The Co-Evolution of Networks and Capabilities in Innovation Systems: Principles for Systemic Policy Design

This paper develops a co-evolutionary network model to analyze how micro-level interactions among heterogeneous organizations generate macro-level structural patterns and performance outcomes in innovation systems. Organizations possess knowledge stocks, absorptive and distributive capacities, and adaptively rewire their connections. We integrate six key mechanisms---capacity-constrained knowledge flows, endogenous capacity accumulation, resource-based collaboration costs, innovation as a growth-structure interaction, strategic repositioning, and adaptive network rewiring---into a formal simulation framework. The model is calibrated using Approximate Bayesian Computation to match stylized facts from the innovation literature and employed in a structured scenario analysis spanning alternative policy-relevant regimes. Results reveal systematic trade-offs with important policy implications. Expanding connectivity without parallel capacity development yields limited gains; isolated capacity building amplifies inequality. In contrast, coordinated interventions targeting both network structure and organizational capabilities produce the most robust and equitable growth. Comparative analysis across four distinct economic environments demonstrates that intervention effectiveness is highly contingent on underlying frictions. The findings underscore the need for innovation policy that is explicitly network-aware and systemic, emphasizing bundled, context-sensitive interventions rather than isolated levers. The model provides a computational laboratory for exploring such policy design principles.