基于偏好DEA模型的中国纺织业效率评价  

Performance Measurement of Chinese Textile Industry Based on Data Envelopment Analysis Model with Preference Information

作　　者：王新宇[1] 吴瑞明[2] 

机构地区：[1]中国矿业大学管理学院,江苏徐州221008 [2]上海交通大学管理学院,上海200052

出　　处：《中国管理科学》2005年第2期142-148,共7页Chinese Journal of Management Science

基　　金：国家自然科学基金资助项目(70371075)

摘　　要：基于具有输入输出指标偏好信息的数据包络分析模型评价中国纺织工业的效率,给出了描述规模报酬不变和规模报酬可变假设的DEA模型及其对偶规划的一种形式。模型克服传统DEA模型应用中常见的决策单元权重为零进而高估决策单元效率的不合理现象,有效地测算了中国31个地区纺织工业的经济运行效率,并分析了地区间效率水平差异的原因。利用各决策单元在经验生产前沿面上的投影点,估计了中国纺织工业的随机生产前沿函数,最后用效率弹性线性递减模型分析了销售收入、人均资产、台港澳和外商投资对各地区效率的微观影响关系。In this paper,a data envelopment analysis model with preference information about input and output targets is set up to evaluate the economic operational efficiency of the textile industry of Chinese 31 regions.A general framework embracing CCR and BCC models is put forth.The phenomenon that because of a lot of zero weights and overestimating DEA efficiency of decision making units in ordinary DEA model can be avoided by using the model.We also analyze what causes the differences among the efficiency of the textile industry of Chinese 31 regions.Then the stochastic production frontier function of Chinese textile industry is obtained with the data set of all decision making units' projection points on the production frontier face by regression analysis model.Finally we find that three economic factors:sale incomes,asset per person,Foreign Direct Investment,have important effects on the efficiency of China textile industry by an econometric regressive model.

关 键 词：偏好信息 数据包络分析 随机生产前沿 效率评估 中国纺织工业 

分 类 号：F723[经济管理—市场营销] C931[经济管理—产业经济]