Evaluating Agent-Based Modeling as a Tool for Economists

Abstract

Agent-Based Modeling has the capability to become a very powerful tool in economics, but in order to achieve this will require a great deal of effort on the part of its current practitioners. Economics at its core studies people and their actions. This is inherently difficult to do since the motivation of people is not always discernible. A method by which the variety of human behavior could be mimicked, thus providing insight into the process of decision-making that underlies the behaviors, would be invaluable to researchers. Agent-based modeling is such a method. In situations where analytical methods are not sufficient to answer the questions posed, agent-based modeling can be used to obtain solutions to interesting problems, and assess their robustness. However, until the majority of economists accept agent-based modeling as a legitimate tool, the adequacy of the results will remain in question. Certainly one area that will have to be addressed in order to expand the potential of agent-based modeling in economics is sheer exposure. However, more important to its acceptance will be advancements in the cohesiveness of the discipline. The purpose of this note is to present a series of question that must be answered before agent-based modeling can become accepted as a powerful economic tool.