On the days their national team plays in this month’s World Cup contests in South Africa, English fans watching on TV will veer between joy and despair — and afterwards some of the men will toss chairs, spit profanity and blacken eyes.
Unlike soccer hooligans, these men rampage behind closed doors against wives and girlfriends.
After seeing domestic violence calls rise nearly 30 percent when England played its matches in the 2006 World Cup, police chiefs in that country last month urged all districts to initiate or step up pre-emptive home visits to repeat abuse offenders when England starts its tournament play against the U.S. Saturday.
Would a similar type of preventative action help on Sundays during football season in the U.S.?
You may think so once you’ve read a study released by the National Bureau of Economic Research. The authors found that domestic violence calls to police spiked by an average of 10 percent in areas where a National Football League team suffered an upset loss. (And while their research may summon memories of the canard that Super Bowl Sunday is biggest day for domestic abuse in the United States, it has never been proven.)
“What our paper says is that when there is an unexpected negative shock, a man is more likely to hit a spouse,” says Gordon Dahl, associate professor of economics at the University of California, San Diego, and one of the study’s co-authors.
First released in preliminary form in November, the study is another contribution of behavioral economics to fields generally associated with sociology and psychology. Dahl and co-author David Card, professor of economics at the University of California, Berkeley, hoped to expand and improve on the existing descriptive research on intimate-partner violence — a costly social plague that surveys show, despite progress, still affects more than a million women in the U.S. each year.
The authors examined data collected over a 12-year period for six teams and controlled for location, point-spread, weather and size of the local viewing audience. They were curious about the role of emotional cues and set out to test the hypothesis that negative ones — benchmarked relative to a rationally expected reference point — set the stage for a man to lose control.
To the authors, the increase in violence reports after upset losses indicates that a man attacks a wife or girlfriend because watching his team lose unexpectedly sets the stage for a breakdown of control, a spasm of anger and frustration.
That conclusion lands the authors in the middle of an incendiary social policy rivalry about what causes and how to stop men from hitting and beating their wives and girlfriends.
Now, Dahl and Card agree that losses of emotional control account for only some intimate-partner abuse.
At least half the time and maybe more, battery follows a pattern in which a man makes a rational decision based on a desire to control a woman, to keep her from leaving the relationship, from talking about her job, from cooking the wrong dish for dinner, for failing to lavish enough care and attention on the man. Men are kings of the castle; women must keep them happy.
During the 1990s, when new laws first helped fund shelters and treatment centers and spur more arrests, battered women who lived in fear were afforded more protection and more men were directed toward counseling based on the control interpretation and social messages that supported it.
Some see shortcomings in any theory of abuse that doesn’t have control and coercion at its heart. According to the influential Domestic Abuse Intervention Project in Duluth, Minn., all tactics and forms of abuse against women involve attempts to exert and maintain power; violent incidents are not isolated breakdowns of control or even cyclical expressions of anger and frustration.
“That’s what men say — ‘I lost control,’” quips Scott Miller, a team leader at the Duluth program.
Other family violence experts say instrumental and expressive violence are often intertwined.
“It is a major leap to go from loss of games to a violence type,” says Edward Gondolf, research director at an addiction research institute at Indiana University of Pennsylvania and the author of many books on battering and family violence. “There are numerous dynamics and situational variable that are at play and could account for the outcome.
If men were truly out of control, according to the Duluth view, why don’t they punch police officers or their bosses?
The Duluth model and its adherents have taken their share of criticism, too, in recent years, especially from a handful of critics who cast doubt on the re-education programs for men. The critics say the Duluth model overlooks the role of mental illness in domestic violence, and they call for more psycho-therapeutic treatment, marriage counseling, fewer mandatory arrests and more attention to women’s violence against men.
Dahl believes there’s been too much “butting of heads” among researchers and advocates of the different causes and approaches to abuse.
“I don’t understand that,” he says. “There’s more than one type of abuse. I’d like to stamp out any form of domestic violence.”
He and Card created the mathematical model for their study using data from the National Incident-Based Reporting System.
The authors limited themselves to states where there is only one “local” team and because of limits on the data studied only games played by the Carolina Panthers, Denver Broncos, Detroit Lions, Kansas City Chiefs, New England Patriots and Tennessee Titans. To define favorites and underdogs, Dahl and Card settled on a Las Vegas point spread of three or more points.
An upset loss, they found, seems to initiate a hot period when a man is more likely to use poor judgment and lose control — just like a parent who does a bad job when he or she is tired or has had a bad day at work.
Watching a team win when it was favored or lose when it was the underdog meant relatively little in the research results. Only after the upsets, in the last hour of play and two hours after a 4 p.m. game especially, did the number of calls rise noticeably. “The data suggest that the spike in violence after an upset loss is concentrated in a narrow time window surrounding the end of the game,” they wrote.
The more at stake for the fans, the bigger the spike in violence, Dahl and Card found. If the home team was still in playoff contention (based on having at least a 10 percent chance of making the playoff), the result of an upset loss is a 10 percent increase in violence. Upset losses to a traditional rival were associated with a 14 percent increase in violence reports.
Upset losses were more potent if the home team took a whipping. If it suffered four or more quarterback sacks, four or more lost possessions (due to fumbles or intercepted passes) or 80 or more penalty yards, the estimated effect of an upset loss was a 15 percent increase in reports of violence.
Could the statistics be interpreted differently?
“One could as easily claim instrumental violence for the same data,” says Gondolf. “Men expect sympathy, mourning time and support when their team loses (or when they experience loss in general). They act to reinforce those expectations and punish those who violate them. For example, a woman violates that inflated expectation when she asks for something, doesn’t kowtow or doesn’t make a sufficient big deal of the loss.”
In a recent phone conversation, Dahl noted that upset losses produced increases in friend-on-friend violence, too, so the emotional sting isn’t a danger only to spouses and girlfriends. But spouses get the worst abuse, he says. “And it happens at home, in places where families are supposed to be protected.”
During the initial mid-1990s campaigns against domestic abuse in England and Wales, according to an England Home Office report, West Midlands and South Wales Police sent letters to offenders advising them of the conditions of their bail and consequences of arrest. Derbyshire Constabulary had a car or wagon specifically for prolific offenders and arrested 10 of the 28 known offenders who had not yet been brought in.
Significantly, the Duluth program encourages arrests and prosecutions because avoiding jail is a major incentive for an abuser to enter a treatment program.
As to whether such methods could ever be used systematically at specific times, such as holidays or during Sunday afternoons in autumn and winter in the U.S., no one can say, and nothing like it apparently has been tried. Miller says it’s not the holiday or the game that brings the abuse — it’s because the man is home and drinking.
It also remains to be seen whether devoting resources to individual days when violence is likely diverts attention from the bigger battle — a battle in which much progress has been made.
Dahl says just like drinking, the disappointment of an upset loss — loyalties to football teams being what they are — peels away some of a man’s judgment, making trouble more likely.