Abstract Industrial cluster theory, one of the most influential theories in academic research and government policy, has recently generated considerable debates. The paper reviews the debates among Western scholars and examines several key issues. First, spatial boundaries of industrial clusters and methods identifying them are confusing. Porter confuses the competitiveness of enterprises with that of regions and countries. We argue that spatial boundaries should not be too large,because of few industrial linkages due to distance and boundary effects. Methodologically, Porter uses the location quotient to identify industrial clusters. However, lo cation quotient shows industrial specialization of regions, not industrial linkages. Second, the paper argues that industrial clusters are influenced by the process of industrial evolution. If a region can not upgrade its industries, the industrial clusters in the region may fail. Third, the paper argues that industrial cluster study is isolated from the research on globalization,global production networks and global city regions.In fact, globalization is interwoven with regional development, and there is no exception for clusters. Clusters have been merged into global production chains.Fourth,the paper argues that there lacks a definite policy framework because of vague concepts and methods.Lastly, the paper suggests that industrial cluster theory has to be used carefully due to the confusions.Industrial cluster is still a ‘grey box’. The mechanisms of formation and its economic outcome are yet to be fully understood. China is embracing the theory of industrial clusters, but we should consider carefully the reality of China and develop industrial cluster theories with Chinese characteristics.
2007, Vol. 22 
