基于2003-2012年面板统计数据,分析区域创新能力指标值及其基尼系数变化,发现期间我国省域创新能力差距明显,且存在扩大趋势；探讨了区域创新能力差异的影响因素,选用面板数据固定效应模型(FEM)建立了回归方程,应用截面加权广义最小二乘法(CSW-GLS)估计法确定了模型参数,基于该模型应用Shapley分解法计算、分析了各种创新能力影响因素对创新能力差异的贡献度,据此提出缩小省域间创新能力差距宜以增加创新滞后地区科技人才储备为重,其次考虑大力推进创新滞后地区市场化程度,其他影响因素的优化为再次。

Based on 2003-2012 panel data, the variance of index value and its gini coefficient of regional innovation ability are analyzed and it shows that the differences of innovation ability among China’s provinces are obvious and show an expanding trend. The influence factors on the differences of regional innovation ability are discussed, the regression equation is established using panel data FEM model and the parameters of which are determined by CSW-GLS estimation method. The various factors’ contributions to the innovation ability differences are calculated and analyzed using Shapley decomposition based on the above equation. Furthermore, the key point to narrow the innovation ability gap among provinces in China is increasing scientific talents reserve for innovation lag areas, followed by promoting market level and developing the other influence factors for lag areas.

F107

全国统计科学研究计划项目“基于空间Durbin模型的我国省域居民消费结构溢出效应研究”（2013LY057）；教育部人文社科研究项目“我国城市化质量空间差异的成因和优化策略”（14YJCZH191）