The Path To Herd Immunity

 The profile of Covid-19's mortality by age suggests an obvious strategy for achieving herd immunity: let the relatively young and invulnerable mix freely, isolate and protect the vulnerable. Obviously I'm not the first person to say this. I am very confused as to why such a strategy hasn't been tried. 

I am a libertarian, and I don't particularly like the government telling people either that they must or must not commingle. Any version of "You must return your children to public school" or "You must report to your college campus" completely creeps me out. People who aren't comfortable with exposing themselves to the virus should be free to arrange their lives so they aren't exposed. (In fact, people who don't want to send their children to public school shouldn't have to, ever, under any conditions.) Some children or parents of children are vulnerable. They may be immune-compromised or have other risk factors. It certainly makes sense that they should be able to keep their households out of any herd immunity strategy that involves a controlled spread of the virus. But I also think people shouldn't require a doctor's note to make these decisions about arranging their lives. Maybe someone has no risk factors whatsoever. The other risks they assume in the normal course of their daily lives might be orders of magnitude larger than their mortality risk from Covid-19. They should be able to exclude themselves and their children from any government plan, for whatever reason or for no reason at all. 

With all that said, infectious disease control is a legitimate role for government. It is a classic externality problem. Your individual efforts to stem the spread are a public good. The benefits are non-rivalrous (everyone gets them; someone's benefiting doesn't preclude someone else's benefiting from the same quantity of spread-mitigation). They are non-excludable (you cannot prevent someone from benefiting from your efforts of spread-mitigation, hoping perhaps to extract a fee from them). I'm an anarcho-capitalist (at least on even days), and even I have to admit that there is a compelling reason to have government do something in this space. (Assuming your particular government can do so competently...admittedly that qualifier is often not satisfied.). Governments should not have the power to close down businesses, but they should be collecting information, issuing guidance and making decisions about how and when to open schools and other institutions that it directly controls. (If a purer anarchocapitalist wants to get mad at me for saying this, fine. But given where we are today, the government "public health" institutions that currently exist are the only game in town. I'm speaking to what they should do assuming they will continue to exist, which they certainly will.)

From the paper linked to above:

The estimated IFR is close to zero for children and younger adults but rises exponentially with age, reaching 0.4% at age 55, 1.3% at age 65, 4.5% at age 75, and 15% at age 85. We find that differences in the age structure of the population and the age-specific prevalence of COVID-19 explain 90% of the geographical variation in population IFR. Consequently, protecting vulnerable age groups could substantially reduce the incidence of mortality.

That paper was apparently posted in late August, but some version of this was known since March, when the lockdowns began. The elderly are vulnerable, and "elderly" is really a proxy for "has underlying conditions." (As in, some elderly people with good lung capacity are not really at risk, just as some young people with respiratory issues are at elevated risk.) There is no excuse for not using this information to guide public policy. The risk to children and young adults is minuscule, orders of magnitude smaller than other risks that they assume (or that their parents subject them to, presumably with their best interests in mind). If we had simply protected the vulnerable populations but allowed the virus to spread among the young and healthy, we'd have some degree of herd immunity by now. We'd have these epidemiological fire-breaks in our public schools and universities. The virus might get in, but it would find few new hosts and ultimately have nowhere to go. 

I understand why people don't like this strategy. The objection is usually some version of, "If young people get the virus, they will spread it to vulnerable people." It's hard for me to put into words just how much I have lost patience with this line of argument. It never made much sense. First off, we need to treat "tolerance of lockdown policy" as a depletable resource. (Robin Hanson says it well in this post. Hanson also has some posts from early March arguing, and backed with numerical simulations, that it makes sense to expose the non-vulnerable.) People eventually tire of living like prisoners, being shut-ins, being denied the services they're used to for arbitrary reasons. They begin to spontaneously disobey the law, then eventually they demand policy changes for a return to normality. Bearing this point in mind, I think we have actually wasted a tremendous amount of time. From mid-March through May, I think most people were extremely diligent about staying isolated. It's only since then that workers started returning to offices, senior citizens started venturing out more, etc. My in-laws spent two months inside their house, not even venturing out for a walk. My wife brought them groceries. That was the time to let our small children return to school. So what if there was an outbreak of Covid-19 in schools and universities? The virus would have spread within a resilient population while the vulnerable were being fully isolated from them. (I wonder how many college students were sent home to live with their elderly parents, or non-elderly parents who comingle with the elderly grandparents. College dorms should have stayed open to give these kids an option to isolate themselves from the vulnerable, but that was botched.) There is even an argument for not closing down the school and not being overly strict about mask-wearing and temperature-checking. (As in, are we going for herd immunity or aren't we?) The vulnerable were being isolated and were, for the time, tolerating it. At this point, my in-laws are making no effort to keep their distance from their grandchildren. People are slipping out of their habits of scrupulous caution, even the ones who are very adamant about lockdowns and mask-compliance. Sheer exhaustion is setting in. 

(Call this anecdotal, but I saw large numbers of graduations parties in late May and early June. Unmasked young adults were comingling, eating together, conversing, and presumably doing other things young people do together. One night I could hear high schoolers partying in a nearby back yard. This matches basically what I've heard from parents with teenage children and news stories of Covid outbreaks spawned by wild parties. The parties are happening. We need to construct public health policy assuming that there will be non-compliance. Paraphrasing Donald Rumsfeld, we implement public health policy with the public we have, not the public we'd like to have. And, once again, it's really not a big deal if these young people themselves get sick, unless of course they spread it to the vulnerable.)

We may have missed the best opportunity to go for herd immunity, but it's still an option. (In my opinion, the best option.) It may require asking the vulnerable to return to levels of isolation and precaution they were experiencing in March through May. 

Someone could dissuade me of the "herd immunity" strategy being a good idea. But it would take some kind of quantitative argument. Perhaps someone could model the "mixing" of vulnerable and non-vulnerable populations and demonstrate clearly that, under reasonable assumptions, there's too much leakage? (I believe this Twitter thread is trying to quantify the leakage issue, so is this one.) Even then, someone would have to articulate a clear path out of this mess. They'd have to answer philosophical questions about what the non-vulnerable's duty is to the vulnerable. What do the vulnerable even want (as in, are they even asking their children and grandchildren to halt their lives)? And what's the point of preserving a virus-free "commons" that nobody is supposed to be using? Even if it doesn't confer significant herd immunity, unleashing the young and relatively healthy is the right thing to do for these other reasons. Focusing mitigation efforts specifically on the vulnerable is rational public policy. You want to allocate resources and efforts to where they're doing the most good. The flip-side of that is that we don't want to incur excessive costs "protecting" people who aren't really threatened. 

There are other objections to a "protect the vulnerable, unleash everyone else" strategy. To many people, I'm sure it sounds like "Let's intentionally subject people to viral infection." But this is the wrong framing. We're talking about allowing people to return to their normal lives, where the virus will be one of many risks they encounter. If that's "intentionally infecting people with the virus", then allowing people to drive is "intentionally subjecting people to fatal traffic accidents." (More to the point, allowing 16-year-olds to drive, knowing they must start learning somewhere, is "intentionally subjecting children to fatal traffic accidents." I have this same "We need to start somewhere" reaction towards people who want us to stay locked in forever.) This isn't just an objection to going for herd immunity, it's an objection to letting anyone take any risks for any reason. It's not serious. There really is an important moral difference between an act of commission and an act of omission. 

Another possible objection is that we shouldn't even subject children to the risk of the coronavirus. (I'm not sure anyone is actually saying this, but it seems to be implicit in some of the arguments I have heard with respect to lockdown policy.) I think this is kind of silly, because we subject children to much larger risks all the time, usually without even thinking about it. They face mortality risks from auto accidents, swimming pools, and trampolines. Put in it's proper context, the risk they face from the coronavirus is a rounding error. It is perfectly appropriate that we are inured to certain background risks, which we have implicitly or explicitly chosen to accept as a fact of life. 

Yet another objection is that we can simply sit tight and wait for a vaccine. Why bother with the carnage of hard-to-control viral spread when a vaccine will ultimately save us? I personally don't think this is reasonable. A working vaccine is still months out by almost any sensible projection. And who knows how effective it will be? A vaccine is a means of generating herd immunity based on extremely mild to asymptomatic infections. Same with "protect the vulnerable, unleash everyone else." Some small number of young people would get extremely ill and die under an "unleash the young" policy, but then again any vaccine could subject recipients to similar risks. (When you get a vaccine, there is usually a liability waiver listing several warnings about who should and shouldn't get it. There's always the chance that someone misjudges which category they're in, or that the vaccine has unknown defects. See the CDC's list of real and imagined vaccine safety concerns.) I also have this "boiling the frog" reaction to the notion of waiting for a vaccine. It might seem like one is in reach at this point, but if you'd told people in March to "live like this until a vaccine comes out in 2021" they would have rightly revolted. 

(Somehow variolation and isolation is off the table? Again, here is Robin Hanson, who was right about many things very early on. Is that another example of people getting squeamish about "intentionally" doing something that's basically inevitable anyway? As if controlling who gets the virus and when is somehow worse than the "non-deliberate" alternative that kills a comparable or larger number of people?)

________________________________________

I know that "herd immunity" doesn't actually mean what it sounds like. It doesn't mean the virus can't infect anyone. It basically means there are enough immune people in the population that the virus stops spreading at an exponential rate. It can still be spreading at a linear rate. I am using the term loosely in this post to mean "some kind of benefit is conferred by having large numbers of immune individuals." I am not referring to a specific phase transition that happens in epidemiological models when exactly 71.5% of the population has immunity. As many commentators have pointed out, this is a squishy concept anyway. It can exist in some  places but not in others, and it is dependent on behavioral responses to the virus, it's not a property of the virus in and of itself. I don't particularly care what exact threshold we need to reach for herd immunity. Quite simply, more immunity is better than less immunity, particularly when it's simply a by-product of doing the right thing (unleashing the young). 

I am seeing atrociously bad faith treatment of the herd immunity concept in the media. See this piece in The Atlantic (or listen to the associated episode on the Social Distance podcast). There is no mention of the "protect the vulnerable" part of the plan, which is crucial. I follow a few lockdown skeptics in my podcast and blog feeds, and they are all constantly criticizing our inability to protect nursing homes from outbreaks. Practically nobody (short of out-and-out virus deniers) is advocating a "let 'er rip" or "yank off the band aid" approach, in which no mitigation measures whatsoever are taken. The attempt to associate a herd immunity strategy with "the right wing" and to smear Scott Atlas are pretty clear evidence of the Atlantic piece's agenda. I may do a longer post dealing specifically with this piece. It was near the top of my Google search, and it seems pretty typical of the language I'm hearing on this topic. So it presents a good foil.