关键词: 产业结构升级, 空间计量, 空间杜宾模型

Abstract: Based on China’s provincial data (including municipalities and autonomous regions) from 1997 to 2012, this paper explores the characteristics of the spatial pattern of China’s industrial structure at provincial level and its main influence factors using spatial autocorrelation analysis method and spatial panel econometric model. The results indicate that firstly, China’s provincial industrial structure exhibits an obvious spatial agglomeration. Secondly, on one hand foreign direct investment, physical capital investment, energy investment, resident consumption, urbanization and technology advance all give impetus to the upgrading of China’s industrial structure, and on the other hand, impetus from government consumption and human resources investment is not strong enough. Of all these variables mentioned, only energy investment, government consumption and resident consumption have obvious spatial spillover effects. Thirdly, the upgrading of China’s industrial structure is mainly driven by investment, with foreign direct investment being the most important influence factor. However, the driving force of consumption demand proves to be insufficient. Fourthly, human resources investment and technological advances fails to produce enough spillover effect to back up the upgrading of industrial structure. Finally, this paper proposes some practical suggestions for policy making.