Research On The Economic Interaction Between China's Provincial-level Regions Based On The Endogenous SLX Model

The problem of economic development gap and spatial coordinated development between China's regions has a profound social reality background,and it also attracts long-term and extensive attention from the economics community.From the perspective of spatial econometric method innovation,this paper introduces the spatial interaction model using endogenous spatial weight matrix,and empirically studies the spatial interaction of China's provincial economy from 1997 to 2016.The SLX model based on endogenous spatial weights matrix has the ability to measure two-way asymmetric spatial overflows between regions.Meanwhile from the empirical point of view,flow patterns,spatial interaction structures and evolution trends of spatial overflow between different regions in China are investigated.The research methods and conclusions of this paper are complementary to the study of regional economic spatial convergence and spatial spillover effect test in China,and have certain reference value for the research of regional coordinated development in China.The main research objectives are detailed as follows:(1)A comprehensive summary of neoclassical growth theory,new economic growth theory and new economic geography theory.Focusing on factor flow and spillover effects involved in above growth theories,the spatial interaction mechanism of inter-regional economy is analyzed and discussed under the new economic geography.(2)By improving the classical spatial panel model,a completely endogenous spatial weight matrix is implemented,and finally an econometric model capble of measuring two-way asymmetry spillover effects is obtained.(3)Combining the results of the previous two parts,analyzing the economic correlations of provincial-level regions in China in recent years,so that the research content of this paper can test the theoretical methods of this paper,and also provide reference for China to further coordinate regional economic development.The research idea of this paper is to construct a completely endogenous spatial weight matrix which is based on the existing spatial interaction theory research,and to study the space interactions between provincial regions in China by using the spatial econometric model which can express two-way asymmetric spatial spillover effects.The effects may contain directions,patterns,and temporal and spatial evolution characteristics of the interaction between provinces.The key research contents of this paper are included as the following parts:(1)Analysis of the mechanism of inter-regional economic interaction.From the view of the neoclassical economic growth theory and the new economic growth theory,this paper sorts out the theoretical basis of inter-regional economic interactions with the factor externalities as the main object.At the same time,this paper proposes a numerical analysis of the economic interaction between provincial-level regions in China under the growth economic theory framework of new economic geography.(2)A spatial panel model construction method of which spatial weight matrix is completely endogenous.This paper analyzes the inadequacies of classical spatial econometric models(SAR,SEM,SDM,etc.)which contain an exogenous spatial weight matrix in the analysis of inter-regional economic interactions then demonstrates that the endogenous spatial weight matrix is a viable path to improve the effects of classical methods.On such a basis,the dynamic factor model method is applied to construct a completely endogenous spatial weight matrix based on the SLX model(called DFM-SLX model).Finally,the DFM-SLX model is proposed with its model selection,estimation method and statistical characteristics.(3)Empirical analysis of economic interactions between provinces in China.Based on the provincial panel data of China from 1997 to 2016,this paper first tests the economic development differences,convergence and spatial correlation between the provincial regions in China in the last two decades,and then applies the DFM-SLX model to measure interactions.The direction and size of the economic effects of inter-regional inter-elemental flow and spillovers in China are deeply tested and measured,and the dynamics,characteristics and development trends of inter-regional economic interactions in China are discussed with empirical results.The main innovation and contribution of this paper is to propose a DFM-SLX panel spatial measurement method based on dynamic factor model which can construct and estimate spatial weight matrix endogenously.Through a dimension reduction processing,the method makes it feasible to fully parameterize a spatial weight matrix,thus breaking through the symmetry and element non-negative constraints of the traditional exogenous spatial weight matrix.The endogenous and parameterized spatial weight matrix can express a more general spatial correlation between samples,and also avoids the problem of arbitrariness in the setting of an exogenous one.Finally,the empirical results of the DFM-SLX model for the spatial interaction of the provincial-level regional economy in China from 1997 to 2016 also show that the model has the ability to identify the spatial asymmetric spillover effects between spatial samples.Based on the above research,the main conclusions of this paper are as follows:(1)the paper measures the global Moran index of per capita real GDP in each province-level region observed during 1997-2016,which is related to China's economic space.The correlation test found that during the observation period,the economic distance correlation in the first half of the year from 1997 to 2007 was relatively stable;after 2008,the spatial correlation showed a trend of continuous decline;(2)The empirical work of the DFM-SLX model on the spatial interaction between provincial regions in China during 1997-2016 shows that the DFM-SLX model performs much better than models such as OLS,SAR,SEM,and SLX on both the goodness of fit and the AIC criterion;(3)Empirical research based on the DFM-SLX model shows that the overall interaction between economic activities in China's provinces during the 1997-2016 period is characterized by positive spatial spillovers in the eastern region and negative in the western regions;the results of that in 1997-2007 and 2008-2016 respectively show that the spatial spillover trend in the first half of the year is consistent with the flow in all periods,and opposite to the flow in the latter half;this shows that inter-regional space spillover effects during 1997-2007 are relatively stronger;(4)The estimation results of DFM-SLX in multiple continuous time kernels show that there exist several stages of evolution of spatial interaction between eastern,central and western regions of China,of which 2003 and 2009 are changing years.At the same time,the central region plays a stable role as a bridge between regional interactions.