Agent-based and system dynamics modeling: a call for cross study and joint research

Abstract

In recent years, the "Science of Complexity", as promoted by the Santa Fe institute, has been recognized by mainstream scholars in prominent scientific disciplines ranging from physics over economics and computer science to the social sciences. In various reviews, it appears, however, as if Complexity Theory and techniques such as agent-based modeling are unique in their capacity of modeling nonlinear systems. These reviews overlook that such systems have been modeled and simulated at least since the late 1950s, e.g., by the research track known as System Dynamics, a discipline with a rich body of literature. This paper gives an overview of the general modeling principles of both tracks, describes their areas of applicability, and discusses their relative strengths and weaknesses. It tries to identify areas in which the two modeling traditions complement each other, and where they overlap. The paper concludes that cross study and joint research are overdue.