A Spatial Econometrical Analysis On China's R&D And Economic Growth

Chinese economy growth pattern is locating in the transition-period from extensive to intensive. Research & development (R&D) and its contribution to the economy growth of China have increasingly attracted attentions of the global academic community. Globalization and digitalization seem to declare the 'death of geography'. Nevertheless, it is suspectable that the knowledge spillover of China is also limitless. Does regional knowledge production have spatial non-stationarity? Does spatial spillover effect occur between large and medium-size enterprise and science and technology (S&T) institute? Does S&T's R&D have impact on China's economic growth? If it does, how far does it effect? Based on Romer's idea of spillover effect of production factor and Griliches-Jaffe's knowledge production function, the above questions can be analyzed by spatial econometrics at two spatial levels.At the provincial level, two kinds of analyses are conducted. First, a deep exploratory spatial data analysis (ESDA) of the dissemination of regional R&D activity is performed. Some global and local indicators for spatial association are presented, the presence of a dependence process in the distribution of R&D activity for different periods are summarized, the spatial difference of regional R&D activity are also examined. Second, the contributions of various factors, such as R&D expenditure, gross domestic product (GDP), as well as the propensity to patent to the regional knowledge product are estimated using Griliches-Jaffe's knowledge production function by confirmatory spatial data analysis (CSDA). Meanwhile, common factor hypothesis is tested. R&D expenditure and GDP are found to be important in regional knowledge production while the propensity to patent is also important. Econometric results imply that the knowledge spillover is limited. Moreover, the production of knowledge by regions seems to be also affected by spatial spillovers due to R&D activity performed in other regions.At the city level, two kinds of analyses are also conducted. First, the spatialdistribution of regional R&D activity is performed using ESDA technique. The spatial association and the aggregate spatial difference of regional R&D activity are examined by global and local moran's I. Second, the spatial variation of the effect of various factors on the regional knowledge production is examined using the technique of geographically weighted regression (GWR). All parameters have robust spatial non-stationarity or spatial variation over the whole study area. In particular, the GWR model is significantly better than the OLR (ordinary linear regression) model.R&D is the engine of regional economic growth. S&T institute has been the main force of China's R&D activity. Based on the an endogenous growth model by Romer (1990), the impact of S&T's R&D on economic growth is analyzed using CSDA technique at the provincial level data over the period 1990 to 2002. Econometric results indicate that a high density of S&T's R&D accelerates regional per-capita income growth, and that R&D at large and medium-size enterprise does not contribute to the regional knowledge accumulation, and as well that regional income growth seems to be also affected by nearby region in the form of economically positive spillovers.In addition to the above discussion, several suggestions for stimulating R&D and the economy growth of China are briefly proposed.