【作者】 李君华；

【导师】 徐长生；

【作者基本信息】 华中科技大学 ， 西方经济学， 2007， 博士

【摘要】 本文的研究目标是在全球化背景下研究产业集聚和布局的一般机理,并试图以此解析中国制造业布局的状况和产业的演变过程。在对影响产业布局和集聚的各种因素进行研究之后,本文对中国制造业分布的集中度进行了测算,对集中状况的影响因素和它对地区产业增长的贡献进行了实证检验。在研究方法上,本文采用了规范的经济学模型和实证分析相结合的方法。在实证研究部分,本文采用了地理集中基尼系数、单位根检验和协整检验、面板数据回归等方法。本文的主要研究结论如下:1.产业集聚可能外生于资源禀赋,也可能内生于各种外部性收益和递增报酬,反过来,集聚又可能增强各种外部性收益和递增报酬的强度,也就是说,集聚与影响集聚的内生因素之间存在因果累积循环效应。2.集聚均衡是由各种向心力和离心力相互作用的结果。向心力因素主要有:知识溢出或学习效应、规模经济、范围经济、多样化经济、需求外部性、地方性公共物品和公共政策、社会资本、比较优势和要素禀赋、人力资本及其产权、外商直接投资等。离心力因素主要有:空间距离、运输成本、拥挤效应和负外部性等。从自给自足均衡、到综合城市均衡、再到地区专业化均衡的过程是由地区内和地区之间的运输成本与学习效应强度系数(或其它自增强的内生因素)的增减变化所推动的。一旦地区专业化人数的规模达到某一临界点,即使各个参数保持不变,生产系统也会进入一个持续的转移动态过程,直到下一个稳定均衡。3.知识溢出或学习效应对于产业集聚具有特别的重要性。知识溢出本身也要受到多种因素的影响,这些因素包括:个人学习和接收能力、地区文化和语言特征、知识溢出方与接收方之间的地理距离、双方专业背景的相似性、地区的知识存量规模、被溢出知识本身的性质和编码化程度、交流频率和交流密度等。4.当地方政府在特定中心地区投入公共物品时,为了获得消费该项公共物品的权利和争取到其周围的有限土地,在人口和厂商向该地区集聚的过程中会存在竞争和拥挤。生产者和消费者从公共物品中获得收益对于距离的的敏感性导致了人口的迁移和产业集聚,而拥挤则保证了均衡的存在。5.全球化与本地化、专业化与多样化是两对并行不悖的概念。越是全球化高的行业,其本地化程度越高;越是多样化的工业生物圈,其核心能力的专业化程度越高。如果想要在本地发展有竞争力的产业,则必须从产业全球化的视角来考察;如果想要在本地发展多元化的工业圈,则必须从核心能力共享和专业化的角度去思维。6.通过运用地理集中基尼系数对1994—2003年间中国17个制造行业的绝对集中度进行测定。结果表明,中国17个制造行业的地理集中度都比较高,其中电子通信设备制造业、仪器仪表及文化办公机械制造业、化学纤维制造业是的集中度最高。从地区看,绝大多数行业的绝大部分都分布在广东、江苏、浙江、上海和山东等东部沿海省市。从时间上看,制造业向东部集聚的趋势一直都在增强,除食品制造、饮料制造、烟草加工和医药制造业之外,其余14个行业的地理集中度基本上都保持了上升。采用行业专业化系数对17个制造行业相对集中度进行测定的结果表明,大多数行业的专业化水平和工业结构的差异程度(相对集中度)都比较低,但烟草加工业、电子通信设备制造业、化学纤维制造业的相对集中度相对较高。从动态的观点看,除烟草加工业之外,其余16个制造行业的行业专业化系数都保持了上升,其中化纤制造、食品加工业上升最快。7.根据1952-2005年各省工业增加值数据,运用截面数据回归对中国各地区之间的产业转移和集聚现象进行分析的结果表明,中国工业在不同的发展时期具有不同的特征:改革开放以前是以产业分散占主流,20世纪80年代以调整为特征,而90年代则出现了从内地向沿海地区集聚的趋势。从单位根检验的结果看,江西、陕西和新疆三省基本保持了平稳的发展态势,而其它省市区之间则存在转移和此消彼长的关系,但西部各省的变化较小。从协整检验的结果看,除了西部内部各省区之间可能存在产业分散和均匀化的趋向外,其它区域内部各省区之间和全国各大区域之间均服从由随机因素和共同因素同时主导的随机性过程。8.利用2002年度中国省区/行业二维数据对中国制造业布局影响因素建模的回归结果表明,比较优势、新经济地理、知识外溢以及地方保护主义等因素对于中国制造业分布都具有重要的影响,其中新经济地理和知识溢出效应在集聚过程中具有因果累积循环性质。9.利用1999—2003年度中国29个省区、20个制造行业的三维动态面板数据对中国各地区行业经济增长的影响因素进行回归的结果表明:行业地理集中、地方专业化对于地区产业增加值的增长具有显著的正向贡献,但对于其人均收入增长的贡献显著性不够高;地区全部工业的集中度与地区产业增长、人均收入增长呈非线性关系;工业多样化对于各地区产业增加值和人均收入增长的影响不明显;企业规模对于两者具有显著的正向影响;各地区产业的竞争活跃程度与地区工业的增长存在显著的正向关系,但对于人均增长的影响不明显。更多还原

【Abstract】 This dissertation aims to probe into the general mechanism of industry agglomeration and distribution under the background of globalization and, based on which, to analyze the evolution process of Chinese manufacturing industries and its distribution conditions. It investigates the determinants of industrial distribution and agglomeration, and calculates current centralized situations of China’s manufacturing distribution, and testifies empirically both the influences of industrial concentration degree and the contribution of centralization situation to regional industry growth. The methodology adopted in this dissertation is to combine normative economic models with empirical analysis. In empirical study, the methodology includes Gini coefficient, the unit roots test and cointegration test, panel data regression.The main conclusions of this dissertation are as followings:1. Industry agglomeration is possibly exogenous from resources endowment, or is possibly endowed in various kinds of external economy and the increasing returns. In turn, agglomeration possibly strengthens the intensity of external economy and the increasing returns. In other words, agglomeration and these endogenous factors have an effect of circular and cumulative causation.2. Agglomeration equilibrium is the result of the interaction of centrifugal force and centripetal force. The centripetal force factors mainly include: Knowledge spillovers or learning-effect, economy of scale, economy of scope, diversifying economy, demand externality, local public goods and public policy, social capital, comparative advantage and resources endowment, human capital and its property right, FDI, ect.. The centrifugal force factors mainly include: spatial distance, transporting cost, congested effect and the negative externality, and so on. The process from the autarkic equilibrium to the city complex equilibrium and to the regional specialization equilibrium is propelled by the changes of transport costs and learning-effect intensity (or other self-increasing endogenous factors) of regions and inter-regions. Once the number of regional specialized population reaches to a certain point, even if each parameter remains unchanged, the production system will enter into a continuous dynamical transition process till the next steady equilibrium.3. Knowledge spillover or learning effect has special importance to industry agglomeration. Knowledge spillover itself is influenced by many factors, including personal learning capability and receptivity, regional culture and language characteristics, geography distance between the sender and receiver of knowledge spillover, similarity of their professional background, volume of regional knowledge stock, the character and coding degree of knowledge-spilled, frequency and density of intercommunion.4. When local governments input public goods in special central-place, in order to obtain rights and scarce land around them, there is competition and congestion among population and manufacturers in the process of agglomeration. The incomes that producers and consumers earn from public goods are sensitive to distance, which leads to population transfer and industry agglomeration, and congestion ensures equilibrium existing.5. Globalization and localization, specialization and diversification are two compatible concepts. The higher globalization the industry is, the higher localization is; the more diversified the industry biosphere is, the more specialized the core capacity is. If you want to develop a competitive local industry, you should consider things at a global view; if you want to develop a diversified industry cycle, you should think from angle of core capacity sharing and specialization.6. Through the geographic concentration Gini index defined by author, this dissertation measures the geographic concentration degree of China’s 17 manufactures from 1994 to 2003. The result shows that the geographic concentration degrees of China’s 17 manufactures are relatively high, among them, electro-communications equipments manufacturing, apparatus, instrument, culture facility machine manufacturing, chemical fiber manufacturing are line in front, distributing mostly in some eastern provinces along sea, including Guangdong, Jiangsu, Zhejiang, Shanghai and Shandong. Seeing about the time-processing, except foodstuff, beverages manufacturing, medical and pharmaceutical products and tobacco processing, the geographic concentration Gini index of other 14 manufactures keep boosting up. Using industries specialization coefficient, the result of the tests of relative concentration degree to 17 manufactures indicates that the difference degrees (relative concentration degree) of specialization lever and industry structure of most industries are relatively low, except tobacco processing, electro-communications equipments manufacturing, chemical fiber manufacturing industry. From a dynamic view, industries specialization coefficient of 16 manufactures excluding tobacco processing industry keep rising, among them, chemical fiber manufacturing, foodstuff process industry increase rapidly.7. The auther applies the panel data regression to analyze the industrial transfer and agglomeration among regions on the basis of the industrial value-added of China’s provinces from 1952 to 2005, and it draws the conclusion that the industrial transfer among China’s regions has different characteristics in different phases——before the reform and opening-up, the mainstream of industrial development was the decentralization, the adjusting in 1980’s and tended to centralize the littoral in 1990’s. In view of the unit roots test, except Jiangxi, Shanxi, and Xinjiang are balancedly rising, there exist transfer and relation of ebb and flow in the rest provinces, but the change degree of the Western provinces is little. In light of the cointegration test, except the tendency of industry decentralization existing among the western internal provinces and areas, the national areas and the internal provinces of other districts obey the stochastic process dominated together by stochastic factor and common factor.8. The auther estimates a model of determinants of China’s manufacturing industry distribution, using two-dimension data of province/industry of 2002. The regression result shows that comparative advantage, new economic geography, knowledge spillover and regional protection ect., have much influence in China’s manufacturing industry distribution, among them, new economic geography and knowledge spillover have an effect of the circular and cumulative causation in the agglomeration process.9. Using three-dimension dynamic panel data of 29 provinces and 20 manufacturing industries from the year of 1999 to 2003, the author analyzes determinants of economic growth of China’s regional industries. It is evident that both industrial geography agglomeration and regional specialization have significant and positive effects on regional industry growth, but insignificant on per capita income. Concentration degree of regional entire industries is non-linearity with regional industry growth and per capita income growth. The industrial diversification has insignificant effects on both regional industrial value-added and per capita income growth, but enterprise scale has significant positive effects on them. Competitive activity grade of regional industries is positively associated with industrial value-added, but is insignificantly associated with per capita income.更多还原