How do player politics affect fan attendance? The answer is complicated
Research inspiration can come from any number of places.
For Oskar Harmon, an associate professor in the Department of Economics at UConn, a recently published paper was inspired by a conversation with a friend.
About football.
"One of my close friends said they were no longer going to watch NFL games," Harmon explains. "I was a bit surprised, because he's normally open to different points of view. So that's what motivated me."
What discouraged Harmon's friend was the widely publicized NFL player protests during televised U.S. national anthem proceedings at the start of games.
Player protests, notes Harmon, became one of the most visible controversies facing the NFL during the 2016 and 2017 seasons, as players like the San Francisco 49ers quarterback Colin Kaepernick sat or kneeled as the anthem was played live on network television.
In 2018, major broadcast networks discontinued live coverage of national anthem ceremonies, greatly reducing visibility of player protests. But the interest in what happened in 2016 and 2017 didn't wane for Harmon.
"I was looking at how television ratings changed during the protests," Harmon explains, "and I was really surprised that no one was asking about what happened inside the stadiums. Do all people feel so strongly that they're not going to go to the games? Because people are very committed to football. And I didn't see anybody in the literature looking at it."
Despite media reports of a decline in television viewership and attendance at NFL games amidst the protests, academic research has been limited.
In a review of published studies, Harmon found that a few looked at television viewership responses to the anthem protests, but none examined in-stadium game attendance.
Harmon, who studies consumer behavior and behavioral economics and is a football fan himself - "And I'm very, very, very disappointed in the Giants," he quips, "but I still watch their games" - partnered with fellow associate professor Jungbin Hwang, an econometrics researcher who uses statistical and mathematical theory to analyze economic data, to see what happened to NFL stadium game-day attendance during the height of the NFL protests in 2016 and 2017.
Their findings were recently published in the June edition of the Southern Economic Journal.
Home-field advantage
For their study, Harmon and Hwang looked at the impact of the protests on ticket sales and used local economic and political conditions to estimate the impact of player protests on game attendance.
"The attendance statistics are produced by the NFL at a website they have," Harmon explains. "And then, to understand political polarization, we looked a two different data sources on Black Lives Matter protests as well as data from the Washington Post that tracks fatal police shootings of unarmed Black males."
As a measure of political polarization, they used county-level voting data from the 2016 U.S. presidential contest between Hilary Clinton and Donald Trump.
They also used sports betting data to see how competitive games were and controlled for things like the day of the week, the time of day, and game-day weather.
"Because that could affect attendance," Harmon says. "The idea is that you've got your main variables of interest, but then you need to control for other factors that might influence attendance. So, you isolate those."
"We were particularly interested in how these protests by NFL players in those two years affect attendance of the games in the stadium, because there are some hot debates about the effect, negative and positive," says Hwang. "And it's not simple. We had to systematically organize the data into panel format and employ causal interference methods for panel data that can control for confounding factors, such as team, game location, week, and seasonal effects."
What Harmon and Hwang found was that some reduction in game-day attendance was consistent with boycott behavior amongst fans.
That reduction, however, was somewhat offset by attendance from MAGA-aligned fans engaged in counter-protest as well as fans who attended in order to support player protests.
"We are very cautious to be decisive in one specific direction," says Hwang. "Claims that political protests are driving fans to or pushing fans away from a stadium are really not that simple. The results are mixed, and fans still have a lot of incentive to be in stadium, either supporting or disagreeing with these protests."
The greatest in-stadium attendance impact, the researchers noted, was the response of fans to away-team player protests, where fans had a more statistically significant response in attendance to protests by away teams visiting their stadium as opposed to protests by players on their home team.
The finding showed that attendance was lower when an away team's players were protesting, as opposed to a home team, an unexpected result for Harmon, who suggested home-team loyalty or local demographics might be the cause.
"The big surprise was the away-team reaction - I thought it would be uniform," Harmon says. "I did not expect that it would be largely in response to away teams."
A new arena for research
Harmon says that the study is limited by the nature of NFL ticket sales.
"Seventy-five percent of NFL tickets are sold the season before," he notes, "so we're only looking at 25% of overall ticket sales. So, there might have been other people who didn't physically go, but their tickets had already been sold."
The study also did not include surveys of individual people.
"We're trying to look at correlations and then interpret them and say, well, this is a plausible scenario," Harmon says.
The data is also limited by the relatively small sample size, Hwang says.
"We only have 32 weeks of game days," he says, "and fans are pretty fixed as far as inflow of demand each season. Sports markets, we believe, are kind of inelastic."
Additional study, they said, that looks at turnstile data - as opposed to ticket sales - might shed greater light on in-person impact.
But at least for Harmon, this new-to-him area of research, embarked upon after a personal conversation, has already led to additional discussions with colleagues and a new area of inquiry for him.
He's planning to take his behavioral economic research back to the sports arena to look at how response to unexpected sporting event outcomes, combined with political sentiment, can impact social and domestic issues.
"There's literature about the Super Bowl and follow-up articles about upset games and their relationship to spikes in domestic violence," Harmon says. "The little part of the literature that I want to tap into is how political polarization plays a role in fan response to game loss and incidents of domestic violence."
