over retirement age to stay home”
          Some people over retirement age live in multiple-generation households. Some live in institutions. Both situations expose them to people who are not staying home and are therefore potential vectors.
          Some of us who do live alone nevertheless inevitably come into contact with others – medical providers, caregivers, service technicians, bankers, retail workers, postal workers. Not everything can be done online or noncontact. And early on, no one knew whether the virus could be transmitted on surfaces.
          General rule of thumb: if you think you have hit on a simple, effective solution to a problem created by COVID, you haven’t.

        • Phil, practically speaking, how do you organize this? You create a phone app, the phone app shows the current price for vaccine dose, and lets you pay for your vaccine with credit card or bank ACH transfer. When you pay, the govt generates a QR code that contains your SSN and a public key signature validating that it’s been paid. To get a vax you show up at a site and have your QR code scanned. They validate the public key signature and register in a computer database that you’ve received your dose and then give it to you. Simple. Literally a week of work for a team of 4 computer programmers or less.

        • >>“we should just be telling people over retirement age to stay home”
          >Some people over retirement age live in multiple-generation households.[etc]

          I am not arguing for this now.

          But my line of thought at the time was that having the oldest people stay home would do at least *some* good, at very little risk of supply chain disruption.

          It’s not that I expected it to be all that effective — but it would be better than absolutely no action, and I was more concerned about supply-chain implications of dramatic action than I was about the disease.

          (The underlying assumption being that the disease itself wouldn’t disrupt supply chains, because in a working-age relatively-healthy population there would be largely asymptomatic or quite mild cases. I’ve seen lower numbers of truly asymptomatic, vs. presymptomatic, cases now – so I am not sure this is correct. I’m just saying that’s what I was thinking in March-April.)

        • >>I absolutely believe that in a well functioning economy without tremendous gaming of the system, chicken prices for example would have been easily as high as fancy fish is normally (ie. like $20/lb). That would have been fine. More pandemic sourdough and eating beets or whatever would have been good and saved a lot of lives.

          I guess part of my concern is, while this sounds perfectly fine… do we really know how far the system can be stretched? I don’t know that we’ve had really major disruptions since the massive post-WWII/50s/60s urbanization of the US (I guess the energy crisis of 1973 maybe…), much less post-everything being just-in-time inventory, etc.

          I am not sure the second-order effects (ie if food A is more expensive then there is more demand for foods B and C, etc.) are really known.

        • >Phil, practically speaking, how do you organize this? You create a phone app, the phone app shows the current price for vaccine dose, and lets you pay for your vaccine with credit card or bank ACH transfer. When you pay, the govt generates a QR code that contains your SSN and a public key signature validating that it’s been paid. To get a vax you show up at a site and have your QR code scanned. They validate the public key signature and register in a computer database that you’ve received your dose and then give it to you. Simple. Literally a week of work for a team of 4 computer programmers or less.

          There are so many smart people on this board, but a few of them are a little too in love with their ideas. 4 days? The above claim seems overstated to me–maybe completely unreasonable. Prices would need to be linked to supply, no? Across multiple manufacturers as vaccine is produced? Or distributors of vaccine? A combination of both? How do we know what’s available at which time so price reduces correctly? The infrastructure seems a little complicated. When would a person be able to get the vaccine they paid for? What if it’s not even available, and by the time it is available, the price already reduced? Or if the price is a nice $100 bucks but the closest place to receive it is 250 miles away? What happens to price if supply dramatically increases? Will the price drop? This sound like an unregulated market–how will you prevent secondary markets from occurring or exploitation? The payment method is credit card–how will this be linked to SSN? How is the government validating the SSN to the payer? Since SSN and credit card/bank transfers are involved, I’m assuming security is important–can this be assured? What about people that purchase vaccine but never claim? Or become incapacitated or get COVID and no longer want the vaccine? Refund? What about false actors that ask for refund if they paid $1000 for the vaccine but it becomes $50 a few months later? Seems like this application would be a prime target for hackers–a few hundred million SSNs/credit card numbers linked to an application that was built in 4 days? Would securing back-end transactions take a little bit a time? IoS/Android compatibility? Web application?

        • Uanon:

          I think you miss the point because a lot of your questions just don’t come up with my proposal.

          First off, we grab every SSN on file and generate a QR code for everyone based only on their age. So by default you get some date way in the future unless you’re 79 years old or whatever.

          We start the auction incredibly high, $100k a dose is not out of the question. The number of people buying initially would be small. What you buy is a cryptographic certificate that you’ve paid and gives you TODAY as your date instead of whatever you were initially assigned. Also, the right to cut in front of the line at any location and get your vaccine in front of people who are “age based”.

          No refunds. If you don’t like it, don’t buy it… you can get your dose in the “regular” line prioritized by age or whatever. If you pay and your date isn’t moved up you can call a hotline or whatever, they’ll look for your financial transaction and ensure your QR code is generated. That’s your only recourse.

          If there are no places within 250 miles of you maybe don’t buy your way into the head of the line? If there are doses at a location, and you show up, you’ll get it first.

          There exists some group of people who could create an app / webpage that would do the technical job of accepting your payment and printing a cryptographically certified QR code in less than a week. I’m confident of that. I mean there are tons of people making apps all the time, there is probably an existing one you could modify only slightly to achieve this. If you put a crypto/security expert on the team, I’m sure this part is very doable. (But you’re right that there are lots of people who would do a bad job, it’d have to be audited by a separate group for security. I’m not saying we could have it running in a week, just that we could design and code it in a week.)

          Do you need to secure the servers well? Sure. I’m not saying “just any old web app developers could do this” but I am saying that there are multiple very competent security conscious groups in the world any of which could in fact do this. If we put a couple million dollar prize up and told the NSA to vet the results and select a winner or whatever, you’d get a working system in short order. There’s just not much to it.

          The order form takes: Name, state of residence, state ID card number, SSN, and price, and prints a QR code that has those pieces of information, a unique identifier, and a cryptographically signed hash of the rest of the contents. You can have it sent to you in the mail, delivered to your smartphone, emailed, or just log into the website and put in your SSN or state ID and it’ll re-print it for you.

          The computational part of this is easy because all of it is known “solved” technology.

          As a start on the design, you’d put the crypto-signing app behind a firewall with zero allowed inbound connections. It’d poll the database and generate the certificates in batches. perhaps every 5 seconds. The database would be behind a separate firewall accepting only database port connections from private IP addresses over TLS connections. The web servers would be in the DMZ accepting only TLS based web connections. You’d write the software in a memory safe language to eliminate buffer overflow type problems. Keep the functions very simple.

          But there’s no point in talking further about this, because it’s not going to happen. I mean, we’re clearly going to continue to do something stupid instead of getting appropriate people involved which is basically computer science + security + programmers. The problem of handing out stuff in priority order is basically queueing theory, the problem of securing it is cryptography and computer security, and the problem of creating front ends and taking payments is web commerce. Google has a team that could do this and have it up and running before the end of February I’m sure. Cloudflare does too. I mean cloudflare has **already** got the “cloudflare waiting room” up and running: https://blog.cloudflare.com/cloudflare-waiting-room/ which is actually MORE complicated technology.

        • It kinda sounds like you’re unnecessarily combining two separate ideas here. You’re saying we can reduce transmission by using the funds raised by your auction to pay people to stay home. But we already have the money for that! We can just do that part outright. And if you’re worried about cost you can impose a wealth tax or raise taxes on high incomes.

          So then the question is how to rollout the vaccine to either further reduce transmission or protect the most vulnerable. Your proposal doesn’t seem to affect either goal – it seems to primarily benefit economic recovery. Which is fine to argue I guess but I don’t think I agree with that goal.

        • No, we can’t just pay people to stay home. If everybody stays home we will all starve to death in the dark. He’s saying that the power of the free market could be put to work to determine who is really truly essential and to get those people vaccinated. If food is getting scarce because food workers won’t go to work unless they get a vaccine, the cost of food will go up and there will be lots of money to be made by paying a ton of money to get your food workers vaccinated. Etc.

          Sounds fine in principle but I think completely unworkable in practice.

        • Yeah this was my concern back in March/April, and I am still not entirely certain the fact that we avoided a supply-chain collapse means the risk of such wasn’t unacceptably high.

          (Kind of like the Cold War – we managed to not blow up civilization, but I think the risks we took were still unjustifiable.)

        • Okay I see, I was conflating Daniel’s argument with another proposal I had seen. In my comment I was meaning pay non-essential workers (e.g. restaurant workers) to stay home. His proposal relies on paying everyone to stay home and then – if I understand correctly – the auction would ideally incentivize firms to pay to vaccinate their workers at an appropriate market rate.

          So then I think the problem would be: would essential workers be sufficiently incentivized to not go to work unless they get a vaccine? And after vaccination would they go back to work? Like if the “UBI” payments are high enough to do the former, then they would probably also be high enough for a subset of people to go get the vaccine and then quit their jobs, no? And if that subset of people is large enough then that can cause huge supply chain issues. Not to mention all sorts of other frictions involved.

          Hopefully I’m understanding this correctly now!

        • Michael J,
          To the extent that your concerns overlap greatly with mine, I think you’re understanding this correctly now. Daniel will probably argue that we are both missing something; let’s see.

        • So I think we’re mixing up a number of proposals… What could we have done last April, vs what should we do today, etc.

          Pre Biden we had a fool and a charlatan running the goddamn country, so we couldn’t expect anything useful to happen.

          As of today, could we auction vaccines to good effect? Hell yes, at least in terms of technically it could be carried out. Would politics prevent it? Perhaps. Are there concerns that require carefully thinking about some details? Hell yes. Do I have the answers to all those details? No. But I can tell you this:

          1) People staying home to prevent infection **should be paid** for the public good they’re creating

          2) Stopping some people from working is a huge loss compared to others. The biotech researcher example above, having them stay home is a huge loss compared to having a guy who answers trouble tickets at the IT dept for a medium size business do his work from home… Vaccine priority should respect this fundamental truth.

          3) Not all doses have to be in the vaccine auction, we can still use rationing as well… it’s a continuous parameter from 0% to 100% auction.

        • The point of the auction in terms of how it distributes the vaccine is that it distributes the vaccine first to those people who are harmed the most by staying home (ie. those who would be “willing to pay” a lot to be able to go back to work etc). The point of paying people to stay home is that they’re producing a public good and so they should be compensated for doing so.

          The point of transferring the money between the high willingness to pay group and the public-good-producers is that it avoids having to “poof” all the money into existence and therefore represents a lower inflationary mechanism and a real transfer of wealth from wealthier to poorer. When we inflate the currency we tend to harm the poorer people, but when we “tax” the wealthy and give it to the poor they actually get more *real* wealth.

          Finally, the group with the highest life-years lost is not entirely clear. If you look at just COVID deaths it seems to be people around 70 years old. But If you include SUICIDE, it may shift to the 20-40 year old range. I haven’t found good data, but the increase in suicides throughout 2020 was pretty dramatic.

        • I would be *very* interested to see suicide data. It seems entirely plausible that it would have risen, but I’ve only really seen that talked about by people who were taking the tack of “just tell old people to stay home, no restrictions on working-age people”.

        • My sister is a Psychiatric nurse practitioner. Her anecdotal data is that suicide among less than 40 year olds probably quadrupled or something like that (meaning she knows about a lot more suicides among 20-40 year olds than in a normal year). She of course is a biased source. So I’d LOVE to see the data as well. I think people between ages 20-30 in particular were super hard hit and a lot of suicides that otherwise wouldn’t have occurred did occur. I’m not political about this at all, just think that it’s a real issue.

        • Daniel
          You will soon be eligible to be included in the ranks of economists (as someone else said, that is not a compliment). I see to huge problems with your proposal, both based on practical issues that doom most traditional economic logic:
          1. I have always liked you UBI proposal – but linking it with vaccine distribution will doom both to failure. Linking complex issues like these with the intent to solve both simultaneously, rarely works. I am not hopeful in this case.
          2. Most economists realize that when you pay people to not do something, it generally is unenforceable and highly inefficient. How are you going to monitor that those that accept payment to stay home don’t then decide to go to the bars? I supppose we could issue ankle bracelets to all those that accept payment, but then we have a different set of issues – a place where I would rather not go.
          Instead of complicating the issues, I’d rather simplify it. We clearly have a severe shortage of vaccines, and this will continue for some time. The traditional options for allocating a very scarce resource include: auctions, random allocation, first-come-first serve, prioritization by rules (made up by whom? the devil is always in the details). I’ve seen my county, after the state and federal governments have failed to do their jobs (often for political reasons) resort to a free-for-all (first come first served). In Florida that did not work well, but at least people could wait oudoors. Given that our temperature here will be -20 degrees this week, our waiting will have to be online. But the online system is defective and it is creating (too) much anxiety. So, I think some meaningful prioritization of susceptable and essential groups makes perfect sense. As soon as the door is opened to all the 65+ ages, the shortage become huge, so that pool should be allocated randomly. Create a waiting list, randomly order it (sort of the way the armed forces draft worked before we went to a volunteer army), and I think most people will accept such a system. This is not rocket science and most people realize there is not enough vaccine to go around. But the way it is being allocated currently is dangerous – it creates anxiety and wasted time and is unnecessary.

        • Dale, my UBI advocacy should be seen as separate from this proposal. The proposal here is to transfer funds from “eager” vaccine users to “willing delayers” by paying everyone until they get the vaccine. It doesn’t matter whether they stay home or not just that they haven’t yet got the vaccine.

          Also as I mentioned there is no reason why we can’t mix auction and various rationing schemes. I’d suggest start by doing 50% auction and 50% rationing by random number generation within allowed categories. Do it all within the unified cryptographically signed QR code system proposed above. Just send out through the mail a QR code to everyone with a SSN based on age and random numbers… no need to sign up. Let the “eager” ones buy their way in front of the line.

        • Well, you mentioned UBI, but okay, let’s put that aside. I like your random rationing scheme and think it is much better than what many places are doing. I don’t get the point about paying people to stay home, regardless of whether they stay home or not. Given that we have a significant portion of the population that either does not think they can get COVID, doesn’t care, and/or is not concerned with the welfare of others, this seems like an invitation to people to be paid for the worst behaviors they can engage in.

          As for the “eager” buying their way to the front of the line, I could be persuaded if all that money was guaranteed to be put to good use – which I have no confidence in. While it is economics 101 to espouse the virtues of a system based on willingness to pay, I think most people would reject this scheme, as do I. While it is naive to believe we will all be treated equally, an overt system that permits people to buy their place in line violates my sense of ethics. I realize it is somewhat hypocritical, given that schemes that purport to treat everyone equally rarely do so. But I’d rather at least try to leave people’s wealth out of the determination of who gets the vaccine and when.

        • Dale. Yes I think we got mixed up in a couple different ideas at different points, because people were comparing my idea now, to what we might have been doing in the past etc.

          Suppose I could show you that with my system in place the median time to vaccination falls, the death rate is lower, and the economic damage is lower? Would you still think it’s not ethical because it involves using money to collect and organize information? Of course I can’t show that. But it seems plausible that it would be, just by virtue of the fact that the system is simple and easy to administer compared to 15000 different sites each doing it differently. For example there are reports online that prioritizing 65+ is causing all kinds of problems because these groups aren’t good with technology and can’t figure out how to sign up online and each state/county/entity has a different sign up system and etc.

          With my basic proposal, there’s no signing up, we just dump the SSN database into some software and batch generate everyone QR codes that contain the date on which they’re eligible… WE can mail these to everyone, plus people can go on the web and get them by entering their SSN… have them emailed, whatever. Done. Everyone knows about what day they’ll be able to get it. If things go sour with this scheme because of delays in mfg or whatever, we regenerate these codes later and invalidate the old ones.

          Next the only thing we offer on top is the ability to jump the line for an initially very high and slowly declining cost.

          Presumably this lets people who are losing a lot of money by not working get earlier vaccinations, and hence reduces economic disaster. Funneling the money back to the people who are waiting in long lines and won’t be vaccinated until later is rational and a “good” use of the money in my opinion. You get paid for waiting.

          The problem of people going about their lives and infecting others can be handled by fining people and proper enforcement. You might be right that our country is too much of a failed state to even be able to do the basics like fine people for endangering others, but it’s not a reason to prefer the failed state to rebuilding a good functioning society.

        • Daniel
          Of course I’ll object if your proposal achieves all those metrics. What those metrics leave out is the distribution of the costs and the benefits; this is something economists like to ignore. What your proposal (if you could demonstrate the things you say) would show is that worrying about the distribution (who gets the shots and when) costs us something in terms of overall measures of the median time to vaccination, the average death rate, and the aggregate economic cost (this last concept if fraught with measurement issues since it is highly sensitive to peoples’ incomes). I’m not saying these would be irrelevant – in fact, they are important and would be valuable to document. But all of those measures ignore which people get left out when you move to a system where money helps determine who gets the vaccine. This is the Achilles Heel of economic recommendations – the political reality is that people often care more about the distribution than the overall impacts. This can lead to great folly, but I’m not ready to abandon concerns about distribution.

        • Please note that in my proposal EVERYONE gets the vaccine at no cost for the actual shot, the only question is *when*. Furthermore, let’s say that everyone gets it presumably before say June 2022, and likely before say Nov 2021.

        • Many children and all undocumented people lack Social Security numbers. Many homeless people lack mailing addresses. Not requiring appointments is likely to result in lines at most-preferred times, with risk of spreading contagion, or discouraging people from being vaccinated. No jurisdiction is likely to enforce stay-home orders on those who have ‘sold’ their priority in the queue; it would be burdensome both to the enforcers and the general public to identify the vaccinated, the unvaccinated who are essential workers, and the cheaters among those out and about. It would be much less acceptable than ‘stop and frisk’.

        • Again, I think you’re missing the point. And maybe this is indicative of a bigger problem, you can have a good idea, and people will think that it’s some other idea, and hate it…

          No SSN? Fine, wait in line to the end, when we’ll open it up to anyone… I don’t see this as a big problem. Undocumented people is ~ 11M people country wide, that’s 11/320 = 3% not zero, but hey maybe you don’t get priority if you’re not legally in the US, just wait to the end of the line.

          >Not requiring appointments is likely to result in lines at most-preferred times:

          Hey you know what, there’s already a multi-mile long line to drive up to Dodger Stadium and get your vaxx… and that’s **with** appointments. nothing new.

          Mailing addresses aren’t required. Anyone can check their date on the website. Homeless shelters and local medical clinics and things can print out your QR code for you. If you know it’s your date, show up without your QR and the checker will simply look at your ID and enter your ID info and check online for you. It’s just not that complicated.

          >burdensome both to the enforcers and the general public to identify the vaccinated, the unvaccinated who are essential workers, and the cheaters among those out and about. It would be much less acceptable than ‘stop and frisk’.

          No it’s not at all. Once you get your vaccine, they register that you got it at the time you got it, and you automatically stop receiving payments from the treasury department. Nothing needed.

        • Actually, I misspoke, what you want to do is in the month after your “available” date, you cease receiving payments whether you got the vax or not. Even easier, because you don’t even need to collect the info on whether someone got it or not.

        • Even better. Instead of giving you a DATE you give out a literal number in line… between say 1000000 and 400000000 (1M and 400M).

          Then the govt website like the Census bureau’s population counter just literally shows a number that changes each morning… If your number is less than the currently shown number, you are eligible…

          If there are long lines etc, we just make that number change slower. If there are short lines, we run it forward a lot faster.

          People who want to buy their way up in line can buy “today’s number” at whatever the auction price is today.

    • This puts a lot more faith in the free market than I have.

      How many essential industry firms work on extremely thin margins and do not have the cash to bid for these expensive vaccinations? Much of the food industry comes to mind

      I can see NFL teams and cash-rich tech firms bidding strongly while some grocery chains or trucking are begging their banks for a loan.

  2. The problem is that (as far as I know) we do not , at this time, have any proven transmission blocking vaccines. The phase 3 trial results were based on symptomatic COVID (subjects were only tested if they displayed symptoms). So we do not know fi vaccinated 18-45 year-olds will just get asymptomatic but transmissible COVID.
    More recent trials are regularly testing everyone so there may be hope for the future.

    • Yes, the tests weren’t designed to test transmission, but isn’t it safe to say (1) Most vaccines for similar diseases block both symptoms and transmission and (2) the accumulated evidence shows asymptomatic carriers are less infectious than symptomatic carriers? If you buy into both of those you’re going to have a strong prior belief that the vaccine will cut down transmission substantially. Not everything has to go through an RCT before it’s actionable.

      • Not an RCT. Just any attempt whatsoever to test people after they are vaccinated. It beggars belief that no actual tests for shedding of virus were performed as part of the vaccine trials. Yet everyone wants to pretend the best case scenario is true. Incredible.

        • “It beggars belief that no actual tests for shedding of virus were performed as part of the vaccine trials.”

          Very good point.

        • After writing the above comment, I heard on radio (and regrettably forget anything that would give me a reference — too much other stuff going on in my life since then) something to the effect that one of the most recent vaccines (Moderna perhaps?) was examined for something that was suspected of reducing virus shedding. I can’t find it, but did get the following, which gives some insight into the question of transmitting after vaccination:

          https://www.npr.org/sections/health-shots/2021/01/12/956051995/why-you-should-still-wear-a-mask-and-avoid-crowds-after-getting-the-covid-19-vac

        • I’m not sure why there weren’t tests for viral shedding in the vaccine trials (at least in most of them, I think AZ was doing periodical asymptomatic testing?). I think it was some logistical concerns or something. Like I think they figured they could get results faster or recruit faster / more widely or something. I don’t know.

          But anyways that’s the past, can’t change it. It would be nice to have direct data about transmission but we don’t. But we do have information about it like Jackson mentioned. We know some stuff about mechanisms and there are analogous examples with other vaccines / viruses. It makes sense to use that information instead of ignore it. So I don’t really get this statement “Yet everyone wants to pretend the best case scenario is true. Incredible.” If you disagree with the strength of the prior information you can argue that but I don’t think there’s any “pretending” here.

        • Michael J,

          You’re espousing the pattern that has held for a solid year now. We’re in a hurry, we can’t do ANY basic evaluation of how the thing spreads and now it’s too late to do it anyway.

          A year from now, two years from now, that’s going to be the party line apparently. The problem is too important to actually study, we need to just react blindly and keep reacting blindly until the problem goes away.

          I personally find that attitude completely unjustifiable.

        • I mean I’m arguing that we *do* basic evaluation of how stuff works. Instead of lamenting the stuff we didn’t do and get stuck in this weird paralysis where we’re just like “we don’t have data so we can’t make policy!”, let’s use what information we have and continue to collect the data we need.

          I guess part of it was my fault, I didn’t explicitly say the last part. So let me be more clear. Yes, it was a mistake to not test for asymptomatic cases during the vaccine trials. No, we actually do have information about if vaccines prevent transmission. Yes, let’s collect data on asymptomatic cases for people who have received a vaccine.

          Decisions and policy need to be made now – keeping in mind that the absence of decisions / policy is itself a decision. That does not mean when we make decisions we cannot adapt to new data. We should always be collecting data to better inform our decisions.

          Saying “we don’t know anything” imposes paralysis. It was why the U.S. waited so long to recommend masks. It was why the U.S. did not acknowledge the role of short-range aerosols for many months.

        • I should add to the last point to complete the rule of three: It is why the U.S. is not delaying the second dose, at least by a randomized trial.

        • Michael J: “this weird paralysis where we’re just like “we don’t have data so we can’t make policy!”

          Precisely this idiocy is what typifies our cast of characters in leadership. And there’s the other variety, which is, “we know the answer, don’t you know, some veterinarian said so on you-tube!”

          Between these poles of idiocy the boat must still be guided a course to sail the uncharted seas ahead.

    • “we do not…have any proven transmission blocking vaccines. ”

      Doesn’t matter. If the choice is to vaccinate a senior citizen in a nursing home or a grocery store worker, even if the vaccine is proven *not* to block transmission, the grocery store worker is the obvious choice: a) the grocery store worker is performing an important activity for the rest of society, while the senior (sorry seniors!) is doing nothing useful at all; and b) a nursing home is an isolated environment where it’s *relatively* easy to control infection.

      Whatever the case we’re no further behind by vaccinating the grocery store worker and if it turns out to block transmission, which is highly likely, we’re way ahead.

      • ha, yes, I guess grocery store workers are “essential” so already getting vaxed, but the argument still mostly applies. nothing is lost by vaccinating younger people and likely much is gained by reducing transmission.

        • Jim said,
          “I guess grocery store workers are “essential” so already getting vaxed”

          This is my impression also. I checked online and got https://www.aarp.org/work/working-at-50-plus/info-2020/covid-vaccine-essential-workers.html , which said (as of December 23, 2020), “According to the Advisory Committee on Immunization Practices (ACIP), the second wave of vaccinations should include frontline essential workers — approximately 30 million people with jobs in grocery stores, public transportation, meat plants, among other businesses — and people age 75 and older. The third wave should include other essential workers (roughly 57 million people), alongside people ages 65–74 and those with high-risk medical conditions.”

          Does anyone know what proportion of the “frontline essential workers — approximately 30 million people with jobs in grocery stores, public transportation, meat plants, among other businesses” have been vaccinated? From my experience as someone in the “age 75 and older” category, it’s been difficult to sign up for the vaccine, let alone get it — there seemed to have been no opportunity to sign up for vaccination until it was offered to “age 65 and older”, and that just puts you on a waiting list at best — or just a list of people who will be contacted when they can be put on a waiting list.

        • Martha,
          My 79-year-old mom got her first dose a few weeks ago (in Boulder, Colorado) and is scheduled for her second one next week. She didn’t pull any strings or anything, she just followed the normal procedure there. So maybe this is an example of what we’ve been hearing about: there are big state-by-state and county-by-county differences in vaccine availability and in how it gets distributed.

        • Yes, there probably are big state-by-state and county-by-county differences in vaccine availability and in how it gets distributed. Indeed, I have heard of big differences between my county (urban and suburban) and the one just east of here (small town and rural). My county has been doing a questionable job, but the one to the east doesn’t even have the resources that we do.

      • Jim:

        And yet just last month there was a big outcry about how the vaccine distribution system was too complicated and it should just be given to older people!

        I guess there’s a moral hazard issue too. If transmission is a concern we should be giving vaccines to California politicians, invaders of the U.S. Capitol, and other people who have a habit of loud maskless conversation in closely-packed crowds.

        • Wouldn’t you assume those groups already have antibodies?

          There is supposedly at least 25 million US cases already, probably 5-10x more. Let us just say around 25-50% already have some form of immunity, even if it has waned.

          Shouldn’t most of those people be the same ones contributing to transmission? The people rarely leaving their house are going to be the ones without immunity.

        • If this were true, I don’t think we’d still be seeing big outbreaks that are (evidently) among people who work in crowded conditions and live in large households. But the huge number of cases in Southern California seems to be mostly these people and their families, if news reports can be trusted on this (which I admit is not a given).

        • Something just doesn’t seem to add up about it. I can think of a few assumptions here (some overlap):

          1) Assumed “risky” behaviors are not really all that risky relative to “non-risky”
          2) Immunity is frequently waning and many early cases (who were never tested last winter/spring) are getting it again
          3) Immunity is frequently not sterilizing (the person can still transmit)
          4) The reported number of cases so far is much higher than reality

        • or 5) The proportion of people being careful is much less than we think…

          CDC estimates 83 million infections through December… probably over 90 million now. But if only say 33% of the US population is really being *effectively* careful that’s still less than half the available population, probably not enough for herd immunity.

        • CDC estimates 83 million infections through December… probably over 90 million now. But if only say 33% of the US population is really being *effectively* careful that’s still less than half the available population, probably not enough for herd immunity.

          That falls under:
          1) Assumed “risky” behaviors are not really all that risky relative to “non-risky”

        • Not really. There’s a difference between “we are wrong about what behaviors are risky vs. safe” vs. “we are more or less right about what behaviors are risky but the majority of the US population has more or less given up on taking precautions and are doing the risky things despite advice”.

        • Riskiness is not an intrinsic characteristic of a behavior; it depends on the infection rate. When the infection rate was 0.1%, people hanging out together in small groups, or going to work in a food processing plant, was not ‘risky’, in the sense that the probability that you’d get infected if you did that was very low. When the infection rate is 5%, that’s a very different story.

          We’ve seen this all over the world, and regionally within the U.S. It’s not complicated.

        • I’m not sure what infection rate means exactly but % positive has been pretty much between 5-10% all year.

          Are you saying that after a year of lockdowns and telling everyone to wear masks an even higher percent of the population is infected now than ever before?

        • Compliance is a normal consideration when choosing an intervention. If side effects seem worse than the benefit the patients will stop doing it.

          Of course everyone knew this, which is why they said it would only be two weeks at the beginning.

        • Anoneuoid –

          > If side effects seem worse than the benefit the patients will stop doing it.

          If people listen to stupid shit on rightwing media and think mask-wearing is an issue of “freedoms” they don’t even start wearing masks.

        • I think the infection rate (by which I mean the number of people who have detectable virus, divided by the number of people) is near its maximum, yes, though it appears to be declining. You )Anon) act like this is surprising somehow, but I don’t see why it should surprise anyone. “the hammer and the dance” was predicted way back at the start: declining cases leads to loosening of restrictions and less careful behavior, which leads to increasing cases and ultimately increasing deaths, which leads to imposition of restrictions and more careful behavior, repeat ad nauseum. We could imagine the overall effect to be flat, decreasing, or increasing on a long timescale. We ended up with “increasing “ but even I, who is famous for being easily surprised by the way things go, have not been surprised by this.

        • My expectations just based on estimating my own time to complete projects is to double it then go up an order of magnitude. So two weeks = 4 months.

          Most people only expected two weeks though. If a doctor tells you the painful treatment will only continue for two weeks, and a year later your symptoms are worse and the treatment is still continuing or even becoming more intense, you will start to question if the doctor is being honest or knows what they are doing.

        • Somebody told you “two weeks” and you believed them? Go ahead, pull the other one.

          The day before the Bay Area’s first lockdown, early last year, a friend said she thought the coronavirus news was stuff and nonsense and that we wouldn’t even be talking about it in a month or two. I disagreed. She asked how long I thought it would be before we were back to normal and I said “two years.” And I am not claiming special powers or insights; I thought it would be at least a year and probably two years because that’s what I had picked up from the news. I don’t know where you saw this “two weeks” nonsense but I certainly never saw anything like that.

        • I don’t know where you saw this “two weeks” nonsense but I certainly never saw anything like that.

          We live in different realities. If you don’t know about 15 days to stop the spread shutting down businesses back in March you have no idea what is going on in the world. The only thing I can think of is its an ivory tower effect.

        • Anon,

          ‘Ivory tower effect’ doesn’t really work I think, I’m a self-employed consultant.

          I get my news from the same news sources most of this blog’s readers probably do: New York Times; NPR; whatever shows up on my Google News feed on a given day; an approximately weekly foray onto Fox News to try to see what news other people are seeing; Reuters; and so on. Most of these sources were explicitly or implicitly talking about the pandemic lasting months or years, back in March 2020. When I say “implicitly” I mean a lot of articles didn’t give specific predictions for when things would return to normal, but they talked about vaccine development and distribution in units of years, and talked about the need to potentially reopen closed rural hospitals and to shut down some manufacturing long enough that stocks of supplies would run out, and things like that would obviously take many weeks.

          Where were you seeing predictions of a couple of weeks? Can you give some examples.

        • “Most people only expected two weeks though.”

          Huh. Here in CA, where I’m part owner of a whale watching company (not my living, though), when the Bay Area first shut down we planned for being closed through the end of summer.

          While I can’t claim to know everyone in our local tourism-oriented economy, no one I knew thought it would only last two weeks.

        • Anoneuoid –

          > probably 5-10x more.

          Probably? First, the CDC says 4-5 times more second, you think probably as many as 250 million Americans have infected? Plus another 20 million vaccinated.

          So 80% of the public with some form of immunity? Really?

        • Joshua,

          Depends on what you mean by “some form of immunity “. It is very much an open question to what extent someone infected in February 2020 is “immune” in February 2021.

        • Brent –

          Sure. Full immunity hasn’t been proven. But I haven’t heard of very many cases of reinfecfion up until most recent variants.

          80+% of the public having been infected seems just a tad implausible given serological analyses.

          There are many examples of studies like this. From November, but still – a relatively high prevalence area:

          https://www.maricopa.gov/CivicAlerts.aspx?AID=1876

        • I suspect the take up rates among anti-maskers and right-wing politicians are going to be pretty low. If their group R-sub-t remains above 1.0, is it worth trying?

        • Jim:

          I was thinking about this the other day when getting my hair cut. Everyone’s wearing masks, I think the risk is low . . . but if you think about tradeoffs, it’s ridiculous to think that I’m risking my health just to get a trim. At least for a politician you can argue that the haircut makes a difference as it affects perceptions, but for me it’s just pure vanity. I guess the take-home message from this particular story is that there’s a role for rules, as we can’t just rely on individual decisions. When getting a haircut was illegal, I didn’t do it, but in the past several months it’s been allowed, so I’ve gotten my hair cut a couple times. I still can’t imagine eating inside a restaurant, though.

        • I’ve had two haircuts since last March, both of them outdoors. I’ve also hacked off some of my own hair a couple of times.

          I dunno, maybe there’s an argument that people should keep going to the barber / hair salon in order to keep the barbers and stylists employed. But there’s no way I’m doing it until vaccination rates are much much higher. My 80-year-old mom is moving in two doors away from me, next week, and we will be treating her as part of our household! But even if this weren’t the case, it just doesn’t seem worth the risk just to get a haircut. When I say “the risk” I mean both the risk to me — very low risk of death, but I don’t want to lose my sense of taste or have any of those other awful side effects — but especially the risk that I will pass it along to someone else.

        • Good luck with your mom. We haven’t seen my mother in law in a year (with the exception of talking to her outside when she drove an hour to deliver christmas gifts to grandkids). And I was able to convince her not to go to thanksgiving with her sister. But she won’t let me help her schedule grocery deliveries.

          Wanted to offer a head’s up to anyone who cares to worry about such things.. my husband (infectious disease doctor in the still very busy northeast) said we should probably start wearing N95 respirators if we are going to be in relatively close contact or in socially distant situation for longish periods of time (like a hair cut, or a busy grocery store, but maybe just regular mask for a 5 minute convenience store trip).

          We happened to have a bunch of respirators from when his practice had to prepare for dealing with rule-out ebola calls a couple of years ago. We used them all up this spring when the hospital nearly ran out (and they only gave out one a week, and even rationed surgical masks… things you only know if someone you know works in a covid center… every day there was a nurse assigned to hand out labeled bags containing your mask for the week).

          Higher transmissibility necessitates better filtration for equivalent protection (I guess no one here needs that spelled out)

          So yeah, if you think you might care about it as prevalence of the transmissibility mutations increase, now is probably the time to buy a few. If you never use them you can always use them when sanding something or petting a cat!

        • Phil said,
          “I’ve had two haircuts since last March, both of them outdoors. I’ve also hacked off some of my own hair a couple of times.”

          I’ve only had one — I did it myself, using a method I found on the web. It’s time for another one. (Fortunately, I wear it long enough to pull back in a band if I choose, so amateur haircuts are adequate. I don’t think I had a professional one until I was in my thirties? forties? )

      • Well, as long as we’re being crass, grocery workers may be performing essential functions, but most are low-skill and easily replaceable. In any case, you vastly oversimplify. Grocery workers may have many <2m encounters in a work day, but most are brief. Other public-contact occupations have far closer, longer-duration encounters (dental hygienists, e.g.). Which factors are more important in contagion? Guesswork is not good enough.

        • I agree that in some sense guesswork is not good enough — but sometimes that’s the best we can do.

          You are correct that, for example, a dental hygienist has far closer, longer-duration contact with a patient than a grocery worker has with a customer.
          But there is more to the story:
          Dental hygienists have contact with fewer patients in a day than, let’s say, a checker at the grocery store.
          And the dental hygienist has strong washing-up protocols, and high quality masks. And a grocery store shopper may also be exposed to the virus through indirect contact with the people who stock the shelves and the people who prepare the deli foods.

          These show that guesswork can be snap-judgment-guesswork (I’d categorize your statement as that) or educated-guesswork, which would bring in the additional factors above (and probably others I have not thought of). When we do not have definitive evidence, educated-guesswork is better than snap-judgment-guesswork.

        • I agree. I was actually not making _any_ judgment, snap or otherwise, but I guess (sic) I did not make myself clear. I meant to say, I am unaware that there is sufficient data to make informed choices. At a minimum, number, duration, proximity of contacts, amount and type of PPE, sanitizing protocols, relative transmissibility via contact, droplet and aerosol, ventilation and filtration equipment effectiveness, are factors, and no doubt more.
          I would only consider guesswork to be ‘educated’ if there is a rationale underlying it that experts can discuss. If it simply springs full-grown from the brow of even an expert, it might as well be a coin toss.

        • Ken Schulz says:

          “Well, as long as we’re being crass”

          I don’t think that’s crass at all. Senior citizens depend on a strong economy to generate their pensions, but they don’t do much to produce revenue.

          But your contention that grocery store workers are easily replaceable is bellied by the help wanted signs on the front of every WalMart, Kroger, Safeway, Home Depot, Costco, Lowes and other stores, not to mention Amazon FedEx and UPS and just about every other retail occupation.

          Just yesterday a ~75yr old dude checked me through wal mart and he was so bloody slow I almost jumped behind the counter and did it for him. He’s probably already in 17 different vulnerability categories anyway though.

          I doubt you could build a model to make such a priority assessment. It would be so complex and riddled with questionable assumptions that it would be a year late and many dollars short of functional. However, a few reasonably intelligent and knowledgeable humans could have started working on a plan in July, vetted it with the public and had it ready to go by the end of November.

        • Of course I could not build such a model, I’m neither an epidemiologist nor an infectious-disease specialist nor an environmental-control systems engineer,etc. But we have such experts. The goal is not high precision nor absolute fidelity, the goal is to capture the most important factors. I don’t know what you envision being in a plan, but when we may only know parameters within ranges, models let us estimate best, worst, and more likely outcomes.
          “Senior citizens depend on a strong economy to generate their pensions, but they don’t do much to produce revenue.” Neither do children. I don’t measure the value of a human life by economic output. Further, we do not live in a subsistence economy; only a fraction of our GDP represents necessities of life. Much of our output is in discretionary sectors, and at that we don’t need the labor of children or the elderly to produce this surfeit. So in a pandemic we can reduce travel and entertainment activity drastically without life-threatening consequences, we can reduce the production and consumption of luxuries, so long as we are willing to support idled employees with the means to eat, keep their homes, pay the utilities, and other necessary expenses. Debt will expand, but by minimizing disruption to the basic structure of the economy, we will be able to restart it faster, and begin to return to normal levels of output.

      • “the senior (sorry seniors!) is doing nothing useful at all”

        I’m 66 and like many people, I work (not out of necessity, though) so my first response is screw you. The world is full of productive people 65+.

        Beyond that, something that more vulnerable people are doing is filling hospitals and dying (generally after hospitalization). Vaccinating more vulnerable people, in particular those 65+, along with HCWs can fairly quickly bring our hospitals back to normal operation. No more situations like LA County putting ambulance arrivals who aren’t faced with an immediate life-threatening situation (like severe trauma) into portable structures with a dozen or so watched by a single EMT, with a 12-36 hour wait to see an ER doc and an average five day wait for admission.

        So clogging hospitals with covid patients also effects people who need emergency care.

        I think the prioritization is fine. CDC recommended moving those 65+ to higher priority precisely because of delays in getting the vaccine out and the impact being seen on hospitals.

        What I see from others reminds me a bit of a pack of feral dogs fighting over a handful of kibbles. It’s not pretty. And whatever one thinks of the situation, the best solution is to find a lot more kibbles.

        The weekly 10.5 million delivery goal for three weeks seems doable – Monday and Tuesday saw just under 6 million doses delivered (CDC tracks by when they’re received, not shipped, now). That’s scheduled to go up in later March by a significant amount.

        My county has gotten to the point where HCWs and first responders who will take it (those who won’t are another problem of course) have all gotten first doses. My gf qualified as a first responder and will get her second shot a week from tomorrow, by mid-February second shots for HCWs and first responders should be done. So we’re on to 75+ and will probably open up for 65+ in a couple of weeks and will be starting on essential workers in a similar timeframe.

        • I’ll second the “screw you.” I’m over 65 and feel I’m being useful – in fact, I even cause damage to some people (some of my teaching evaluations are evidence of that – luckily others balance that out). If we want to measure how much “usefulness” there is in different people or groups of people… well, I think I don’t want to go down that path.

      • To complicate matters more when discussing seniors and vaccine scarcity: One ideally also needs to consider (when possible) something like “QALYs” (Quality Adjusted Life Years) when making decisions about allocating vaccine that is in short supply . Real life example: An 80-year old relative in poor health got her first COVID vaccine shot two days before she was diagnosed with metastatic cancer and opted for hospice care. ( I don’t think she’s going to get the second shot.)

  3. This will not be the last pandemic. We have fifty states. Central federal leadership should have allocated three or four different rollout strategies to clusters of states to get real data on this question. Lots of smart people thinking very hard does not work better than real world experience. Since we did not set up the a priori study, we should at least take a good look at how things are playing out. That will require cool objectivity, no preformed ideas, and absence of political motivations.

    • My position is that we could have modeled this problem well in advance of the availability of vaccines, and obtained much of the needed empirical data in advance; for example, we should have collected data on viral shedding in vaccine trials. We also need much better data on contact rates and durations by age, occupation, household characteristics, etc.; data on risk-reducing behaviors. Such information would at least have constrained the problem space, and reduced risk for all participants in any rollout trials.

      • Ken has it right. It is now one full year since we knew this was at least a potential world-changing pandemic. And yet I can’t find any evidence a year into it that we’re trying to collect the most basic epidemiologic data concerning the mechanism of spread.

        All the behavioral and policy models attempt to work backwards using secondary data like mobile-phone mobility or positive tests among self selected test seekers. I don’t think anyone can put a convincing number on the proportion of infections that are passed by any given type of contact (i.e. grocery store worker, people eating in restaurants, people wearing masks, people NOT wearing masks).

        Everybody thinks the correct policies and behaviors are “obvious” or “no-brainers”. SO I guess nobody really wants data to be collected that might, heaven forbid, be seen as contradicting their pet no-brainer solution to the pandemic.

        • I have been writing high and low for months; that the accumulating mountain of case histories supports backtracking to clusters and inference of relative risk for different routes / sites of transmission. That information evolves dynamically, and probably regionally, and should be the basis for concrete, dynamically evolving guidance to the public. I can download the spreadsheets from the county and see at a glance that the big box stores are risky places — for employees, without a doubt — but what about for customers? Well, I have to *presume* that is the case, absent any concrete data. But the concrete data, through which some fraction of cases which this week may be associated — with confidence — to some particular site of transmission last week or the week before … this concrete data *exists* and should be put to use. It. Is. A. Scandal. That. It. Is. Not. But we are dulled to scandal aren’t we now? So let’s just sit on our hands and watch netflix.

        • “And yet I can’t find any evidence a year into it that we’re trying to collect the most basic epidemiologic data concerning the mechanism of spread.”

          This suggests one of two possibilities

          1. epidemiologists have been sitting on their asses for the last year

          -or-

          2. you need to look harder.

        • the accumulating mountain of case histories supports backtracking to clusters and inference of relative risk for different routes / sites of transmission. That information evolves dynamically, and probably regionally, and should be the basis for concrete, dynamically evolving guidance to the public. I can download the spreadsheets from the county and see at a glance that the big box stores are risky places — for employees, without a doubt — but what about for customers? Well, I have to *presume* that is the case, absent any concrete data. But the concrete data, through which some fraction of cases which this week may be associated — with confidence — to some particular site of transmission last week or the week before … this concrete data *exists* and should be put to use.

          But, No, sir; it is *not* being put to use, not in this county. It could be. But it is not.

      • “we could have modeled this problem well in advance of the availability of vaccines”

        Why would we? We don’t need a plan that’s perfect to the fourth decimal place. We need a plan that is generally effective and acceptable to the public. For the last part, for sure, you need human input.

        If people want to model something so bad why not model something that humans aren’t capable of, like model how fast the virus could propagate through a large gathering via each of the five or six proposed modes of transmission? This has been doable since May but I haven’t heard of it being done.

        Just breaking it down, you’d have to start by modeling how people interact at a party; then analyze the exact mechanisms and probabilities of transmission via aerosol, oral droplet, nasal droplet, human touch, inanimate surfaces etc. Big challenge. Not doable by humans with pencils.

  4. It is moral to want to protect the most vulnerable. It is practical to seek the most efficient way to end the pandemic. It is pragmatic to recognize that we have neither the vaccine supply nor the distribution systems, at present, to vaccinate the 20-49 y.o. population. It may be that, if we are limited to vaccinating smaller groups for the foreseeable future, the elderly and at-risk are the best small groups to target.

    • The foreseeable future …. is it complete fantasy to suppose that the manufacturing and distribution be treated with all available dispatch? Perhaps it is fantasy …. “we” cannot even figure out how to put the effort into making the stupid damn masks.

      • These vaccines aren’t something you cook up on your stovetop. It takes the building of industrial plant which takes time.

        The production of the mRNA vaccines for the US is scheduled to take a sizable leap several weeks from now. In the interim, providing low dead space syringes is allowing for the getting of six rather than five doses per Pfizer vial, and Moderna is seeking FDA approval to increase doses per vial to 15 rather than 10 because in their case filling/capping of vials is a bottleneck.

        • When Gen. Groves made inquiries of the US Treasury, asking if they could provide so-many hundreds of tons of silver for (the secret Uranium separator at Oak Ridge) the initial response was, “The unit of account here is the troy-ounce”. That attitude had hung the project up hitherto and that’s why a tyrant like Groves (and a 1st rate engineer) had been put in charge. It paid off.

  5. Why didn’t they monitor virus shedding in the trials? I think it’s the cost and difficulty of repeatedly testing 30 thousand trial participants. It couldn’t be done with a small subsample either, since most trials accumulated 100-200 cases of Covid. Even if one doubled that to account for asymptomatic infections, it’s just one % or so of the total.

  6. I am not a big fan of this, since

    a) currently vaccine supplies are low, and most of the deaths are in a small percentage of the population
    b) we are a lot more certain that the vaccine prevents severe illness/death than that it prevents further transmission

    With say 60 million shots we could prevent most of the deaths, especially if the British are right that the first dose is largely protective.

    I actually think we should have vaccinated the whole nursing home population and everyone over 80 before most of the “front line” groups, excepting *only* medical workers who *actually work with COVID patients*.

    (Actually, I think the balance of risk/benefit would have been in favor of vaccinating nursing home patients after Phase II trials. Probably something like 5-10% of the entire nursing home resident population has died of COVID, though good numbers are difficult to come by; any potential side effects are just irrelevant on that scale, and evidence of immunogenicity should IMO have been enough to say we can reasonably expect at least some efficacy.)

  7. A few observations:
    (a) first, people who think the rollout of vaccines isn’t going well because it’s hard to get a reservation for a shot are prioritizing process over results. The only metric is throughput. Wasting vaccine to try and align vaccines with some notion of either distributive justice or in some prioritization against spread or against vulnerability is insane. Just measure throughput at the state level and you’ll see what states know what they are doing and what states don’t.
    (b) on a personal level, I am ineligible as of today, but will become eligible when I turn 65 next month. However, I have cleverly evaded the vaccination restrictions by actually getting the disease last month. That’s one way to get a few months’ breathing room. I guess, while vaccine stockpiles grow. Fortunately, my case was ridiculously mild. What’s interesting to me (ex post) is that the arguments about whether or not this is “worse than the flu” rarely confronts this unfortunate heterogeneity: for 90 percent of patients it’s better than the flu, and for 10 percent of patients it’s 1000 times worse. While that makes it much worse on average, this is a case where the mean is not all that interesting.

    • >>this is a case where the mean is not all that interesting.

      Yes, exactly.

      Which I think contributes to some of the “oh it’s no big deal” stuff – for someone who had a mild case, or all the people they know who have had it had very mild cases, that may be more convincing to non-statistics-oriented people than national-level hospitalization and death statistics.

      Especially if one doesn’t trust the government providing the numbers…

    • I live in Connecticut, which is methodically following its prioritization scheme, and is one of the best-performing states, in terms of doses administered per population, and proportion of received doses administered. Prioritization need not hinder throughput.

  8. Haven’t read the comments, but obviously what must be factored on is the likelihood that cohort will show up to get jabbed.

    It’s like when people said minorities shouldn’t be prioritized over old people. Again, all of this depends on who shows up to get vaccinated.

  9. I’m a little confused by the paper’s assertion “School reopening has not resulted in substantial increases in COVID-19 attributable deaths” when they also write “School opening is associated with an estimated 25.7% [14.5%-40.5%] increase of COVID-19 infections and a 5.9% [3.4%-9.3%] increase in COVID-19 attributable deaths.”
    5.9% might not sound like a lot but given that there has been 450k+ deaths in the US so far, their estimate suggests school opening could be associated with ~27k deaths. (probably less so since only some schools are open)
    Correct me if I misunderstood something.

  10. The key of vaccination at this stage is to protect people, i.e., the vulnerables and essentials etc.

    Once we’ve done this, then the hospitalization and mortality rates will be much lower than before. and then, we can vaccinate young people to stop the pandemic.

    From a realistic public health point of view, a pandemic with very few people dying is not a big deal. A pandemic of a mild disease is not of concern.

    Think about 2009 H1N1. People stop worrying about the swine flu pandemic in 2009 because the mortality is very low. I don’t remember many people get vaccinated against swine flu.

    • Yes. And your assessment is in accord with the available science, which has established that vaccines protect the vaccinated. We don’t have much data on whether they protect others by reducing transmission.

      • Moderna did some swab tests and those showed the possibility of a significant reduction in the rate of detectable infection of about 60% (not sure of the CI but it would be large).

        Preliminary analysis of field data in Israel suggests around 50%.

        While anecdotal, a local hospital has shown a steep post-vaccination drop in the rate of infection in its staff.

        It will reduce transmission, there’s very little doubt of that. The question is “by how much”?

      • We sit around all day talking about how to model with informative priors, combining data from disparate sources, and doing multilevel approaches.

        This is a respiratory disease caused by a coronavirus. We can leverage both data from other diseases in the same family, as well theories and mechanistic models. Furthermore, we have data on asymptomatic spread (not as contagious as symptomatic). Yet, so many people around here seem stuck with uniform distributions just because we lack direct observations.

        I don’t get it.

        • This. I don’t know why people reading a blog that so strongly advocates including prior information in analyses throw their hands up in the air and go full-on uniform prior (or even a prior weighted *against* an effect!) when it comes to whether or not the vaccine will reduce spread. We broadly know how coronavirus infections work. We know a lot about how our immune system works. We have lots of experience with other vaccines. Each of those three topics are complex and have quirks and knowledge gaps, so the prior should have some probability on vaccination having no effect on spread. But knowledge from these fields indicates that most of the mass should be well within the “vaccination substantially reduces spread” region. Determining if the reduction is 70% or 95%, would require data. But given our prior knowledge it would be quite surprising if the reduction in spread was only 1%, or even only 10%.

          Of course Moderna/Pfizer/everyone else should have done regular swabbing of trial participants. But they didn’t. Doesn’t mean we have to go “we have no idea if it prevents spreading”. It reminds me of the “no evidence masks work” thing at the beginning of the pandemic.

        • The only thing that is known for sure is that vaccines will prevent most people from getting violently sick, should they get infected. Most, not all. One in five or so will still get sick, despite the jab.

          However, sterilization is another thing and it appears that so far the vaccine doesn’t help with that. I believe virologists already figured that out.

          Basically, we are all disease vectors, just like mosquitoes. When vaccinated, we can still spread it, but we don’t get as sick.

Leave a Reply to dhogaza Cancel reply

Your email address will not be published. Required fields are marked *