Building a spatial lag model and spatial error model on the premise of geographical spatial weights matrix and eco⁃nomic spatial weights matrix respectively, we discuss the dynamic process of industrial structure’s convergence based on 25 cities’macroscopic statistical data from 1996 to 2014 in the Yangtze River Delta Economic Zone. The outcomes show that no matter what kind of spatial matrixes, the primary industry doesn’t have spatial autocorrelation and convergence;influenced by geographical and economic distance, the secondary industry and the tertiary industry show absolute convergence, of which the speed of tertiary industry is faster than that of secondary industry, and the secondary industry shows a conditional convergence, which is significantly affected by technology, labor and capital;the tertiary industry’s spatial autocorrelation under economic spatial weights matrix is significantly stronger than that under geographical spatial weights matrix;ignoring the differences of industrial structure’s convergence between different spatial weights matrixes will result in errors of the research on the region⁃al development convergence. Based on this, it is proposed that the establishment of regional economic development strategy should be based on the research of industrial structure’s convergence under different spatial weight models, the convergence of regional development should be refined, the regional economic cooperation should be strengthened, and the gap in industrial development among cities should be gradually narrowed so as to promote regional industry’s coordination and economic inte⁃gration.%文章根据长江三角经济区25市1996-2014年宏观统计数据，基于地理空间和经济空间权重矩阵构建空间滞后模型与空间误差模型，探讨研究区三大产业的β收敛性动态变化特征，结果表明：研究区第一产业不存在空间相关性和收敛性；在地理和经济距离影响下，研究区内二、三产业呈现绝对β收敛，而第三产业收敛速度高于第二产业，其中科技、劳动力和资本要素显著影响第二产业的收敛性；经济空间权重矩阵前提下第三产业的空间相关性远高于地理空间权重矩阵；基于上述研究结果，发现忽略空间权重矩阵下产业结构收敛性差异将导致区域发展研究误差，区域经济发展战略的制定应以不同空间权重模型下产业结构收敛性研究为基础，加强区域经济合作，促进区域产业协调和经济一体化。
展开▼
