当变量存在空间相关时,利用传统计量方法会使得结果产生偏倚。运用探索性空间数据分析和空间计量模型,以专利申请量作为区域创新衡量指标,发现1998--2010年我国各省域专利申请量和FDI流入均存在显著的空间正自相关,三大区域并不存在传统意义上的俱乐部收敛现象。但引入FDI作为控制变量之后,发现收敛系数小于零,认为我国区域创新存在以FDI为条件的俱乐部收敛,FDI对中国区域创新的俱乐部收敛具有显著促进作用。
When there exists spatial correlation in variables, traditional methods will produce biased results. With the exploratory spa- tial data analysis and spatial econometric model, patent applications as the indicator of regional innovation, the results showed that there was significant positive spatial autocorrelation in the patents applications and FDI of all the provinces from 1998 to 2010. There was no significant traditional club convergence in the three areas of China. After the introduction of FDI as a control variable, the coefficient became less than zero. So it can be sure that there exists club convergence of regional innovation with FDI as the condition. FDI promote the formation of club convergence of regional innovation in China significantly.