(Updates below)
Something has been strangely lacking in the national discourse about vaccination mandates to deal with COVID: how does recovery from having the virus compare to being vaccinated when it comes to future personal risk and ability to spread the virus to others. The implicit assumption throughout the media is that anyone who does not want the vaccine has never had COVID. The benefits of the vaccine for the individual and the effect on transmission as described by the CDC mention nothing about the effect of natural immunization1, other than an August 6 press release regarding a disappointingly awful study claiming to have found that KY residents who were unvaccinated were over twice as likely to be reinfected2.
Yet, as a recent article in Lancet discusses, natural immunity seems to be at least as good as vaccination in terms of reinfection and resistance. So much so that the article states [footnotes removed]:
Although those studies show that protection from reinfection is strong and persists for more than 10 months of follow-up, it is unknown how long protective immunity will truly last. …We are limited by the length of current reported follow-up data to know with certainty the expected duration that previous infection will protect against COVID-19.
In other words, so far as we know, natural immunity lasts as long as immunity through vaccination, and is equally strong. If you had to choose between vaccination and getting an unknown dose of COVID and hoping for the best, yea, get vaccinated But, if you have already had COVID and recovered…
Given the evidence of immunity from previous SARS-CoV-2 infection, however, policy makers should consider recovery from previous SARS-CoV-2 infection equal to immunity from vaccination for purposes related to entry to public events, businesses, and the workplace, or travel requirements.
Now, sure, this is one article looking at a handful of studies and summarizing them. It might turn out it is understating, overstating, misstating, or just stating the facts, and we should be cautious.
The real problem here is that we didn’t really need this to question the value of vaccination of those who already lived through the disease, but no one seemed capable or willing to discuss the vaccination gap in those terms. Let’s consider briefly the cost/benefit analysis of vaccination given you have natural immunity due to previous infection.
When doing cost/benefit analysis of a course of action, it is necessary to define BOTH the course of action’s end state and the base state. The benefits are in terms of how much better the end state (call it S1) is than the base state (call it S0) and then how much it costs to get to S1 from S0. This is an application of marginal thinking3, the little difference between S0 and S1. We don’t consider “Everything is perfect!”4 nor “Everything is awful!” but rather “This is pretty bad, and that is pretty bad, too,” and then try to decide if the difference between the two is net positive, and therefor worth changing from one to the other.
So if we are considering whether or not to get vaccinated, our pretty bad S1 outcome looks something like “I have X% less chance of getting (re)infected, Y% of dying if I do, and Z% chance of transmitting the disease to someone else if I get reinfected. If I do transmit it, that person has A% chance of getting it, B% chance of dying, and C% chance of transmitting it.” X and Y are the purely personal concerns, Z, A, B and C are the social externality concerns5. What those values happen to be, that’s an empirical question. We will get back to those. The most important outcome variable as we shall see is Z, which is not zero (the vaccines do not sterilize you) because A, B and C all depend on it, as does the justification for forcing you to get vaccinated.
What are the benefits of vaccination?
For someone who has never had COVID, getting vaccinated then makes a fair bit of sense, depending on your other characteristics. You are less likely to be noticeably infected if exposed, less likely to die if you get infected, and somewhat less likely to transmit the disease if infected (how much is questionable, the CDC giving a range of 0-50% less likely to transmit). Now, if you are under 65 and in good health, you have a very low risk of death anyway, so the benefit is minor there, but you might get a less serious case which is nice. (People get flu shots despite sometimes getting sick, so I am assuming they consider avoiding a week’s sickness as a benefit.)
Your risk of getting infected in the first place is questionable; I haven’t seen great numbers on how likely people are to get a first time infection after vaccination when exposed to COVID compared to unvaccinated. Unless we start doing human tests where we actively expose people to a known amount of COVID we are probably never going to have those numbers. The best we can go with is probably cases of COVID people happen to get that are serious enough to get them tested again, as opposed to asymptomatic infections.
Conditional on being infected, you are somewhere between 0% less likely to transmit to 50%, based on estimates from viral load. That helps a bit, as viral load seems to determine how sick someone gets upon being infected. We don’t know how much, and again, the vaccine does not sterilize you, so you can still harbor and transmit the disease.
So for those who never had COVID, vaccination offers personal benefits of lower severity and lower likelihood of death, and some social benefits of less likely transmission, although how less likely is unknown, and the numbers are not high. Notably, however, for segment of the population that is unlikely to die from COVID at all, the personal benefits are not nearly so high as for the more vulnerable members.
If we are to believe the new Lancet article, however, for those who have had COVID and recovered, all those improvements disappear. Which… makes perfect sense, really. Vaccination is simply the safer version of “Get the disease so your immune system knows how to fight it,” training your body against a lesser or dead version of the disease (or mRNA tricksiness) so that it is ready for the real thing. If you already have had the disease and gotten through, what is the lesser version going to train your immune system to do that it doesn’t already know? A vaccine is preferable to taking your chances with a very nasty disease, but once you have had the disease it isn’t clear what benefit there would be.
Note that the vaccine wouldn’t be expected to confer transmission benefits to those who have had the disease, because to the extent it works, it works by lowering viral load, not by sterilization. That is, if you have less virus bouncing around your system because your body is smashing it up effectively, you have less virus to transmit. Both the vaccine and natural immunity result in your body being able to lower the virus load when it comes to making you sick, and so lower the viral load when it comes to transmission.
So, on the benefit side, those untouched by COVID get benefits from the vaccine that roughly match having gotten the disease by normal transmission: upon (re)infection, lower severity, less chance of death, and possibly lower chance of transmitting to others. The benefits of the vaccine to those who have already recovered, however, are apparently close to zero.
What are the costs of vaccination?
Fortunately, these are pretty similar whether you have had COVID already or not, so it is a pretty short list.
You might get sick with COVID like symptoms for a few days. About half the people I know who got vaccinated reported this. One nurse told me she had COVID early, then got the vaccine and got symptoms worse than her original case. Others were apparently fine, or just felt icky for a day or so. So there’s a chance to get sick.
Some other side effects we don’t know about yet. Even “safe” medications have possible side effects, which is why we don’t just take a few dozen different medications every day just in case, Pascalian Medicine as Scott Alexander put it. I personally would say this is a very low expected value cost, but then again I avoid taking medicine whenever possible. I don’t even take aspirin or ibuprofen if I can avoid it, and I know when I have a headache or muscle pain. No one I know died from the vaccine, but there is at least one website that tracks reported side effects, I am told.
Taking a knee to overbearing officials making demands outside their legal purview. Governments, and those who are a big fan of them, seem to overlook this one a lot. Just as there are always people who find a button labeled “Do Not Press” irresistible, so there are people who will stop doing something upon being told it is mandatory. Particularly when those doing the telling have no proper authority to make those demands.
There are no doubt other costs out there, like not liking to get shots, but it seems to me these are the big three. Maybe big two, but I think 1 and 2 should be separate since we know getting sick is a side effect (because that is pretty much how it works) and other side effects are much less certain.
Trading off costs and benefits
So what is the trade off here?
If you don’t have natural immunity, the private benefits of vaccination seem pretty large compared to the costs. If you are in the age range where deaths are more common, or have health problems that make a serious case of COVID risky, vaccination makes a great deal of sense. Less so if you are young and healthy; the benefits to children seem to be close to zero compared to just getting the disease since the risks are so low, but adults probably come out ahead on the costs and benefits unless they find standing up to unreasonable mandates very appealing.
On the social side, vaccination is less of a slam dunk when it comes to benefits because of the very moderate benefits against transmission, and the low costs of getting infected for most people. To put it another way, transmission only matters a great deal if others cannot protect themselves; if most everyone with a greater than 1% chance of death can be protected via vaccination or natural immunity, and those who really don’t want to get sick can get vaccinated and so be protected, then transmission is not that big of a deal6.
If you do have natural immunity, however, the benefits of vaccination are effectively zero according to the Lancet, so ANY costs will outweigh the benefits, to both you and society. Any side effects, whether getting sick for a few days or more exotic outcomes, mean society is worse off for your getting vaccinated7. Even without medical side effects, simply adding one more illegal dictate that you accept is a cost to both yourself and society that should not be ignored.
So when consuming your favorite media and seeing people argue “Vaccination obviously fantastic; only crazy people are not willing to get vaccinated, and we should make them,” consider that some of the unvaccinated are those with natural immunity, for whom vaccinations are as likely as not a net cost. It would irrational for those with natural immunity to get vaccinated unless they strongly believe that their natural immunity is less effective than vaccines. The evidence doesn’t seem to support that, nor does a basic understanding of how vaccines work; I bring up the latter because the state of science and dissemination of information is such that trusting what is presented as evidence is rather hard these days.
At the very least, a charitable and honest discussion of vaccination should recognize that those who don’t want the vaccine have a better reason than just not liking shots, hating Joe Biden, or being anti-vaxxers.
Now, if only we had good tests to see if people had COVID previously or not…
How many people in the USA have natural immunity, anyway?
Just a little factoid, since it doesn’t really fit anywhere else, asking Duck-Duck-Go how many documented COVID cases are there in the US gives an answer somewhere between 48-49 million. Of course, many people get asymptomatic cases, or cases that don’t require hospitalization or are mistaken for a cold or flu and so don’t get documented. I have seen estimates for total cases including these undocumented cases as high as 160 million, but who knows. (Man, wouldn’t it have been nice if there had been really cheap tests available?) At any rate, perhaps somewhere between 13-50% of Americans have had COVID already, and thus natural immunity.
How does that line up with how many are vaccinated? Damned if I know, and I am pretty sure no one does. If we roll with assuming 150 million people have had COVID, then chances are pretty good that the 35% of adults who haven’t gotten vaccinated are mostly naturally immune. If the number is closer to 20-30% of the population have had COVID, then there could be zero overlap. We just don’t know how many people had COVID over the past two years to make a useful inference.
My guess, is that 160 million cases is actually underestimating the total number. A highly virulent respiratory disease (the Delta variant has an R0 of 5-6, that is, every person infected infects 5-6 more on average) with symptoms ranging from death to asymptomatic spreaders, that seems to me like the sort of thing nearly everyone gets after a year or two, but most don’t realize it. I wouldn’t be surprised that an actual infection case count of 60-95% of the total population turned out to be the case after a year and a half, with the majority of those infections being in the sort of people who “never get the flu”8 and so don’t even realize they were hosting the virus. The documented spread of the disease despite lockdowns, mask mandates and all that suggests to me that COVID transmits a lot faster than we think.
Update: Shout out to the Common Sense Substack posting a story on this topic the day after that covers much of the same ground, although with actual journalism things like interviewing people. It also links to a paywalled Wall Street Journal article that apparently links to a study showing that natural immunity is 27 times better at preventing infection than vaccines? If anyone has the link to the study, or a full text, I would appreciate it!
Note on terminology: I am using “natural immunity” to mean something akin to “Vaccinated by getting the actual disease.” Neither getting vaccinated or recovering from COVID seems to make you immune to getting it again, nor do either “sterilize” you against spreading the disease should you get it again.
Link here. The article doesn’t give you the information necessary to get to their 2.34 times more likely to get reinfected number; the closest I can come from Table 2 is 39% of unvaccinated getting reinfected compared to 23% of vaccinated. The discussion section includes a lot of mitigating factors as well, such as mismatches in the vaccination data base meaning the actual count of vaccinated people is undercounted. That makes me rather skeptical of the 16% increase in reinfection rate. Perhaps I shouldn’t be surprised that a CDC sponsored study written by CDC employees would fully back the CDC’s preferred message, but it smacks of “Well, our data is bad and untrustworthy, but its small result supports our preferred outcome, so we will go with it.”
Man… it would have been great to have already written up that post from Econ’s Greatest Hits. I need to get on that…
That being the Nirvana Fallacy.
Yes, even if you have a family that loves you, whether or not you get sick or die is your business alone. You should probably take those other people’s feelings into account when you make decisions, but if we are going to assume people own themselves, their bodies and their lives, no one else gets a claim on how your decide to spend your life.
“What about those with compromised immune systems who can’t get vaccinated?” I hear you cry out. Well, compared to what… what was your plan otherwise? Vaccinated people can still transmit, so those with compromised immune systems are still going to be exposed. I also find myself wondering what people with compromised immune systems do to avoid catching the normal corona viruses, you know, those that cause cold and flu in people. One hopes that will work against COVID, because it is clear that non-sterilizing vaccines won’t help.
Always remember, when considering social benefits and costs, include your own benefits and costs. I am amazed at how often I have had to remind students “You are part of society. Your costs and benefits count!”
Looking at you, dad.