Abstract
This paper explores whether, under what conditions, and why supervisors treat rule breakers favorably. Using field data from six seasons in the National Hockey League (NHL), we find that coaches select rule breakers for play more often than rule-abiding players. However, this preference is bounded in several ways. It reverses when a player’s rule breaking becomes extreme, holds for minor (but not major) rule breaking, is absent for both players and teams in the lowest quintile of rule breaking, is observed in regular season (but not playoff) games, and is amplified when teams are on losing streaks but disappears after repeated wins. An experiment replicates this bounded preference for rule breakers and identifies that those in a position to reward rule breakers do so in part because they perceive them as more committed to team success. However, at extreme levels of rule breaking, concerns about the liability rule breakers represent eclipse these positive perceptions, reversing the preferential treatment they enjoyed when their rule breaking was more moderate. Together these findings illuminate one unexpected reason why rule breaking is so rampant and unethical behavior remains so pernicious in organizations.