Unveiling the dynamics and determinants of world city network: Insights from global airline data

Existing research on world city networks (WCNs) has predominantly focused on macro-structure analysis, leaving a gap in understanding the dynamic changes from a multi-scale perspective and underlying factors shaping WCNs. Therefore, this study utilizes airline data to explore WCN dynamics at both macroscale and mesoscale, and employ a weighted exponential random graph model to investigate determinants of WCN formation. The research findings reveal that WCN displays hierarchical and regional tendencies. At the macroscale, the evolution of urban hierarchy demonstrates both stability and change, and at the mesoscale, WCN morphology follows a core-periphery spatial arrangement with strong regional patterns. ERGM results highlight the influence of both endogenous network structures and exogenous city characteristics on WCN formation. Preferential attachment emerges as a crucial driving force, with disassortativity facilitating connectivity between peripheral and core cities. Additionally, GDP, language, institutional proximity and path dependence positively contribute to WCN formation, while population and distance have negative effects.