Abstract for: Comparing Behavioral Dynamics Across Models: the Case of Copper

In many public policy issues diverging understandings of the system can be encountered. These diverging understandings can reside in the mental models of the different actors involved, or even be codified into structurally different models of the system. For an analyst it can be of great use to get insight into how and under what conditions the behavior of the models is different. In this paper, we address this problem. We present a general approach for comparing two or more structurally different models in the presence of additional uncertainties. This approach can be used to get insight into how different the results of two or more models are, and the conditions under which the models produce different results. The approach uses Exploratory System Dynamics Modeling in combination with the dynamic pattern feature metric. We demonstrate the approach using a case study. This case study focuses on the future dynamics of the copper system. Here, there are experts favoring a top down way of modeling the system and there are experts favoring a bottom up way of modeling the system. We use both and find that for some outcomes of interest adopting either perspective makes no difference, while for other outcomes of interest only under specific additional assumptions about other uncertainties there is a difference in behavior.