MEDIA VIOLENCE IS GOOD FOR YOU

Media; social learning; journalism; assault; inter-personal violence; cinema

Mar 26, 2025

Ask any psychologist about the effects of media violence, and their thoughts will turn to one experiment before any other: Albert Bandura’s Bobo Doll experiment. They’ll probably talk your ears off.

The Bobo Doll experiment famously shows that children often behave violently themselves after exposure to ‘violent models’ – by which we mean adults who also behave violently. Any parent can tell you that children copy adults, of course, but in fact the process Bandura isolated was a little bit more subtle. When they witnessed violence in real life, the children simply added new behaviour to their repertoire. They could and would reproduce the violence themselves, but only when conditions were right. The children weren’t programmed like machines: they just added new options.

Robert Kennedy wsas assassinated. Television journalists covered the story relentlessly. Media violence quickly became a matter of public concern. Bandura himself was invited to speak to various impressive committees. The Federal Trade Commission adopted new standards in advertising (no more films of children bopping each other on the heads with mallets) and news magazines published photographs from Bandura’s laboratory.[i] That was a bit of a coup for a psychologist.

Psychology textbook unanimously told us that media violence was a Bad Thing and it had Bad Consequences. On the occasions when those consequences failed to show up in laboratory research, well, the textbook authors found it relatively easy to blame the bland stimuli that psychologists are forced to use (we’re not allowed to show video nasties in the lab), small sample sizes, or some other frustration. Meanwhile, the Seventies became a byword for television violence. When they weren’t on the sofa enjoying Kojak, psychologists lamented.[ii]

But the thing is… Bandura’s experiments were about children hitting an inflatable doll with a hammer. For sure, those who saw adults behave violently were more likely to behave that way themselves, and, for sure, the effect persisted even when the children watched films rather than saw it happen in real life. Even so, doubts remain.

‘By proving that a child will imitate the behaviour of an adult model, the experiment showed the power of examples of aggression in society’.[iii] So reads one typically grandiose summary of Bandura’s discovery…but is this really what we mean when we think about ‘media violence’? Let’s be honest: no country in the western world was exactly suffering from epidemics of school children rushing into toy stores and beating up inflatable dolls, especially not with hammers. No Bobo dolls were crying out about unfair victimisation. The dolls had no support groups and no protests; none of them pinned protest badges to their chests (partly because, being inflatable, they’d have burst).

If you’d like to read more about the Bandura experiments, you’ll be pleased to know that I wrote a newsletter on the subject. Just click here!

And then there’s this. Children are one thing. Adults are another. I don’t know about you, but I am more frightened of violent adults than violent children. Adults are much more menacing, if only because they don’t get sent to bed when they’re naughty. There is a certain amount of disagreement about the effects of media violence on adults[iv] – and one remarkable study that, well, I reckon every psychologist in the area wishes they’d thought of first. I certainly do.

One small technical point here: Claims about the effects of media violence are just that – claims about effects. They imply that real-world violence is caused by violence on screen. I’ve written elsewhere about the tremendous difficulties involved in establishing causal relationships like this. I shan’t repeat it all here, but feel free to Click if you’d like to read more about this. In a nutshell, here’s the problem: we can only establish causal relationships by means of laboratory experiments. But you can’t bring a cinema or living-room into your laboratory. Hence you can’t directly test the effects of violent movies or video-games.

You’ll be relieved to know that there’s another option. It’s not perfect, but what is? We can take advantage of naturally-occurring situations that work a bit like experiments and see what we can learn from them. (One famous example is a study of flashbulb memories after the Twin Towers attacks.)

A pair of economists took advantage of the obvious fact that violent movies are not released on the same date every year. I say ‘obvious’, but of course everything becomes obvious once someone else notices. That meant they had a perfect, naturally-occurring research design, complete with experimental and control groups, plus a straightforward measure – the number of assaults that occurred on the weekend (or, rather, the amount the authorities knew about- not a perfect measure but not bad).

Let’s make the design clear: the economists compared the number of assaults reported to the police during, say, the first weekend in July under two different conditions - years in which violent movies were released that weekend and years in which they weren’t.

If media violence causes real-world violence, you’d expect the number of assaults to increase on ‘violent movie’ weekends. If, however, it has the opposite effect, you’d expect it to decrease.

Researchers sometimes worry about what we call ‘statistical power’ – that is, the probability of discovering an effect, even if it really exists. If your sample (that is, the number of people taking part in your research) is small, this is a real risk: a relatively small effect may not be detectable if you study only a few people.

Reassuringly, that was hardly a danger in this study. The researcher estimated that about 25 million Americans enjoyed a mildly-violent move, and half that number enjoyed a really violent one.[v] That’s as big a sample as you’re ever likely to see anywhere, and indicates one great advantage of not running a laboratory experiment. You’re not going to fit 25 million Americans into a lab. If violent movies had an effect, given the sample size, these researchers would definitely expect to see it.

‘To be honest, we didn’t know what we’d find,’ said one of them in an interview. ‘Our idea was really, “Oh, we’re probably going to see an increase in crime”. And that’s exactly the opposite of what we found’.[vi]

What they found, in fact, was this: about one thousand fewer assaults on weekends when violent movies were released. One thousand. That, as we scientists like to say, is a shitload. ‘Between the hours of 6.00 pm and midnight, for every million people watching a strongly violent movie, the number of violent crimes decreased by 1.3%.’[vii]

Such a finding ran counter not only to most researchers’ predictions, but (and here I’m speculating) their hopes as well. It’s not what they’d have liked to learn. Naturally, then economists had to look back at the data and try to figure out what was going on. In retrospect, the data actually fits quite well with some of our intuitions about media violence – but, as I say, everything becomes obvious once someone else notices.

We can explain the findings quite easily if we think about exactly who commits most assaults. What are their demographics? Well, for one thing, they’re male, right? And young (not as young as the participants in the Bandura experiment, to be sure, but at the age when the testosterone rules the neocortex and bad decisions get made). They’re also drunk. Drunk young men – they’re the ones you want to avoid on a night out, however much your limbic system may be begging you to please steer clear of that hen party over there.

Who are the main audience for violent movies? Young men, full of testosterone, of course. And what are they not doing while they’re passing a couple of hours in the cinema?[viii] Drinking, of course. Let’s sum it up in one sentence: Violent movies keep young, testosterone-addled young men off the streets and out of the bars to the tune of one thousand assaults a year.

Admittedly, this isn’t quite the argument that the movie industry wants you to hear. Quentin Tarantino, for instance, is a proponent of the theory of ‘catharsis’ – the idea that when we identify strongly with a character in a movie, and live their life through them, we leave the cinema emotionally depleted and lacking the will to start a fist-fight. Nope, it’s not that argument at all. Nevertheless it is quite a convincing one.

Of course, our cinema-goers may head straight to the nearest bar and start drinking. But suddenly that doesn’t matter so much. Thanks to the cinema’s rigid no-alcohol policy, they’re already several drinks behind their usual schedule and unlikely to get drunk enough before closing to start anything questionable.

If you are inclined towards theory, you may be interested to know that there link here is to a view of punishment called social defence. The term was invented by the nineteenth-century criminologist, Cesare Lombroso, who believed that he’d identified the ‘born criminal’. Some people, Lombroso thought, were simply doomed to commit crime. The only thing to do prevent them was lock them away where they couldn’t harm us. Getting potential criminals into the cinema is much better than getting them into prison, of course. It’s voluntary, for one thing. And cinema, unlike prison, is no school for crime.

Go see a violent movie. It’s good for everyone.

Want to know more about Criminal Psychology but don’t know where to start? Here are two suggestions:

Here is a piece on an effective police interview technique.

This piece is all about the effect of chance and probability in the criminal-justice system.

Before you leave, please remember that I overheard the blue buttons insulting your soufflé. Bash ’em. Bash ‘em hard.

MEDIA VIOLENCE IS GOOD FOR YOU

Media; social learning; journalism; assault; inter-personal violence; cinema

Discussion about this post