Third-party punishers do not compete to be chosen as partners in an experimental game.

Tommaso Batistoni,Nichola J Raihani,Pat Barclay

doi:10.1098/rspb.2021.1773

Abstract

Third-party punishment is thought to act as an honest signal of cooperative intent and such signals might escalate when competing to be chosen as a partner. Here, we investigate whether partner choice competition prompts escalating investment in third-party punishment. We also consider the case of signalling via helpful acts to provide a direct test of the relative strength of the two types of signals. Individuals invested more in third-party helping than third-party punishment and invested more in both signals when observed compared to when investments would be unseen. We found no clear effect of partner choice (over and above mere observation) on investments in either punishment or helping. Third-parties who invested more than a partner were preferentially chosen for a subsequent Trust Game although the preference to interact with the higher investor was more pronounced in the help than in the punishment condition. Third-parties who invested more were entrusted with more money and investments in third-party punishment or helping reliably signalled trustworthiness. Individuals who did not invest in third-party helping were more likely to be untrustworthy than those who did not invest in third-party punishment. This supports the conception of punishment as a more ambiguous signal of cooperative intent compared to help.

Highlights

Punishment refers to the act of paying a cost to inflict a reciprocal cost on a social partner [1] and it has been proposed as a key factor supporting the evolution of cooperation among non-relatives [2,3,4,5,6,7]
We were interested in the signalling value of punishment compared to helping behaviour: whether investments in punishment are less prone to escalate when there is competition to be chosen as a partner; and whether punishment investments are viewed by observers as less reliable indicators of an individual’s cooperative intent
We found that investments in both punishment and helping behaviour were higher when these would be observed by another individual, this effect was stronger for helping than for punishment and individuals generally invested more in helpful than in punitive behaviour

Summary

Introduction

Punishment refers to the act of paying a cost to inflict a reciprocal cost on a social partner [1] and it has been proposed as a key factor supporting the evolution of cooperation among non-relatives [2,3,4,5,6,7]. One route to obtaining return benefits from punishment is if the target of punishment behaves more cooperatively in future interactions with the punisher (as originally suggested by [1]). Such outcomes seldom seem to occur in experimental settings (reviewed in [7]). [9,10]) have shown that people often punish in situations where they act as third-parties, who were not the primary victim of the cheat and do not expect to interact with either the victim or the cheater in future interactions Both laboratory (e.g. [8]) and field experiments (e.g. [9,10]) have shown that people often punish in situations where they act as third-parties, who were not the primary victim of the cheat and do not expect to interact with either the victim or the cheater in future interactions

Objectives

Results

Conclusion