Spatial characteristics and dynamic changes of provincial innovation output in China:An investigation using the ESDA
我国省域创新产出的空间特征和时空演化——基于探索性空间数据分析的实证

Innovation activities in each region not only depend on their own characteristics,but also on those of the regions that form the neighborhood to which it belongs.Regional spillover as a spatial interaction is important in explaining agglomeration,innovation and regional growth.A great deal of literature has deeply dealt with the issue from a spatial perspective since the 1990s,especially in the context of urban and regional studies.Unfortunately,the traditional approaches to regional innovation suppose that each region is independent from others.This paper uses spatial statistical techniques to establish the statistical relations among data according to the geographical locations.It aims to understand the spatial dependence and autocorrelation related to geographical locations.Using the methods of exploratory spatial data analysis(ESDA) and spatial analysis software,this paper analyzes the spatial distribution of innovation outputs,measured by the number of patient applications,throughout 31 Chinese provinces from 1997 to 2008.The visual patent distribution plot has shown the distribution of innovation outputs at the provincial level and its spatial dynamic changes.A significantly high level of spatial concentration of innovation outputs among Chinese provinces has been captured by the computed spatial Gini coefficient and the Concentration Ratio,and the concentration level has increased steadily over the past 10 years.The analysis using the Moran’s I statistics gives the strong evidence of spatial autocorrelation in innovation activities among provinces,while the concentration pattern of innovation activities among provinces and its changes over time have been revealed by using the local Moran’s I and the Moran scatter plot,which indicate the clustering nature of the spatial distribution of provincial innovation activities.Spatial Gini coefficient and Moran’s I index have indicated that innovation activities of Chinese provinces are not randomly distributed.Our findings suggest that innovation activities are spatially differentiated among Chinese provinces over the 10 years,and innovation activities at the provincial level are highly localized.This study can provide a scientific basis for the intuitive expression of the spatial correlation of innovation outputs among provinces,and puts forward that the spatial statistical analysis could present some references valuable for analyzing spatial structure and patterns and policy-making.