Right now, 1.5% of all Americans have been confirmed as having COVID-19. That’s an amazing number—4 million confirmed cases—far more than any other nation. In fact, despite having just 4% of the globe’s population, the United States still has about a quarter of all cases of COVID-19—a remarkable measure of the Trump White House’s failure to provide anything that resembled leadership in the face of this enormous crisis. However, on Wednesday the Centers for Disease Control and Prevention (CDC) published analysis that matched the results of another recent study, indicating the true number of cases in the United States is likely around 10 times the number of confirmed cases. That would mean that as much as 15% of the population may have been exposed at this point, with some areas having a rate of exposure as high as 35%.
In some ways this is terrifying—the number of active cases of COVID-19 in the United States right this moment is likely around 20 million. That’s the kind of number that is daunting to the very idea of case tracing and isolation. If the full 4,300 person staff of the Census were to work on nothing else, they would each be dealing with 4,650 cases—and that’s if everyone just stood still long enough to let them work it through. But is there also good news in this? Doesn’t it bring the United States closer to “herd immunity?” And what’s with the widely circulated notion that COVID-19 immunity can be achieved with a relatively small number of infections?
There’s not a magic number that represents herd immunity. Actually, herd immunity is not even a thing—nothing makes the remaining population one bit more immune to a disease just because some other percentage already has it. 99.999% of the population might have been past COVID-19, but were you seated across from someone with an active case for some period of time, and you were part of that 0.001% who had not been infected … odds are you’d be on your way to a case of COVID-19.
All that herd immunity really amounts to is a decrease in the rate of transmission. Since the start of the pandemic, the rate of transmission for COVID-19 (often expressed as R0) has been somewhere between 2.4 and 3.0. That’s the average number of people infected by the average person with an average case of COVID-19 (on, you know, average). But that number is far from constant and far from universal. It’s already been demonstrated that some people are behind dozens of cases. Others infect no one. That can vary because of population density, housing, infrastructure, and social distancing rules, just for a start.
If you had a densely populated city where everyone tended to travel by public transport, lived in multi-family dwellings, and local authorities were doing nothing to interfere with the spread of the disease, the R0 of COVID-19 could easily reach the double-digit values usually assigned to diseases like measles or smallpox. On the other hand, a rural area where everyone travels in their own cars, lives in separate homes, and is lucky enough to have authorities enforcing a mask mandate might well have an R0 below 1—the point at which the outbreak can no longer sustain itself and gradually fades away.
That is, of course, the goal: Pushing the R0 below 1 so that each infected person infects fewer than one other person, and the number of active cases gets smaller and smaller. Much of the United States may have actually attained that status during April or May when stay-at-home orders closed many businesses and allowed large numbers of Americans to make a good run at finishing off the catalog of Netflix. Then, unfortunately, with Donald Trump banging the drum for reopening, governors swiftly reopened, pushing the number back up, then up some more, until the rate of transmission was closing in on unchecked.
Where herd immunity comes into this is just limiting the opportunity for the virus to spread. For example, if someone carrying COVID-19 meets only five people during the whole period when they are contagious, that’s five opportunities to pass along the disease. If one of those people has already been infected, then they have only four opportunities.
So if the United States were evenly spread with 15% of people who had already been infected by COVID-19, and a currently contagious person is meeting people randomly, their opportunity to pass along the disease is also reduced by 15%. If the R0 of the disease was at 3.0 and nothing else changed, 15% of people being infected would reduce the effective R0 to around 2.6. With herd immunity alone, it would take around 70% of people being infected to reach the point where R0 fell below 1.0, the epidemic could no longer sustain itself, and the number of active cases declined even if everyone just behaved “normally.”
Of course, there’s a basic problem with that. Even if the United States really has logged 40 million cases—which is far from sure—it’s also logged about 145,000 deaths. Reaching 70% infection would easily result in over half a million deaths (680,000 if the case fatality rate remains unchanged). In addition, studies have shown that 90% of COVID-19 patients still have at least one symptom six weeks after “recovering.” Close to 20% are left with lasting and possibly lifelong heart, lung, kidney, or brain damage. And that’s the short list.
There’s another big problem. Those cases are not distributed randomly across the populace. Cases are structured in knots of family, clusters of friends, and accidental strings of meetings. The percentage of people who have already been infected is often tightly connected to a church or neighborhood. Meaning that everyone else outside that cluster is just as subject to infection as they always were. That “lumpiness” in the data means that a 15% infection rate is really far from 15% effective in reducing the rate of transmission.
One more thing: Even if the CDC is right and there are 10 times as many cases out there than we know, and the overall rate of death from COVID-19 is below 1%, reaching herd immunity would still cost half a million lives and cause millions of lasting disabilities. Quite possibly tens of millions. That’s the optimistic view of herd immunity.
So why then are there so many articles circulating around the internet claiming that herd immunity with COVID-19 can be attained with only a 20% infection rate? If that were true, the United States is almost there! The pandemic will go away in a week!
Yeah, but … no.
This idea seems to have come from looking at some specific instances, such as the number of cases aboard the Diamond Princess cruise ship. The end number of confirmed cases on the literally ill-fated cruise was around 20% of the passengers and crew. A similar number of people were infected in London before the number of cases declined there, and in New York City during the horrible first months. Maybe there’s some poorly understood factor that means just 20% is the real “herd immunity” for COVID-19.
Unfortunately, that’s not the case. On the Diamond Princess, in New York, and in many other cases, the level of cases in the surrounding area made some contribution to knocking down R0 just as we discussed above. But it was all the other things done in both cases—social distancing, cleaning surfaces, wearing masks—that made the major contribution to reducing the rate of transmission and reducing the scale of these outbreaks.
In fact, it’s self-evident that the 20% “ceiling” is entirely coincidental. Not only have many locations (South Korea, New Zealand, Japan … it’s a long list) successfully wrangled the epidemic without ever coming close to hitting 20% of the population infected, there is an equally long list of specific situations where much higher percentages were infected. This includes a number of prisons where the rate of infection has reached 60%. That kind of number shows two things: First, these prisons really are approaching a value that could provide significant herd immunity; but more importantly, it shows how little care prison authorities have taken, because with even minimal effort to prevent infection, this kind of percentage would never have been achieved.
20% is not a magic number. In fact, if the CDC is right in their estimates, there are at least seven whole states that are already above this value. Even if the rate of deaths and hospitalizations really is one-tenth of what the confirmed count indicates, ending the pandemic through achieving herd immunity would be horrific, and would leave the nation with a populace weighed down with ill health that may last decades. Not to mention the mass graves.
Fortunately, we don’t have to do that. When both parties in a situation are wearing masks, masks are about 85% as effective as immunity through infection. If both parties just stay the hell away from each other, it’s 100% as effective. Social distancing, stay-at-home orders, and universal mask mandates can provide the effect of herd immunity without the cost. Until a vaccine arrives to provide the real thing.