Spatial patterns and determinant factors of population flow networks in China: Analysis on Tencent Location Big Data

Abstract

Despite the rising interest in understanding the population flows across cities, few empirical studies have paid close attention to their network characteristics and determining factors. Using the Tencent Location Big Data, this paper first analyzes the spatial patterns of interregional population mobility among 328 Chinese cities and aims to identify the influential factors associated with population flows based on the exponential random graph model (ERGM). Supported by social network analysis method, we found that the spatial distribution patterns of population flows were relatively stable, showing a diamond shape with 7 central points, with the most attractive cities located in east developed urban agglomerations. Nevertheless, the proportion of cities with a positive population inflow in the west increased. The subgroup structure gained by Walktrap algorithm suggested that the intra-province flows comprised the majority. Furthermore, variants of ERGMs were used to determine geographic, demographic, economic, natural, and linguistic factors, as well as others related to the level of facilities. The results are in line with the neoclassical economics theory and the amenity-led theory, while amenities' factor played a more significant role. Our research can provide important insights for strengthening network connections among cities to achieve more balanced regional development.