This study analyzes deterrence schemes and their impact on stealing. The results confirm Becker’s deterrence hypothesis. Moreover, crowding out of pro-social behavior occurs due to deterrence incentives: when deterrence incentives first exist and are removed later on, subsequent behavior is more selfish than without this deterrence history. This study offers evidence that (part of this) crowding out takes place via change of emotions. Without deterrence incentives in place, in a variant of the dictator game, players with pro-social emotions steal less. When players face expected punishment pro-social emotions are deactivated and do not decrease stealing; in this case self-centered emotions get activated and motivate greater stealing. This study provides support for theories on emotions in behavioral criminal law and economics and offers new insights for deterrence policy.