Myths of Violence: research on guns, media, & subcultures of violence

The four most common explanations for school shootings are mental illness, gun ownership, media violence, and a “culture of violence”. There is wide agreement that a “culture of violence” exists, and that it’s bad, but curiously little debate about exactly what the phrase means. Explanations of gun ownership and media violence, however, are hotly contested, in no small part because the policy prescriptions are so varied.

In the past, I’ve researched the transformation of catastrophic events into media myths. Briefly: if an event is shocking enough, the concept of the catastrophe de-links from the historical event and becomes a metaphor for explaining other literally unrelated events. Consider the framing of the Deepwater Horizon oil spill as “Obama’s Katrina” (Cohen, 2010Vogel, 2010) or the persistent warnings of a “Cyber 9/11” (Martinez 2012, Bloomberg 2012). Hackers with 0day exploits are no more similar to hijackers with a box cutters than is an oil spill like a hurricane, but we understand these phrases as metaphors.

One of the characteristics of media myths is that they collapse the explanations for a catastrophe within the name applied to the event so, after a period of negotiation of “how this happened” and “what we call it”, a singular black box is produced. While there may be widespread consensus as to the existence of the black box, disagreement may remain about what exactly is within. While everyone may understand what “Katrina” or “9/11” refer to, there remains virulent conflict between subcultures about what lessons the should be taken from the catastrophe. School shootings form another class of myths, but one which re-open the black box with every successive catastrophe, exposing the enduring conflicts of interpretation.

It is a peculiar features of modern American punditry and politics that the popular opinions and policy which prevail often have little (if any) relationship to relevant research. I’m of the opinion that, when living in the Age of Information, it is perverse to ignore empirical evidence. In this post I’m going to relate some of the scholarship on the effects of violence media, gun ownership, and the (sub)culture of violence. I’ll leave questions of mental health and Constitutional interpretation to others.

Guns and electronic entertainment are both technological artifacts embedded in American culture and, when studying their effects, I’m not convinced they should be evaluated differently. Implicit in this post are assumptions about the relationship between humans and their tools, technology and the society, which will are often left unexpressed. I don’t want to get sidetracked by a rebuttal of instrumentalist views of technology but it’s worth stating I agree that “all media are extensions of some human faculty — psychic or physical”  (McLuhan, 1967). The literature I survey uses a more constricted (and popularly recognized) definition than McLuhan, limiting it to electronic entertainment media like TV, radio and video games. I will not refer to guns as “media” any further.

[W]eapons proper are extensions of hands, nails, and teeth, and come into existence as tools needed for accelerating the processing of matter… Since our new electric technology is not an extension of our bodies but of our central nervous systems, we now see all technology, including language, as a means of processing experience, a means of storing and speeding information. And in such a situation all technology can plausibly be regarded as weapons. (McLuhan, 1964)

To argue that guns or electronic media have no effect upon humans is to deny the radical changes in warfare, social organization and economics made possible by each.  Any reasonable definition of “culture” includes the artifacts produced, used and consumed; guns and media programming included. It follows that changes in technology use will be accompanied by a change in culture: “We shape our tools, and thereafter our tools shape us” (McLuhan, 1964). The question, for me, is not whether these tools have altered our cultures, but how. Or, more directly to the topic of this post: what evidence do we have about what kind of effect one or another technology has on the rate of violence and mortality?

Media Effects on Aggression & Violence

In the aftermath of the Newtown massacre, the public debate re-opened over the societal effects of violent entertainment. “All for curbing weapons of war. But shouldn’t we also quit marketing murder as a game?” tweeted David Axelrod, one of the President’s top aides. Senator Jay Rockefeller (D-WV) introduced legislation to direct research funding to “investigate the impact of violent video games and other content on children’s well-being,” and called on the FTC and FCC to “take a fresh look” at restricting access to such entertainment (press release). MPA Chairman Chuck Dodd issued a statement committing the film and TV industry to support “President’s efforts” (then un-announced), be part of the national conversation and “help America heal”.

The discipline of media concerned with media effects is so well established that it has birthed subfields which now threaten to become disciplines in their own right. While media researchers may be encouraged by Senator Rockefeller’s threat to fund more research, they have good reason to be mystified about what new information policy makers could possibly want. The “discontinuity between news reports and the actual state of scientific knowledge” is the source of considerable annoyance among psychologists, who largely blame the commercial misinformation and he-said/she-said journalism (Bushman & Anderson, 2001).

The scientific debate over whether media violence has an effect is basically over, and should have been over by 1975″ (DA Gentile, 2003). While the original research concerned the influence of TV and movie violence, we now have over twenty years of accompanying research on the influence of video game violence.

“Heavy diets of violent television and movies clearly have a detrimental effect on children… Although there is less research on the effects of violent video games than there is on television and movies, the preponderance of evidence looks very similar to the research on violent television. In particular, violent video games appear to increase aggressive thoughts and feelings, physiological arousal, and aggressive behaviors, as well as to decrease prosocial behaviors” (DA Gentile, 2003).

By 2010, the scientific debate over whether violent video games have an effect was basically over: they increase aggressive thoughts and behaviors among children, adolescents and adults, in the short- and long-term, across cultures (Anderson et al, 2010a).

While most of the discussion of media effects focuses on the pernicious impact of the violent content which dominates commercial programming, researchers found hope in studies of prosocial effects:

“Both in experimental settings and at home, children who watched prosocial content behaved significantly more positively or held significantly more positive attitudes than others… the results suggest that television is no more prone to fostering violence than it is to fostering prosocial behavior… the conclusion is that television has the potential to foster positive social interactions, reduce aggression, and encourage viewers to be more tolerant and helpful” (Mares & Woodard 2005)

Studies on prosocial video games have found significant increases in helpful/cooperative/altruistic behavior across linguistic and cultural samples (Gentile et al, 2009). “It appears that doing nice things to game characters during a video game generalizes to real-world social interaction in the form of more helpful behavior,” furthermore “the simple absence of antisocial media content has positive effects on pro- and antisocial outcomes” (Greitemeyer, 2011). Lest I be accused of letting the RIAA off the hook: similar prosocial/antisocial effects have been found for music lyrics. One study tested prosocial, neutral and antisocial music in a restaurant and found that lyrical content dramatically altered the size and frequency of tips (Jacob, Guéguen & Boulbry, 2010).

Scientific debates often continue even after they are “basically over” (Kuhn, 1962). The most notable detractors of media effects don’t deny the phenomenon, but rather the definitions, methods, evidence, and theories advanced by others. Some researchers argue that most studies are invalid because they define “aggression” as anything less than violent — a significant increase in temper-tantrums, disrespect, bullying or bickering being insufficient — or didn’t establish media effects strong enough to cause psychosis (Ferguson 2009, p.42). Theoretical arguments can seem like a distinction without a difference to the outsiders: “[E]nvironmental stressors are catalysts for violence; they don’t cause violence, but they may stimulate specific violent acts in a particular individual who is already prone to violent behavior.” According to this dissidents’ Catalyst Model, media violence inspires copy-cat behavior among people who would otherwise still act violently (Ferguson 2009, p.45). This particular model is a popular argument among gun rights advocates (“Then they’ll just use knives! Should we ban knives too?!?!1!!”) and those who find explanatory power in “evil” incarnate but, as we’ll see in subcultures of violence literature, it ran rather serious problems when tested in previous decades.

There is good cause to be skeptical of arguments which extrapolate mainstream media effects research into a direct causes for specific incidents. Violent media viewing is not predictive of criminal behavior in individuals (Heath, Kruttschnitt & Ward, 1986) or on a population-basis (Savage 2003 and Savage 2008). Proponents of media effects are very express about this:

“We have never claimed that national violent crime data are a good test of media violence effects… media violence researchers do not claim that violent media are the most important risk factor for aggressive and violent behavior. However, of all the risk factors linked to aggressive and violent behavior, exposure to violent media may be the easiest factor for parents to control.” (Anderson et al, 2010b)

These findings are counter-intuitive for both proponents and opponents of the mainstream “violence (TV/music/video games) cause violence” formulations. While consumption of media violence causes aggressive thoughts and behaviors in individuals and groups, such thoughts and behaviors are insufficient to predict individual or social criminally. To the extent that mass shootings and other violent crimes are the problems requiring a policy response, we should not expect media labeling regimes (or even content censorship) to yield measurable gains.

Gun Effects: research on gun ownership, violence & mortality

If the scientific community has reached a consensus about the relationship between violent media and aggression, it remains uncertain over the relationship between gun ownership and violence. In contrast to media effects, “gun effect” remains an immature field which has all but stagnated for a generation.

[I]1996, pro-gun members of Congress mounted an all-out effort to eliminate the National Center for Injury Prevention and Control at the Centers for Disease Control and Prevention (CDC). Although they failed to defund the center, the House of Representatives removed $2.6 million from the CDC’s budget—precisely the amount the agency had spent on firearm injury research the previous year…

In 2009, Branas et al published the results of a case-control study that examined whether carrying a gun increases or decreases the risk of firearm assault. In contrast to earlier research, this particular study was funded by the National Institute on Alcohol Abuse and Alcoholism. Two years later, Congress extended the restrictive language it had previously applied to the CDC to all Department of Health and Human Services agencies, including the National Institutes of Health. (Kellermann & Rivara, 2012)

The irony of post-massacre calls for more research into media effects is near total silence on the need for more gun effects research.

“US residents of all ages and both sexes are more likely to die from suicide when they live in areas where more households contain firearms” Miller et al, 2007) and international comparisons find gun ownership so strongly correlated with suicide that the latter can be used as a proxy for the latter (Killias, 1993a). International comparisons show positive correlations “between the rates of household gun ownership and the national rates of homicide and suicide as well as the proportions of homicides and suicides committed with a gun,” and no evidence for the argument that people kill with other weapons if guns are unavailable (Killias 1993b). While leaving open the possibility of an exogenous variable which rendered gun ownership mood, such a variable would “need to produce high gun ownership levels and high homicide rates using firearms, but only marginally increase homicides through other means” (Killias, 1993a). In news media and punditry, that mystery variable is called the “culture of violence” (or alternatively, “evil”).

* Studies which have found a positive correlation between high gun ownership and rates of violence have been unable to separate gun ownership from other social factors (i.e. “a culture of violence”).
* While it appears that past gun control legislation has had little efficacy in lowering rates of violence in the United States (Kleck & Patterson, 1993), those experiences revealed some of the mechanisms of gun violence which could inform better policies.

Culture Effects: Subcultures of Violence

Among sociologists, “subculture of violence” designates a specific set of theories and studies popular in the 1970s and early 1980s. This scholarship sought to explain relatively high rates of homicide and gun ownership in the South, finding explanations in historical and socioeconomic factors. In later permutations, this theories was later applied to African American communities (especially young black men), explained with cultural legacies of Southern slavery and segregation. Neither the “Southern Subculture of Violence” hypothesis nor the “Black Subculture of Violence” fared well under statistical analysis.

To test the “Southern Subculture of Violence” hypothesis, Dixon & Lizotte used questions from the General Social Survey (1973, 1976, 1980, and 1984) to categorize respondents with Defensive and Violent attitudes. Those with Defensive beliefs endorsed using violence to confront a man beating a woman or child, or breaking into a home; those with Violent attitudes endorsed using violence to confront a protester or clumsy drunk.
They found regional variation in Defensive beliefs, but not in Violent beliefs, and both were more prevalent among white males than black males. Violent beliefs were related with youth, rurality, and lower levels of income and education. Defensive beliefs, in contrast, increased with exposure to the Southern region, educational and income, but also decreased with age. They found no significant relationship between gun ownership and Violent attitudes, regardless of the type of gun owned.

Note further that we are not suggesting that a subculture of violence does not exist. The results of our research, for example, demonstrate the existence of a non-regional subculture of violence related to age, education, race, income, and city size. We are merely asserting that any southern subculture that exists is not based on values supportive of violence and that any subculture of violence that exists is not southern, independent of structural factors. Our findings also cast doubt on the premise that gun ownership is a defining characteristic of subcultures of violence. In our analysis, gun ownership, whether of pistol, rifle, or shotgun, is in no way related to either a regional or nonregional subculture of violence. While defensive attitudes are related to firearms ownership, these attitudes are unrelated to the violent attitudes indicative of membership in a violent subculture (Dixon & Lizotte, 1983).

Defensive and Violent beliefs were were more common among whites than blacks. but whites were and Violent beliefs were related to youth, poverty, rurality and low levels of education. They found no support for a the hypothesized southern subculture of violence
In 1997, Cao, Adams & Jensen used the same methods to test the “Black Subculture of Violence” hypothesis. Using GSS data from 1983 to 1991, they also found that white were more likely to hold Defensive beliefs than blacks, and found no
(CaO, Adams & Jensen, 1997)

“defending women, children, and property from unwarranted aggression” versus “hitting a protester” .

temp text: The US has spent more money on the media effects of Hollywood and videogames than on the effects of gun ownership. We’ve been trying to prove that TV and other media causes crime and violence for about 70 years, and you know what we’ve got? Nothing. There is no measurable effect. We can’t say the same about guns because the 1996 Republican Congress banned federal funding for research.

Guns are the #2 cause of injury death for kids under 9, and have probably surpassed automotive injuries as a cause of death (2012 data not out yet); these deaths are heavily concentrated in the ~30% of American households with guns. We know that owning a gun increases the chance of being shot, but we don’t know why. We know that having a gun in the house increases the chance of kids being killed, but we don’t know why. We know that gun owners are more likely to commit suicide (and ~55% of gun deaths are suicides), but we don’t know why. We don’t know how to reduce the number of these deaths because the NRA and GOP has basically prohibited any qualified scientist from asking the question (some states have made it illegal – a crime! – for doctors to ask if a gun was involved in an injury).

If I go back to the early Clinton years to dig up research, I can suggest a gun control law which uses criminal background checks to target 80% of the people who use guns in commission of a crime (20% of deaths). It wouldn’t touch anyone without a record. We can’t find out if that research is still valid and we can’t discuss such a policy. Why? I don’t get it, but any time anyone notices that guns are used in gun crimes, the NRA freaks out and the noisiest gun owners start hopping around barking at Hollywood, Imaginary Hitler and the New World Order. And we take this seriously.

The result is that we get Congress proposing new money to study videogames and not one penny to ask if assault and battery is a precursor to armed robbery or if a non-fatal gun accident is a precursor to a fatal gun accident. It is within the President’s power to allow the research anyway, and that’s one Executive Order I think we desperately need signed.