A Spatial Econometric Modeling Of Chinese Economic Growth.

Regional researchers have always been interested in the regional development of mainland China.The strategic choices of regional economic embodies China's modernization and pressing need for the long-term development,it also the practice and development of regional economic strategy.However,Western scholars have fundamental different views with respect to the problem of regional development of China(Tsui,1991).Along with the international society's interest to China growing strongly and the related research increasing rapidly,China's leadership found that developing coastal regions first is in accord with China's national interest(Yang, 1990).Although the regional special policy may widen the gap between the coastal and inland regions in a short time,this is the cost of China's modernization and overtaking the newly industrializing countries.As China's leadership hopes,once the coastal regions develop substantially,it can become locomotive and promote China to the development path of the modern economy.Consequently,China's leadership suggest such diffusing pattern of regional economy(Zhao,1984).In the succeeding 1960's and 1970's,the development of the Third World became a major concern and the main field of the research,and the hypothesis of "core-periphery" was the most popular research subject.As most of the developing problem can be described by the well-developed "core" and the underdeveloped "periphery",the theory of "core-periphery" has been extensively applied in the field of social sciences study.At the past half century,the technique of examining the hypothesis of Myrdal-Hirschman's "core-periphery" has developed rapidly.However,the empirical evidences derived from empirical study remain dissatisfactory.The cause of the predicament is likely that the existing research is bounded by the analytic approach; they try to measure the properties and strength of the spatial interaction of the "core-periphery",using the the non-space model,which depicts the function and causality between variables.Recently,Ying(2000) confirmed the diffusion-polarization process with different properties and strength,based on the newly developed methods of exploratory,spatial data analysis."Spatial autocorrelation",a spatial correlation model which is proposed and constructed by Cliff and Ord(1973) based on the early research of Moran(1948),Geary(1954), Whittle(1954),is the new monument in the geography and regional science,and it can be used to test the correlation properties and strength between spatial location. Moreover,Moran's I and Local Moran may be the most convenient approaches to test spatial autocorrelation.According to the above-mentioned analysis,we analyze China's economy using the global autocorrelation and local autocorrelation model.Based on the per-capita GDP of 30 provinces over the period 1952-2001 in China and the Local Moran model, we get the spillover distribution of the per-capita GDP of 30 provinces in China.The spillover effect is significant,hence if the significance of Local Moran is statistically significant(p≤0.05),the growth is not a independent stochastic event in space,but rather relating to the spillover transferred between provinces(adjacent) under the "core—periphery" structure of China's economy.The positive Local Moran value suggested a spreading process,thus the negative value suggested the pattern of "Backwashing effect".