An Agent-Based Model of Trade with Distance-Based Transaction Cost

This paper describes an application of agent-based modeling to investigate the effect of a distance-based transaction cost on trade. Long-distance trade is rapidly increasing, but may ultimately be constrained by our ability to move material goods between sellers and buyers. Unlike information exchange, trade in material goods is dependent on the price of oil and vulnerable to future scarcities of oil. In addition, there are growing concerns about greenhouse gas emissions from long-distance transportation. Our purpose in this study is to take the first step in understanding the impact of a distance constraint on free global trade using a simple artificial economy. We use the perspective of agent-based computational economics to model two different scenarios of random initial allocations of goods among traders, and investigate the response of the economy as a distance-based transaction cost is applied. We show that a geographically skewed initial allocation of goods performs poorly, while a more uniform initial distribution responds in a highly resilient way as the transaction cost is varied. Underlying this resilience is the emergence of a stable trade network that has some of the properties of scale-free networks.