Cooperation induced by random sequential exclusion

¹ Center for Systems and Control, College of Engineering, Peking University - Beijing 100871, China
² Center for Intelligent and Networked Systems, Department of Automation and TNList, Tsinghua University Beijing 100084, China

^(a) longwang@pku.edu.cn

Received: 7 May 2016
Accepted: 6 June 2016

Abstract

Social exclusion is a common and powerful tool to penalize deviators in human societies, and thus to effectively elevate collaborative efforts. Current models on the evolution of exclusion behaviors mostly assume that each peer excluder independently makes the decision to expel the defectors, but has no idea what others in the group would do or how the actual punishment effect will be. Thus, a more realistic model, random sequential exclusion, is proposed. In this mechanism, each excluder has to pay an extra scheduling cost and then all the excluders are arranged in a random order to implement the exclusion actions. If one free rider has already been excluded by an excluder, the remaining excluders will not participate in expelling this defector. We find that this mechanism can help stabilize cooperation under more unfavorable conditions than the normal peer exclusion can do, either in well-mixed population or on social networks. However, too large a scheduling cost may undermine the advantage of this mechanism. Our work validates the fact that collaborative practice among punishers plays an important role in further boosting cooperation.