Agent-Based Simulation in Path Dependence Research: A Network-Theoretic Model

I suggest a model to research standard diffusion and path dependence in networks of actors (e.g. organizations, organizational units) adopting technologies. The model includes several existing models as special cases. It should enable exploring path dependence in greater detail. The model distinguishes between three phases. First, it looks at standard diffusion in a static network. It afterwards simulates the – potentially path-dependent – growth process. Then path-breaking or -reinforcing interventions in the structures of the network take place. The model sets itself apart from other models by having different forms of network initialization and network growth strategies. First experiments show that I can replicate several existing path dependence models, e.g. Arthur’s canonical model, as a special case. This builds confidence in the models’ validity. Along the way, I constructed a flexible simulation laboratory ready-to-use for other path dependence researchers. Future directions are described.