Kill Pixels, Not People

Exploding the fake scientific consensus on violent video games

For decades it has been a shibboleth among some social psychologists that increasingly violent media-television, movies, and video games-increase the risk of violence in society. In a 2001 review article for American Psychologist, the Iowa State psychologist Craig Anderson and the Ohio State psychologist Brad Bushman claimed that media violence is nearly as significant a risk factor for social violence as smoking tobacco is for lung cancer. "Research on violent television and films, video games, and music reveals unequivocal evidence that media violence increases the likelihood of aggressive and violent behavior in both immediate and long-term contexts," Anderson and some colleagues asserted in 2003. And in 2007, University of New Mexico pediatrician Victor Strasburger estimated that 10 percent to 30 percent of the violence in society was attributable to media content.

As recently as October, Bushman and two colleagues reported the results of a poll of media psychologists and mass communication scientists in the journal Psychology of Popular Media Culture that there is a "broad consensus" among media psychologists and mass communication scientists that violent media increase aggression in children. Earlier this year, Bushman and a colleague denied being in the thrall of a "moral panic" over violent media, instead accusing dissenting researchers who "use violent media themselves" of being "biased by the force of cognitive consistency and experience a 'reactance' of 'regulatory panic.'"

What is the evidence linking media violence to aggression? A lot of it comes from experiments in which undergraduates view violent scenes or play shoot-'em-up video games for 15 minutes and then are tested for aggression in various ways. Other undergraduates view mild content or play nonviolent games. Typical tests for post-play aggression include how loud a noise blast a player administers to an unseen (fictitious) subject; how much hot sauce he or she adds to food that an unseen subject will eat; and questionnaires designed to find out if the viewers or players are having aggressive feelings or thoughts. Many of the studies do find that viewers of violent content and players of violent games will blast noise a bit louder, dollop a bit more hot sauce, and cop to having slightly more aggressive feelings and thoughts than those who view mild content or play nonviolent games. Interestingly, the researchers rarely pause to wonder if providing the opportunity for aggression actually licenses its commission in their experiments. According to proponents of this theory of media violence, these lab results are relevant to the real world.

Their basic theory linking media violence to real violence can (somewhat unfairly) be summarized as "monkey see/monkey do." They believe that media consumers have difficulty distinguishing between real and fictional mayhem. Violence on movie or video screens supposedly supplies behavioral scripts that viewers and players later act out. Reel violence leads to real violence.

But now the old guard is being challenged by a new generation of researchers who are calling their theories, methods, data, and sweeping assertions into question. Stetson University psychologist Christopher Ferguson is one of the chief antagonists. In their drolly titled 2013 commentary, "Does Doing Media Violence Research Make One Aggressive?," Ferguson and his colleague, German researcher Malte Elson, invite readers to contemplate a thought experiment as a way to think about the plausibility of the "monkey see/monkey do" theory. "Take 200 children and randomize 100 to watch their parents viciously attack one another for an hour a day, the other 100 to watch a violent television program an hour a day," they suggest, "then assess their mental health after one month is over." Surely they are right when they assert that "to suggest the mental health outcomes for these children would be even remotely identical is absurd." As the thought experiment makes clear, ordinary folks do recognize that people, including children, can distinguish between real and fictional violence and will react accordingly.

Recent research bolsters this common-sense view of how people actually experience media. In October 2014 the Villanova psychologist Patrick Markey and colleagues published a study comparing trends in onscreen violence to America's murder and aggravated assault rates between 1960 and 2012. They report that movie violence has dramatically increased in the past 50 years, and that depictions of gun violence in PG-13 movies have tripled in the last 27 years. Controlling for possible confounders such as age shifts, poverty, education, incarceration rates, and economic inequality, they report, "Contrary to the notion that trends in violent films are linked to violent behavior, no evidence was found to suggest this medium was a major (or minor) contributing cause of violence in the United States." In November 2014, the FBI reported that the violent crime rate has fallen by nearly 50 percent over the past 20 years.

With video games, players are not merely passive viewers but active participants in pixelated carnage. In the December 2014 Computers in Human Behavior, a team of researchers at the University of Queensland in Australia used the standard 15-minutes-of-play format widely adopted by video aggression researchers to assess whether playing ultra-violent, violent, and nonviolent video games had any post-play effect on two measures of pro-social behavior. In one, players are paid $5, asked to fill out a brief questionnaire about a local children's charity, and told they can donate some money on their way out. In the second, players are told that they are choosing the level of difficulty of a puzzle that another subject has to finish in a limited time in order to earn money. The hypothesis was that the more violent the game, the harder the puzzle and the lower the charitable donations would be. Instead, the researchers reported that there was no difference among the three groups with regard to pro-social behavior, although the players of the ultra-violent games did donate more. "There is now growing reason to suspect that playing violent video games does not impact prosocial behavior in a normal population," concluded the researchers.

In the November Journal of Communication, Ferguson writes, "If media violence is a precursor to societal violence the introduction of violent video games in the United States would be expected to precipitate increased youth violence rates." Yet as violent video game consumption has increased nearly eightfold since 1996, the violence rate among Americans ages 12-17 fell from 35 to 6 per 1,000 people.

How did social science go so wrong? Ideology. As one parses the research, it becomes apparent that well-intentioned liberal social science researchers engaged in inquiries they hoped would result in restrictions that would prevent school shootings, reduce the murder rate, usher in strict gun control, and, one suspects, elevate their fellow Americans' lowbrow tastes in entertainment. They continue to decry the alleged deleterious effects of violent media even as U.S. violence rates continue their steep decline. The old guard actually cannot see how their experiments and studies are a massive exercise in confirmation bias.

Fortunately, younger social scientists are questioning the ideology that underpins so much prior media violence research. Ferguson and Elson observe that media moral panics eventually abate, in part because the kids who grew up with new media become adults who are less inclined to identify them as a source of social ills.

As the old panic paradigm falls apart, Ferguson and Elson observe, "some scholars actively and aggressively attempt to quell dissenting views, disparage skeptics, question the motives of those who disagree with them, and enforce a highly ideological view of this field." In the April 2014 issue of Pediatrics, Bushman and his colleagues somewhat plaintively asked, "Why is it so hard to believe that media influence children and adolescents?" Ferguson and Elson's reply in the winter 2014 issue of the journal European Psychologist: "The most parsimonious answer to this question is, in fact, 'Because the data are not convincing.'"