基于两阶段共享投入DEA的中国各省市自治区产学研效率评估

Evaluation of Industry-University-Research Efficiency in China’s Provinces,Municipalities and Autonomous Regions Based on Two-stage Shared Input DEA

数据包络分析是衡量决策单元效率的一种方法,其中决策单元内部结构被视为一个“黑箱”.两阶段网络结构DEA是将第一阶段的所有产出都作为第二阶段的中间变量.采用网络DEA的方法,构造了一种新的两阶段同时共享投入和中间变量的DEA模型.与现有的两阶段共享DEA模型相比较,新模型在第一阶段过程中增加了额外产出,具有更强的普适性.运用构造的新模型评估了我国各省市自治区的产学研效率,结果表明,两阶段共享投入DEA模型是合理的.

Data envelopment analysis is a method to measure the efficiency of decision-making units,in which the internal structure of decision-making units is regarded as a“black box”.The two-stage network structure DEA takes all the outputs of the first stage as the intermediate variables of the second stage.This paper uses the network DEA method to construct a new two-stage DEA model that simultaneously shares inputs and intermediate variables.Compared with the existing two-stage shared DEA model,the new model has added extra output during the first stage,which is more universal.By applying the constructed new model to evaluate the efficiency of industry-university-research in China’s provinces,municipalities and autonomous regions,the rationality of the two-stage shared input DEA model is tested.