People are not really afraid of coronavirus, and for good reason: The odds of dying of COVID if you catch it are no higher than your normal odds of dying within a year of something else.
by Will Boisvert
In my last piece on COVID-19 I waaay underestimated the death toll it would have. It has now passed 2,529 deaths per million in the U. S. (as of December 29, 2021), four times more than the 674 dpm I anticipated based on the 1957-1958 Asian Flu pandemic. In some locales—New York City, Peru, Hungary—the per capita mortality might get close to the level reached in the 1918-1919 Spanish Flu pandemic, history’s deadliest (or at least it might have without the vaccines). So, in light of the massively greater mortality, have I changed my mind and accepted lockdowns as a necessary response to the virus?
Well, no. Bad as it has been, the pandemic has not been bad enough to justify the harsh lockdowns. That’s not just what I think, it’s what everyone tacitly thinks—except for the public health authorities who have decreed the lockdowns.
That’s because COVID isn’t scary unless you look at it from a very particular perspective. That’s the perspective of medical elites who see COVID as a tidal wave that’s clearly visible in epidemiological models, testing results, statistical aggregates and the pressure gauge of ICU occupancy. Looking down from on high through these statistical lenses, they see the big picture of death on a scale that’s vast but imperceptible to laymen on the ground. But the view from on high isn’t the only valid perspective, and it may seriously conflict with our broader understanding of human well-being.
***
Imagine that you live in a village of a thousand people, and one day the state health commissioner comes to a town meeting to sound a warning about a viral pandemic that’s on the way.
He says:
“This is the worst pandemic in a century. The death toll will be catastrophic unless you take drastic measures to prevent infections. You’ll have to work from home unless your job is absolutely essential. Everyone will have to stay at home except for essential tasks, and non-essential businesses will have to close for months on end depending on infection rates. Everyone will have to wear masks everywhere. You’ll have to close schools for a year or longer, ban gatherings in homes, ban church services and cancel Thanksgiving and Christmas.”
Then you say:
“Jeez, doc, that’s terrifying! How many deaths should we expect from this virus?”
And he replies:
“It will be a nightmare! You’ve got about a thousand people in town, and if you don’t impose these restrictions then you’re looking at two, maybe three deaths over the next year or so, probably among very old and sick people because the virus rarely kills kids and younger adults. In a normal year you would see eleven or twelve old, sick people dying in town of natural causes, but this year you could see a total of thirteen or fourteen old, sick people dying after you add in pandemic deaths. That’s how bad it could get if you don’t lock this village down.”
Then you might reply:
“Uh, doc, that doesn’t sound very bad. I don’t think it justifies a lockdown.”
A lot of people at the meeting would agree with you. Parents wouldn’t want to close school for a year, businessmen wouldn’t want to close their shops, young people wouldn’t want to shut down socializing and night life, nobody would want to wear masks. Even old people at highest risk might not want to forego family gatherings or cancel Thanksgiving and Christmas, knowing that, at their age, they may not have many more to celebrate.
So there’s a disconnect between the epidemiologists’ aggregate statistical view of COVID and the perspective of individual risk that a layman has on the virus. In the first view the pandemic looks terrifying; in the second it looks banal.
If we broaden our gaze from a small village to the whole country, the risks still look modest. Consider the decline in life expectancy caused by the pandemic and lockdown measures in the United States. The CDC estimates that life expectancy at birth fell in 2020 to 77 years, down 1.8 years from 2019’s 78.8 years. The media trumpeted the drop—the biggest since World War II!—but let’s put the risk perspective: the 1.8-year drop in life expectancy brought us back to the bad old days of 2003. So to reframe the question, should we undertake harsh lockdowns, stay-at-home orders, masking and distancing just to avoid the mortality risk that we blithely accepted in 2003, when we had the same death rates that we did in 2020? Most of us remember 2003, and if we are honest we would have to say no, since back then we didn’t consider implementing any of the much milder public health measures, like banning cigarettes and alcohol, that could have permanently raised life expectancies by 1.8 years and then some.
In fact, that headline life-expectancy drop hugely overestimates the increased risk of death, because the calculation assumed that death rates will stay the same for the entire lifetime of the 2020 birth cohort, which they will not once COVID runs its course. If the COVID pandemic ends up killing one million Americans before it subsides, the average drop in life expectancy will be all of eleven days. That’s a tiny risk, and no one would embrace enormous lifestyle changes to avoid it—except for epidemiologists.
Thus the fundamental conundrum of the pandemic—the clash between the view from on high of the public health authorities who urge drastic measures to avoid massive death tolls, and the ground-level view that can’t discern any scary or even significant risk from COVID. And thus the ongoing political conflict, with health officials, and the politicians and media figures who defer to them, berating a recalcitrant public that resists lockdowns.
To be sure, most of the population supports restrictions in polls, and doesn’t actively resist them in practice. Mostly, they follow the orders of government officials.
But the public still has a visceral sense of ease over COVID that manifests in many ways. Harsh lockdowns and mandates have provoked major grass-roots protests in the United States and Europe, while lax measures in places like Sweden, Florida and South Dakota have not. The great U. S. COVID referendum of 2020 saw nearly half the electorate opting for COVID-complacent Donald Trump. Politicians with a wary eye on public discontent are quick to ditch restrictions during lulls, despite Fauci’s hand-wringing. And then there’s the sea of people who tacitly resist COVID restriction with rules violations and secret partying. That tacit resistance occurs across the political, class and intellectual spectrums, from Sturgis bikers to cheek-by-jowl BLM protesters to the unending parade of politicians and health officials caught socializing without masks or distancing in violation of the orders they impose on everyone else.
All these signs point to a basic truth: no matter how much public health experts try to make COVID frightening, we’re just not scared of it.
You can see the logic behind both warring perspectives. The state health commissioner is looking not just at one village of a thousand people but at many such agglomerations whose tiny numbers of deaths add up to thousands across the state. His eye is on county hospitals where imperceptible dribbles of cases from small villages get funneled into a raging torrent that swamps ICUs. He’s mesmerized by the mountain range erupting in the middle of the aggregate mortality stats. Most of all, he’s a doctor, and his consuming priority is to prevent deaths, with all other considerations secondary. That view from the top has become the dominant perspective, drummed in by a media that fixates on public-health pronouncements, the spiraling aggregate case-o-meter, and tearful war stories from frontline nurses.
But the perspective of ordinary people on the ground also has a logic. The picture of COVID in their vicinity is far from terrifying, and they reflexively balance the risk with other life priorities, just as we do with other low-level mortality risks whenever we light up a cigarette or get in a car.
And by the numbers, that nonchalant ground-level view of COVID is pretty rational. At any age, the risk of dying from COVID after you catch it is no higher than the ordinary risk of dying that we live with all the time.
To see this, look at the table below, comparing the age-specific infection-fatality rates from COVID with ordinary yearly mortality risk at comparable ages, from the 2019 U. S. Social Security period life table.
| Age Band of COVID infections | COVID IFR, % | Age in SSA period life table | Odds of dying within one year without COVID, men, % | Odds of dying within one year without COVID, women, % |
| 0-4 years | 0.00 | 2 years | 0.0260 | 0.0212 |
| 5-9 | 0.01 | 7 | 0.0127 | 0.0103 |
| 10-14 | 0.01 | 12 | 0.0145 | 0.0116 |
| 15-19 | 0.02 | 17 | 0.0676 | 0.0278 |
| 20-24 | 0.03 | 22 | 0.1327 | 0.0491 |
| 25-29 | 0.04 | 27 | 0.1612 | 0.0642 |
| 30-34 | 0.06 | 32 | 0.1926 | 0.0933 |
| 35-39 | 0.10 | 37 | 0.2318 | 0.1236 |
| 40-44 | 0.16 | 42 | 0.2828 | 0.1596 |
| 45-49 | 0.24 | 47 | 0.3831 | 0.2332 |
| 50-54 | 0.38 | 52 | 0.5808 | 0.3529 |
| 55-59 | 0.60 | 57 | 0.9047 | 0.5425 |
| 60-64 | 0.94 | 62 | 1.3248 | 0.8006 |
| 65-69 | 1.47 | 67 | 1.8280 | 1.1158 |
| 70-74 | 2.31 | 72 | 2.6249 | 1.7892 |
| 75-79 | 3.61 | 77 | 4.2159 | 2.9837 |
| 80-84 | 5.66 | 82 | 6.9190 | 5.1658 |
| 85-89 | 8.86 | 87 | 11.7771 | 9.1578 |
| 90 + | 17.37 | 92 | 19.7287 | 15.9491 |
(COVID infection-fatality rates from NF Brazeau, R Verity, S Jenks et al. “COVID-19 Infection Fatality Ratio: Estimates from Seroprevalence.” Imperial College London (29-10-2020), doi https://doi.org/10.25561/83545. Annual odds of dying from the U. S. Social Security period life table, 2019, https://www.ssa.gov/oact/STATS/table4c6.html
The annual odds of dying is a good benchmark of risk perception: if you learn that you have a 75 percent chance of dying from cancer within a year, that will definitely concentrate your mind on your mortality. The table above shows that, at all ages, if you catch COVID you’re no more likely to die of it than you are to die of something else within a year.
(You could argue that your total risk of dying within a year roughly doubles during a COVID infection until you recover, as 99 percent of people do. But that doubling is an overestimate because people who die of COVID are less healthy than average, so their annual odds of dying are larger than those of the general-population SSA cohorts in the table. Also, the COVID infection-fatality rates are from 2020; new treatment protocols and drugs have lowered the infection fatality rates since then.)
Thus, COVID mortality risks sits squarely within a familiar range of other mortality risks that we live with as a matter of course. If you’re not losing sleep over your odds of dying within the year, you probably won’t lose sleep over your odds of dying of COVID. That’s true even of the very old: their risk of dying from a COVID infection is quite large, but no larger than the risk that something else will do them in within a year.
People don’t explicitly realize how small and unexceptional a risk COVID poses because they overestimate infection-fatality rates. One survey of risk perception among Americans put the median IFR guesstimate among the public at 5 percent, and the average at 15 percent—many times higher than the actual IFR of 1 percent or so established by serology studies. The KEKSTCNC COVID-19 Opinion Tracker survey found that people in Britain estimated the percentage of the population killed by COVID at 7 percent, while Americans put it at 9 percent, two orders of magnitude higher than the real death toll. Opponents of mandates, on the other hand, harp on the low IFRs. Of the many divisions among the public over COVID policy, one of the biggest may simply be that mandate skeptics have a lower, more accurate assessment of the risk the virus poses.
But most people don’t use statistics to form their intuitive sense of the danger posed by the virus; instead, they rely on cognitive heuristics. While we hear news reports of thousands of people dying of COVID, most of us don’t personally see anyone dying of it. Three of my family members have recovered from COVID (or what they assumed was COVID), but I don’t know anyone who has died of it. The same is true for the large majority of Americans. No matter our estimation of risk in the abstract, our visceral sense is that the risk is very low, and for good reason.
That’s why so many people of all ideological stripes, including lockdown zealots, are personally so relaxed about the risks of COVID and continued socializing and protesting and refusing to vaccinate. They wouldn’t do that if the pandemic were frightening. The Black Death killed off one third to one half of affected populations, and 19th-century cholera epidemics sometimes killed five percent or more of a city’s inhabitants. Those pandemics were terrifying because everyone knew people who had died, including young people, and had a keen sense of personal danger. If COVID were that dangerous, there would be no vaccine resistance.
But COVID isn’t anywhere near that dangerous. Neither was the 1918 Spanish Flu, often ranked as history’s worst pandemic. It was, by the epidemiologists’ aggregate death toll, but the risk it posed—it killed about 0.7 percent of the U. S. population—was far from terrifying. That’s why it made very little impression on society. Public health measures against the flu were haphazard and contested, and it was remembered mainly as a footnote to World War I even though it killed ten times as many Americans as the war did.
COVID is less lethal than the Spanish Flu, but it’s had a much bigger impact. That’s because the epidemiological perspective has grown more comprehensive and much more influential over the intervening century. Epidemiologists can see COVID better because they have tests and global surveillance networks to spot and track it. We now have ICUs and nursing homes to concentrate desperate cases and deaths and make them conspicuous. The media are better coordinated in their coverage of the illness and their dissemination of epidemiological orthodoxy. And the state has more capacity now to intervene in the economy, regiment society and confine the population than it did in 1918—a capacity that epidemiologists commandeered in the public-health revolution of March, 2020.
You could view that revolution as a scientific triumph, a milestone in progress towards a more rational, better organized, healthier society. But you could also see the new ascendancy of the epidemiological perspective as an extreme ideological distortion of science, one that has foisted on us a level of hysteria and social restrictions comparable to what prevailed during the Black Death—for a disease risk that’s one hundred times less serious.
The flip side of the distorted epidemiological perception of risk is a blindness to costs. Public-health authorities imposed lockdowns with little thought for their negative health effects.
We don’t know yet the full extent of those effects. The most obvious was a jump in drug overdoses in the United States of about 21,000 in 2020, or 30 percent over 2019’s toll. Those 21,000 extra overdoses are just a small fraction of 2020’s 385,000 COVID deaths, but their relative impact in years of life lost is substantial. Studies put the average years of life lost to a COVID death at roughly ten years (probably an overestimate), while the average years of life lost to a drug overdose—which typically occurs among younger adults—is about 40 years. So 2020’s rise in overdoses represents about 840,000 years of life lost, or about 21 percent of the 3.9 million years of life lost to COVID in 2020. Then there are traffic fatalities, which rose in 2020 by 7.2 percent over 2019’s toll, reversing a downward trend despite a sharp drop in vehicle-miles traveled; experts attribute the uptick to emptier streets tempting people to drive more recklessly. Since traffic deaths also cause an average loss of about 40 years of life, the extra 2,584 deaths in 2020 add up to another 100,000 years of life lost. Cigarette and alcohol consumption also rose substantially in 2020, reversing downward trends, which means more deaths from cancer, cirrhosis, stroke, heart attack, lung disease and car accidents in years to come.
But there are other costs besides sickness and death. COVID lockdowns derailed the life projects of millions of Americans, disrupting educations and careers, bankrupting businesses and shutting down social lives. Those costs tend to get dismissed because no one is dying of them—at least not yet—and epidemiologists don’t reckon costs that don’t show up in mortality or morbidity stats.
But loss of life isn’t entirely incommensurable with the loss of life chances, and people can and do balance those losses against each other. Let’s try another thought experiment. Suppose you’re a fifty-year-old restaurant owner, and you’re given a choice: 1) see your restaurant go bankrupt now because of a COVID lockdown, and have to scramble to find a new career and try to rebuild your finances just when you’re starting to put your kids through college; or 2) keep the restaurant but catch COVID at the age of 82 and face a 6 percent chance that your life will be cut short by a few years. Which would you choose? Many people, maybe most people, would choose the second. If you’re 21 years old, would you choose 1) to spend two years out of school, stuck on unemployment or working a dead-end “essential” job, unable to go on dates because bars and restaurants and parties are shut down; or 2) have a normal youth but catch COVID when you’re 72 and face a 2.3 percent chance of dying from it? Again, a lot of people would take the COVID risk over the lockdown.
It makes sense that people would want to take such risks, because building a good life is worth risking a small chance of dying. Financial security is profoundly important to middle-aged people, and losing it can darken their lives for decades and weigh more heavily than the loss of a few years at the very end. Starting a meaningful career is profoundly important to people in their twenties; they don’t have infinite opportunities, and finding the right ones can shape their whole lives. And for young people socializing isn’t selfish frivolity, it’s serious business; they have to party and date if they want to find spouses and start families.
Now suppose that you’re 87 years old and you have a choice: 1) be cut off from your family for a year because of a lockdown; or 2) live the way you’ve always lived, risk catching COVID, and run a 9 percent chance of dying within a month or two if you do catch it. Many people would choose the latter, because it doesn’t feel all that different from their 10 percent chance of dying within a year anyway, and because they don’t have time to make up for missed opportunities to live a little. A lot of elderly people would—and did—want to gather with family at Thanksgiving, maybe their last, despite the risk of COVID. They wanted to enjoy one of the few consolations of old age: to see that a son’s business is thriving; that a granddaughter has a new boyfriend; that what they started in life will go on after they die.
That impulse to discount the risk of COVID in order to go on living was not uncommon among the elderly, and not just among fat old Republican politicians and white-bearded Sturgis bikers, but among ordinary people too. Epidemiologists inveighed against the risks they were running, but by their standards—quite rational standards—they were being reasonable. The simple cost-benefit calculation that people were making was invisible from the epidemiological perspective.
It was also dismissed on the grounds of collective responsibility, which is the moral core of the epidemiological perspective: we must lock down not only to protect ourselves but to respect the right of others not to have their lives put at risk by viruses we spread. It’s a powerful argument, and one that was hard to gainsay during much of the pandemic.
But it’s obsolete. We now have good drug treatments for COVID, and vaccines that are still quite effective at preventing hospitalization and death, although their effectiveness against initial infection has waned. Omicron is much less virulent than previous variants and much more transmissible, following the usual evolutionary trajectory of respiratory viruses towards a state where it’s endemic but innocuous. Omicron is also readily spread to and by the fully vaccinated, which means that existing vaccines don’t confer much collective protection against catching the virus, only individual protection against severe disease. With Omicron now the dominant version of COVID, the case for collective action against the virus has evaporated.
Unfortunately, the epidemiological demand for collective responsibility—and collective punishments—has only grown harsher as the rationale for it has collapsed. Mask mandates persist, and unprecedented new vaccine mandates have set up a medical apartheid state by throwing millions out of work and banning them from ordinary business and social activities. The epidemiological perspective has grown increasingly authoritarian even as it has departed radically from the science.
There’s no remedy for the flaws in the epidemiological perspective, nor should there be: it’s an essential viewpoint, and its insights and recommendations should be taken seriously. But it can’t be allowed to dictate policy, because it’s inevitably a partial and blinkered viewpoint. Looking down from on high, it can only see a collectivity; it’s incapable of negotiating or even perceiving the complex trade-offs individual people face in balancing the risk of disease with the other risks and tasks they contend with. Only ordinary people, freely navigating conditions on the ground, can do that.