fb-pixel Skip to main content

How a spate of shark attacks in 1916 might predict how the new coronavirus outbreak will affect Trump’s re-election chances

A series of shark attacks off the coast of New Jersey may have dampened the vote for President Woodrow Wilson.The Boston Globe - The Boston Gl/Boston Globe

It’s only a matter of when, not if, the new coronavirus, Covid-19, causes “severe” disruption in the United States, or so said the Centers of Disease Control and Prevention last week. With the first case of infection by “unknown exposure” in California, and the fast rate of spread elsewhere in the world, Americans could be looking at months of school closures, lots of telecommuting, an end to mass gatherings, and a potentially significant number of fatalities in a worst-case scenario.

But what about the potential impact on the nation’s politics? Strangely enough, a look back at what happened when sharks ravaged the coast of New Jersey more than a century ago offers some clues — and those clues don’t bode well for President Trump.


In their 2017 book, “Democracy For Realists: Why Elections Do Not Produce Responsive Government,” two political scientists, Larry M. Bartels and Christopher H. Achen, examined the political impact of a spate of shark attacks in early July 1916 in northern New Jersey that killed several local teenagers.

With sharply declining tourism revenue, and no system in place to prevent shark attacks, local communities turned to the federal government for help. President Woodrow Wilson, who had previously served as governor of New Jersey, convened a Cabinet meeting, but his own Bureau of Fisheries told him there wasn’t much to be done. By the time he mobilized Coast Guard resources to patrol the waters, the sharks had moved on and the fear and controversy had faded.

But local voters, adversely affected by the economic consequences of the attacks, still took their anger out on the president. According to Bartels and Achen’s calculations, in the communities most impacted by the attacks, Wilson underperformed by 10 percentage points.

Ironically, as the authors note, responding to the local issue may have fed the notion that the federal government should have done something to handle the situation.


Now, it should be pointed out two prominent political scientists have raised methodological concerns about the study and questioned whether the impact of the shark attacks on Wilson’s electoral showing was as strong as the authors claimed. But Bartels and Achen stand by their research.

And they argue the New Jersey case wasn’t a one-off phenomenon. “Voters," they write, "consistently and systematically punish incumbents for conditions beyond their control.” The two authors looked at a century of data on droughts and floods and found that “wet and dry conditions” cost the incumbent party 0.7 percentage points, and extreme “droughts or wet spells” cost them around 1.5 points. They estimate that in 2000, extreme weather conditions cost vice president and Democratic candidate for president Al Gore 2.8 million votes in states that were either “too wet or too dry,” likely costing him the presidency.

So does that mean that if the new coronavirus has a highly disruptive impact on Americans it will hurt Trump politically?

I posed this question to the authors. According to Bartels, when the Spanish Flu, which took the lives of an estimated 50 million people around the world, hit the United States in 1918, the political impact was largely non-existent. The flu received surprisingly little national attention and no one, says Bartels, was complaining, "Why isn’t President Wilson doing more to save people?”

That may seem hard to imagine from a 2020 perspective. But in 1918, there was not a broad expectation that the federal government would get involved in this sort of public health problem. Victims, moreover, were scattered around the country, so connections to the influenza ravaging Europe were not necessarily made. And few public commentators linked the deaths to Wilson’s actions.


Bottom line, Bartels argues, there wasn’t a clear link “between the president and the disaster.”

“In most cases,” Bartels told me, “incumbents do get blamed, whether deserving or not, because someone succeeds in tying the disaster to them.”

The more the new coronavirus outbreak, and the government’s response to it, becomes politicized, the worse things are likely to turn out for Trump. Since Trump has already gone out of his way to blame Democrats and the media for exaggerating the potential impact of the virus, the politicization train appears to have left the station.

Trump’s unsuccessful efforts to minimize the impact of the rapidly spreading virus for fear that panic will negatively affect the economy may backfire spectacularly. Now his actions and response are tied directly to the crisis.

If Covid-19 causes major dislocations, Trump will inevitably pay a price for any resulting economic slowdown, even if blame for the underlying pandemic can hardly be placed at his feet.

But what would compound the political price is if the administration’s response is seen as inadequate. Early indications are that the administration’s lack of preparedness, general incompetence, and past cuts to global health security programs could have precisely that impact.


It goes without saying that no American should be rooting for this to happen. The loss of life, livelihood, and normalcy will take a heavy toll on the lives of ordinary Americans.

Combined with the extraordinary disruption of school closures and Americans being home-bound for weeks or months, coronavirus has the potential to do more to deny Trump a second term than anything that happens on the campaign trail this fall.

Michael A. Cohen’s column appears regularly in the Globe. Follow him on Twitter @speechboy71.