Simulating the Ecology of Oligopolistic Competition with Genetic Algorithms

Abstract.

In economics, the n-person oligopoly game and the n-person IPD game are often considered close in spirit. Our analytical framework shows that this is not the case owing to the path dependence of the pay-off matrix of the oligopoly game. By simulating the evolution of a three-person oligopoly game with genetic algorithms, we explore the significance of the path dependence property to the rich ecology of oligopoly. The emergent behavior of oligopolists in the simulations indicates how the path dependence nature may shed light on the phenotypes and genotypes coming into existence. The features shown in this research can be further exploited in more practical contexts so that nontrivial policy issues in industrial economics can be seriously tackled.