On Monday, the United States District Court for the District
of Minnesota struck down the Minnesota Restricted Video Games Act. See Entertainment Software Assoc. v. Hatch, No. 06-CV-2268 (D. Minn.
2006). The Act prohibited minors [below 17] from knowingly renting or purchasing a video
game carrying an ESRB rating of AO (Adults Only) or M (Mature). As of this
moment, only 23 titles carry an AO rating (including only one game of any note,
Grand Theft Auto: San Andreas),
but 989 titles carry an M rating (including my own current favorite, Splinter Cell: Chaos Theory). The
penalty imposed on a minor for violating the Act was a $25 fine. I'll focus on the use of the empirical evidence in this case.
The state relied primarily on a meta-analysis by Professor
Craig Anderson of
First, as more studies of violent video games have been conducted, the significance of violent video game effects on key aggression and helping-related variables has become clearer. Second, the claim (or worry) that poor methodological characteristics of some studies has led to a false, inflated conclusion about violent video game effects is simply wrong. Third, video game studies with better methods typically yield bigger effects, suggesting that heightened concern about deleterious effects of exposure to violent video games is warranted.
The magnitude of these effects is also somewhat alarming. The best estimate of the effect size of exposure to violent video games on aggressive behavior is about 0.26[.]
The court did not mention Anderson’s findings in any detail, but it did find his article “completely insufficient to demonstrate an empirical, causal link between video games and violence in minors.” (p. 6) The Eighth Circuit actually called for demonstrating a link between violent video games and psychological harm to minors, not violence in minors. Whether or not the court meant something different than the Eighth Circuit in its characterization of the issue, it found Anderson’s methodology unpersuasive:
Even assuming the methodology employed by Dr. Anderson to be correct, (n.1) Dr. Anderson’s meta-analysis is far too slight to bear the weight of the State’s argument.
(n.1) Dr. Anderson's meta-analysis seems to suggest that one can take a number of studies, each of which he admits do not prove the proposition in question, and "stack them up" until a collective proof emerges. It is fair to say that his article does not, on its face, demonstrate the validity of this thesis. In making this observation, the Court sees no present need to undertake a Daubert analysis concerning the article's admissibility -- especially when the article itself identifies empirical flaws which keep it from actually supporting the State’s purported interests. See Daubert v. Merrell Dow Pharm., Inc., 509
U.S. 579 (1993).
Entertainment Software, slip op. at 6-7.
The court clearly doubted the value of using a meta-analysis.
The court even sounds dismissive of
The court’s view, however, is out of sync with the more complicated world of empirical research. A single study that provides the definitive answer to some question is rare, especially in the social sciences. It’s more common to find a pile of useful studies, some of which are, much to everyone’s frustration, contradictory. When there are not even enough studies to make a pile, one must be cautious about relying on what is available. One point of a meta-analysis is to avoid putting undue weight on any one study. We usually want a variety of people researching a question, using different samples and different methodologies. We want multiple angles on the problem -- and then we want a sensible way to assess the collective results. A meta-analysis is a widely accepted means of making this assessment. See generally Mark W. Lipsey and David B. Wilson, Practical Meta-Analysis (Sage 2000); Morton Hunt, How Science Takes Stock: The Story of Meta-Analysis (Russell Sage 1997).
On the topic of video game violence, there is a small pile
The court offered several other objections to the study, one of which was problematic. It was that “the body of violent video game
literature is not sufficiently large to conduct a detailed meta-analysis of a
specific feature.” Entertainment Software,
slip op. at 6. What does this mean? A specific feature of what? Based on the
court’s discussion, it’s not at all clear. You have to look at
Despite my concerns about how the empirical evidence was used
in this case, I still think the court’s result was correct.
The court also said it could not tell “whether violent video
games cause violence, or whether violent individuals are attracted to violent
video games.” Entertainment Software,
slip op. at 7. Fair point. We need to know more about the experimental studies
in the literature. For example, did the researchers recruit participants
without regard to their pre-existing interest in video games? Notably, the
state conceded it could not show any sort of causal link between violent video games and harm, broadly defined. Id. at 7. But there are some relevant
experimental studies --
Based on the court’s discussion, it looks like the state failed to provide some important information. Of course, even if the state provided all of the necessary information to fully understand what the empirical evidences shows, a very tough issue remains: how much evidence of how much of an effect is sufficient? Understanding the empirical evidence does not answer this question.
*There was also an issue in this case about an improper delegation
of authority to ESRB, which provided an independent ground to strike down the
**Hat tip to William Childs at TortsProf for posting the full text of the opinion and for calling attention to the court’s discussion of meta-analysis.