“Is It Only a Game? Video Games and Violence”, 2021-08-01 (; similar):
Popular media often links violent video games to real-life violence, although there is limited evidence to support this link.
I analyze how adolescent boys’ violent behavior is affected by the releases of new violent video games in the US Variation in children’s exposure to the releases comes from variation in video game release and interview dates and thus is plausibly exogenous.
I find no evidence that child reported violence against other people increases after a new violent video game is released.
Thus, policies that place restrictions on video game sales to minors are unlikely to reduce violence.
…The measures of violent behavior are obtained from the Child Development Supplement (CDS) to the Panel Study of Income Dynamics (PSID). I examine the effect of violent video games on 2 types of violence—aggression against other people and destruction of things/property. Violence measures are obtained from children themselves and their parents. The main sample is restricted to boys aged 8–18 years, a subgroup of children who are most likely to play violent video games. Data on the release dates of violent video games have been collected specifically for this study from an online video game database MobyGames. In addition, video games sales data from the VGChartz database are used to identify the most popular violent video games, which are expected to increase children’s video game hours to a sufficient extent so that changes in violence levels could be detected in the data.
…I first use the exposure to the releases of popular violent video games to predict 8–18 year old boys’ daily video game hours in the 6 months before the survey. I find that boys’ weekday video game playing increases by 15–20 min per day (32–39% with respect to the mean) following a release of a popular violent video game, but not immediately after the release. Statistically-significant effects are found 4 to 5 months after the release. These effects are more pronounced for older (12–18 year old) boys and regular video game players. There are no effects of violent video game releases on girls’ video game hours.
I then regress child and parent reported violence on the predicted video game hours. On the one hand, I find no evidence that a release of a popular violent video game increases violence, as self-reported by children themselves. On the other hand, parent reported destructive behavior is found to increase following a release of a popular violent video game in some subsamples of children. Taken together, these results suggest that all 3—direct, substitution, and selection—effects are important. An increase in parent reported destructive behavior following a release of violent video game shows that children may act aggressively after playing violent video games. The likelihood of violence against people, however, does not increase, which suggests that the substitution and selection effects outweigh the direct effect. The results are consistent with more violence-prone boys being attracted to violent video games and/or video games substituting for other violence related activities. The importance of the substitution effect is supported by the finding that children spend less time away from home after a popular violent video game is released.