韩兆洲,程学伟.中国省域R&D投入及创新效率测度分析[J].数量经济技术经济研究,2020,(5):98-117 | 中国省域R&D投入及创新效率测度分析 | Measurement and Analysis of R&D Investment and Innovation Efficiency in China | | DOI: | 中文关键词: 创新效率Bootstrap-DEAσ收敛R&D投入 | 英文关键词: Innovation Efficiency Bootstrap-DEA σ Convergence R&D Investment | 基金项目:本文获国家社会科学基金重点项目“粤港澳大湾区协调发展机制体系统计测度和决策支持系统研究”(19ATJ004)的资助。 | | 中文摘要: | 研究目标:测度我国2007~2016年省域创新效率,分析各省域、三大地区和八大经济区域的创新效率的动态趋势和变化以及其时空收敛性质,探究R&D投入与创新效率之间的关系。研究方法:基于专利产出为创新产出的视角,运用DEA、SFA以及Bootstrap-DEA的面板模型测度区域创新效率,并采用σ收敛模型检验其时空的收敛性质。研究发现:传统的DEA模型存在明显的高估偏误,未加入环境因素的SFA模型参数偏误较大,Bootstrap-DEA测出的创新效率值具有更小的偏误;R&D投入与创新效率之间存在一定程度的正相关关系,但这种正相关关系在R&D投入达到一定数值时,有边际递减的倾向;2007~2016年全国平均创新效率呈现“U”形过程,东部沿海地区的创新效率遥遥领先,但地区与经济区域之间存在“追赶效应”;全国的创新效率收敛指数基本呈现“L”形过程,省域创新的效率差距在逐渐缩小,区域创新活动有一定的空间集聚现象。研究创新:首次运用Bootstrap-DEA的面板模型测度省域创新效率,对效率估计值的偏差进行修正,并给出其置信区间和模型偏误。研究价值:为各省域、各经济区域的创新效率的提高提供政策依据。 | 英文摘要: | Research Objectives: This paper is to measure the provincial innovation efficiency in China from 2007 to 2016, analyze the dynamic trends and changes of innovation efficiency in each province, three regions and eight economic regions, and their spatio-temporal convergence properties, and explore the correlation between R&D investment and innovation efficiency. Research Methods: Based on the perspective of patent output as innovation output, the panel models of DEA, SFA and Bootstrap-DEA are used to measure regional innovation efficiency, and the σ convergence model is used to test the convergence property of time and space.Research Findings: There are obvious overestimation errors in traditional DEA model, and the parameters of SFA model without environmental factors have larger errors. The innovation efficiency value measured by Bootstrap-DEA has smaller deviation. There is a certain degree of positive correlation between R&D input and innovation efficiency, but this positive correlation tends to decrease marginally when R&D input reaches a certain value. The average innovation efficiency of China presents a U-shaped process curve, and the innovation efficiency of eastern coastal areas is far ahead, but there is a“catch-up effect” between regions and economic regions; the convergence index of innovation efficiency of the whole country basically presents an L-shaped process, the efficiency gap of provincial innovation is gradually narrowing, and regional innovation activities have a certain spatial agglomeration phenomenon.Research Innovations: For the first time, the panel model of Bootstrap-DEA is used to measure provincial innovation efficiency, and the bias of efficiency estimation is corrected, and its confidence interval and model error are given. Research Value:This paper can help to provide policy basis for the improvement of innovation efficiency of provinces and economic regions. | 查看全文 |
| | |