Spatial Econometric Model And Applications To Chinese Economy

On the basis of systematically combing the classic spatial econometric analysis framework, estimation and testing methods, this paper gives a more detailed interpretation and discussion of several advanced development of spatial econometrics, such as the spatial structure transformation model, Non-stationary spatial econometric model, the spatial filtering methods, and spatial Shift Share analysis. After that, this paper uses these methods to the practical problems of China’s three applied research, this paper provides some foreshadowing and references to carry out follow-up studies and further research.First, based on the development of spatial econometrics context, this paper gives a brief survey about the main spatial econometric model form, estimation and testing methods and the main applications of spatial econometric. And from the cross-sectional data and panel data point of view, this paper discusses a set of several classic space econometric models and maximum likelihood estimation method in detail, and uses the classic models to study China’s economic convergence, considering spatial correlation and spatial heterogeneity of circumstances at the same time, re-estimates China’s regional economic convergence rate and convergence mechanisms. The study found that there are structural changes in the spatial error model more in line with the reality of our country; there are different convergence mechanisms in the eastern, central and western regions. The estimated results of Spatial Durbin model show that the regional economic growth has spillover effect, the average growth rate of per capita GDP of a region and its adjacent the region’s average growth rate is positively correlated, and then this paper briefly describes several forward development directions of spatial econometrics.This paper gives a more detailed interpretation and discussion of two cutting-edge development directions in spatial econometrics:the non-stationary spatial econometrics and spatial filtering methods. Non-stationary spatial econometrics was begun from Moran I statistic of spatial unit root test method, and followed by two-step LM spatial unit root test which is more reasonable to distinguish between stationary positive spatial autocorrelation spatial unit root. And then discusses the spatial pseudo-regression and the spatial co-integration theory which is similar to the E-G ideology in time series analysis. Further, we discuss a parameter, nonparametric and semi-parametric methods of spatial filtering. Conducted a detailed interpretation of the non-parametric G-statistic method and eigenvector method, and gave a detailed description of the semi-parametric methods about how to extract the eigenvector. Finally we use non-parametric G-statistic spatial filtering method to conduct an empirical study about regional economic convergence in China with cross-sectional data. The study found that the method can eliminate the spatial correlation between individuals to a certain extent, regression analysis with the classic space measurement method results were compared with the filtered data’s regression.Based on the traditional method of Shift Share Analysis and for its inadequacies, we leads to the dynamic Shift Share Analysis, and further consider the space effect s between the regions, leads to dynamic spatial Shift Share Analysis. The paper uses the method to study9industry in our country regional, taking into account the national effect and industrial structure impact at the same time, we also take into account the spatial correlation between regions more accurately in order to reflect the interaction between the region, and reveal the regional industrial competitiveness, the dynamic process of change and spatial interactions, the study found that the traditional agricultural and industrial provinces should continue to develop its traditional advantages, and further use the positive impact of the surrounding area’s industrial development, a reasonable expansion of investment in the tertiary industry,. Provide some policy recommendations for the regions to optimize the industrial structure and promote balanced regional development.