Is Inequality Inevitable? An Agent-based Bargaining Model of Network Structure and Nominal Characteristic Tags

• In Event: 260 - Section on Mathematical Sociology Paper Session. Advances in Mathematical Sociology

Abstract

How does network structure affect the stability of inequality? Although the Nash bargaining game is well-studied in analytic game theory and agent-based models, no previous work has put bargaining games in networked populations with the potential for emergent agent classes. This present work, a marriage between literatures of evolutionary game theory and computational modeling, produces systematic stratification at the aggregate population level through the repeated, decentralized, pairwise interactions of agents with nominal characteristic tags. While equality (i.e. fair division) is the stochastically stable equilibrium in bargaining models, the specification of two agent types in this model underscores network structures sufficient to maintain states of inequality between groups on surprisingly long time scales. The operating mechanisms of neighborhood size and local clustering are unpacked, underscoring that this result of intergroup inequality is not dependent on strict assumptions of in-group favoritism, ruling out alternative explanations of homophily or density. Instead, this sufficiency claim relies systematically on the ability for agents to interact either in local communities or the population at large. This work concludes that states of inequality are much stickier and less escapable in tractable time scales than previously thought.

Authors

• Natalie Herbert, Annenberg School for Communication, University of Pennsylvania

• Soojong Kim, University of Pennsylvania

• Damon M. Centola, Annenberg School