Why Chasing Herd Immunity (Without a Vaccine) is a Bad Idea
Updated 1/31/2021
At a December briefing devoted to the rollout of vaccines to combat COVID-19, President Trump said “You develop immunity over a period of time, and I hear we’re close to fifteen percent [infected]. I’m hearing that, and that is terrific. That’s a very powerful vaccine in itself.” Taken at face value, this figure would have been on the low side, and the President actually may have been referring to the total number of infections found through testing, then around fifteen million.
Trump’s enthusiasm for higher numbers, however, suggested the influence of a belief that more infections could accelerate the end of the pandemic through naturally-occurring herd immunity, when a disease comes to a halt because it can no longer spread through sufficient new hosts.
Though he didn’t explicitly embrace this approach, Trump’s comment alludes to a view holding that other than for those at the highest risk—the very old and those in group and nursing homes—there should be minimal public health restrictions and few if any disruptions to regular daily life and activities. Schools, restaurants, bars, and gyms should generally stay open. Though infections will multiply, herd immunity will be reached more quickly and the resultant social and economic benefits will outweigh the short-run, immediate cost of a higher number of infections and fatalities.
This de facto support for seeking herd immunity even before vaccines make a dent in case and mortality numbers comes from several directions. Former Trump advisor Dr. Scott Atlas has been a strong proponent of limiting and ending most public health efforts at mitigation and prevention. A group of scientists and doctors (like Atlas few of them epidemiologists) have backed this approach in the so-called Barrington Statement.
While few proponents believe that individuals should actively seek out infection (as in chickenpox parties of old) most believe in a strategy in which herd immunity is the byproduct of an approach that minimizes restrictions on daily activities and individual behavior. Many of those who hold this view, for example, hold attitudes toward encouraging or mandating the wearing of masks that range from skeptical to hostile. Openness to herd immunity typically goes hand in hand with a certain degree of fatalism toward the spread of infectious disease. Mark Meadows, President Trump’s chief of staff, for instance, said bluntly that “We are not going to control the pandemic.”
This herd immunity strategy shouldn’t be dismissed out of hand. Indeed, for infectious diseases without a vaccine on the horizon, and that have established a persistent foothold in a population, there is no alternative either than reaching collective immunity through natural means or imposing a prolonged and drastic mitigation in hopes that the disease will run out of hosts or mutate in a less virulent direction. And even under these latter circumstances the primary goal is often to stagger the rate of growth of a disease and recoveries so that hospital systems are not overwhelmed.
What could it mean—in terms of infections and fatalities-- to embrace a strategy of herd immunity, in other words to roll back even the existing patchwork state efforts at public health mitigation before a vaccine has been widely administered?
As 2020 ends, we are confronting all-time highs in the U.S. in cases and impending near-certain highs in sustained fatalities, with the latter already reaching near-record levels and the impact of both winter and the holiday surge yet to run its course. Current fatalities are largely the result of the Thanksgiving surge; given the roughly monthlong lag time between infection and fatalities, deaths from the virtually inevitable Christmas/ New Year’s surge are likely to peak in late January and subside by mid-February or so-- absent an unexpected change in behavior, changes in the virus, or a vaccination schedule that deviates widely from current expectations.
We can be pretty certain that there is still ample room for Covid-19 to spread--even leaving aside the possibility of reinfection. Roughly twenty million Americans have tested positive for Covid-19: let’s make the reasonable assumption that this is one-quarter of the actual total, so that there are roughly 80 million Americans who have or will have some measure of immunity.
Currently, with modest efforts at mitigation, the U.S. is recording a record 200,000 cases a day. Without curbs on movement, some restrictions on business openings, and efforts to encourage the wearing of masks, this total could easily reach 250,000 a day for a sustained period. It is doubtful that it would go much higher, as some degree of local saturation will take place and individuals typically take more precautions as hospitals near peak capacity and deaths rise, but this is a reasonable ballpark figure.
For the low average over the next four months with Covid-19 mitigation in play, 100,000 new infections a day is plausible, since the upcoming months will average much higher totals and then presumably will start to taper off by mid-February or so as holidays recede, the weather improves, the new administration issues more advisories, and vaccination starts to make a measurable impact.
The difference between the two scenarios over four months is roughly 18,000,000 additional infections under the non-intervention scenario. If the case fatality rate for these new infections were .25 percent, far below the current nominal rate but reflecting the deaths or vaccination of many of the most vulnerable, this would still result in some 450,000 additional deaths over this period, more than doubling the national death toll from Covid-19 up to this point. (This back-of-the-envelope number is consistent with most of the major models and estimates of how many lives might be saved, for instance, through more consistent and proper wearing of masks.)
And even with this new surge of infections the country would still be far short of the roughly 65 to 70 percent of individuals believed to be needed to reach natural herd immunity, since the optimistic hopes that herd immunity could be reached at a far lower percentage have largely been dashed.
On its own this potentially high death toll doesn’t tell us whether a de facto herd immunity policy is a good or bad idea. The likely outcomes need to be balanced against the speed and effectiveness of vaccination, the consequences of allowing high infection rates, the degree of mitigation that public health measures can offer, how onerous that impact is on everyday lives and the economy, the durability of acquired immunity, and the impact of Covid-19 on the large majority who weather the illness, among other factors.
Without going into detail, most of these balancing factors would seem to lean against the strategy of a full reopening with few or no restrictions that the deliberate pursuit of herd immunity would suggest.
The vaccines which have already been approved for use on an emergency basis (Pfizer and Moderna) appear to be safe and unusually effective both in preventing illness and transmission. Others are in the pipeline—the Johnson & Johnson vaccine, which is expected to be approved by late February, will have a lower degree of effectiveness on paper but will still protect very substantially against severe bouts of COVID-19, which is the principal consideration. This latter vaccine, moreover, unlike its predecessors, can be stored at room temperature and requires only a single dose, key factors in a mass vaccination effort. Barring an unexpected failure of distribution this would mean at least 100 million Americans could be vaccinated by late spring and enough vaccine for all Americans could be produced by the fall.
To be sure, the early rollout of the vaccine, predictably, has come in fits and starts, but this should improve as more sites are opened, new vaccines are approved, production is stepped up both for vaccines and their components, like vials and syringes, and President-elect Biden, once inaugurated, is able with his team (led by Jeff Zients, who helped reverse the initial failures of the “Obamacare” federal website) to coordinate the federal effort across departments.
Moreover, a higher number of infections increases the chances of mutations, one or more of which could elude currently or soon-to-be available vaccines and cause more fatalities, prolonging the damage caused by the virus.
Except in a few locales for short periods, the United States has never imposed “lockdowns” of the kind that existed in much of Southern Asia, Australia, and Europe. What we have had are, at most, lightly-enforced shelter-in-place orders, which have had only a modest impact on the virus—at best containment rather than control—while causing substantial disruption to many sectors of the economy.
However, there are countries and provinces in the world that have largely contained COVID-19 without ever resorting to true lockdowns but through a combination of personal responsibility, quarantines for travelers, and basic public health tracking and tracing. Japan is the most obvious example, but closer to home these include the state of Vermont and Canada’s Atlantic provinces, Nova Scotia in particular.
Despite a massive fall surge, Germany—despite having a much older population (ten years on average higher than the U.S., higher population density, and an earlier exposure to the virus)—still has fewer than half as many fatalities as the U.S. on a per capita basis.
The good news is that we know much more than during the first outbreak of the novel coronavirus in spring 2020 how to mitigate its spread while not shutting down the economy, assuming that businesses and individuals will follow safety protocols. For instance, allowing stores to open at 20 to 40 percent of regular capacity appears to be a sweet spot for maximizing sales while minimizing infections, especially for smaller businesses. We know that certain types of events contribute to maximum spread…family gatherings, any indoor gatherings in large numbers, bars, restaurants, and gyms.
Likewise, as many countries have demonstrated, it is possible to reopen schools and keep them open given ample resources and a careful plan: the obstacles in the U.S., while real, have more to do with politics and previously crumbling infrastructure than with the particular challenge of COVID-19. In general, while it is impossible to mitigate the impact of COVID-19 without harming the economy to some degree (and putting undue pressure on some sectors and certain people, especially low-income women), the tradeoff is much less stark than polemicists on all sides of the issue believe.
(I’ve written elsewhere about the misunderstanding of mask “mandates,” which are designed explicitly to nudge citizens in the direction of individual responsibility rather than to bring the heavy hand of the state to bear in earnest. Americans in fact have adapted more quickly to wearing masks than to many other public health measures most of which in fact pose a greater challenge to individual liberties, such as mandatory seat belts or smoking prohibitions, and generally with far fewer benefits. The impulse to resist the arbitrary authority of the state is admirable, but the downside of not wearing masks for the well-being of others, especially in crowded indoor spaces, is potentially enormous.)
Another key factor is that COVID-19, for many who have survived it, is an illness that is hard to shake. Thousands of “long haulers” continue to experience daily symptoms and the long-run consequences to cardiovascular and respiratory systems are unknown, though they are almost certain to be worse than those associated with respiratory diseases like the flu. While it is likely that reinfections are likely to yield milder bouts than earlier exposures, we don’t yet know this for certain.
Moreover, though COVID-19 is a disease that afflicts primarily the old, allowing huge numbers of infections inevitably means that more younger people succumb. While we wouldn’t put society on full pause to combat a new disease that killed annually some 25,000 Americans between 20 and 50 alone (the rough number of related deaths in that span of time from COVID-19, and probably now the single biggest cause of death in this age range), we would certainly hesitate to take measures that might accelerate this toll.
Sweden, while never embracing the “herd immunity” concept officially, took a different path from the remainder of Europe and most of the world. This took the form of not closing schools, businesses, and generally refraining from taking any nationwide, state-backed precautions. Buoyed by its demographics—a large percentage of younger citizens, many of them living alone—highly-educated, with a model safety net—Sweden chose, in effect, to sail a well-rigged and prepared ship into the heart of the storm.
Assessing the results is complicated, since the full measure of response can only be measured over many years and on many dimensions, including long-run psychological health, educational attainment, and long-term economic performance, but the short to medium-term results have been discouraging. Sweden has incurred roughly ten times the fatalities of its neighbors, has reverted back to a conventional strategy during the most recent, winter surge, and has an economic performance and outlook no better than its neighbors or its peers.
In the United States, South Dakota poses an interesting counterpart. Like Vermont, it is rural and far from global travel routes, and unlike the Green Mountain state South Dakota was removed from the initial U.S. outbreak of COVID-19. Nevertheless, Governor Kristi Noem has made a point of keeping the state open and South Dakota now improbably ranks among the state leaders in deaths per capita, some five times greater percentage-wise than Vermont which translates into some 1200 additional deaths. There is some evidence that case and death rates are beginning to decline sharply due presumably to the sheer number of infections.
Perhaps a bigger problem, however, than the fatalities themselves in South Dakota is that the state has been a seedbed of infection throughout the upper Midwest. A large motorcycle rally in Sturgis, for example, demonstrably caused a considerable number of follow-up infections throughout the region. And South Dakota also depended on its state neighbors to handle hospital cases that exceeded the state’s capacity. With these substantial caveats (which point out, in passing, the huge impact of mobility and an absence of mobility restrictions driving the COVID-19 crisis in the U.S.) it is hard to make the case for South Dakota as a model for the success of a local herd immunity strategy, though this has not prevented Governor Noem from taking an unwarranted victory lap.
Finally, that COVID-19 has primarily afflicted the elderly has surely made COVID-19 more bearable while it remains a human tragedy of immense proportions. In the herd immunity strategy, sometimes more or less directly, there is a strong echo of the Social Darwinist perspective that equates survival with strength and the death of the weaker as a natural phenomenon of life, even a cleansing.
Early in the pandemic, for example, President Trump reportedly asked his advisor, Dr. Anthony Fauci, “Why don’t we let this wash over the country?” In Northern California, the chairman of a local city planning commission in Contra Costa County compared COVID-19 to a forest fire that burns off “old trees, fallen brush, and scrub-scrub sucklings,” leaving behind a more balanced environment in its wake.
Response to the latter comment, predictably, was community outrage and the hapless commission member was dismissed from his position. It would have been better had he remained in office and to have had a real public conversation about the flaws of this Social Darwinist approach, namely that abandoning (or downplaying) the weak tends in practice to produce a slippery slope in which successive groups—and increasingly those who are not “weak” but politically disenfranchised or considered otherwise “undesirable”-- tend to be considered expendable, ostracized or much worse. No one ever expects to be on the wrong side of this divide, until they are.
Moreover, it is a strength rather than a weakness of a humane society to value individuals beyond their utilitarian value. Even our formal cost-benefit calculations tend to value years of life very generously, based on this impulse. Rescuing or prolonging a known life or closing a highway to determine a cause of an accident represent priorities that no utilitarian calculation could justify. Likewise, in choosing to “over-value” many older lives threatened by COVID-19 we are both tacitly affirming the bond between generations, and the sacrifices made for us by those who are now nearing the end of life, while making a strong statement about the value of an individual life independent of current circumstances.
1/31/2021 update. While most of the original points remain valid, the most significant development of the past several weeks, in addition to the change of federal administration, is the identification of further and potentially more infectious mutations, in particular the 1.1.1.7 that emerged in South Africa, and their discovery in different U.S. states. Because the U.S. does far less genomic sequencing tests than in many countries, just how prevalent these mutations are and how quickly they will dominate is unknown. We hope to reach a sweet spot in which the number of vaccinations plus acquired immunity (the silver lining, as it were, to the lackluster public health response that cost so many lives) prevents the rapid spread, including to many who may be reinfected by a new variant. On the plus side, the impending emergency authorization of the Johnson & Johnson vaccine, with others still in the pipeline, may tip the balance in favor of control. In any case the rollout has now reached a pace at which waving the white flag on mitigation measures would be self-defeating, both because it could encourage the onset of new mutations and because it could undercut the effect of vaccinations themselves, which become fully effective only after a considerable lag time of several weeks.
In India in late January, the Financial Times is reporting that some urban areas are seeing dramatically slowing infection and death rates presumably thanks to the local spread of naturally-induced herd immunity. However, as one doctor in Tamil Nadu quoted in the story notes, “We will never know how many people died—no one is counting.”