Does watching football lead to domestic violence?
Later this week, families across the country will sit down for their annual turkey dinner. In many households, this will be followed immediately by another Thanksgiving tradition: switching on the TV to watch grown men bash one another to near-unconsciousness on the football field.
That this ritual may be hazardous to the players' health is obvious, but new research reported in a National Bureau of Economic Research study argues that just watching football can be harmful to fans and their families. Based on domestic violence police reports from the years 1995-2006, the report finds that when an NFL game ends in an upset, the home state of the losing team experiences a sudden, brief uptick in domestic violence.
There are two main schools of thought about how domestic violence happens, and they are by no means inconsistent with each other. Such acts may be the explosive result when years of simmering tensions finally come to a boil. Or they may be an impulsive lashing out at whichever victim is close at hand in a moment of anger (perhaps followed shortly by shame and regret). To the serious football fan, a painful loss may trigger just such a moment of anger.
To assess the football-violence connection, economists Gordon Dahl and David Card collected 12 years' worth of NFL game histories, matching up the teams' records to data on "intimate partner violence" from the National Incident Based Reporting System. The NIBRS data, which include information on all crimes reported to the police, are often only available state-by-state. So the researchers focused on states where there is only one "local" team, to avoid confusion over what happens in California when, say, the 49ers win and the Raiders lose. After further limiting their analyses to states for which there's adequate crime data, the researchers ended up with six teams: the Carolina Panthers, Denver Broncos, Detroit Lions, Kansas City Chiefs, New England Patriots, and Tennessee Titans.
How much grief would we expect each loss to give its fans? If the loss is entirely expected, probably not much: For this year's 1-and-8 Lions, each new loss is received with resignation. By contrast, last week's loss by the Steelers, favored by a seven-point margin over the Bengals, probably rankled many fans in Pittsburgh. To quantify fan rage, Card and Dahl collected the Vegas point spread, which provides an indication of which side was expected to win and by how much. A less-expected loss equals angrier fans. They classify teams with a spread of three or more points as favored to win.
Card and Dahl found that on Sundays during the regular season, losses by favored teams—that is, painful losses—are associated with an 8 percent increase in intimate partner violence. (For, say, the Commonwealth of Massachusetts, with a population of nearly 6.5 million, this translates into an extra seven incidents on a Sunday when the Patriots unexpectedly lose.) These extra cases appear in the hours immediately following the game—3 to 6 p.m. for games with a 1 p.m. start time, and 6 to 9 p.m. for those with a 4 p.m. start—further bolstering the case that postgame rage may take the blame. The spikes in violence are nearly twice as big in emotionally charged matchups between traditional rivals, like the annual Bears-Packers matchup, and also in games with lots of turnovers and penalties.
It's important to note, however, that Card and Dahl's findings don't necessarily imply that football games increase violence overall. In earlier work, Dahl, together with Berkeley economist Stefano della Vigna, found that violent crimes actually decrease when graphic movies like Hannibal and Scream 2 are released. They argue that this is the result of would-be aggressors spending the night at the movies rather than out on the streets committing felonies. If the NFL also has its share of violent fans, football may keep potential criminals off the streets for at least a few hours on Sunday afternoons. It's possible, in other words, that football leads to spikes in domestic violence but dips in other kinds of violent crime.
And while a tough loss for the home team may touch off abuse, that doesn't mean football is the root cause of postgame violence. More likely, the loss merely serves to set off an attack that was already waiting to happen. In a world without football, acts of abuse might merely get postponed, only to be brought on later by some other source of anger. In the long term, rather than blaming football, we may be best off focusing on addressing the more fundamental problems underlying abusive relationships.
Recent work by Brown economist Anna Aizer points to one possibility. In a study forthcoming in the American Economic Review, Aizer shows that potential female earnings are a critical determinant of spousal violence. Such violence declines as wages in female-dominated sectors—such as hospitality services—increase relative to male-dominated ones (e.g., construction). Aizer argues that high-earning women feel more empowered to walk out on potential abusers, since they're less dependent on a man's paycheck to get by. (Men will also be less inclined to chase away an extra breadwinner.) Providing a way out for battered women—through the support of shelters or direct financial assistance—would similarly provide an escape from imminent abuse, not just shift it to a later date. And, like higher wages, these measures would also make the threat of departure more credible, causing potentially violent partners to think twice before lashing out impulsively. Still, as I settle in to watch a little post-turkey football on Thursday afternoon, I'll be rooting for Green Bay over Detroit and hoping that—whatever the outcome—fans remember that it's only a game.
Ray Fisman is the Lambert Family professor of social enterprise and director of the Social Enterprise Program at the Columbia Business School. He is the co-author, with Tim Sullivan, of The Org: The Underlying Logic of the Office. Follow him on Twitter.
Photograph of football players by Jed Jacobsohn/Getty Images.