Recently I came across perhaps the most mind-bending chart about the pandemic I’d seen over three-plus years. Originally published two years ago in The British Medical Journal, it shows how Covid affected age-standardized mortality in England and Wales — a statistic that controls for demographic change in measuring death rates, so that a country doesn’t look as if it’s getting sicker just because it’s getting older.
Two things about the chart jumped out. First, at the onset of the pandemic in 2020, there was a dramatic spike in age-standardized mortality. For men, the increase was 14.6 percent, according to the Office for National Statistics; for women, 11.9 percent.
Second, though: In historical context, that jump did not appear all that large. It only brought age-standardized mortality to the level it had been in the year 2008, meaning that, correcting for age, the English and the Welsh were no more likely to die in 2020, in the midst of a once-in-a-lifetime global health crisis, than they were 12 years before, in what did not seem like a particularly deadly year at the time.
Novelty matters, and the sudden arrival of a masterfully infectious and deadly virus was justifiably alarming and galvanizing, adding a large amount of death on top of a mortality baseline we’d all wish was much lower. But progress in improving that baseline matters, too. And from that vantage, the mortality setbacks of 2020 looked smaller than the apparent gains of the previous 20 years — and not just in England and Wales. Across much of Western Europe and North America, even the horrible pandemic peaks only brought age-standardized death rates as high as they were in normal-seeming years around the turn of the millennium.
At first, this seemed hard for me to believe, given that more than 20 million have died worldwide from the disease, with the pandemic experience in every country quite far from normal. But while the two points may seem contradictory, ultimately, they reinforce each other: The better you appreciate the gains of the last several decades, the larger the disaster of Covid appears. Somewhat incredibly, the pattern applies to many pandemic-era phenomena beyond mortality, with Covid-19 playing the role of a statistical time machine, one that returned us not to the era of the Black Death or even the 1918 flu but to a much more recent and less alien-seeming time just a decade or two in the past. Which raises an implicit question: How do we come to terms with 20-year setbacks in a time of pretty rapid, if uneven, progress?
Let’s start with mortality. Across other nations in Europe, the rough pattern appears similar to the English one: an abrupt jump in 2020 mortality that nevertheless did not reach the age-standardized levels experienced around the year 2000. In the United States, the setback was larger, but only slightly. According to the Centers for Disease Control and Prevention, the rate here was around 882 per 100,000 in 1999, and, after falling to 715 per 100,000 in 2019, rose only to 835 per 100,000 in 2020 and 880 per 100,000 in 2021.
There are many ways of thinking about this fact, as is often the case with pandemic data. One is that, well, the country has indeed gotten older, and Covid mortality was vastly disproportionate among older people. Adjusting for that may suggest the landscape of American death in 2020 would have been “normal” in, say, 2000, but it doesn’t change the fact that 475,000 or more died that year who would not have died otherwise.
A second observation is that, adjusting for age, we experienced what seemed to be normal death rates in 2000 as essentially unthinkable in 2020. For many of us, there was simply no way to intuitively contextualize the mass mortality of the pandemic, with most Americans measuring its brutality not by the standards of the medium-distance past but against much more recent memories — and the expectations for the future we carried forward from those past few years. The result: a 20-year setback that felt more like a once-in-a-century, or even unprecedented, social rupture.
A third observation, related to the second: It took an astonishing reversal to reveal astonishing gains we’d come almost entirely to take for granted. Between the end of World War II and the onset of the pandemic, American-adjusted mortality rates had fallen by about half. Between just the years 1999 and 2019 — thanks to declines in smoking and pollution, more successful cancer therapies, better heart disease outcomes and other lifestyle factors — we had improved mortality outcomes by about one-fifth. These were the gains that the pandemic undid.
This is not just the story of age-standardized mortality. In fact, the time-machine framework applies to a variety of other impacts of the pandemic, many of which, when measured, show the same pattern: a setback of about two decades.
Take pandemic learning loss. On the gold-standard National Assessment of Educational Progress tests, 9-year-olds performed in 2022 as well as 9-year-olds had in the year 2004 for reading and math. Average scores for fourth graders fell to 2005 levels in reading and to 2000 levels in math.
It applies, too, to the pandemic-era crime wave. In 2020, the country’s murder rate increased by approximately 29 percent from the prior year, according to the Brennan Center for Justice, to 6.5 murders per 100,000 people. The spike was startling, the biggest jump in several decades and enough to generate a whole wave of commentary, campaigning and even policy retooling.
But the murder rate had been 6.3 per 100,000 Americans in 1998; in 1997, it had been 6.8 per 100,000. I don’t want to suggest that anyone should be happy to return, abruptly, to one of the most violent periods in modern American history. But the crime rates that seemed to send a shock through the country’s politics in the midst of the pandemic were also not unprecedented for most Americans old enough to drink.
Some commentators, mostly on the right, have lamented pandemic-era social disarray by highlighting homelessness in addition to crime, but there the setbacks were of an even smaller time scale: National rates of homelessness grew during the pandemic only to levels last seen in 2013. A similar pattern applies to rates of death by car accident, a much talked-about phenomenon of the early pandemic, which brought per capita vehicle deaths back to levels last seen in the United States in 2008.
Because some of these measures had been improving more rapidly than others before the pandemic, the setbacks are of different scale, as well. And the time-machine concept doesn’t apply uniformly to some other measures of social disarray. Overdose deaths continued their rapid growth through the pandemic, for instance, increasing by 50 percent between 2019 and 2021. Suicides slightly declined, for the first time in many years, and then began to rise again. The second year of the pandemic produced the country’s first baby boom in nearly a generation (though a small one). Unemployment first exploded spectacularly and then slowly but steadily declined, so that by this year, overall unemployment was at a 53-year low and Black unemployment at an all-time low. Inflation, too, exploded, though more than a year later, and has since been softly and steadily gliding downward. Thanks in part to pandemic policies, childhood poverty fell dramatically, and then, thanks to their end, more than doubled in a single year. Economic inequality actually shrunk for the first time in a generation.
But at the highest level many of the central narratives of the American pandemic experience conform to this basic framework: The experience was like a 10- or 20-year setback.
What should we make of all this? The pandemic was a brutal and destabilizing shock. Beyond the million-plus deaths and the millions of hospitalizations in the United States, many of the second- and third-order effects are still being processed by regular people, sorting through experiences as varied as social isolation and marriage dissolution and new jobs, and still being measured by social scientists. And a lot depends on what happens next — with mortality trends, crime and academic achievement — and whether these two-decade setbacks are passing or lasting. Traveling back in time for a year or two is one thing; rewinding the clock of progress more permanently is another.
In the meantime, even our incomplete data sets tell us some things — not just about the relative brutality of the pandemic itself, or the impact of our efforts to contain it, but also about how we mark progress and judge setbacks, tell stories and point fingers. That is one thing that makes it a world-historical event: We’re still too close to see it very clearly. Perhaps too close to see the year 2000 all that clearly, either.
David Wallace-Wells (@dwallacewells), a writer for Opinion and a columnist for The New York Times Magazine, is the author of “The Uninhabitable Earth.”