Research On The Spatial Correlation And Variation Of Regional Labor Productivity

Since the reform and opening-up policy, China’s economy has developed rapidly. But the differences in regional economic are bigger and bigger with development. The existing and expanding of differences have an enormous impact on each country or region of sustained economic development and social stability, and the level of labor productivity represents the level of regional economic development on some extent, therefore, the detailed analysis of regional economic disparities to find the corresponding solutions is an urgent need.On the basis of reviewing the research about spatial correlation and variation of regional labor productivity, this paper introduced the basic theory and model of labor productivity. The research was maed with31provinces and autonomous regions (Hong Kong, Macao and Taiwan are exclueded). First, the paper explored the global spatial autocorrelation with global Moran’I, the result shows that labor productivity in different area exits the significant positive spatial autocorrelation, in space it shows characteristics of spatial concentration. Following that, the paper investigated the mode of spatial concentration and variation with Moran scatter map, LISA signification map and LISA cluster map. The paper also used spatial autoregressive model to explore the labor productivity, once again proved the existence of spatial correlation, and concluded that Spatial lag model is more suitable for spatial correlation of labor productivity research.This paper made an intensive study of Geographically Weighted Regression (GWR) Model, including their foundations, weight acquisition and fitting method. Thus, GWR model was applied into the study of spatial variation of regional labor productivity. In this case,4factors were selected and quantified. Thus, the GWR Model was built to analyze the quantitative relations between labor productivity and the influencing factors, and their relations were presented by means of map visualization. The result revealed that, the per capita investment in fixed assets, investment in human resources, per capita expenditure, per capita export trade all showed the positive correlations with labor productivity, meanwhile, the influences varied among different geographical locations. By model comparison, it showed the GWR Model fitted labor productivity much better than the ordinary linear regression model did, the GWR model made a more satisfactory in researching spatial variation of regional labor productivity. Finally, this paper putted forward some recommendations according to empirical results, and some future research topics are given.