Idle hands are the devil’s workshop

Professor Gordon Dahl (my colleague here at UCSD) and Berkeley Professor Stefano DellaVigna have an interesting new research paper titled Does Movie Violence Increase Violent Crime?

Dahl and DellaVigna note that there is substantial week-to-week variation in how many people in America are watching violent movies based on when particular movies get released. In recent years, the first weekend of a blockbuster new release can capture a huge audience share, as seen in the following graph taken from their paper:


There is also substantial week-to-week variation in the number of violent assaults in the U.S, inspiring Dahl and DellaVigna to look for a possible correlation. They found one, but perhaps not what you would expect– the greater the number of people in movie theaters on a given day watching violent movies, the lower are assault rates, both in the hours people may be in the theaters themselves (6:00 p.m. to midnight) and in the early morning following (midnight to 6:00 a.m.). This finding appears to be robust to the seasonal factors that influence the movie companies’ decisions about timing the release of a hoped-for blockbuster, which seasonal factors conceivably would also have some connections with assault rates for reasons unrelated to the effects of the movies themselves. Movies with the most violent content (8-10 on the rating scale) appeared to be most successful in suppressing violent crime.


The authors note that there is a large clinical literature that has arrived at the opposite (and more natural) conclusion that watching violent scenes can induce more violent behavior. But they argue that one factor left out of those clinical studies is the incapacitation effect– as long as you’re in the theater itself, you’re not mugging somebody. Even after you leave the theater, a pattern for the evening may have been set that involves less alcohol consumption and less likelihood of getting into trouble. The more violent the content of the movie, the authors reason, the more likely it is to bring violent people off the streets and into movie theaters.

The conclusion I take away from this is not that we want to encourage Hollywood to cram even more blood per minute into their releases, but rather that there may be substantial public benefits from helping make sure there are constructive activities for young people from at-risk populations to be involved with.

Come to think of it, that might not be such a bad strategy for Iraq as well.

Technorati Tags:

8 thoughts on “Idle hands are the devil’s workshop

  1. Dirk Nachbar

    Interesting research, but I guess there is a timinng problem. One could use a decaying function to account for the lasting effect of the movie on crime.

  2. DickF

    When I noticed the subject of your thread I anticipated the result. As is pointed out in your post violent entertainment removes violent people from society for a time reducing the incident of crime. But for a study of this type to be meaningful there would need to be a study done on the long-term correlation of increased violence in entertainment and the increased violence in society as a whole. For example the per capita murder rate doubled in this country between 1957 –when the FBI started keeping track of the data–and 1992. What was the correlation of violence in entertainment: movies, TV, video games?
    It is common knowledge that the military uses simulations and techniques to desensitize soldiers to killing. Note the following.
    During World War II, US Army Brig. Gen. S. L. A. Marshall had a team of researchers study what soldiers did in battle. For the first time in history, they asked individual soldiers what they did in battle. They discovered that only 15 to 20 percent of the individual riflemen could bring themselves to fire at an exposed enemy soldier.
    That is the reality of the battlefield. Only a small percentage of soldiers are able and willing to participate. Men are willing to die; they are willing to sacrifice themselves for their nation, but they are not willing to kill. It is a phenomenal insight into human nature, but when the military became aware of that, they systematically went about the process of trying to fix this “problem.” From the military perspective, a 15 percent firing rate among riflemen is like a 15 percent literacy rate among librarians. And fix it the military did. By the Korean War, around 55 percent of the soldiers were willing to fire to kill. And by Vietnam, the rate rose to over 90 percent.
    See other articles on the same site.
    Generally, I doubt that such entertainment causes someone to kill or engage in violent activities if they have no natural inclination. But there is little doubt that such entertainment feeds the desires of those who have an inclination to violence. The question is do we believe in the effectiveness of education and advertising to mold behavior. If we do then there can be no doubt that violent entertainment is effective in molding behavior.

  3. JDH

    UPDATE: Mark Thoma calls my attention to speculation that a broad trend of improving adolescent security might be related to the increase in violent video games.

  4. David

    Like Dirk and DickF, I wonder if there is some delayed effect not being measured here. It would be interesting to see what they would find if they ran a VAR or some other time series model on the data.

  5. Just Wondering

    I apologize for posting an off topic comment, but I have learned a lot from reading this blog over the past few years, and was hoping that the good Dr. Hamilton may consider a future post about this topic. I have learned that inflation is directly related to money supply (especially the M2 money supply). The following quote “The Russian central bank prints massive amounts of rubles each year, in exchange for Euros and US dollars that are flooding into the country. The central bank said its M2 money supply grew by 52.7% in the 12-months thru April 1st. ” was found by reading this link (
    My question is, would it be possible for Russia to indirectly create worldwide hyperinflation by artificially inflating its own money supply (the ruble) while simultaneously demanding payment in rubles for its oil if in fact oil becomes more scarce in the future?
    Just wondering.

  6. Alex Khenkin

    My question is, would it be possible for Russia to indirectly create worldwide hyperinflation by artificially inflating its own money supply (the ruble) while simultaneously demanding payment in rubles for its oil if in fact oil becomes more scarce in the future?
    – No, it wouldn’t.
    Small Investor Chronicles

  7. trampjuicerocks

    Surely the other way around. A violent society creates violent movies.
    1960s – New York Movies – Sophisticated comedies.
    The Oddcouple, Breakfast at Tiffany’s…
    Then the 1965 (MASS) immigration act resulting in millions of unskilled immigrants turning many cities into criminal hellholes.
    1970s/80s – Deathwish, Serpico, Assault on Precinct 13, Meanstreets etc..

Comments are closed.