Statistic configurations in the ERGM.

Based on the establishment data of provincial-provincial, city-city, provincial-city, city-provincial Hometown Chambers of Commerce (HCCs) in China by the end of 2022, this paper combines social network analysis and exponential random graph model to explore network spatial patterns and determinants of China’s HCCs. Findings indicate that: 1) Point degree centrality in eastern China is high, but spatial patterns vary across network types: provincial-provincial and provincial-city types exhibit a “rhombus-net” pattern, the city-city type follows a “small-ring-line” structure, and the city-provincial type forms a “large-ring-net” type. Guangdong, Fujian, Zhejiang, Jiangxi, and Shandong constitute core areas of China’s HCC network, and provincial-provincial and city-city types mainly flow within the core area and from core to periphery, while provincial-city and city-provincial types flow mutually between core and periphery areas. 2) Stronger government intervention and earlier HCC connections significantly inhibit both provincial-provincial and city-city HCC establishment. However, this inhibitory effect is not statistically significant for provincial-city and city-provincial types. Only city-city HCCs exhibit economic sensitivity, with less-developed origin cities favoring stronger operation cities, while other HCC types remain economically neutral. Road, dialect, and urban cluster distances significantly affect HCC establishment at the city-level place of operation but show no significant impact on the provincial-level place of operation. This research not only expands the theoretical perspective on the spatial study of social organizations but also provides scientific evidence for breaking down regional market fragmentation and optimizing cross-regional governance systems.