现代服务业集聚形成机理空间计量分析

在纳入空间效应前提下,构建现代服务业集聚形成机理空间面板计量模型,对我国28个省域相关数据实证研究表明:我国现代服务业集聚在省域之间有较强的空间依赖性和正的空间溢出效应。技术差异在时间维度上对现代服务业集聚促进作用显著,在空间维度上并不显著;交易费用与现代服务业集聚有显著的负相关性;知识溢出、规模经济、政府行为对现代服务业集聚促进作用显著。

SPATIAL STATISTICAL ANALYSIS OF FORMATION MECHANISM OF MODERN SERVICE INDUSTRY AGGLOMERATION

Space related issues are widely explored in regional economics. In the sense of the space-time dependence characteristic of data generating procedure, combining with classical statistical models, spatial panel econometrics models can get more comprehensive data information. In the neoclassical growth theory, economy is assumed to be independent. However, technological advances in one area might be transmitted to other area and so the closed economy assumption might not be valid. Therefore, the authors take the possible spatial correlation into account, both in cross data settings and panel data settings. From the econometric point of view, spatial dependence leads to unreliable statistical inference if the spatial effect is present but omitted. The theory is called spatial econometrics. Spatial econometrics is a subfield of econometrics that deals with the spatial interaction (spatial autocorrelation) and spatial structure (spatial heterogeneity) in regression models for cross-sectional and panel data. Based on the spatial econometrics theory, the paper tries to establish an econometric model that includes spatial interdependence among every area in china. In other words, based on the spatial panel econometric model, this research conducts an empirical analysis of influencing factors of modern service industry agglomeration across 28 provinces in china. The analysis proceeds in three stages. Firstly, the authors use the Moran index of spatial statistics to test the spatial correlation, and the results show that there is strong spatial dependence among every areas in china.This is a precondition for the next analysis. Secondly, the research conducts a spatial econometrics model including Spatial Autoregressive Model (SAR Panel model) and Spatial Error Model (SEM Panel model) to analysis the factors of modern service industry agglomeration. The results show that modern service industry agglomeration has a strong spatial dependence and a positive spillover effect. The technological differences had significantly positive effect in the time dimension while effect in the space dimension is not significant. Transaction costs, knowledge spillover, economies of scale and government action have positive impact on modern service industry agglomeration. Finally, the authors point out that from practical perspective, the results have great theoretical significance and practical significance.