【作者】 李科；

【导师】 王少平；

【作者基本信息】 华中科技大学 ， 数量经济学， 2013， 博士

【摘要】 2006年以来，节能减排成为中国可持续发展战略的重要组成部分。从中国现实看，产业结构是经济能耗的重要来源，而产业结构调整和升级是实现节能减排的重要措施和政策手段。从文献资料看，有关中国产业结构与节能减排关系的研究主要采用线性回归（协整）模型或指数分解模型。然而，中国经济的周期波动和政府宏观调控使产业结构反复波动，具有典型的非线性特征，这导致已有的研究结论常与现实不一致，其政策建议也存在局限性。本文针对中国能源变量的数据特征和能源经济相关理论，应用前沿且适用的能源经济和计量经济的理论和方法，展开对上述问题的研究，以期产生具有明显的现实与政策意义的结论。全文的研究工作和结论及其创新意义概况如下：（1）应用不同的理论和方法的研究结果均表明，中国不同时期的产业结构（以工业增加值占GDP的比重来度量）对能源强度产生的效应不尽相同。以中国产业结构与能源强度的数据演变特征为依据，构建了一个非线性阈值协整模型，研究结果表明，在产业结构处于40.435%处，产业结构对能源强度的长期效应发生非线性转移，并可由逻辑函数刻画其转移特征。具体而言，在工业增加值占GDP的比重向下调整的1983-1994年、1998-2002年和2009年间，产业结构对能源强度具有微弱间断的结构红利；而在1980-1982、1995-1997和2003-2008年间，产业结构体现了明显耗能的特征。这一结论表明，始终不渝的调整结构、加快经济增长方式的转变，是实现结构具有持续节能效应的长期战略；而在短期内，建议将工业占GDP比重调整到40%以下。与已有研究文献相比，本文的研究结论明确了产业结构调整的方向与幅度。（2）基于Kaya恒等式所反映的理论内涵和变量关系，尤其是针对中国经济增长和结构调整的非线性波动特征，应用非线性模型揭示了结构调整对碳强度的影响受制于经济增速。具体来说，当经济增速高于9.053%时，产业结构和能源消费结构的调整均不利于碳强度的下降。在此基础上，基于估计的模型而应用依赖于蒙特卡洛模拟的情景分析，结论表明，中国碳强度年均下降速度的可能值是4.34%（2011-2015年）和3.51%（2016-2020年），这意味着碳强度2020年同比2005年预期将下降41.19%；而将经济增长速度调控到适度区间（约为7%~8.4%），有助于碳减排的可持续性。显然，非线性模型的设定及其实证分析充分体现了中国的经济背景和现实特征，而基于仿真的情景分析，消除了以往文献中简单的三种情景分析中的主观性因素，其分析结论具有稳健性和合理性，因此，上述研究结论对中国未来实施碳减排具有重要的应用价值和实际意义。（3）中国不同省份的资源配置方式具有显著差异，已有文献在分析各省的能效水平时常忽略了这一差异。本文充分考虑各省资源配置方式的异质性，以产业结构合理化水平为阈值变量，运用阈值效应随机前沿模型分析了各省的全要素能源效率。检验和估计结果表明，各省的经济增长存在三个技术俱乐部，且产业结构逾合理，全要素能源效率值逾高。不同技术俱乐部下产出增长率的分解结果显示，产业结构逾不合理，要素投入，尤其是资本投入对经济产出的贡献度逾高，而技术进步对经济产出的贡献度逾低。上述结论表明，改变依赖要素投入的粗放型经济增长方式，依靠技术进步而提高产业结构合理化水平，是提高能源利用效率的有效途径。（4）考虑能耗产生的碳排放污染，本文采用超效率DEA模型而估计环境方向距离函数以改善效率前沿面省份的节能减排效率值；基于估计结果运用面板数据模型考察投资驱动的经济增长方式、产业结构调整对节能减排效率的影响。结果表明，投资驱动的经济增长不利于节能减排效率的提高；劳动力要素的产业间流动有助于提高节能减排效率，而资本的产业间转移则相反；制造业结构变动和升级无助于提高能源效率，但有助于提高节能减排效率。这些结论表明，转变经济增长方式、破除产业间壁垒而促进要素流动和转移、坚持引进技术的消化吸收和自主创新而提高制造业结构的高度化水平是可持续性节能减排的长期途径。与已有文献不同，本文从产业结构合理化和制造业结构高度化为实施节能减排的结构调整政策提供了新视角和新证据。更多还原

【Abstract】 Since2006,―energy conservation and emission reduction‖has become an essentialpart of China’s sustainable development strategy. From historical data, the heavierindustrial structure is a significant source of energy consumption, and industrialadjustment and upgrading is an important entry point and main measure for governmentsto implement―energy conservation and emission reduction‖. From the previousdocuments about the relationship between industrial structure and energy/carbon intensityof China, the main methods are linear regression model and index decomposition methods.However, because of economic cycle and macro-control, China’s industrial structure has atypical non-linear characteristics, which led to the existing research conclusions are ofteninconsistent with the reality, and its policy recommendations have limitations. Based onthe data characteristics and the theory of energy economy, this thesis applied latest andproper econometrical models to analysis the relationship between industrial structure and―energy conservation and emission reduction‖of China. This research is expected toproduce some conclusions in line with the actual situation of China, and give someinspiring policy recommendations. The main research contents and conclusions and itsinnovative profile are as following:(1) Based on the results of different theories and methods, the effects of industrialstructure (measured by the ratio of industry added value to GDP) on energy intensity(measured by energy consumption per unit of GDP) are variously in different periods ofChina. According to data features between industrial structure and energy intensity during1980-2009, this paper applied a threshold cointegration model to analysis the nonlinearrelationship between them. The results exposes that industrial structure has nonlineareffects on energy intensity when industrial structure is around40.435%: during the periodsof1983-1994,1998-2002and2009, the industry/GDP was declined, and industrialstructure produced negative effects on energy intensity weakly and discontinuously; during the periods of1980-1982,1995-1997and2003-2008, the adjustment ofindustry/GDP was not conducive to the reduction of energy intensity. This conclusionsuggests that readjusting the industrial structure and transforming the pattern of economicgrowth are long-term strategies for sustainably pushing the decline of energy intensity. Inshort-term, it suggests readjust the industry/GDP to below than40%. Compared with theexisting research literatures, the conclusions of this study clearly shows the direction andmagnitude of structural adjustment in China.(2) Based on the Kaya identity and the data characteristics of China’s economicgrowth and structural adjustment, a nonlinear model was tested and estimated in order toreflect whether the effects of structural adjustment on carbon intensity were subject toeconomic growth or not. The results show that if the rate of economic growth is higherthan9.053%, industrial structure and energy mix were not conducive to decline in carbonintensity. Furthermore, the scenario analysis which depends on the Monte Carlosimulation show that the expected values of the average annual decrease rate of China’scarbon intensity are4.34%(2011-2015) and3.51%(2016-2020). It means the carbonintensity in2020is expected to decline by41.19%compare to2005. It also suggests thatthe rate of economic growth in the interval of7%to8.4%can contribute to carbonemission reduction. Obviously, the positive analysis of the nonlinear model is fully reflectsChina’s economic background and carbon emission‘s characteristics, and simulation-basedscenario analysis eliminate subjective factors in similar studies, so the conclusion is robustand rationality. Therefore, the conclusion of the study has an important application valueand practical significance for China‘s carbon emission reduction in the future.(3) There is a significant difference of resource allocation in different provinces ofChina, but past papers about energy efficiency often ignore it. Take account of theheterogeneity of resource allocation in provinces, this paper takes rationalization ofindustrial structure as a threshold variable, and uses a threshold effects stochastic frontiermodel to analysis the total factor energy efficiency. The test and estimate results show thatthe economic growth of the different provinces has three technological clubs, and the more reasonable of industrial structure, the higher total factor energy efficiency. Furthermore,the decomposition results show that the less rational industrial structure, the highercontribution of factor inputs, especially the capital investment, on economic output, andthe lower contribution of technological progress on economic output. The aboveconclusions mean that change the economic growth pattern, and enhance the level ofrationalization of the industrial structure by technological progress are effective ways toimprove energy efficiency.(4) Because carbon emission generates by energy consumption, this paper usesenvironmental directional distance functions, which is estimated by super-DEA model toimprove the efficiency frontier provinces‘estimators, to calculate the total-factor energyefficiency (TFCE). Then, it uses some panel data models to examine the relationshipbetween economic growth pattern, which is characteristic by investment driven, theindustrial structure adjustment and the total-factor energy efficiency. The results indicateChina‘s economic growth pattern is not benefit to improving TFCE; labor flow betweenindustries produces "structure bonus" on TFCE, and capital flow between industriesproduces "structure burden" on TFCE; the change and upgrade of manufacturing industrystructure doesn‘t benefit to improve energy efficiency, but it is help to enhance TFCE.These conclusions indicate that it is long-term ways to improve energy/carbon efficiencyby transforming economic growth pattern, and getting rid of the barriers betweenindustries, and enhancing the optimization of manufacturing structure by digestion andabsorption of new international technology and self-innovation. Compared with theexisting literatures, this article provides a new perspective and a new evidence forimplement―energy conservation and emission reduction‖through industrial structure fromthe rationalization of industrial structure and the optimization of manufacturing structure.更多还原