Regulatory protective measures and risky behavior: Evidence from ice hockey

We provide evidence supporting the Peltzman effect, by which individuals required to wear protective gear end up taking additional risks that potentially offset the intended aim of the device. We take advantage of the fact that wearing a visor-a protective device in Ice Hockey-is mandatory in European, minor, and junior leagues but not in the NHL. This allows us to estimate the impact of wearing a visor by comparing the behavior in the NHL and other leagues of players who always wear a visor with that of players who wear one only when it is required. We find that when players are forced to wear a visor their behavior becomes more risky, earning an additional 0.19 penalty in minutes per game (compared to the average 1.14 penalty in minutes in our sample). We estimate an even larger effect of visors when we focus on players who were forced to use one during the 2004 season, when the NHL canceled its regular season and players had to move to European leagues temporarily. These estimates are not driven by players' observable attributes, playing style, or other differences across leagues.