Air transport networks and urban economic activity: Dynamic relationships across major global airports

This study examines the dynamic relationships between air transport network structure and urban economic activity across the world’s top 50 airports by passenger volume in 2024, using monthly data from 2014 to 2024. As the aviation indicator, we adopt the Global Airport Connectivity Index (GACI), a composite measure of network measurements; as the economic indicator, we use Visible Infrared Imaging Radiometer Suite (VIIRS) Nighttime Lights (NTL), providing a globally consistent, high-frequency proxy for urban economic activity. Short-run causality is assessed via the Toda–Yamamoto procedure, and long-run equilibrium relationships are tested using the Autoregressive Distributed Lag (ARDL) bounds testing approach. The Toda–Yamamoto test reveals that approximately 90% of airports exhibit no statistically significant short-run Granger causality in either direction, indicating that monthly network changes and economic fluctuations do not move in lockstep. In contrast, the ARDL bounds test detects long-run cointegration at 45 airports. The predominant pattern is aviation-led at 33 airports, with bidirectional relationships at 11 airports and economy-led at only one. Among the 11 bidirectional airports, the speed-of-adjustment analysis reveals that NTL adjusts toward equilibrium faster than GACI in every case, reflecting the structural rigidity of network reconfigurations relative to economic responses. These findings demonstrate that changes in air transport network structure predominantly precede adjustments in urban economic activity in the long run, offering policy implications for airport infrastructure investment.