Gun Violence Is Not Declining, Despite Drop in Homicide Rates

    In a new paper published in Human Communication Research, researchers at the Annenberg Public Policy Center (APPC) and the Ohio State University show that gun injury rates are a more sensitive indicator of the trend in gun violence than gun homicide rates. The finding comes as a response to a challenge published in the same journal which argued that, despite the rise in gun violence in PG-13 movies reported by the APPC and Ohio State researchers in Pediatrics in 2013, there is no evidence that actual gun violence has increased in recent years.

    In the new paper, Dan Romer and Patrick E. Jamieson at APPC and Brad Bushman at Ohio State argue that the increasing ability of hospital trauma centers to prevent deaths from gun injuries makes the use of gun death rates problematic. The researchers present data collected by the Centers for Disease Control from emergency room visits showing that despite declining gun homicide rates, gun injury rates have increased since 2000, when the index was first created.

    In 2013, researchers at APPC and Ohio State published a study in Pediatrics showing that by 2012, gun violence in the top-grossing PG-13 movies had more than tripled since 1985, when the PG-13 rating was established. The study also found that in 2012 the trend in gun violence in the top PG-13 movies exceeded that in the top R-rated movies. The authors (Bushman, Jamieson, Weitz, and Romer) suggested that gun violence in popular movies could be a source of what has been dubbed the “weapons effect,” the finding that the mere presence of guns can increase aggressive thoughts and behavior. The report was based on APPC’s Coding of Media and Health Project (CHAMP), which has cataloged trends in the portrayal of health risk behavior in movies and television since 1950. Other work based on the CHAMP database includes studies on long-term trends in tobacco and alcohol use and suicide in movies.

    In a challenge published in Human Communication Research in 2014, researchers at Villanova and Rutgers universities argued that there is no evidence that trends in gun violence in movies are related to actual homicide rates in the United States and that the authors of the Pediatrics engaged in sensationalism by even suggesting that such a link might exist. To support this claim, they compared annual changes in CHAMP movie gun-violence rates with annual changes in adult homicide rates and found no relationship. The authors did not challenge the central finding of the original study that gun violence has dramatically increased in PG-13 movies, which are marketed to children, and in fact said that it is “even more troubling” that the level of gun violence in PG-13 movies is as prevalent as in R-rated films. The challenging authors acknowledged that “research suggests a link between violent media and relatively minor forms of aggression [although] whether these generalize to severe forms of violent behaviors, such as homicide and aggravated assault, is unclear.”

    We reject the claim that APPC sensationalized the potential relation between movie violence and violence in youth. As we initially stated, the increased popularity of PG-13 movies exposes youth to considerable violence, and the presence of guns on screen “may increase the aggressive behavior of youth.” Indeed, we initially suggested that future research should examine “whether [media] violence with guns is more likely to increase aggression in youth than violence without guns” and “whether films containing gun violence provide viewers with scripts on how to use guns.” However, we noted that additional research is needed to determine the existence and magnitude of movie gun violence effects on youth.

    A rejoinder, also published in Human Communication Research in March 2015, did not challenge our responses. Instead, the challenging authors again used national homicide rates (this time in adolescents), which have also declined since their peak in 1993, to argue that those rates are inconsistent with any influence of media violence, even though their initial challenge argued against using such raw trends for this purpose. The challenging authors scoured the literature to find examples illustrating how both researchers and politicians have drawn a relationship between media violence and violent behavior. These statements have little to do with our original study as reported in Pediatrics.

    We stand by the original findings and the need for further study to understand the effects of movie violence on youth.

    For more information: Michael Rozansky, Director of Communications, Annenberg Public Policy Center | mrozansky@asc.upenn.edu | 215.746.0202.