On Tuesday, GayGamer reported on the news that gaming may, in fact, reduce violent crime. This idea, which started in a BBC article and was spread quickly around the internet, was presented as a just a quick blurb, but was expounded upon greatly. Thanks to the efforts of talk-backer Travis (whose comment has been, unfortunately, removed - I'm assuming most posts with links in them are automatically deleted by our ever-busy spam police), I've been able to review the original research suggesting that violent video games may actually help suppress violent crime. It's titled "Understanding the Effects of Violent Video Games on Violent Crime", and it presents a compelling and provocative heterodoxical take on the subject. The original research can be found here or here (or by Googling the title), and it's authors are A. Scott Cunningham of Baylor University in America, Benjamin Engelstätter from the Centre for European Economic Research (aka. the ZEW) in Germany, and Michael R. Ward of the University of Texas at Arlington.
The gist of the authors' argument is that almost all current research on aggression and video games lacks external validity. This is a statistical term meaning that the research doesn't test the effect it seeks to understand in real-world situation. Testing aggression related to video game violence by asking participants to play a game for half an hour, in a lab, and then rating their levels of aggression does not nearly resemble the argument that politicians (and some scientists) are making: That someone who plays too much Halo - for example - is going to then go and commit a violent crime. The authors argue that trying to link - for example - increased heart rates or increased competitiveness after having played a video game in artificial surroundings to the commission of violent offences is overreaching.
Rather, the authors argue that the time spent playing a violent video game leads to what is called an "incapacitation effect." To put it simply: Gamers are too busy gaming to commit violent crimes.
They're not arguing that violent media do not increase aggression. They're not arguing that people with violent tendencies aren't drawn to violent video games. They're arguing that games aid in preventing violent crimes by keeping people on the couch. And, they have numbers to back it up.
Adjusting for general seasonal trends in both gaming (the flurry of purchases around December) and crime (more crimes are committed in the warmer months), the researchers plotted US sales of violent games against US crime statistics. Allowing for the time that it may take to complete a game, the researchers found that sales of violent and nonviolent video games predicted a "small" reduction in the overall crime rate.
Only a "small" reduction? Regardless of its size, this finding is directly counter to the argument that violent video games increase crime rates. Even no increase in violent crime following the release of violent games would be evidence against that theory.
The article is not without its flaws. Measuring large trends against each other is notoriously messy, and the study is a secondary analysis of existing data - not a true experiment. It is also not clear to me whether it has been peer-reviewed. Nevertheless, the authors themselves point out that the intention was not to perform an experiment, but to examine the issue with the maximum possible external validity. Real-life data may be messy, but they have the greatest external validity possible. Superimposing this research on that which is already available in the field adds an interesting twist to the controversy - one that makes sense and deserves to be examined further.