中国市域间日常人口流动特征及影响因素

基于腾讯位置大数据，分析了2015—2018年中国368个城市间人口流动的空间格局，并基于指数随机图模型（ERGM）识别了与人口流入、流出相关的影响因素。研究发现：① 人口流动的空间分布格局相对稳定，形成了以京、深、沪、穗、蓉、莞为“中枢”的菱形空间结构。② 通过社区划分得到的城市子群结构表明，社区间呈现明显的地理临近和省际分异特征，既形成了以省会城市为核心、受省界制约明显的中心–腹地结构的小型城市子群，也形成了跨越省界的多中心结构的大型城市子群，但大部分城市以省界为主要流动圈层，省域内人口流动更为密切。③ ERGM模型确定的人口流入、流出网络影响因素与新古典经济学理论相一致，人口规模、城市化水平、时间成本、经济成本等市场因素和经济因素在人口流动中仍然具有主导性作用。④ 城市对外来人口的吸引力更大程度上取决于自身的属性特征，而城市人口外流更依赖于外部网络关联要素，一定程度上验证了推拉理论中城市“拉力”的主导力量，以及各类距离因素的综合作用。

Characteristics and Influencing Factors of Daily Population Flow Among Cities in China

Using Tencent location big data, this study analyzes the spatial pattern of population flow among 368 cities in China and identifies the influencing factors related to population inflow and outflow based on an exponential random graph model (ERGM). 1) From 2015 to 2018, the spatial distribution pattern of population flow was relatively stable, forming a rhombic spatial structure with Beijing, Shenzhen, Shanghai, Guangzhou, Chengdu, and Dongguan as the ‘center’. The densely populated nodes and channels are mainly concentrated to the east of the Hu Huanyong Line. The significance of this study lies in further determining the core cities and main pillars in the population flow network. 2) The urban subgroup structure obtained by community division shows obvious geographical proximity and inter-provincial differentiation among communities, which form not only small urban subgroups with the provincial capital city forming the core and bordered by the provincial boundary, but also large urban subgroups with a multi-center structure spanning provincial administrative boundaries. However, for most cities, the provincial boundary delimits the main flow circle, and population flow within the same province is more frequent. 3) The influencing factors of the population inflow and outflow networks determined by the ERGM model are consistent with the predictions of neoclassical economics. Market and economic factors such as population scale, urbanization level, time cost, and economic cost still play a leading role in population flow. 4) The attraction of a city to the floating population depends on its individual attributes, while the urban population outflow depends more on the external-network-related elements of the city. To a certain extent, this study verifies the predominance of the urban “pull” in the push-pull theory and the comprehensive effect of various distance factors.