City Size,Industrial Agglomeration Pattern Selection,and Urban Economic Efficiency
Does the continued expansion of large cities indeed hinder the improvement of urban economic effi-ciency?Previous literature suggests that at the early stages,urban economic efficiency increases with the expansion of city size.However,after surpassing a certain threshold(referred to as the optimal city size),urban economic efficiency begins to decline with the further increase in city size,indicating an"inverted U"shaped relationship between city size and urban economic efficiency.However,this conclusion that applies to mature cities'development in Western countries may not be consistent with the development of Chinese cities.The economic development of large cities in China seems to further increase with the expansion of city size.Taking into consideration the strong drive for urbanization in China,what kind of industrial agglomeration patterns should China's cities of different sizes select?Does the paradigm of"large and diversified,small and specialized"still apply?To explore these questions,this paper first reviews the viewpoints in previous literature regarding the agglomeration effects and congestion effects of the expansion of city size.It then proposes that"at the current stage,the marginal ben-efits of increasing city size in China may not have begun to decrease and the marginal costs may not have begun to in-crease",and provides explanations from aspects of public service supply capacity,potential for the development of the digital economy,and industrial structure.Finally,this paper constructs a balanced panel dataset of 281 cities at or above the prefecture level in China from the China City Statistical Yearbook from 2003 to 2019 for empirical testing.The theo-retical analysis of this paper demonstrates the nonlinear relationship between city size and urban economic efficiency,with different industrial agglomeration patterns exerting moderating effects.Therefore,to address the problem of omitted vari-able bias,this paper constructs an econometric model with quadratic terms and multiple interaction terms.The research results are as follows.Firstly,the impact of city size on urban economic efficiency generally presents a"U"shaped rela-tionship,with the turning point where industrial specialized agglomeration and industrial diversified agglomeration are at their average level and the negative impact turns positive occurring at a city size of 1.74 million people.Secondly,the in-crease in industrial specialized agglomeration strengthens the"U"shaped relationship between size and efficiency.When the city size is less than 1.29 million people,the inhibitory effect of size on efficiency is enhanced.When the city size is greater than 1.29 million people but less than 2.63 million people,the inhibitory effect is weakened,and when the city size is greater than 2.63 million people,the promoting effect is strengthened.The increase in industrial diversified agglom-eration weakens the"U"shaped relationship between size and efficiency.When the city size is less than 2.37 million people,the inhibitory effect of size on efficiency is weakened.When the city size is greater than 2.37 million people but less than 2.63 million people,the inhibitory effect is strengthened,and when the city size is greater than 2.63 million people,the promoting effect is weakened.Thirdly,the direct impacts of specialized agglomeration and diversified ag-glomeration on urban economic efficiency are respectively represented by quadratic functions with an upward or down-ward opening.The above conclusions remain valid after robustness tests,including replacing the explained variables,replacing the explanatory variables,and eliminating the impact of special samples.Additionally,considering the potential endogeneity problem in the model,this paper first constructs the Bartik instrumental variables for city size,specialized agglomeration,and diversified agglomeration using shift-share design,then uses river density and surface roughness as additional instru-mental variables for city size and specialized agglomeration respectively,and finally solves the endogeneity problem through two-stage least squares method,with results consistent with those of the benchmark regression.The research conclusions provide empirical evidence and policy implications for how to more rationally allocate population resources,leverage industrial agglomeration advantages,and improve urban economic efficiency.
City SizeIndustrial Specialized AgglomerationIndustrial Diversified AgglomerationUrban eco-nomic Efficiency
万方数据
维普
