Spatial Analysis of Industry Clusters Based on Local Spatial Statistics:A Case Study of Beijing’s Manufacturing Industry Clusters

Guoping Li
2012-01-01
Abstract:Incorporating local spatial autocorrelation analysis into the identification of regional industry clusters,this article presents a useful method for regional industry cluster analysis,which can clearly identify the spatial distribution patterns of regional industry clusters.We use the national benchmark industrial clusters as the template,and apply local spatial autocorrelation analysis based on the local Getis and Ord G statistic to analyze the spatial distributions of manufacturing clusters in Beijing.Our data are drawn from the Beijing Economic Census and the National Economic Census in 2008.We find out that the manufacturing clusters of Beijing are mainly located in the suburban areas,and new investment and development areas are the main locations attracting the manufacturing industries to concentrate and form industry clusters.Through analyzing the spatial patterns of Beijing’s manufacturing clusters,we find out that 1) historical path-dependency,natural resources conditions,transportation accessibility,regional policies and so on are the factors driving the formation of manufacturing clusters in Beijing’s suburbs;2) manufacturing clusters related with industrial linkages are often located near to each other;3) manufacturing clusters are scattered throughout the suburbs in Beijing,and different clusters have various location patterns;and 4) the functional divisions of industry sectors within the cluster over space are not distinct,and many clusters are duplicated in industry components at different locations,which leads to the overlap of functional divisions of areas.Therefore we suggest to further improve the agglomeration of manufacturing clusters in Beijing,to clearly indentify the regional functions and to guide manufacturing clusters through proper regional policies,so as to enhance the regional divisions and specializations of manufacturing clusters based on the regional functions and specialty.
What problem does this paper attempt to address?