Understanding the structural evolution and driving mechanisms of urban network using firm-level big data and TERGM modeling

Abstract

As regional relationship studies shift from the central place paradigm to a network paradigm, understanding the structural patterns and evolutionary mechanisms of urban network is crucial for network-oriented regional planning. This study leverages firm-level big data, comprising 105,123 headquarters and 253,535 branches, and applies the Temporal Exponential Random Graph Model to analyze the structural evolution and driving mechanisms of the urban network in the Yangtze River Delta (YRD). From 1995 to 2020, the YRD exhibited a clear trend toward polycentric and networked development, accompanied by significant regional disparities. The YRD urban network is characterized as a scale-free network, exhibitting distinct hierarchical patterns and a tendency for preferential attachment. The network comprising both horizontal connections between cities of the same tier and vertical connections between cities of different tiers and higher-tier cities show stronger enterprise connections, while lower-tier cities prioritize linking with higher-tier cities over those of the same or lower tiers. The evolution of the urban network in the YRD is driven by mechanisms such as size-based agglomeration and dispersion effects, temporal dependence, network self-organization, preferential attachment, assortative mechanism, and multi-dimensional proximities. This study enhances our theoretical understanding of the structural patterns and evolutionary mechanisms of urban network.