Abstract for: Measuring Dynamic Complexity: A Standardized Measure for System Dynamics Simulation Models

Dynamic complexity is an important aspect of the realm of management. It is present, i.e., when an action has one set of consequences locally and a different set of consequences in another part of a system. Dynamic complexity is best captured and analyzed by ordinary differential equation (ODE) models of the system dynamics type. Measuring an ODE model’s degree of dynamic complexity approximates the dynamic complexity of a real system. However, no approach exists for measuring this concept. Here, the paper contributes with three measures. These measures yield the following benefits: First, one can inspect the degree to which a model is able to endogenously generate dynamic behavior and therefore capture the degree of dynamic complexity. Second, one can evaluate and compare the property of a quantitative model to endogenously generate its behavior. This feature enables researchers to filter out more parsimonious models. Third, these measures can be used to enhance the validity of simulation models, e.g., by reporting it in publications. And finally, they can enliven discussions about dynamic complexity. The paper develops these measures by means of a cascade of examples. The conclusion discusses limitations as well as future research.