Effects of violent media content: Evidence from the rise of the UFC

We document the effect of violent media on crime. Specifically, we evaluate the effects of The Ultimate Fighter, a hit TV show that features fighters competing in violent mixed martial arts and which brought Ultimate Fighting Championship into the mainstream. We estimate the effect of exposure to the show's earliest episodes using panel data from police agencies across the United States and a strategy that uses network ratings prior to the show's premier as an instrumental variable. We show that this exposure significantly reduced crime: these effects are particularly evident for assault, began in the month the show premiered, and persisted for many years. These estimates do not reflect systematic differences across geographic areas in their trends in crime rates prior to 2005. To complement our main results, we also investigate the effects of "UFC Main Events," which air in bars and on Pay-Per-View. This analysis additionally suggests reductions in violence caused by viewership.

Legal access to alcohol and criminality

Previous research has found strong evidence that legal access to alcohol is associated with sizable increases in criminality. We revisit this relationship using the census of judicial records on criminal charges filed in Oregon Courts, the ability to separately track crimes involving firearms, and to track individuals over time. We find that crime increases at age 21, with increases mostly due to assaults that lack premeditation, and alcohol-related nuisance crimes. We find no evident increases in rape or robbery. Among those with no prior criminal records, increases in crime are 50% larger-still larger for the most socially costly crimes of assault and drunk driving.

Gender and the influence of peer alcohol consumption on adolescent sexual activity

I consider the alcohol consumption of opposite-gender peers as explanatory to adolescent sexual intercourse and demonstrate that female sexual activity is higher where there is higher alcohol consumption among male peers. This relationship is robust to school fixed effects, cannot be explained by broader cohort effects or general antisocial behaviors in male peer groups, and is distinctly different from any influence of the alcohol consumption of female peers which is shown to have no influence on female sexual activity. There is no evidence that male sexual activity responds to female peer alcohol consumption.

Saving babies? Revisiting the effect of very low birth weight classification

We reconsider the effect of very low birth weight classification on infant mortality. We demonstrate that the estimates are highly sensitive to the exclusion of observations in the immediate vicinity of the 1,500-g threshold, weakening the confidence in the results originally reported in Almond, Doyle, Kowalski, and Williams (2010).