摘要：Recent mass shootings by persons seemingly afflicted with serious mental illness (SMI) have received extensive news media coverage and prompted national dialogue about the causes of, and policy responses to, mass shootings. News media framing of SMI as a cause of gun violence may influence public attitudes about persons with SMI and support for gun violence prevention proposals. We analyzed the content of a 25% random sample of news stories on SMI and gun violence published in 14 national and regional news sources from 1997 to 2012. Across the study period, most news coverage occurred in the wake of mass shootings, and “dangerous people” with SMI were more likely than “dangerous weapons” to be mentioned as a cause of gun violence. In recent years, mass shootings—in Blacksburg, Virginia (Virginia Tech); Tucson, Arizona; Aurora, Colorado; and most recently Newtown, Connecticut—have received extensive news media coverage and prompted a national dialogue about the causes of and appropriate policy responses to gun violence. Although mass shootings are rare compared with other forms of gun violence—65 000 persons in the United States are shot in criminal attacks each year 1 —extensive news media coverage of mass shootings draws public attention to the problem in a way that “everyday” gun violence does not. Public concern about mass shootings is likely heightened by the apparently random nature of such events, which often occur in seemingly safe public places, such as schools. 2 The discussion of mass shootings in the news media has addressed a wide array of issues that may contribute to gun violence, including widespread availability of guns in the United States, 3 violent video games, 4 bullying, 5 gang involvement, 6 childhood neglect and abuse, 7 and serious mental illness (SMI). 8 The Virginia Tech, Tucson, and Aurora shooters appear to have had SMI, 9–11 a category that includes conditions such as schizophrenia and bipolar disorder. 12 Even though these shootings have drawn public attention to an apparent link between SMI and gun violence, the vast majority of persons with SMI are not violent 13,14 and a small proportion of gun violence is committed by persons with SMI. 15 The relationship between SMI and violence is complicated by factors such as substance abuse and trauma. 13 Furthermore, existing risk-assessment tools do not accurately predict violence among persons with SMI, making it challenging to target policy initiatives to the small subpopulation at heightened risk of violence. 16 Nonetheless, some policymakers at the federal and state levels have targeted policy interventions to reduce gun violence at persons with SMI, proposing policies to improve screening and treatment 17 and prevent persons with SMI from having guns. 18 In the wake of the Newtown shooting, several states—including Maryland 19 and New York 20 —passed laws restricting access to firearms among those with mental illness. These states also passed gun laws unrelated to mental illness, including assault weapon bans. The recent focus on SMI as a potentially important factor in mass shootings has generated concern among some segments of the public health community. Mental health researchers have long suspected that news media portrayals of violent persons with SMI contribute to negative public attitudes about persons with serious conditions like schizophrenia. 21–23 This suspicion is supported by results of a recent experimental study, which found that respondents who read a news story describing a mass shooter with SMI reported higher perceived dangerousness of and desired social distance from persons with SMI, compared with respondents randomly assigned to a control group. 24 Communications research suggests that public attitudes about groups of people are heavily influenced by news media portrayals of specific individuals, particularly when the public has little experience with the group in question. 25 Given that the majority of the public has no personal experience with SMI, 26 news stories describing acts of gun violence by persons with SMI may exacerbate negative attitudes about an already stigmatized population. Although stigma surrounding depression and anxiety has decreased in recent years, 27 widespread negative public attitudes toward persons with conditions like schizophrenia have remained steady 21,28 or by some measures increased. 21,29 Experts have linked these negative public attitudes to the pervasive social problems affecting persons with SMI, including undertreatment, 30 poverty, 31 and homelessness. 31 Mental health advocates have expressed concern that the news media’s focus on mass shooters with SMI could lead the public to view SMI as a common cause of gun violence. 32–34 Social psychology research suggests that the torrent of news stories describing mass shooting events by persons with SMI could lead the public to wrongly view SMI as a primary causal factor in such shootings. In a series of randomized experiments investigating news media coverage of poverty, for example, Iyengar found that event-focused news coverage emphasizing specific individuals or events led respondents to blame impoverished individuals themselves for the problem of poverty. 35,36 By contrast, thematic news coverage of poverty—which described the issue in broad, general terms—led respondents to more frequently attribute responsibility for poverty to societal factors. Importantly, Iyengar found that individual versus societal attributions of responsibility for social problems led to differing support for public policy responses. 35,36 For example, respondents who blamed societal factors for poverty were more likely to support poverty reduction initiatives. 35,36 In commentaries published in several prominent public health journals, experts have asserted that a misguided focus on SMI as an important cause of gun violence could lead the public to support policies that restrict the rights of persons with SMI without meaningfully reducing gun violence. 32–34 In particular, experts have expressed concern that policymakers might promote policies to prevent persons with SMI from having guns, rather than policies to strengthen gun laws more broadly, as an appropriate means to reduce the likelihood or severity of gun violence in the United States. 34 In a 2011 commentary in JAMA , Gostin and Record argued that to reduce gun violence in the United States, policies should target the “dangerous weapons” used in all gun crime, not “dangerous people” with SMI, most of whom are not in fact violent. 34 Although dangerous-people causal framing implies that responsibility for gun violence lies with a select group of individuals with SMI, dangerous-weapons framing implies that responsibility lies with widespread access to guns in US society. In line with previous work suggesting that societal attributions of responsibility for social problems may raise public support for policy solutions, 35,37 framing dangerous weapons as a principal cause of gun violence may heighten public support for meaningful reform of US gun laws. To better understand how the relationship between SMI and gun violence has been framed in the public discourse, we conducted a content analysis of news media coverage of SMI and gun violence from a large sample of news outlets over a 16-year period from 1997 to 2012. To assess how news media portrayals of gun violence by persons with SMI may influence attitudes about this group, we evaluated the volume and content of event-focused news coverage describing specific shooters and shooting events, compared with thematic news coverage describing the issue of SMI and gun violence in broad terms. We hypothesized that thematic news coverage would be more likely than event-focused news coverage to mention key facts related to SMI and policy proposals to reduce gun violence, respectively. To determine whether news media coverage differed during periods of heightened public attention to gun violence following mass shootings, we compared framing of causes of and policy responses to gun violence by persons with SMI in the weeks following mass shootings with news coverage during other periods. We hypothesized that news stories in the 2 weeks following mass shootings would be more likely to mention policy proposals to reduce gun violence than stories in other time periods. Finally, we examined whether news coverage mentioning specific causes of gun violence was correlated with coverage of specific types of gun violence prevention policy proposals. We hypothesized that news media mentions of the dangerous-people causal frame would be positively correlated with mentions of SMI gun restriction policy proposals, and that mentions of the dangerous-weapons causal frame would be positively correlated with mentions of other (non-SMI) gun policy proposals.
