The Political Relevance of Irrelevant Events (2016)

Can the result of a sporting event swing an election?

This is Throwbacks, a newsletter by me, Michael Weinreb, about sports, history, culture and politics, and everything in-between.

If you like what you read, please click the button below, join the mailing list for FREE and please share, on social media or through e-mail or however you feel comfortable sharing.

And if you’ve been reading for a while, please consider a paid subscription to help keep this thing going—you’ll also get full access to the archive of over 200 articles. (And right now, you’ll get 20 percent off either a monthly or annual subscription for the first year.)

Note: I’m out of town this week, but now seemed like a good time to re-run this post, with few tweaks/updates.

---

I.

On November 2, 2016, the Chicago Cubs defeated the Cleveland Indians, 8-7, in Game 7 of the World Series. Six days later, Donald Trump was elected president of the United States, winning the state of Ohio by an eight-point margin. These facts are seemingly unrelated, but they didn’t feel unrelated to Ethan Busby, who is now an assistant professor of political science at Brigham Young University, and was then a Ph.D. student at Northwestern University. Maybe it wouldn’t have made the difference for Hillary Clinton in Ohio if the Indians had won the World Series, Busby thought, but how much of a difference could it have made?

Busby is an established professional in a serious field who understands that this is a peculiar correlation to make. But Busby has also come to believe that voter behavior is not entirely rational; he believes, as do many of his colleagues, in the theories most prominently advanced by two eminent political scientists, Christopher Achen and Larry Bartels. Achen and Bartels popularized the notion that, essentially, our voting behavior is influenced by the kind of stuff that’s completely irrelevant to the decision we’re actually making. And Busby believes that at least one of those irrelevant factors could be the results of an entirely unrelated sporting event.

In 2017, Busby and two of his Northwestern colleagues, James N. Druckmann and Alexandria Fredendall, published a paper entitled, “The Political Relevance of Irrelevant Events.” It was conceived on a simple premise: How would people with a vested interest in a crucial game alter their opinions of the president of the United States in its wake? In other words, how would an event that is entirely irrelevant to the performance of a president actually change someone’s opinion of the president, if at all?

Busby and his colleagues surveyed Ohio State and Oregon students to gauge their approval ratings of Barack Obama before the teams met for the national championship in 2015. Then, after Ohio State beat Oregon 42-20, they surveyed those students again. And they found a correlation: The students at Oregon had a lower opinion of Barack Obama after the loss. The students at Ohio State had a higher approval rating of Obama after the win.

 

It was not a huge effect: It equated to roughly half a point on a seven point approval scale. It also tended to disappear over the course of a week, and it vanished if the subjects were actually informed of how their thinking had changed. But still, it was statistically significant. And in the final days of a close election, when everything seems to matter, Busby wondered, maybe that meant wins and losses on a field could make a small difference. And if that was the case, why couldn’t other irrelevant factors make a difference? And could all those irrelevant factors add up to make a larger difference?

“It needs to be a race that’s close in a place where it could make a difference,” Busby says. “But when elections are close, I think it could make a difference.”

Busby’s paper was not the first to tie the results of sporting events to electoral politics. Back in 2010, Neil Malhotra, Cecilia Hyunjung Mo and Andrew Healy of Stanford’s Graduate School of Business published a paper entitled, “Irrelevant events affect voters' evaluations of government performance.” Malhotra, Hyunjung Mo and Healy (who now works as an analyst with the Cleveland Browns) studied college football fan bases in the days before an election, to get a sense of whether they’d be more likely to vote for the incumbent if their team won. And they, too, found a correlation. Again, it was small, and it was fleeting, but it was just enough to be statistically significant--and it briefly lit up the Internet when it was published.

“It wasn’t our point to say that all these election results are being decided by college football games,” Malhotra says, because most of the time, the effect wouldn’t be large enough to actually alter the results. “Our point was to say, O.K., outside of a laboratory setting, is there evidence that emotional reactions to events can influence voting?”

---

II.

At first, it’s kind of a wild idea to get your head around. How could any sporting event cause us to think differently about a choice that we fundamentally know has nothing to do with it? And there are skeptics who feel that way, too--one of them, Anthony Fowler, published a paper in 2015 that refuted the Stanford group’s work, attributing it to a “false positive.” In the Washington Post, Andrew Gelman, a professor at Columbia, issued a mea culpa for touting the original study after reading Fowler’s work and declared that “voters aren’t as ‘irrational and emotional’ as is sometimes claimed.” 

“If voters can be easily swayed to change their vote depending on who won the last college football game, then they’re probably making their voting decisions based on all kinds of idiosyncratic factors,” Fowler says. “It’s not a completely preposterous idea. I just think that it’s unlikely to have a lot of bearing in real-world elections.”

But a year after Fowler published his paper, the 2016 election occurred. And this caused us to question pretty much everything. The idea that emotions and mood shifts in the days before an election could impact the vote started to feel less preposterous, because the winning candidate built his entire campaign around appeals to voters’ emotions and mood shifts. Why wouldn’t people--particularly undecided voters without any strong partisan leanings--carry their mood into the voting booth with them? And why wouldn’t seemingly irrelevant events like a football game have the potential to impact their choices?

Both Busby and Malhotra have replicated their own studies--Busby at the 2016 national championship game between Clemson and Alabama, which Alabama won, 45-40. (While Alabama fans didn’t change their view of the president, Clemson fans did have a more negative view of Obama.) And in June of 2020, Malhotra and Mo published a replication of their original study that produced essentially the same results--that sports may be a small factor in decision-making, but that they can be a factor. 

A lot of the dispute between Fowler and these other political scientists gets into complex statistical reasoning--at one point, I found myself hopelessly attempting to decipher the Wikipedia page for something called the bias-variance tradeoff--but the central dispute is much easier to understand. And it gets at who we are as human beings. Are we ultimately rational in our political choices, or are we inescapably emotional?

“I respect what Fowler and many others say in trying to push back on this work,” says Mo, one of the Stanford study’s authors who is now an associate professor of political science at Berkeley. “But I still rest on the question of, ‘Why would a mood affect other decisions and not affect political decision-making?”

---

III.

There is, of course, a long history of political figures utilizing sports as a political cudgel. Ex-athletes like Gerald Ford and Bill Bradley became prominent politicians thanks in part to their reputation from their playing days; Boston University history professor Bruce J. Schulman points out that Ford even ran a campaign ad touting his football career at Michigan. Woodrow Wilson utilized his baseball fandom to his advantage, as did George W. Bush. Theodore Roosevelt intervened to “save” college football when it was suffering through a crisis of violence in the early 1900s, Richard Nixon--who won the endorsement of Jackie Robinson in his 1960 campaign--later utilized southern college football as part of his race-baiting Southern Strategy, and Donald Trump, clinging to that lineage, has regularly used protests within sports to energize his base.

But all of that is part of a larger, overarching strategy. All of those occurrences were conscious choices and strategic decisions. And that’s what sets this relatively new scientific literature apart--we’re talking about decisions could be happening without us even realizing they’re happening.

Once you read the studies, you start seeing patterns everywhere you look. For instance: On November 6, 2016, the Green Bay Packers lost to the Indianapolis Colts, 31-26. It was the Packers’ third loss in four games, and it dropped them to 4-4 on the season. A headline in the Appleton Fox-Cities Post Crescent read, “Listless Packers seeking spark.” Two days later, Donald Trump won Wisconsin--the tipping point state of the 2016 election--by 27,257 votes. 

How many of those 27,257 people were concerned that Aaron Rodgers was past his prime and chose to vote for the maverick Republican challenger instead of continuing with a Democrat in the White House after eight years of Obama? How did their yearning for something different on the field affect their yearning for something different on their television screens every night for the following four years? (That Packers team won its final six games of the regular season before losing to Atlanta in the NFC Championship game.)

“You know, it could be important,” Busby says. “Especially a place like Wisconsin, where it’s really close.”

---

IV.

There are, of course, a million caveats here. Fowler told me that he and his colleague, B. Pablo Montagnes, looked for a statistical correlation with the NFL and didn’t find any. Malhotra told me that he didn’t look at the NFL because, in big cities, there are so many factors that it’s difficult to isolate the impact of a single sporting event. That’s why both Busby’s and Malhotra’s group focused on college teams (Malhotra’s group also studied March Madness)--because their fan base is easier to geographically pinpoint, and because the passion tends to radiate throughout the small towns and campuses in which many of them are located. 

But then, aren’t the Packers the closest thing the NFL has to a college team? And if there is an effect in pro sports, wouldn’t it be that much larger than what we see in college sports, given the size of the fan bases?

And if it’s true in sports, Busby wonders, why shouldn’t it be true for entertainment? What if, say, a highly touted Star Wars movie were to be released in the days before the election--and what if it turned out to be a flop? Could that impact the voting decisions of thousands of light-saber wielding acolytes in swing states? 

Busby hasn’t found a good way to gauge that idea through scientific study; because sports are zero-sum, with winners and losers and geographically defined fan bases, they’re easier to gauge. And that’s also why this will likely remain a contentious issue in the political science community: Because it isn’t easy to get into people’s heads, and to understand their thinking, when it comes to something as utterly personal as voting.

Yet Busby also pointed out that the impact of these events only mattered if people didn’t realize that they were playing a role in their decision making. Because Trump has politicized everything--including sports--it may actually be harder for people to view anything as an “irrelevant event,” including a game.

But still, I can’t help but wonder: On November 2nd, Ohio State will play at Penn State in one of the most high-profile games of the season. Here is a swing state that could potentially decide the outcome of an election, if that election turns out to be closer than the polls indicate. Trump will likely be there in person, which could be a boost or an annoyance. It may not seem rational, but maybe what we’ve learned over the course of recent history is that we’re not as rational as we think we are--that our emotions are always there with us, even in the voting booth.

“I think the answer is that we’re both emotional and rational,” Malhotra says. “But the more interesting question is, ‘What explains when people act emotionally, and what explains when they act rationally?’”

---

This newsletter is very much a work in progress. Thoughts? Ideas for future editions? Contact me via twitter or at michaeliweinreb at gmail, or leave a comment below. If you enjoyed this newsletter, please subscribe and/or share it with others.