An agent-based retail location model on a supply chain network

Clusters of business locations, which considerably impact daily activities, have been prominent phenomena. Yet the question of how and why Þrms cluster in certain areas has not been sufficiently studied. This paper investigates the emergence of clusters of business locations on a supply chain network comprised of suppliers, retailers, and, consumers. Krugman (1996) argued that urban concentration involved a tension between the centripetal and the centrifugal forces. Based on that notion, this research proposes an agent-based model of retailers' location choice in a market of homogeneous products. In this game, retailers endeavor to maximize their proÞts by changing locations. Retailers' distribution patterns are measured by entropy and cluster density. Simulation results reveal that as more retailers engage in the game, clusters autonomously emerge and the entropy of clusters increases. Once retailers exceed a certain number, average density of clusters begins to decline; all discrete clusters gradually merge to a large cluster, spreading out uniformly. This research thus Þnds that the centripetal force attracts retailers to supplier locations; with even more retailers entering the market, the centrifugal force disperses them. The sensitivity results on model parameters and consumers' demand elasticity are also discussed.