Does violence against women rise 40% during the Super Bowl?
In the media buildup to this year's Super Bowl I saw a mention of the old story that there is a 40 percent increase in violence against women on Super Bowl Sunday due to testosterone-jacked men taking it out on the women in their lives. I seem to recall that this story has been debunked but couldn't find anything definite and look to you to sunder the mists of ignorance.
Don't expect miracles. This myth was debunked three days after it first broke in the media in 1993, but seven years later it's still making the rounds. I don't know who's worse — advocates for good causes who make baseless claims or the chumps in the media who report them as fact.
The whole thing began when Fairness & Accuracy in Reporting (FAIR), a liberal media watchdog group with a vocal feminist wing, decided to draw attention to the problem of domestic violence by persuading NBC to run an anti-wife-beating spot during the 1993 Super Bowl. To bolster its case, FAIR sent out a press release saying the day of the Super Bowl was "one of the worst days of the year for violence against women in the home." A separate statement sent to FAIR activists said "women's shelters report a 40 percent increase in calls for help during Super Bowl Sunday."
FAIR and other women's advocacy groups held a press conference a few days before the Super Bowl in Pasadena, California, the site of the game. One speaker, Sheila Kuehl of the California Women's Law Center, cited a study by researchers at Old Dominion University in Virginia showing an increase in police reports of beatings and hospital admissions in northern Virginia following games won by the Washington Redskins during 1988 and 1989. The Associated Press subsequently reported that there was a 40 percent increase in calls for help following the Super Bowl and a similar increase after Redskins victories. The AP reporter later said he got these figures from Kuehl and from a FAIR spokesperson.
FAIR's attempt to draw attention to domestic violence was a huge success. NBC agreed to run the anti-wife-battering spot, and a flurry of reports appeared in major media citing the 40 percent figure, giving the impression that the football/violence link was a proven fact. On ABC's Good Morning America a Denver psychologist said she'd been collecting data for ten years and that she too had found an increase in wife battering during the Super Bowl. A press release distributed on behalf of the University of Buffalo echoed the earlier claims and warned women, "Don't remain alone with him during the game."
It was all baloney. Ken Ringle, a skeptical reporter for the Washington Post, called around and found that there was no evidence of increased violence against women during the Super Bowl and that claims about violence following Redskins victories had been exaggerated. Janet Katz, one of the authors of the Old Dominion study, said that emergency room admissions of women for gunshot wounds, stabbings, lacerations, and so on were slightly higher on days when the Redskins won, but not 40 percent. (Having read the study, I'd say even that claim is dubious. See below.) Ringle called several women's shelters and found no evidence of increased violence against women on Super Bowl Sunday. Domestic violence experts said they knew of no research supporting such a claim.
On the day of the Super Bowl, Ringle's story debunking "Abuse Bowl" claims appeared on the front page of the Post. Some newspapers subsequently backpedaled on the credulous stories they had run earlier. The Wall Street Journal published a scathing editorial, and Rush Limbaugh took the opportunity to ridicule the "feminazis" once again.
Furious, FAIR denounced the Post story in a long complaint to the newspaper. But the Post's ombudsman and later the American Journalism Review did some checking of their own and found that while Ringle had made a few mistakes, his central claim was correct — there was no reliable data documenting an increase in violence against women during the Super Bowl. FAIR now says it never claimed to have hard evidence, but when you read the AJR account of the controversy it's clear the group bears most of the blame for the misleading stories. For example, the 40 percent figure cited in the memo sent to FAIR activists, which was passed on orally to many reporters, was based on a statement in Donna Ferrato's book Living With the Enemy (1991), the evidence for which evaporated on close examination.
The reality of violence against women is shocking enough — why make stuff up? Unfortunately, this isn't the first time the facts have been distorted in defense of women's rights. That's an issue I'll return to next week.
About that study
The folks at FAIR were kind enough to send me the Old Dominion study ("The Impact of Professional Football Games Upon Violent Assaults on Women," G. F. White, J. Katz, and K. E. Scarborough, Violence and Victims, vol. 7, no. 2, pp. 157-171, 1992). Over a two-year period 680 women were treated in emergency rooms for assaults, lacerations, etc. — an average of less than one per day. The statistical analysis is dense and difficult to follow, but a footnote contained the following:
On the day of a win the actual mean number of women admitted is 1.05, the average expected number is 0.75, one day after a win, actual 0.80; expected = 0.58; two days after a win, actual 0.70, expected 0.63.
On "win" days, therefore, violence increased 40%. But there were only 20 such days during the two-year study period. Doing the arithmetic, we find that 15 cases were expected, while the actual number was 21 — a difference of six cases. You can trick this out with all the statistical jargon you like, but it seems foolish to base any grand conclusions on such trifling numbers.