Our History of the Pandemic Is a Mess
We are entering the fourth year of the pandemic, believe it or not: Freshmen are now seniors, toddlers now kindergartners and medical students now doctors. We’ve completed two American election cycles and one World Cup cycle. Army volunteers are nearing the end of their active-duty commitment. It’s been a long haul but in other ways a short jump: Three years is not so much time that it should be hard to clearly remember what happened. And yet it seems to me, on many important points our conventional pandemic history is already quite smudged.
You could write columns about any number of misleading pandemic fables. (For my sins, I have: about America’s Covid-19 exceptionalism, about “red Covid,” about pandemic learning loss.) And some misunderstandings have been etched into our collective memory: over aerosol spread or the value of masking, ventilators and ivermectin (to name a few). But as time rolls on, the bigger point feels even more important to me. Though the fog-of-war phase of the pandemic is over, we are still struggling to see clearly many of its major features, captive instead to narrative formulations we’ve imposed on even messier realities, perhaps as a way of avoiding the harder questions they might raise.
Which do I mean? Below are a few examples that sketch that bigger phenomenon. This is not at all a comprehensive list, nor is it meant to be. But I hope it is an illustrative one, itemizing several ways in which huge swaths of the country see the experience of the past few years through prisms of anxiety and partisanship, self-justification and self-interest.
This is bad for future preparedness, of course. If we’re hoping to draw lessons from the past few years, it may be worth knowing that we might pay relatively more attention to the pandemic’s second year, for instance, and perhaps relatively less to its first. If we are trying to assess China’s “zero Covid” policy, we should have a clear picture of its vaccination failures rather than attributing the brutality of its current wave to decisions made three years ago. If we’re hoping to adjudicate what seems like a forever war between lockdowners and let-it-rippers, it probably helps to recall what first-year pandemic policy looked like — and how much of what we might remember as policy was really just pandemic.
It matters for present-tense level setting, too. If you’ve spent the past month worrying over pediatric hospital wards overwhelmed by the country’s tripledemic, you may have gone hunting for a narrative explanation — that masking and other pandemic restrictions had produced an immunity debt among children or that immune damage from Covid-19 itself had created worse outcomes across the population. But flu diagnoses have already peaked nationally — quite early, by historical standards, but no higher than in average seasons — and respiratory syncytial virus diagnoses have been falling for weeks. (And there were fewer pediatric deaths from flu so far this year than just before the pandemic.)
There is also a distressing historiographic lesson, which preoccupies me more. We need to learn from our failures if we hope to get future pandemics right, experts have warned for several years now. But policy questions aside, it doesn’t even seem to me we’re getting the history of this one right, though we just lived through it. You might think time would bring more clarity, but it seems that just as often, a more distant perspective allows misunderstandings to calcify.
First, the United States never had lockdowns. (Not like elsewhere in the world, at least.)
China sealed residents inside apartments in 2020; two years later it sealed workers inside factories. For much of the early pandemic, Peru permitted only one member of each household to leave the home one day each week for groceries or medical care. It wasn’t until this March that travelers to New Zealand could enter the country without first spending 10 full days locked in a hotel room.
In contrast, the United States had state-by-state shelter-in-place guidance that lasted, on average, a month or two, and that was not policed in a very draconian way. Roads were open without checkpoints, streets were free to walk, and stores that remained open were, well, open, for anyone to visit.
The disruptions were significant, of course. Many millions quickly lost their jobs, though much of that blow was softened by pandemic relief, and many public-facing businesses closed, as did schools and parts of hospitals. White-collar offices adopted work-from-home policies, large gatherings were canceled, and there were some accounts of people being ticketed in particular localities for gathering in parks or on beaches.
But in the global context, if anything, American restrictions were remarkably light. Consider a tool developed by the Blavatnik School of Government at Oxford, and published by the Financial Times,to compare the stringency of pandemic policy over time. For a brief period in March 2020, the United States appeared to have imposed restrictions roughly at the global average, with many nations stricter and many looser. But almost immediately, the rest of the world’s lockdown measures became stricter, while the United States’ remained the same. And by May, just two months after restrictions began, the United States was among the least strict places in the world. Mitigation policies were, of course, imposed here, but the U.S. response was not an outlying extreme then or at any point later in the pandemic.
So when Elon Musk, shortly before declaring that his pronouns were “Prosecute/Fauci,” shared a meme showing Dr. Anthony Fauci, the now-former head of the National Institute of Allergy and Infectious Diseases, whispering to President Biden, “One more lockdown, my king,” Musk may well have been giving voice to a widespread American frustration with the length of the pandemic. But it’s unclear what policy or even policy guidance he was referring to. Sure, there were long school closures in many places, as well as mask mandates or recommendations, widespread testing and, in some venues in some parts of the country, vaccine mandates, too. But in retrospect, to the extent that the country as a whole was ever governed by shelter-in-place orders, it was under the previous president, not this one, and they were lifted almost everywhere by early summer of 2020. (The director of the Centers for Disease Control and Prevention, Rochelle Walensky, even called masks “the scarlet letter” of the pandemic.) To call the mitigation measures of the past two years lockdowns is to equate any policy intrusion or reminder of ongoing spread with a curfew or stay-at-home-order — which is to say it is a striking form of American pandemic narcissism.
Most governors during the pandemic seemed to benefit politically.
The year 2020 was one of pandemic lionization. By that April, the average approval rating for American governors was 64 percent. The following election season featured a couple of high-profile races that have shaped the narrative about pandemic politics and the costs of mitigation, with Glenn Youngkin, a Republican, defeating the Democrat, Terry McAuliffe, for the Virginia governorship in part by channeling public frustration with Covid restrictions, and New Jersey’s governor, Phil Murphy, a Democrat, barely hanging on against a little-known Republican challenger yelling about lockdowns. But a report from the Brookings Institution suggested that of the 10 governors with the biggest popularity declines from mid-2020 to mid-2021, eight were Republicans. (The other two were Democrats in red states.)
And by this November, the political fallout seems to have very clearly settled down, at least at the state level. Ron DeSantis, a Republican, famously won re-election in Florida in part by campaigning against Covid mitigations. But Democrats J.B. Pritzker in Illinois, Gretchen Whitmer in Michigan and Tony Evers in Wisconsin, won, too, each having deployed aggressive statewide mitigation efforts and each winning larger shares of the vote than they secured in their previous races. In Colorado, Gov. Jared Polis positioned himself as a reopening Democrat and won, and in Ohio, Gov. Mike DeWine positioned himself as a cautious, Covid-conscious Republican and won, too. In fact, in only one state, Nevada, did an incumbent governor lose re-election in 2022 — and that race pitted the incumbent, a Democrat who didn’t win a majority in his previous race, against a Republican challenger who didn’t win a majority in this one.
There are a number of possible ways to read these results, including that the pandemic simply retreated as an issue over time. But it is hard to look at a slate of 36 elections in which only four governorships changed party hands and conclude that pandemic backlash remains a dominant force in electoral politics.
The most consequential year of the pandemic in the United States was probably not 2020 but 2021.
Works of serious retrospective history lag works of journalism, inevitably, but one risk of real-time reporting is that we never get around to reckoning with turbulent events with anything like proper hindsight. Instead we are left with accounts focused almost exclusively on the story’s first act. That is where we are now: The list of books devoted to the American pandemic response in 2020 is quite long, and the list of books — or even authoritative long-form reporting — devoted to the following years is minuscule.
This is especially problematic because — judging both by total mortality and by America’s relative performance against its peers — 2021 was far more telling in its failures. In the first year of the pandemic, the United States performed somewhat worse than some of its peers in the wealthy West but not that much worse. We failed to stop the virus at the border, but so did most other countries in the world, and by the end of 2020, the country’s Covid-19 per capita death toll was near the European Union average. The country spent that first year obsessing over mitigation measures and the partisan gaps that governed them: school closures and indoor dining, mask wearing and social distancing. But it was in the pandemic’s second year, in which mortality was defined much less by mitigation policies than by vaccination uptake, that the country really faltered.
Mass vaccination, though miraculously effective, didn’t usher in a lower overall death toll.
To judge by cumulative deaths, the midpoint of the American pandemic so far is April 2021, when 550,000 Americans had died and more than 100 million Americans had been fully vaccinated. We’ve had more deaths since the end of the initial Omicron surge, this past winter, than the country had experienced by late May 2020, when The New York Times proclaimed the death toll of 100,000 “an incalculable loss.”
This is not because vaccines don’t work, of course. But especially with the initial Omicron wave, infections became so widespread that they effectively canceled out the population-scale impact of vaccination . If you get a vaccine that cuts your risk of dying from Covid by 90 percent, for instance, but infections grow five times as common, you are only twice as safe as you were before — and the same math applies to the country as a whole.
Of course, without vaccination, current infection rates would have produced a much higher toll. But overall, though the death rate has decreased, year over year it hasn’t decreased all that dramatically. There were about 350,000 Covid deaths in 2020, about 475,000 in 2021 and about 265,000 in 2022.
One word for this pattern is “normalization,” and it is undeniably the case that as a whole, the country is less disturbed by those last 265,000 deaths than it was by the first 350,000. But we did quite a lot to keep the toll as low as 350,000 in that first year and have chosen to do successively less in the year of vaccines and then the year of Omicron that followed. We have effectively recalibrated our mitigation measures roughly around the mortality level of 2020 — as though that death toll was not an anomaly but a target.
Barring a major new variant, 2023 should be less brutal. But to this point, even widespread vaccination (two-thirds of the country as a whole and over 90 percent of American seniors) hasn’t been enough to dramatically change the trajectory of pandemic death in this country. And if we are building our understanding of social risk simply from the infection-fatality rate, which tells us the risk of death given an infection, we’re missing half of the critical information — how likely that infection is to begin with.
China’s vaccines are probably not much worse than ours; it just did a poorer job vaccinating the elderly.
Especially as “zero Covid” protests began in China this fall, Western commentators emphasized that the Chinese vaccines offered considerably less protection than the mRNA versions developed in and preferred by countries like the United States. These days, it’s much harder to measure vaccine effectiveness, in part because of growing immunity from vaccine doses and infections.
Most of our best data shows that, especially after one dose but also after two, the mRNA vaccines do more to protect against severe hospitalization and death than do the Sinovac and Sinopharm varieties developed and manufactured in China. But most Americans who are up-to-date with vaccinations are already past three shots to four. And after three doses, the difference may be quite negligible, with some studies showing only a somewhat modest mRNA advantage. According to one high-profile study published in The Lancet: Infectious Disease, among the most vulnerable — those over 80 — three doses of the Chinese vaccines may offer slightly better protection.
But an alarmingly high number of China’s oldest citizens, perhaps one-third, have not been vaccinated. This means the relative share of China’s older population that remains entirely unprotected is as much as six times as large as that of the United States, and of course, in absolute numbers, the vulnerability is even larger. Which makes that vaccine gap, though quite significant, less a matter of science and technology than of political and social factors — chiefly the matter of why China has done so poorly to protect its most vulnerable citizens.
Many hypotheses have been offered to explain this shortcoming, from worries about side effects to troubling history with past vaccination campaigns and confidence that “zero Covid” would eliminate disease spread in perpetuity. But among the less-talked-about possibilities is that the vaccination program may have been designed less to save lives by protecting the most vulnerable than to preserve the work force by focusing on the young and middle-aged. In theory, this could also explain what seems to outsiders like a whiplashing policy reversal, from “zero Covid” to zero surveillance. Even limited testing and mitigation measures, which would slow the spread of the disease, could cause more economic disruption than was considered acceptable (or medically necessary, given the age of the work force).
The world’s worst pandemic was probably not in the United States or Britain, Italy or Spain, China or India but in Eastern Europe — notably in Russia.
Because medical record keeping varies so much from country to country, official Covid death tolls are a misleading measure of pandemic impact. In wealthy countries, where more testing has been done and causes of death are recorded somewhat more systematically, the numbers appear relatively higher, and in poorer countries, with less testing and somewhat less scrupulous death certificates, they are lower.
Excess mortality statistics tell a more reliable story, though because they essentially compare total deaths against recent historical averages for a country, they rely on statistical modeling and the availability of older data. The Economist maintains the best running excess mortality database, and the story it tells about the global toll of the pandemic is very clear. Of the 106 countries included in its data set, the 12 hardest hit were in Eastern Europe, as were 17 of the worst 20. Many of these are small countries; the Economist estimates the two most brutal pandemics in the world were in Serbia and Bulgaria, each with populations under seven million. The third-worst pandemic was in Russia, where there were more than one million excess deaths in a population of more than 140 million, an excess per capita death toll two and a half times as heavy as the American one. (Interesting time to launch a war of choice.)
Long Covid is definitely real, but it’s also becoming less common.
In 2020 the United States treated reports of long Covid almost as a ghost story — anecdotes at the spooky margins of our collective nightmare and ones we didn’t know how much to trust. Three years later, thanks in part to the tireless work of patients and advocates, the phenomenon is a much more central part of the pandemic story told by public health officials, politicians and the media. But just as we’ve grown slowly to accept long Covid, it is also becoming less and less common. Growing research shows that risks are declining. Vaccination and previous infection, though imperfect, appear to reduce vulnerability for long-term consequences, and the severity of early cases of long Covid, like the severity of early cases of acute Covid, appears to reflect the immunological naïveté of the population as a whole, which has been steadily declining ever since.
We’ve moved past interventions like masks as a country, but that doesn’t mean the Great Barrington Declaration advocates were right.
Arguments against pandemic restrictions were made almost as soon as the first schools and offices were closed, typically by conservatives (though many liberals came around to the cause when vaccines arrived). But the case for relaxing restrictions was made most famously in a 2020 document called the Great Barrington Declaration. Written chiefly by Dr. Jay Bhattacharya of Stanford, Sunetra Gupta of Oxford and Martin Kulldorff of Harvard, it proposed that pandemic policy was doing more harm than good, that most people should live normal lives to build up immunity through infections and that the most vulnerable members of society could be protected in much more targeted ways than the one-size-fits-all approach that had been deployed to that point.
It was a bundle of scientific claims and policy proposals, in other words, which itself is telling. Today you might be inclined to think about the question of mitigation simply at the level of policy, asking what restrictions were necessary or helpful, given a shared base of knowledge about Covid-19. But the debates early on were not just debates over policy trade-offs. They also concerned basic science. And on many of those critical points, those pushing against mitigation measures were wrong.
Dr. Bhattacharya, for instance, proclaimed in The Wall Street Journal in March 2020 that Covid-19 was only one-tenth as deadly as the flu. In January 2021 he wrote an opinion essay for the Indian publication The Print suggesting that the majority of the country had acquired natural immunity from infection already and warning that a mass vaccination program would do more harm than good for people already infected. Shortly thereafter, the country’s brutal Delta wave killed perhaps several million Indians. In May 2020, Dr. Gupta suggested that the virus might kill around five in 10,000 people it infected, when the true figure in a naïvepopulation was about one in 100 or 200, and that Covid was “on its way out” in Britain. At that point, it had killed about 45,000 Britons, and it would go on to kill about 170,000 more. The following month, Dr. Bhattacharya and Dr. Kulldorff together made the same point about the disease in the United States — that the pandemic was “on its way out” — on a day when the American death toll was approaching 600,000. Today it is 1.1 million and growing.
This is not to say that these voices should have been silenced or driven from public debate. Some questions they raised were important matters of ongoing contestation, especially in the pandemic’s earliest days. As should be obvious three years in, pandemic policy did involve unmistakable trade-offs; the large, ongoing mortality under Mr. Biden is one reminder that mitigation was never as simple as just hitting the science button. But making arguments about those trade-offs using bad data or inaccurate timelines distorts the picture of the trade-off, of course. And to treat these arguments as merely political debates is to forget how much of the argument for reopening was based on bad science — and how much harder it would have been, at the time, to persuade many people using what turned out to be accurate data.
As for the policy advice of the Great Barrington Declaration? The economist Tyler Cowen recently revisited the case for focused protection — the idea, emphasized in the declaration, that the most vulnerable members of society could have been shielded more aggressively while life continued mostly as normal for everyone else. (A study by his colleague Alex Tabarrok suggested this policy would have been hard to achieve, given that the death rates in the country’s best-resourced and best-run nursing homes were not better than the rates experienced in much more negligent settings. Mr. Tabarrok estimates there were larger missed opportunities in vaccinating nursing homes more quickly.)
Mr. Cowen argued that actions that would have genuinely qualified, in retrospect, as protecting the vulnerable would have included preparing hospitals for patients in January 2020, accelerating vaccine rollout and uptake, and pushing for development of new treatments and promoting widespread testing. “If you were not out promoting those ideas, but instead talked about ‘protecting the vulnerable’ in a highly abstract manner, you were not doing much to protect the vulnerable,” he wrote.
“Publishing papers suggesting a very, very low Covid-19 mortality rate, and then sticking with those results in media appearances after said results appeared extremely unlikely to be true,” he added, “endangered the vulnerable rather than protecting them.”
The great success of the pandemic was Operation Warp Speed, but we’re learning the wrong lessons from it, emphasizing deregulation rather than public funding and demand.
The rush to develop, produce and deliver vaccines is the signal American achievement of the pandemic — so consequential, it is a pretty persuasive rebuttal to anyone decrying the country’s failure to stem the pandemic or pinning that failure on some narrative of national disarray. The vaccines were designed in just days, produced in just months and delivered within a year of the country’s first confirmed case, saving at least many hundreds of thousands of American lives and probably many millions globally.
But in the public narrative of the pandemic, Operation Warp Speed plays a remarkably small role, likely because of the partisan complications. The accelerated development was overseen by Donald Trump and shepherded by Jared Kushner, so even very pro-vaccine liberals are not all that likely to credit the program. But liberals embracing the vaccines have made it somewhat harder for conservatives to claim it as a policy victory. (One wonders how differently this dynamic might have played out if the vaccines had been approved before the 2020 elections, as was originally expected.)
In the public square, then, the job of celebrating the success of Warp Speed has fallen to a somewhat motley alliance of progress-minded technocrats, making the argument that reviving and extending the program may well be the most important public health imperative to emerge from the pandemic. And last summer the White House began an initiative to try to recreate the program’s success — announcing another Operation Warp Speed to develop new vaccines and treatments that could protect the country against future waves of the virus.
But the immediate aftermath of that announcement is telling, with the project sputtering without real funding and no new vaccines or treatments available and few being developed. The White House team had done what it could to learn a certain set of lessons from Operation Warp Speed, including coordinating the development of promising vaccine candidates and accelerating the timelines of clinical trials. But it hasn’t secured money to support the project, nor did it give any concrete reason to believe that there would be significant demand for the new drugs when, if ever, they came online. (The declining American interest in Covid booster shots seems to suggest that demand could be very small.)
On balance, then, we are seeing a test play out in real time. How much additional innovation can be unlocked simply through cutting red tape, and how much requires something more? That is: guaranteed money or guaranteed demand or both. And while it’s certainly true that bureaucratic streamlining played a role in the rapid development of vaccines, it seems to me that the giant size of the market was almost certainly a more important driver — billions of people here and abroad desperate for vaccine protection and deliverance from the pandemic and a world of governments willing to cover the full cost of the shots and their distribution.
It is worth remembering the supply-side lessons of Operation Warp Speed — that public-private enterprise can be streamlined and that legacy regulations may well slow new drug innovation and production (with tragic consequences). But let’s not forget the demand side or what that tells us about future R. and D.: While bureaucracy may well slow development and rollout, removing those obstacles is not nearly as productive as conjuring up a market. In the absence of a new pandemic, it may be that government guarantees are the only tool that might create comparable ones.
How surprising is all this? Early in the pandemic, we were treated to a raft of meditations on the 1918 flu epidemic, each invariably mentioning how little tribute was paid in the years that followed, despite a global death toll in the hundreds of millions, many times larger than the world war it punctuated.
That does not seem all that likely to be our fate this time. Much of the country is happy to move on, of course. But people on both sides of the pandemic aisle seem still invested in prosecuting arguments about mismanagement, so it is hard to imagine the death and disruption of the past few years losing political and social salience anytime soon.
But salience is not the same thing as lucidity, and in the years ahead, as the world begins revising its histories of the pandemic, as it always does in the aftermath of great disruption and trauma, we may find ourselves polishing these simplistic just-so stories into talismans so smooth, they’ve lost all shape.
Perhaps this is inevitable. And yet I’m surprised by it. The country has just passed through the most brutally tumultuous experience in at least a generation, in which more than one million Americans died and everyone else’s lives were dramatically disrupted. The whole time, the shape and near future of the pandemic seemed of absolutely central cultural interest and paramount importance, a top-shelf preoccupation of the news media and a running conversation subject on social channels. Three years ago, that sort of experience might have seemed to be too large for anyone to misperceive. Perhaps that was pandemic narcissism, too.
David Wallace-Wells (@dwallacewells), a writer for Opinion and a columnist for The New York Times Magazine, is the author of “The Uninhabitable Earth.”