尺度重构视角下中国城市多元网络结构研究

Multi-Flow Network Characteristics of Cities in China under the Rescaling Perspective

采用2019年中国战略新兴产业100强企业网络数据和百度指数数据,运用链锁—隶属网络模型、社会网络分析法,采用"自上而下"的研究路径,从全国—城市群间—城市群内三个相互嵌套的空间尺度,解析尺度重构视角下的中国城市网络空间结构特征;应用随机森林和逐步回归法分析城市网络中心性的影响机制。研究发现:(1)信息流视角下,中国城市网络联系紧密,呈菱形空间结构;企业流视角下,网络联系较弱,呈放射状分布;二者均呈现"东密西疏"分散分布。(2)不同尺度的城市网络内部均存在小团体现象,核心城市的集聚力普遍大于其对外的辐射力;中国城市网络层级结构特征显著,择优链接效应明显;城市群间联系小团体特征明显,层级性显著;信息流视角下城市群间网络联系远强于企业流视角下城市群网络联系,京津冀城市群具有较强的控制力,长三角城市群具有较强的吸引力;城市群内部均存在小团体与互惠对特征,网络结构呈现单核心和双核心两种类型。(3)政治资源、知识资本、专业技术服务、人力资本是城市网络中心性的关键影响因素。

This article aims to analyze the multi-flow network structural characteristics of cities in China under the rescaling perspective. Based on the location data of headquarter and branch of Chinese Top 100 Strategic Emerging Industries in 2019,this article applies the interlocking-ownership network model to build the urban network and the Baidu index to obtain the 312×312 value city network,analyzes the spatial structure characteristics of urban network adopting "top-down" research path from the national-inter-urban agglomeration-urban agglomeration under the framework by using social network analysis method, and explores the influence mechanism of urban network centrality with random forest and stepwise regression. Three conclusions are drawn: 1) From the perspective of information flow, China urban network is closely connected, with a diamond-shaped spatial structure;From the perspective of enterprise flow, the network connection is weak and radially distributed. Both of them show "east close, west sparse" dispersed and uneven distribution,which it is higher in the eastern China than the western China. 2) There exists clique phenomenon in different scale urban network. The gathering force of the core city is generally larger than its external radiation force. The central hierarchical structure of Chinese urban networks is remarkable,and the preferential attachment effect is obvious.The network connection between urban agglomerations shows obvious small-group and hierarchy characteristics, the information flow between urban agglomerations is much stronger than the enterprise flow. It has strong control in the Beijing-Tianjin-Hebei urban agglomeration and has strong attraction in the Yangtze River Delta urban agglomeration. The characteristics of small-group and mutual-benefit exist in urban agglomeration. The network structure presents two types:single core and double core. 3) Political resources, knowledge capital, professional and technical services, and human capital are the key factors influencing the centrality of urban networks.