互联网对中国城市内部就业分布的动态影响

This paper aims to reveal the net effect and its dynamic changes of the Internet on intra-urban employment distribution. Firstly, we incorporate the relative changes of the Internet's dispersion and agglomeration force into the analytical framework of the dynamic impact of the Internet on intra-urban employment distribution, and propose corresponding theoretical hypothesis. Secondly, we use the micro-data of 22.47 million enterprises in 289 prefecture-level cities in China's Economic Census in 2004, 2008 and 2013, and take postal districts as the basic spatial unit to construct the employment agglomeration index. To alleviate endogeneity, we construct instrumental variables with the "8-horizontal and 8-longitudinal" optical cable trunk network and the functional form of the exogenous logistic curve, and re-estimate the results. The results show that, on average, the Internet can significantly promote intra-urban employment agglomeration, but with the increase of Internet penetration rate, its promotion effect is characterized as "marginal diminishing", while the spatial multi-centricity of urban employment continues to increase. For cities with a high economic development level, the increase in Internet penetration rate promotes temporary dispersion of intra-urban employment and then "marginal increasing" agglomeration. Thus, this paper concludes that along with the continuous penetration of the Internet and the development of cities, the increase of the agglomeration force of the Internet will exceed the increase of the dispersion force, thereby reinforcing the city's strengths as a center of production and exchange to be a facilitator of urban employment agglomeration, while driving the evolution of urban employment towards a polycentric structure. This paper contributes to the existing literature by revealing, for the first time, the changing characteristics of the Internet's impact on the spatial structure of urban employment with the continuous penetration of the Internet, as well as the heterogeneity of its manifestations in cities at different levels of development, and offers a new explanation for the inability of existing studies to reach consistent conclusions on the diversified impact of the Internet with a rigorous analytical paradigm and a dynamic perspective. The findings of this paper are of great significance for predicting the spatial distribution of economic activities in future cities and guiding spatial planning.