Abstract
Humans are outstanding in their ability to cooperate with unrelated individuals, and punishment - paying a cost to harm others - is thought to be a key supporting mechanism. According to this view, cooperators punish defectors, who respond by behaving more cooperatively in future interactions. However, a synthesis of the evidence from laboratory and real-world settings casts serious doubts on the assumption that the sole function of punishment is to convert cheating individuals into cooperators. Instead, punishment often prompts retaliation and punishment decisions frequently stem from competitive, rather than deterrent motives. Punishment decisions often reflect the desire to equalise or elevate payoffs relative to targets, rather than the desire to enact revenge for harm received or to deter cheats from reoffending in future. We therefore suggest that punishment also serves a competitive function, where what looks like spiteful behaviour actually allows punishers to equalise or elevate their own payoffs and/or status relative to targets independently of any change in the target's behaviour. Institutions that reduce or remove the possibility that punishers are motivated by relative payoff or status concerns might offer a way to harness these competitive motives and render punishment more effective at restoring cooperation.
Highlights
In what has become a classic study, Clutton-Brock and Parker (1995) defined punishment as occurring when: (a) one individual cheats by performing an action that lowers the partner’s payoff, relative to alternative actions; (b) the harmed individual incurs a temporary payoff reduction to reduce the payoffs of the cheating individual; (c) the target behaves more cooperatively in future interactions with the punisher and causes the punisher’s payoffs to increase
While the focus in this paper is on punishment in humans, it will be helpful to keep these examples in mind, as we shall return to the ways that punishment in humans might differ from what we see in other species in our concluding remarks
In laboratory games, humans willingly invest in punishment in scenarios where the act is not observed by others, where the punisher will not interact with the target again in future, and/or where the benefits will be shared by other individuals, including non-punishers
Summary
In what has become a classic study, Clutton-Brock and Parker (1995) defined punishment as occurring when: (a) one individual cheats by performing an action that lowers the partner’s payoff, relative to alternative actions; (b) the harmed individual incurs a temporary payoff reduction to reduce the payoffs of the cheating individual; (c) the target behaves more cooperatively in future interactions with the punisher and causes the punisher’s payoffs to increase. These classic studies showed that people will pay for a sanctioning mechanism (Yamagishi 1986) or will take it upon themselves to punish co-players who do not contribute to a public good (Fehr and Gächter 2000, 2002a), even though the costs of punishment are personalised and the benefits of punishment (in terms of increased contributions to the public good) are shared among punishers and non-punishers alike The findings of these studies lent credence to theoretical arguments that punishment can (sometimes) be viewed as a second-order public good (Boyd et al 2003; Boyd and Richerson 1992), subsequent work suggested that punishment might more often fit the payoffs of the volunteer’s dilemma (Raihani and Bshary 2011), where investments can be self-serving rather than altruistic. We discuss the different empirical predictions that can be made based on different functional accounts of punishment
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
