Violent Media, Magic, and Homicide

One last post on gun control, I promise.

Something that always surprises me is why the contention that violent video games or movies are responsible for violence is given so much credence, both in the media and by otherwise intelligent and well grounded individuals, when there is so little data to back this belief up. The data on the subject is mixed, it would be safe to say that there is a lean in the literature towards concluding a relationship between violent media and aggression but this is a far from overwhelming lean, like we see with gun ownership and homicide/suicide.*

Even admitting this, however, the aggression correlated with video games tends to be relatively minor, hair pulling and minor scuffles among children, the link between violent media and criminal behavior is much less well established, and international correlations don't show a relationship specifically between violent video games and homicide, as with guns (state level was the best I could find). There is a stronger relationship between viewing media passively, like TV in general, and lower social capital which is associated with more violent crime. However, this relationship isn't what is being written about by those blaming violent media for violence.**



So, the question is why are people so willing to believe that there is a relationship, despite the weakness of the evidence, while they doubt so strongly the relationship between guns and homicide? Motivated reasoning is no doubt part of it, but there is doubtless a large overlap between gun owners and those who watch violent movies or play violent video games, so while motivated reasoning no doubt explains some outliers who are gun owners but dislike violent media it seems problematic to ascribe people's willingness to believe the association between violent media and gun homicide when this group is doubtlessly rather small. There may also be a piece having to do with motivated biases propagated by powerful, incumbent actors like the NRA and other aspects of the gun lobby but it is doubtful to me that these lobbies are that much more powerful than the rather well funded media lobbies.

So, if the usual motivational factors are unlikely, what is it that explains people's credulity to the violent media suggestion? The best I've come up with is some rather old anthropological concepts written about in the Golden Bough.*** Specifically, sympathetic magic and Frazer's Law of Similarity. The Law of Similarity posits, essentially, that like leads to like. Because violent media bears a superficial similarity to violent action there must be a relationship between the two, just as primitive medicine would suggest a link between the mandrake root and man or a plant that resembles a phallus would act as an aphrodisiac. This neatly precludes the necessity to establish a mechanism by which violent media would make violent action more likely, as in standard causal reasoning. It is also simpler cognitively than trying to establish what would actually make violence more likely by trying to figure out what mechanisms enable violent action, which increase it's likelihood of success, and which actions prevent it and make it less likely.

But if we try to use standard causal reasoning, instead of magical thinking, the relationship looks spurious. What skills, exactly, does a video game teach that are anything like shooting a gun (and yes, I have shot a gun, though not recently)? Is holding a controller anything like holding a rifle? Does it produce real-world responses? Does playing video games make someone a more successful killer? Other than a superficial visual resemblance there is no clear relationship between a sedentary activity like holding a controller or sitting at a keyboard and actually physically moving through space and using a firearm. Violent media does nothing to change the incentives and obstacles faced by an individual, as firearms do. So why is it that we take the violent media argument seriously, despite the weakness of the evidence for it, if not for primitive magical thinking?

*To name some ungated studies:
Ayres and Donohue. Shooting Down the More Guns Less Crime Hypothesis.
Cook and Ludwig. Principles for Effective Gun Policy.
Cook and Ludwig. The Social Costs of Gun Ownership
Duggan, Mark. More Guns, More Crime.
Kellerman, Arthur A. Treating Gun Violence Before the 911 Call.
Killias. Gun Ownership, Suicide, and Homicide: An International Perspective.
Miller, Azrael, and Hemenway.  Rates of Household Firearm Ownership and Homicide Across US Regions and States, 1988 - 1997.
Assorted links on homicide.

Assorted links on the self defense argument.
Ferguson, Garza, Jerabeck, Ramos, and Galinda. Not Worth the Fuss. (on violent video games)
For the other side of the argument there is this, though I would note that Anderson cites an awful lot of his own studies, and it is almost 10 years old.

** Essential caveat here, gun control and violent media are most definitely not the only two factors at work. As any gun control opponent will point out, countries such as Mexico and Russia have both far higher murders and less guns than the US. However, these countries have more violent crime overall, not just more homicides. There certainly are other factors, like social capital, income inequality, access to mental health services (and health care more generally), rule of law (especially important when states like Mexico, Russia, or South American states are under discussion), prevalence of a black market (drug trade here, but other goods in some nations are important), etc. that are more important in violent crime rates than guns are. However, given a level of violent crime, guns appear to have a strong relationship with homicides and suicides in a society. It would probably be more valuable to target the causes of violent crime more generally than it would be to focus on gun control, but relative to other factors gun control, hard as it is, is the low hanging fruit compared to income inequality, the drug trade, or isolated lifestyles and low social capital.

*** Yeah, it's more than 100 years old and I am aware that anthropology has moved a rather far way from this, but as a simple observational point people do seem to reach for this kind of simple relationship of resemblance = cause even though it is completely spurious reasoning. There is no logical reason that two things that resemble each other should have a causal relationship, for a causal relationship to exist the cause should have some mechanism to facilitate the effect, and mere resemblance does not give us any data on why this relationship would exist.