一种基于极大熵准则DEA的区域科技投入效率评价方法：以广东专业镇为例*

A Novel Analytical Method of R D Input Efficiency Based on MEC-DEA: Take Guangdong Industrial Clusters as the Instance

摘要:

传统基于DEA模型对区域科技投入效率的评价,存在环境变量和随机因素难以估计和测度的问题,本文提出一种新的基于极大熵准则(MEC,Maximum Entropy Criterion)的三阶段DEA模型,尝试部分减少环境变量和随机因素的影响,并将该模型应用到广东省专业镇的科技投入效率评价。论文的实证部分选择了2007-2012年广东省专业镇科技创新投入产出的面板数据进行测算和分析。实证分析结果表明,规模效率和纯技术效率对专业镇科技创新效率的整体提升具有显著影响,但是全省专业镇科技投入产出的综合技术效率水平偏低,不同城市专业镇的科技投入效率差异性较大,珠三角和粤东地区相对较好,粤西和粤北地区在科技资源配置方面则显得更为粗放,存在一定的冗余现象。分析结果还表明广东省专业镇的科技创新效率受到地区经济发展水平、劳动者素质、政府政策支持的多种环境变量和随机因素的影响,面向专业镇科技创新的相关政策和制度建设有一定的改进和提升的空间。

关键词:创新效率评价;数据包络分析;极大熵准则;专业镇

Abstract:

For the difficulty of estimation and measuring the effects of environmental variables and random factors, we present a three-stage DEA model based on the maximum entropy criterion, and apply it to evaluate technological efficiency of industrial towns in Guangdong. The panel data of empirical analysis is from 2007 to 2012. The analysis results show that, the scale efficiency and pure technical efficiency has a significant effect on technology innovation efficiency, but the professional town investment in science and technology, comprehensive technical efficiency level of output is low. The Pearl River Delta and the eastern region are relatively better, resources configurations in the West and north areas are more extensive, redundant phenomena to a certain extent. The results also revealed that the technological innovation efficiency is affected by the level of economic development, the quality of workers, government policies and a variety of environmental variables and random factors.

Key words:Evaluation on Innovation Efficiency; Data Envelopment Analysis; Maximum Entropy Criterion; Industrial Town