The Once-in-a-Century Impact of Covid-19 Deaths
The Once-in-a-Century Impact of Covid-19 Deaths
The Covid-19 pandemic has been traumatic for New York City. How traumatic, by historical standards? Well, city health authorities started keeping more or less consistent track of deaths in 1804. Every year, the city Department of Health and Mental Hygiene publishes a remarkable chart on the cover of its vital statistics report of mortality rates since then, labeled “The Conquest of Pestilence in New York City.” Although the full report for 2020 won’t be ready for a couple of years, there’s enough data available now to estimate how Covid-19 might rank among the pestilences that have beset the city through the centuries.
OK, so it’s not cholera. But it’s clearly the worst thing to hit New York City in a long, long time — and that’s assuming the city’s epidemic is more or less over. I estimated 2020 numbers by taking total deaths in the city through Aug. 1, as reported by the US Centers for Disease Control and Prevention, and assuming that deaths for the rest of the year will equal the August-December average from 2017, 2018 and 2019. This results in an estimated death toll of 81,524 for the full year, about 27,000 more than the 2017-2019 average. That excess-deaths estimate is more than the 23,641 deaths that the city health department has so far attributed to Covid-19, reflecting some mix of missed Covid cases, deaths from other causes that can be ascribed to the pandemic and the countermeasures taken against it, and random variation. To calculate the mortality rate, I assumed that the city’s population declined from 2019 to 2020 at the same rate that it has (according to the Census Bureau’s annual estimates) since hitting an all-time peak in 2017. Anecdotal reports of pandemic-related departures from New York indicate that it has probably declined by more than that, and for this and other reasons I expect that my estimate of 9.9 deaths-per-thousand for 2020 will turn out to be an underestimate. But I wanted to err on the side of conservatism.
One other note about the data behind the chart: It’s available back to 1936 from online city vital statistics reports. But after trying and failing to get the city health department to send me older annual data, I owe the rest of the numbers to the work of Rutgers University at Newark economist Jason M. Barr, who offers a handy guide for where to find them all.
There are multiple ways to place the projected 2020 death toll in historical context. The 81,524 deaths would fall short of the city record of 91,169 set in 1968, and 9.9 deaths per thousand would merely put the mortality rate back about where it was in the early 1990s. One has to go much further back, though — to the influenza pandemic of 1918 — to find an increase in mortality from the previous year as big as my 2020 estimate of 3.4 deaths per thousand. Meanwhile, the percentage increase in mortality is the biggest since cholera arrived in North America in the early 1830s.
One factor complicating such comparisons is that Covid-19 has been deadliest for the elderly, which wasn’t always true for past epidemics. In 1918, 75% of the deaths attributed to influenza and pneumonia in New York City were among those 39 and younger; in 2020, only 4% of the city’s Covid-19 fatalities have been 44 or younger. Measured by the number of life-years robbed from New Yorkers, the 1918 influenza pandemic was thus much worse — and New York wasn’t even all that hard hit in 1918 relative to the rest of the country. Also, other quite-deadly infectious disease epidemics aren’t immediately visible on the mortality charts because they took years to play out. AIDS, for example, has killed a city-estimated 19,840 New Yorkers since 1982, most in the late 1980s and early 1990s.
Still, in terms of the burden on hospitals and the sheer shock to the city and its economy, Covid-19 really does rate among the greatest public-health challenges (or disasters, if you prefer) of New York’s history. Readers of the city’s annual vital statistics reports will be remarking upon it for centuries, assuming that chart keeps running on the cover — and that the city continues to exist, of course.
Most of the rest of the US hasn’t been nearly as hard hit by Covid-19 as New York, so the nationwide mortality-rate increase has been smaller. In contrast to New York’s, though, it is still rising rapidly. The semi-official US death toll from the disease is 176,809, but the number of deaths from all causes was already 236,000 above the recent years’ average through Aug. 8, according to CDC estimates. Falling confirmed-case numbers indicate that deaths will probably begin slowing soon, but even if they do — and there is no major resurgence of the disease this fall or early winter — it’s getting harder and harder to imagine a scenario in which overall deaths this year aren’t at least 300,000 above the recent norm.
Nationwide mortality data is available back to 1900, so here’s what that chart looks like with 300,000 more deaths than normal in 2020.
By this measure, Covid-19 is again the biggest deal since the 1918 influenza pandemic, but it’s nowhere close to as big a deal as that. The estimated mortality increase this year is also only slightly bigger than during influenza epidemics in 1928 and 1936. Influenza viruses were first isolated by researchers in the mid-1930s, and vaccines began to develop not long after that, which helps explain why the 1957-1958 and 1968 influenza pandemics are barely detectable on the chart.
The CDC also calculates age-adjusted mortality, which attempts to correct for changes in the age distribution through the years. The percentage of Americans 65 and older is a lot higher than it was a century ago, and that age cohort is, well, more prone to death than the others. So age-adjusted mortality represents what mortality rates would be if the age distribution remained roughly constant through the years, in this case with 2000 as the index year. The CDC hasn’t released full-year age-adjusted mortality rate estimates for 2019 yet, so I used its estimate for the four quarters ending Sept. 30. For 2020, I did my own adjusting using 2019 population estimates and the age distribution of deaths reported so far this year to the CDC.
The 2020 mortality increase here is again the biggest since 1918 in percentage terms. But it only returns the age-adjusted mortality rate to where it was in 2006. I guess one could argue that this means we all ought to chill out — the US was accustomed to this level of mortality as recently as a decade and a half ago, after all. Given that some other countries have been able to get through the Covid-19 pandemic so far with no detectable increase in mortality at all, though, that seems like a cop-out.
Finally there’s that gap, mentioned in passing above, between the increase in overall deaths and the number of deaths that have been attributed to Covid-19. Most of these deaths represent undercounting of Covid-19 cases, especially early in the pandemic when few tests for the disease were available. Some, though, may be collateral damage: suicides, shootings, medical conditions that weren’t identified or treated in time. Some critics of Covid-related lockdowns have blamed this on the lockdowns, and in some cases they’re probably right. But by this point doctors’ offices are open again, even if some patients are afraid to go, and the continuing economic and societal damage seems to be emanating more from the failure to control the pandemic than the measures taken to thwart it. Yes, some of the restrictions imposed in March and April may have been overkill, but the thing that’s killing people now — directly or indirectly — is Covid-19.