Crime, Punishment, and Recidivism

Abstract

Among the fundamental elements of the sport of ice hockey are the on-ice rules violations occasionally committed by players and the penalties assessed for those violations. During the 1998-99 season, the National Hockey League (NHL) for the first time experimented with the deployment of two on-ice referees for a selection of games instead of the customary single referee, significant in that only referees have the authority to call penalties. In this article, that experimental 1998-99 season provides the empirical setting for a test of the economic model of crime, which suggests that economic agents allocate time to legal and illegal activity by considering the benefits and costs of these activities. Here, those economic agents are NHL players. Empirically, relatively nonviolent illegal activity appears significantly influenced by benefits and costs, but particularly violent acts appear to occur more randomly. Particularly violent penalties increase when a second referee is deployed, suggesting a dominant “apprehension effect” rather than a dominant “deterrence effect” of what amounts to an increase in the presence of police.