【作者】 贺胜兵；

【导师】 唐齐鸣；

【作者基本信息】 华中科技大学 ， 数量经济学， 2009， 博士

【摘要】 改革开放三十年以来，我国取得了举世瞩目的巨大成就，经济持续快速增长，国家综合实力显著增强，人民生活水平有了很大的提高。但是长期以来，我国经济增长主要沿袭了一条“高投入、高消耗、高污染、低效益”的粗放型增长模式。一方面，随着我国逐渐成为世界制造业中心，特别是近年来重化工业的迅猛发展，导致我国资源供给全面紧张；另一方面，不断增长的环境污染物和温室气体排放不仅造成了巨大的经济损失，而且使我国面临的全球环境压力不断加大。随着人口总量的继续增长、城市化的快速推进、居民消费水平的提高和工业行业的持续增长，未来我国面临的资源环境约束势必更加严峻，因此改变当前不可持续的增长方式，转向经济增长与人口、资源、环境相协调的可持续发展模式，已经变得比以往任何时候都更加紧迫。在宏观层面上，我国政府已经确立了可持续发展的国家战略，并提出了科学发展观和构建和谐社会的理念，积极推动国民经济又好又快地发展。以上这些，构成了本文的研究背景。由于我国仍然是一个发展中国家，需要足够的发展空间以解决13亿人的温饱问题和就业问题，因此研究中国的环境问题，必须将经济增长和环境保护结合起来考虑。本文将能源、环境因素纳入到生产率分析框架，采用多种前沿分析方法估计考虑能源投入的省级技术效率，采用方向性距离函数法计算考虑环境因素的省级全要素生产率指数并将其分解为技术效率指数和技术进步指数，进而采用内生分组的面板平滑转换模型(PSTR)考察省级生产率的影响因素。通过实证分析，主要得出以下结论：(1)省级能源效率呈上升趋势，能源投入对省级技术效率有显著的影响。作为重要的生产要素，能源在为经济社会发展提供动力的同时，也带来生态破坏和环境污染问题。本文计算了1995～2007年各省的全要素能源效率，并采用CSSW、BC、KSS、DEA等四种模型计算考虑能源投入的省级技术效率，结果显示，样本区间内各省全要素能源效率总体上呈上升趋势，但是2006年以后出现了下降。无论是否考虑能源投入，我国省级技术效率都呈下降趋势。考虑能源投入之后，各省技术效率排名发生了变化，新的排名有利于更加准确地反映经济增长的质量。通过对不同模型估计结果的比较分析发现，KSS和DEA模型表现出了较好的时变性和稳健性特征。(2)经济增长的环境绩效持续改善，全要素生产率对经济增长的贡献偏低。采用方向性距离函数和DEA模型，计算了不考虑环境因素、单独考虑COD、SO2、固体废弃物及同时考虑这三种污染物时的省级TFP指数，并将TFP指数分解为技术效率指数和技术进步指数。在上述5种情形下，样本区间内我国各省TFP平均增长率分别为—0．2％、1．36％、1．04％、1．23％、2．30％。考虑环境因素后TFP增长率明显上升，这与各污染物的排放强度下降的情况相一致，表明近年来我国经济增长的环境绩效提高了；当考虑环境因素时，TFP对同期GDP增长(9．38％)的贡献率分别为14．50％、11．09％、13．10％、24．52％，说明经济增长主要依靠要素投入的增加。TFP指数分解的结果显示，技术进步是TFP增长的主要动力，样本区间内各省技术效率呈下降趋势。处在生产前沿上的省份多数位于东部地区，中西部地区的省份很少。(3)无论是否考虑环境因素，省级生产率均不存在全国范围的绝对收敛。分别对同时考虑三种污染物和不考虑环境因素的省级生产率进行绝对收敛检验、俱乐部收敛检验和条件收敛检验。研究发现，省级TFP指数存在发散趋势，省级技术效率指数表现出了收敛趋势，省级技术进步指数没有表现出收敛趋势。无论是否考虑环境因素，代表东、中、西三大区域的虚拟变量均不显著，表明不存在俱乐部收敛；当所考虑的控制变量为基础设施水平、工业化水平、经济开放度、能源强度时，所有的控制变量皆不具有统计显著性，表明不存在条件收敛。基于核密度估计的累积相对生产率的分布动态演进进一步印证了收敛系数检验的主要结论。(4)省级生产率具有明显的异质性，各变量对生产率的影响具有非线性特征。在TFP指数估计结果的基础上，采用线性固定效应模型和非线性的PSTR模型，考察劳均资本存量等9个因素对省级生产率的影响。以工业结构变量为转换变量的PSTR模型估计结果表明，我国省级全要素生产率具有显著的异质性特征。总的来看，对外经济交往增加、国民收入增长、基础设施改善、地区大中型企业例提高有助于提升地区TFP，资本有机构成提高和重工业比例增加会降低地区TFP，而能源强度在不同规模结构、第二产业比例在不同的轻重产业结构下对TFP的作用存在明显的不同。基于以上实证分析结论，本文主要提出以下建议：(1)提高能源利用效率，积极发展清洁能源和可再生能源；(2)加快产业结构调整，大力推进企业技术进步和自主创新：(3)推动跨区域的经济交流，缩小生产率的地区差距；(4)全面贯彻落实科学发展观，因地制宜地提高地区全要素生产率。更多还原

【Abstract】 Since the Reform in 1978, China achieves great success in rapid economic growth, in the strengthening of China’s comprehensive power and in the improvement of people’s life. But for a long time, China’s economic growth mode is high input, high consumption, high pollution and low efficiency. On one hand, China has become world manufacturing center, and her resources supply is under tense situation, especially for the rapid development of heavy and chemistry industry. On the other hand, the ever increasing environment pollutions and greenhouse air emission make a great economic loss, and China faces more pressure from the world for the environment pollution problems. China’s resources and environment restrictions are much harsher in the future, just for the growing population, the rapid urbanization, the households’ growing consumption level and the industry continues to grow. So it is more urgent than ever to change the unsustainable growth mode, and to turn to the sustainable development mode which harmonizes economic growth with population, resources and environment. On the macro level, Chinese government has established the country strategy of sustainable development, and has put forward scientific development theory and the concept of constructing harmonious society, to ensure a sound and rapid development. All these constructing the research background of the dissertation.As one of the development countries, China needs enough development space to ensure that the 1.3 billion people have adequate food and clothing and have employment of the 1.3 billion populations. So we must combine economic growth with environment protection in researching the topic of China’s environment. The dissertation introduces resources and environment into provincial productivity research, and employs frontier analysis methods to investigate the energy input’s influence on technical efficiency; The dissertation also employs directional distance function to compute province-level’s all-factors productivity index, and decompose it into technical efficiency index and technical progress index; Then, the dissertation employs panel smooth transition model which is characteristic of endogenous grouping to investigate the influence factors of province-level’s productivity. Based on the positive analysis, the dissertation makes conclusions as follows:(1) China’s provincial energy efficiency is in a upward trend, and the energy input has a significant influence on provincial energy efficiency. As an important production factor, energy not only provides the driving force for economic and social development, on the other hand, but also brings forth ecology destruction and environmental pollution problem. The dissertation calculates the provincial total factor energy efficiency from 1995 to 2007, and applys four approaches, i.e., CSS, BC, KSS and DEA, to calculate provincial technical efficiency which contains the energy input variable. The result shows that China’s provincial technical efficiency is generally in a downward trend within the sample period, but it has declined since 2006. Provincial technical efficiency is in a downward trend whether including energy input or not. The provincial technical efficiency rank changes after including energy input. The new rank is good for reflecting the quality of economic growth more accurately. The dissertation make a comparison among the results of different models, the KSS and DEA model shows good time-varying and robust characteristics.(2) China’s environment performance of economic growth has improved persistently, and the TFP’s contribution to the economic growth rate is quite low.The dissertation employ directional distance function and data envelope analysis to calculate total factor productivity index under three circumstance: First, calculates the provincial total factor productivity index without considering any pollutants; Second, calculates the provincial total factor productivity index incorporating COD, SO2 and solid pollutants respectively; Third, calculates the provincial total factor productivity index incorporating all the three environmental pollutants. And then the dissertation decomposes the TFP index into technical efficiency index and technical progress index. Under all the three above circumstances, China’s provincial TFP rate is respectively -0.2%, 1.36%, 1.04%, 1.23%, and 2.30% during the sample period. The TFP index increased obviously after incorporating the environmental factors. This result is consistent with the fact that the intensity of pollutants’ emission declined, which shows that China’s environmental performance improved in recent years. The TFP’s contribution to the economic growth rate of the corresponding period is respectively 14.50%, 11.09%, 13.10%, 24.52% when incorporates environmental factors, the main source of economic growth is the increase of factor input. The decomposition of TFP index indicates that technical progress is the primary driving force of the TFP rise, and provincial technical efficiency is in a downward trend. Most of the provinces which are on the productivity frontier lie in the eastern areas, few provinces lie in the middle or western areas.(3) There exists no country level absolute convergence of the provincial productivity whether incorporating environmental factors or not.On the basis of provincial TFP index and its decomposition, the dissertation furthermore investigates the provincial productivity’s regional difference. It carries out absolute convergence test, club convergence test and conditional convergence test on the provincial productivity incorporating all the three environmental pollutants and incorporating no environmental pollutants respectively. It also applies kernel density estimation method to investigate the accumulative relative productivity and the dynamic evolvement of its distribution. The result shows that there exists no country level absolute convergence of the provincial productivity whether incorporating environmental factors or not. The provincial TFP index is in a divergence trend and the provincial technical efficiency index has a convergence trend when incorporates environmental factors. The distribution dynamic evolvement which is based on kernel density estimation further verifies the main conclusions of the coefficient convergence tests.(4) Provincial productivity has obvious heterogeneity, and the relationship between productivity and some variables has nonlinear characteristics.On the basis of TFP index estimation, the dissertation employs linear fixed effect model and nonlinear panel smooth transition model to investigate the influence of the 9 variables on provincial productivity including average per capita substance capital, etc. The PSTR models which regard industry configuration variables as transition variables indicate that China’s provincial TFP has significant heterogeneity. In conclusion, the increase of foreign economic relations, the national income increase, infrastructure improvement and the increase of the large and medium enterprises’ rate is beneficial for improving regional TFP, the improvement of organic composition of capital and the increase of heavy industry’s rate will reduce regional TFP. The energy intensity under different industry scale structure and the secondary industry’s rate under different light-heavy industrial structure have different effects on TFP.Based on the empirical study, the dissertation puts forward the following measures: (1) Improving energy efficiency, vigorously developing the clean energy and renewable energy. (2) Speeding up the adjustment of China’ s industrial structure, promoting technical progress and autonomous innovation of enterprises. (3) Promoting cross-regional economic exchanges, narrowing the productivity gap between regions. (4) Carrying out and fulfilling the concept of scientific development, improving regional total factor productivity in the light of local conditions.更多还原