中国省际工业全要素能源效率——基于四阶段DEA和Bootstrapped DEA

Provincial Industrial Total Factor Energy Efficiency in China-Based on Four Stages of DEA and Bootstrapped DEA

将能源和二氧化碳纳入工业企业的全要素能源效率评价体系中,基于四阶段DEA和Bootstrapped DEA方法,在控制了外生环境变量和随机冲击的影响下,对我国2010年中国30个省、市、区规模以上工业企业的全要素能源效率及其分解变量进行了实证分析。主要结论如下:(1)初始D EA模型、四阶段DEA模型以及Bootstrapped DEA模型计算得到的效率得分存在显著差异。不对外生环境变量和随机冲击的影响作控制,得到的全要素能源效率值是有偏的。(2)tobit回归模型显示:各地区工业企业的技术研发投入是全要素能源率提高的有利因素,且对能源和碳排放减少的贡献比例最大;环境保护支出对全要素能源效率影响微弱;国有化程度的提高是全要素能源效率提高的不利因素。(3)剔除环境变量影响后,全国工业企业的平均综合技术效率由0.803下降到0.747,平均纯技术效率由0.825上升至0.951,而平均规模效率则由0.974下降至0.786。规模报酬递减省份均调整为规模报酬递增状态。(4)运用Bo otstr apped DEA对四阶段DEA得到的效率得分进行偏误修正后,所有地区的全要素能源效率得分均有所下降。

This paper incorporates energy and CO2 into industrial total factor energy efficiency evaluation system, based on four stages DEA and Bootstrapped DEA method, by controlling the effects of external environmental variables and random shocks, and analyzes the total factor energy efficiency of Industrial Enterprises above Designated Size of 30 provinces in 2010. The main conclusions are as follows： （1） Efficiency scores calculated with the initial DEA model, four stage DEA, Bootstrapped DEA method are different, and the total factor energy efficiency value is biased without controling the external environment variable and stochastic shocks; （2） tobit regression model shows that RED is favorable factors to improve the total factor energy efficiency, that contribution proportion is the largest about the energy and reduction of CO2 emission; environmental protection expenditures on the effect about the total factor energy efficiency is weak, and that nationalization degree is the adverse factor to the improvement of the total factor energy efficiency; （3） Elimination of environmental variables that influence, the average technological efficiency of industrial enterprises comprehensive decreases from 0. 803 to 0. 747, the average pure technical efficiency increases from 0. 825 to 0. 951, while the average scale efficiency decreases from 0. 974 to 0. 786 ,and the provinces of decreasing reward to scale were adjusted for increasing reward to scale; and （4） after the bias correction by Bootstrapped DEA method in four section of DEA method to get the efficiency scores, the total factor energy efficiency of all regions declines.