What Pittsburgh 1918 teaches us about easing pandemic measures too soon
Politics overruled public health judgment, contributing to the worst death rate of any major American city.
A month ago (it feels like a year, no?), we looked back at why some cities fared better than others during the 1918 flu, to help understand the power of local decisions in driving health outcomes during a pandemic. As Covid’s impact becomes clearer, it’s remarkable how much these century-old lessons stood the test of time.
In 1918, cities that implemented strict social distancing measures early on fared best; today, San Francisco’s quick start has driven its success in containing the outbreak. In 1918, cities closer to coal-fired power stations had measurably worse outcomes than those farther away; today, air pollution again seems tied to higher Covid mortality rates. In 1918, social disparities worsened the flu’s impact; today, studies are finding similar inequities in New York, Milwaukee, and other U.S. cities.
Some of America’s initial Covid hotspots have shown promising signs of flattening their disease curves. With these encouraging trends have come natural discussions around when to ease up on social lockdowns. As that discourse grows, it’s worth reminding ourselves of another lesson from 1918: cities that ended their restrictions too soon often experienced a fresh surge of cases. Even from a strictly economic perspective, cities with longer restrictions saw benefits.
Pittsburgh’s experience during the 1918 flu offers a cautionary tale on both counts.
Motivated largely by political factors and local business interests, Pittsburgh leaders pushed for an end to its gathering ban before the state health department believed it was safe to do so, according to multiple histories of that period. Following that move, Pittsburgh suffered severe case flare-ups, contributing to what ended up as the worst rate of “excess” deaths (beyond a typical flu season) among American cities: 807 per 100,000 people.
What’s more, Pittsburgh’s early release didn’t lead to brighter economic trends, according to new research linking stricter — and longer — social distancing measures with more manufacturing employment and banking assets. Compared to nearby industrial counterpart Cleveland, Pittsburgh had a shorter primary intervention period (by a week) and far fewer total intervention days (53 vs 99). It also had lower manufacturing job growth from 1914–1919: 18 percent vs 42 percent.
To be clear, Pittsburgh’s decision to curtail restrictions wasn’t the only — or even necessarily the leading — reason for these poor health and economic outcomes. But the decision certainly played a role, and given many of the echoes between its situation in 1918 and the one facing cities today, it’s worth taking a closer look.
The politics of pandemics
While Philadelphia gets a lot of attention for having terrible outcomes in 1918, Pittsburgh actually did worse by most measures. Its excess death rate, mentioned above, exceeded that of Philadelphia, which came in at 748 per 100,000 people. All told, writes historian James Higgins of Lehigh University in a 2010 review, “the flu killed fully 1 percent of Pittsburgh’s population.”
Many of the factors for Pittsburgh’s poor health outcomes were set in motion long before the 1918 flu outbreak struck. The city had very high pollution rates, a byproduct of the coal industry. Its population suffered from a number of preexisting conditions known to worsen the impact of respiratory disease. And the city had poor living conditions, including dilapidated housing and open sewers.
Its public health apparatus reflected this state of affairs. The city’s health director, William H. Davis, “was a party stalwart with no medical background,” according to Higgins. Mayor Edward Vose Babcock was a businessman who entered office promising fiscal restraint and didn’t make public health a priority, even at a period in history when such investments were far more common.
In fall 1918, as the flu pandemic swept across the U.S., Pittsburgh’s board of health ignored early signs of an outbreak — convinced (with no supporting evidence) that a weaker strain would hit the city than hit places like Boston or Philadelphia. On September 29, despite clear signs of emerging flu cases, the city held a celebration of 40,000 people to promote war bonds. Into October, health chief Davis was saying things like: “What constitutes an epidemic is a matter of opinion.”
Pittsburgh did agree to implement a state-imposed public gathering ban on October 4. But the city kept public schools open until October 24 — even then only closing when attendance was so low it made no sense to keep them open. Local leaders also undermined the spirit of social distancing by sending out a fleet of volunteers to collect war bonds door to door, with the mayor conveniently overlooking a gathering of 50,000 people to hear the results of that effort on October 19.
All the while, local industries lobbied the mayor “to weaken, and finally end, the gathering ban before the state lifted it or the situation warranted its termination,” writes Higgins. These pitches started as early as October 10, less than a week into the intervention. No doubt the economic pain was real, but graft was also a factor: alcohol distributors were notorious for their kickbacks to city officials, who turned a blind eye to hotels and bars that kept selling to crowds.
As the ban persisted, inching closer to an early November election, politics entered into play. Tensions rose between Mayor Babcock and the state health commissioner, Dr. Benjamin Franklin Royer, a trained physician and public health veteran. In an exchange that will sound familiar to those following the discourse today, Royer was accused of taking drastic measures “for political effect,” to which he responded that he had “no thought other than that of saving life.”
On October 29, not yet a month into the interventions, Babcock met Royer in person and pitched an end to the ban. To help the mayor’s case, a local coroner published a letter claiming to know “officially” that the outbreak was slowing. In fact, records show the city was only then nearing its peak rate of excess deaths. Higgins continues:
Pittsburgh officials, principally the mayor, wished to lift the ban before the state believed it advisable. The weeks-long fight that resulted undermined the already-weak epidemic-fighting measures.
The impact of ending early
Failing to get clearance from the state, Mayor Babcock acted anyway, announcing his intention to end the ban himself. Sure enough, on November 3, Babcock reopened bars and theaters — nearly a week before the state was scheduled to lift the ban, on November 9.
Some establishments did quickly close up again at the state’s behest, limiting the fallout. But Babcock had set a clear tone: rather than ease back into things on a timetable in line with leading public health guidance, pre-pandemic behavior should prevail as soon as possible. On November 11 — the date the armistice was signed ending World War I — Mayor Babcock sparked a massive gathering, despite the fact that dozens of people were still dying daily from the flu. Here’s Higgins:
On November 11, with the armistice signed, Mayor Babcock led a parade that “threaded its way through walls of humanity in the downtown districts” and drove even “staid citizens” to such “rollicking abandon” that “it seemed as if the celebration would never cease.” For the next twenty-four hours, with the mayor’s sanction, revelers packed streets and bars.
In the end, Pittsburgh was fortunate not to suffer a big second wave of the 1918 outbreak, as some cities (such as Denver) did upon ending their restrictions. But its rush back to action coincided with mini surges in the number of deaths and cases. In another study of Pittsburgh, archivist Kenneth A. White shows that new cases flared up shortly after the parade, eclipsing 200 a day four times, with authorities attributing the spikes in part to “crowding during the armistice celebrations.”
White’s chart of daily deaths (below) shows a distinct flare-up on November 8, a few days after the preemptive reopening; another around November 14, a few days after the mass celebration; and yet another from November 19–22:
The decision to cut things short may also have contributed to prolonging the epidemic. According to Higgins, Pittsburgh experienced “much higher than normal morbidity and mortality rates” for months after the restrictions were lifted. Its severe fall wave was followed by “stable but high infection rates, with flare-ups in February and March.” The epidemic didn’t end in full until May. Higgins concludes:
Pittsburgh’s leadership made decisions over the course of months that pushed the epidemic beyond even Philadelphia’s mortality rate and contributed to the longest outbreak during America’s influenza epidemic.
Again, it must be stressed that a great many factors contributed to Pittsburgh’s exceptionally high mortality rate in 1918. The early ending of restrictions alone isn’t to blame; the stage was set by the city’s high pollution, its delay in closing schools, and the generally weak state of public health preparedness. And of course there’s every reason — economic, social, mental health, and otherwise — to want to end pandemic interventions as quickly and safely as possible.
All that in mind, the 1918 flu history suggests there’s at least some wisdom in precaution. Over the long run, extending pandemic restrictions can help protect cities against subsequent disease flare-ups and, by keeping the overall population healthier, potentially even strengthen economic recovery. If this lesson holds as true as the others from 1918, the longer interventions may just prove worth the pain.
Follow Sidewalk Labs with our weekly newsletter and our podcasts.
