Violent Movies and Clueless Consequentialism
posted by Frank Pasquale
Researchers have recently “proven” that “violent films prevent violent crime by attracting would-be assailants and keeping them cloistered in darkened, alcohol-free environs.” In fact, they estimate that about 52,000 less assaults occur each year because of the showing of violent movies. It’s all part of the “Super Crunchers” trend, where economists “crunch . . . numbers to evaluate matters like cheating among sumo wrestlers or the effects of a crackdown on cocaine.” Pity that the urge for “clean identification” and headline-grabbing results produces a bizarre overconfidence about the study’s extrapolability.
The key mechanism identified here is a substitution of movie-watching for more dangerous activities:
“Economics is about choice,” [one study author] said. “What would these people have done if they had not chosen to go and see a movie? Whatever they would have done would have had a greater tendency to involve alcohol. If you can incapacitate a large group of potentially violent people, that’s a good thing.”
I’ll be the first to agree that the study helps demonstrate that, say, those with a propensity for violence are less likely to go out and be violent after watching such a film. But what about the next night, or the night after that? These aggregative, atomized studies do nothing to explore the long-term psychological impact of viewing violence. As Craig A. Anderson, director of the Center for the Study of Violence at Iowa State University, notes,
There are hundreds of studies done by numerous research groups around the world that show that media violence exposure increases aggressive behavior. People learn from every experience in life, and that learning occurs at a very basic level of brain function.
I’d trust individualized psychological research far more than I would follow brute aggregation here. Admittedly, I’m no expert, and I can envision a plausible theory whereby viewing violence substitutes for, rather than encourages, doing violence (though I would guess the “spillover” effect of viewing is far stronger than the “compensation” effect, to use Jon Elster’s terms). I can only congratulate the recent study authors on the limited finding that people are less likely to brutalize others on the very night they watch violent films–not on giving us much insight at all on the overall cultural impact of violent films.
January 7, 2008 at 10:01 am
Posted in: Economic Analysis of Law, Law and Psychology
Print This Post
Responses (10)
James Grimmelmann - January 7, 2008 at 10:54 am
Frank: “But what about the next night, or the night after that?”
The study also seems to have more in common with your take on violent media than this blog post would let on:
Frank - January 7, 2008 at 11:02 am
I agree that the “few nights after” data is useful here. But I still think the economists and the psychologists are talking past each other. The latter are interested in the cumulative effect of viewing thousands of images of violence. The former are trying to engage that debate by looking at effects of, say, one violent movie for one night or a few days or weeks afterward.
Josh Wright - January 7, 2008 at 11:32 am
Frank:
The fact that these authors limit their focus to the short-run effects does not mean that economists aren’t interested in the long-term effects of violent movies (or long term effects generally). The authors note that this is the primary limitation of their research design, describing long term effects as “important,” and noting that this limitation means that the paper “should not be used to inform policy on the effects of limiting the level of violence allowed in the media.”
Of course, data sometimes do not allow us to answer everything at once. Sometimes economists can identify short-term but not long term effects, for examples. But surely you would agree that increasing our empirical knowledge of the short run effects is a contribution with some value, even if marginal (you at least acknolwedge the contribution as “useful”).
But I remain puzzled by the wording of the title of your post. Is “clueless” supposed to describe the authors? It seems to me that you mean the authors (they are, after all, the economists). If so, is it because of they use econometrics? Or is it because they don’t tell us about long term effects and appropriately limit their claims? But I could be wrong. Perhaps it is the readers who misread the scope of the results? Or the journalists who write about them?
Who’s clueless and why?
Frank - January 7, 2008 at 11:49 am
Josh, I’d criticize anyone quick to extrapolate this extremely limited, narrow study into an important comment on the effects of violence.
I have long argued that many commentators are tempted by scientism–a belief that an assertion that is quantifiable is somehow ipso facto more valid than one that is not. Giving this study so much play strikes me as a classic example of that error, given that it has virtually nothing to say about the intuitively most plausible explanation of the bad effect of violent media: namely, a lifelong conditioning toward accepting such acts as normal or normative.
The clueless are those who do not accept the full range of explanatory strategies social science provides. Charles Tilly provides a nice survey of that range:
http://www.newyorker.com/archive/2006/04/10/060410crbo_books
If the paper really “should not be used to inform policy on the effects of limiting the level of violence allowed in the media,” what’s the point of it?
Josh - January 7, 2008 at 12:32 pm
“I’d criticize anyone quick to extrapolate this extremely limited, narrow study into an important comment on the effects of violence.”
I’m still not sure who this applies to, other than perhaps some journalists. Not the authors as I read the paper. But you seem to leap from the short-term focus of the study to the argument that these authors reject other forms of knowledge from social science. To me, that’s an odd reaction to a paper that is a complement to existing social science literature, not a substitute.
Are the authors clueless under your definition?
“If the paper really “should not be used to inform policy on the effects of limiting the level of violence allowed in the media,” what’s the point of it?”
The point is learn more about the causal effects of media violence on crime. More is better when it comes to knowledge about this potential causal relationship, no? And perhaps learning about the short-term effects will help us understand better the mechanism through which the long term effects operate.
You have asserted a mechanism that you believe is the most plausible explanation for the long-term relationship and are confident enough in this view that you view efforts to study other channels as “classic errors.” But for those who do not have access to your foresight about what the evidence will show, one generally doesn’t know this until they engage in the research. I would have thought that you would have applauded research efforts to study the detrimental influence of media (even if not on the dimension you think the most important) rather than describe the authors and their study this way.
Frank - January 7, 2008 at 1:12 pm
“I would have thought that you would have applauded research efforts to study the detrimental influence of media (even if not on the dimension you think the most important). . . ”
What I think is essential here is the implicit effort to reframe this debate as one about the “consequences” of viewing violent media rather than about the goodness or badness of violent media (viewing) itself.
The psychological research here appears to me to dovetail well with a generally virtue-oriented approach to media and politics generally. Certainly, the religious figures (and wing of the Republican party) that have most effectively advanced that approach have come under attack from libertarians and others who believe in an absolutist version of “content neutrality.” But a virtue-oriented approach posits that, in the main, consequences are far too hard to determine “objectively” to count for much in the political debate here. The real issue is the quality of the content itself, and the experience of viewing it.
So to the extent the “aggregate empirical data” talks past the important psychological dynamics here, or marginalizes the far richer normative/virtue-oriented discourse, I think it’s as much a distraction as a contribution.
As for the authors of the study being “clueless:” I did not intend that description for them at the outset, only for those who misused their results. But if they really think that a finding of “52000 lives saved a year due to violent movies” is not going to affect the policy debate, maybe its extension would be appropriate.
Maryland Conservatarian - January 7, 2008 at 2:54 pm
I’m suspicious of any study that doesn’t somehow support an ever-increasing role of government or at least strengthen the conventional wisdom among the political left. I suspect Al Gore agrees with me.
Seth Finkelstein - January 7, 2008 at 8:35 pm
I dunno. I’ve found it odd that a common cultural image is a man dying from death by torture in an ancience Roman method of execution. What’s the effect of being constantly told this is sacred? Surely that’s worth as much study as any movie
.
viol - January 10, 2008 at 10:18 am
a study that gives information about the next 2- to 3-week time frame and can say nothing about longer term effects seems so obviously useless as not even to be worth publishing in the first place.
I have just completed a study showing that hardly anyone was motivated to buy a Cadillac during the 2 weeks after they viewed a commercial by the auto maker! Oh, was anyone motivated to buy one during the course of the next year, after watcing these commercials repeatedly? cain’t say. Wasn’t trying to tackle that one.
Bob - February 4, 2008 at 7:33 pm
I heard one of the authors of this study interviewed on the radio. His final thought was that if entertainment activities are attractive to would-be criminals, then perhaps we should promote something like midnight basketball. Now, I don’t mind the recent incursion of economists into crime (although Becker chimed in long ago), nor does the “black-box” data analysis bother me (as long as there is a necessary limitations section) even the conclusions of the study don’t bother me as a person who knows the crime prevention literature pretty well 9it makes sense on some level)…but it is the complete lack of knowledge of the field of criminology (or that somehow economists don’t need to read the criminology literature) that does bother me. Mid-night basketball has been long dismissed as an ineffective social program that fits into the easy-feel good social policy tool box. When he said that, I was able to dismiss him as guy with a an econometric hammer trying to drive a criminology screw.
Leave a Reply