网络关联、空间溢出效应与中国区域经济增长——基于腾讯位置大数据的研究

以腾讯公司根据手机位置得到的中国地级及以上城市之间人口流动数据为基础,考察人口流动网络下地级市经济增长的空间关联模式及其演变,结合网络分析方法,测度地级市间经济增长的空间溢出效应。主要结论有：① 符合现实的空间权重矩阵应该是非对称的,且随着互动关系的改变而发生变化。② 人口流动网络等反应地级市之间互动的因素在经济增长空间溢出中发挥了重要作用。③ 基于网络的空间溢出效应是地级市经济增长的重要因素。

Network Association,Spillover Effect and China's Regional Economic Growth Based on Tencent's Location Big Data

Based on population mobility data between all prefecture-level cities in China acquired from Tencent mobile phone locations, this article examines the evolution of the spatial correlation model of China's prefecture-level economic growth under the population mobility network. And the spatial spillover effect of economic growth among these cities is also measured in this article by utilizing the network analysis method. The main conclusions obtained in this article are: 1) The construction of the spatial weight matrix should be based on the data that characterizes the interaction and degree between regions. The spatial weight matrix that conforms to reality should be asymmetric and change with the modifying of interaction. 2) Factors such as population mobility networks that interact with prefecture-level cities play an important role in the spillover of economic growth. China's prefecture-level cities show a more obvious ‘center-periphery’ structure. Five central cities: Beijing, Shanghai, Guangzhou, Shenzhen and Chengdu form five fulcrums from north to south and from east to west. The prefecture-level cities with which these five cities have population relationships are concentrated in the eastern and central regions in terms of spatial distribution, and slightly less in the west and northeast. Meanwhile, sub-central prefecture-level cities have formed a hexagonal population mobility network pattern from northeast to southwest. Changchun, Lanzhou, Hangzhou, Dongguan, Nanning and Kunming are vertices of this network. Population mobility is very frequent in this hexagon network and is relatively infrequent outside this hexagon. 3) Network growth effect is an important factor in the economic growth of prefecture-level cities. Using the population mobility network, the population factor has a negative network spillover effect, while foreign direct investment has a positive spillover effect on economic growth. The population factor also has a negative spillover effect on local economic growth, while the fixed asset investment and total retail sales of consumer goods have a significant positive spillover effect, if the geographical proximity network and geographical distance network are used. Finally, according to the research in this article, the policy recommendations are proposed to narrow the economic differences between prefecture-level cities and to achieve coordinated development.