Schwarzenegger Lowers Crime

California Governor Arnold Schwarzenegger is one politician who can credibly claim that he is truly responsible for reducing crime, at least if you believe a new study by economists Gordon Dahl and Stefano DellaVigna. It isn’t his policies as governor, however, that he can take credit for, but rather his acting roles.

In their new paper entitled “Does Movie Violence Increase Violent Crime?” they actually find that assaults are lower when a blockbuster violent movie is released, and their results have attracted the attention of the media. Although it might be tempting to an economics world outsider to react with knee-jerk negativity to this research and discount it as being flawed or nonsense, Dahl and DellaVigna are both outstanding young economists who do excellent work. I have only read the paper quickly, and I did not find anything obviously wrong with it. In fact, it seems like they have done the analysis with great care. The one thing I find strange about the paper is that they put tremendous emphasis on violent movies, but really they find pretty similar magnitudes from people attending non-violent or mildly violent movies as well. More or less, it looks like going to movies is good, clean, wholesome fun that substitutes for other activities, such as heavy drinking, that tend to lead to violence.

matthew heintz

Eh, it's not the movies that reduce the crime, its just the fact that are in a location they can't commit crimes. After school programs, bowling, basketball, w/e would show the equivalent declines.

I think it's very misleading to say that watching violent films reduces crime...which carries with the idea that watching something violent somehow releases violent urges that would've be displaced else where...which I found no proof for at all.


Besides the obvious criticism that G-rated movies apparently do not have the opposite effect of increasing violent crime, there are some things that I doubt were factored in....

1) Anytime people are "out of harm's way"--whether at the mall, the theater, the stadium, or church--it is likely that violent crime will diminish, since the pool of victims is going to be smaller than normal.

2) Since violent movies are typically released right along with light-hearted fare, how do we know that there is just not a general feeling of the warm and fuzzies that descends upon the thugs after watching "The Sweet Story of the Littlest Princess"?

3) Oh, and here's a big one: What happens to the bad guys in movies? Hmmm? I've seen semi-trucks fall of bridges on top of them. Seen them hammered through the chest with a pipe. Seen them fall into wood chippers, and the such like. Or, at the least, arrested and imprisoned. Maybe that gives our thugs a pause that carries over the weekend? I mean, even in monster movies like "Predator" the poor alien is blown to kingdom come!

4) I would think that it has to do with "re-direction" in any case. Either having something to do beside rob the neighbors, or at least thinking and talking about something beside plotting for a victim.

5) If this study is true, I'm betting crime goes up just before violent movies are released, since they have to get money to get into the theater!

6) To be fair, it may not simply be that the thugs are otherwise involved by going to movies, but rather that there is some psychological "release" in seeing a violent film that serves to vent immediate hostilities (though it may surface later). I mean, when I leave a funny movie, I don't walk outside and say, "Hey, I've had it up to here and now I'm going to be hilarious!" No, I've had my dose of funny, and so there is no need for me to attempt to fill the void.

I know Freakonomics is often about finding the un-intuitive answer, but I don't see this conclusion holding up. It's not the "violent" movies. It's the movies, period.



One simple explanation would be that violent movies attract the type of audience that is more likely to be off committing crime, otherwise. (Primarily, young men.) So violent movies could draw potential criminals to theaters (and away from shenanigans) in far greater numbers than chick flicks, etc.


Great to see that after the first article was published on this research and heavily criticized by readers we need to read another post defending the paper. Sure, not everyone is a bright economist but people actually are entitled to having interesting opinions on economists' research. Besides, the point readers had is simple enough: it is not about the data used, nor about what types of movies are used, it's about the lack of policy relevance and analytical implications of this research. Anyone can see and comment on that, no need to be pedantic about it I would think.

I heard a similar comparison drawn between rap music and violent crime. The person advocated the violence and debauchery of rap music as being an outlet for angry young men who would otherwise commit real crimes. The proponent claimed that the crime statistics upheld his theory. Color me skeptical when it comes to violence being good for society, whether in music or film.

Gene Shiau

"[R]eally they find pretty similar magnitudes from people attending non-violent or mildly violent movies as well," observed Levitt. I will put forward a corollary hypothesis: break-ins would be higher when a blockbuster movie is released. After all, everyone would be in the cinema, and the house would be left empty, eh?

The trouble is, break-ins don't make an eye-catching headline. No economists would care to publish a study on that.


News Alert ! News Alert !

Brilliant economist, using impeccable methods,
discovers that prison rehabilitation doesn't work!

"All this money being spent to rehabilitate and
teach job skills to prisoners is a complete waste!"

The brilliant economist compared two prisons with
similar types of inmates. In the first prison, no
money was spent on rehabilitation, while in the second
prison copious dollars were spent on teaching literacy
skills, job skills, interpersonal skills, etc...

Using statistics, the brilliant economist discovered
that the exact same number of prisoners were committing
crimes outside the prison. In fact, even during two days
after being released the prisoners from both prisons
committed a comparable number of crimes...

This should put the debate over prison rehabilitation
to rest!

For more on the effects of media violence see:



"Eh, it's not the movies that reduce the crime, its just the fact that are in a location they can't commit crimes. After school programs, bowling, basketball, w/e would show the equivalent declines."

"1) Anytime people are "out of harm's way"-whether at the mall, the theater, the stadium, or church-it is likely that violent crime will diminish, since the pool of victims is going to be smaller than normal."

These arguments really do not hold weight with me based on my own personal experiences. I know that in high school and college, the places where my friends and I witnessed (and were sometimes involved in) violent crime were exactly these kinds of places: theaters, stadiums, malls, and clubs. These places are magnets for the people who are most likely to commit the kind of impulsive crimes that this study deals with: young people in desperate need of something to do.

A related study that I think would be interesting: looking at peoples' mental states immediately after a violent movie (say, the first hour or two). I have noticed that in the immediate aftermath of watching a violent movie, a lot of people in the theater seem to looking for a fight. Watching the prerelease of the 300 comes to mind.


Mark L

This begs the question, and I'm sure this has been studied before, but, if violent movies deter the viewer from committing violent crimes, do pornographic movies deter the viewer from engaging in similar sexual behavior?

I know it's not a perfect analog, but it's another type of intense movie experience that many would argue alters the viewer's behavioral patterns. I wonder if other readers had any insights on this.


Then, perhaps, if one were to lower the cost of going to the movies (perhaps through a government program), would crime drop as well?


Recent studies (conducted in Europe, of all clpaes!) have concluded that the United States of America, with all it's gun toting cowboys, has a lower average violent crime rate than the European nations and their near-ban on all firearms outside the police and military.But you want to know how to lower our already low violent crime rate?1) Deal with illegal immigration. Most illegals, not all, are associated with gangs and drug cartels and bring their work with them here when they (illegally) cross the border. However, our government is afraid to tackle illegal immigration properly, because of groups like the ACLU that sue whenever someone tries to deal with illegal immigration. Hell, our own government turned on Arizona when it tried to implement a nigh-identical immigration policy to what the Federal government has.2) Deal with gang violence. Most of the violent crime in America is tied to gangs, which predominantly occupy inner cities. However, most of these gangs are of black and Hispanic composition, so if a police department begins to actually make progress in taking down a gang it's not long before the ACLU steps in and claims it is racial profiling and undoes everything the police department managed to achieve via legal action.3) Loosen gun control laws for law abiding citizens. Many politicians seem to think that stricter gun control laws will reduce violent crime, yet they don't seem to realize that the only people following these laws are the law abiding citizens. The gang members and criminals aren't going to give up their guns because of stricter gun control laws. It's also been a proven fact that the areas with the strictest gun control laws have the highest amount of violent crime, while the areas with the most lax gun control laws have the lowest amount of violent crime. Your typical thug isn't going to risk his life for a wallet or purse if there is a chance the person he is going to rob is armed and could kill him in self defense. They like nice, easy, unarmed targets. And they'll always hit someone or someplace they know is unarmed rather than chance a hit on someone or someplace they know is armed or believe may be armed.

[WORDPRESS HASHCASH] The poster sent us '0 which is not a hashcash value.