Geographically And Technologically Mediated Knowledge Spillovers Between Chinese Provinces

Nowadays the Empirical Analysis of China’s regional knowledge spillover is mostly only studied at the level of geographical proximity. By taking the panel data of31provincial high-tech industry of Mainland China in the period of2001-2010as a sample and using the Spacial Lag Panel Data Model, this paper has obtained the effects of regional knowledge spillovers with geographical proximity, technological proximity and integrated proximity, which taking the number of patent and the output value of new products as the product function, respectively. In this paper, we use the highway mileage between the provincial capitals as the index of geographical proximity to construct its spacial weights matrix, the modified Jaffe index as the index of technological proximity to construct its spacial weights matrix, and the product of geographical proximity and technological proximity as the index of integrated proximity to construct its spacial weights matrix. By the Moran’s I test, the mean value of Moran’s I of patent in the period of2001-2010is0.3069, and the one of new products is0.3078, which indicates there is a significant spatial autocorrelation in China’s high-tech industry distribution. By the unit root test to the raw data, the variable sequence stationarity is obtained.Through the OLS parameter estimation with the extended and modified double logarithmic production function model, we have observed that the high-tech R&D investment funds is positively correlated to the number of patent, and the high-tech R&D input of human resource is positively correlated to the innovation of products. The results of Space Lag Panel Data Model show that the high-tech R&D investment funds is positively correlated to the number of patent; the high-tech R&D input of human resource is positively correlated to the innovation of products; geographical proximity, technological proximity and integrated proximity make positive effects on regional knowledge spillovers. Specifically, when taking the number of patent as the output variable, technological and integrated proximities make remarkable effects on the regional knowledge spillovers than geographical proximity does; while taking the output value of new products as the output variable, these three proximities make almost the same effect on the regional knowledge spillovers.