【作者】 袁鹏；

【导师】 陈圻；

【作者基本信息】 南京航空航天大学 ， 管理科学与工程， 2008， 博士

【摘要】 现有对我国地区差异的研究文献虽已较为丰富,但仍然没有形成一致认识。已有研究在分析层次、研究方法和研究角度等方面还存在一些局限性,进一步研究仍有必要性。本文在制造业层次上分析了劳动生产率地区差异的演变趋势,并着眼于从增长绩效、结构因素、集聚效应等新视角来解释地区差异的演变成因,为认识地区差异问题提供了相对微观层面的经验证据,也获得了一些有意义的结论和政策启示。首先,本文分析了制造业劳动生产率的增长收敛性。利用动态经济空间权重矩阵,检验发现制造业劳动生产率存在显著的空间依赖性,为此,采用了空间计量收敛模型和基于空间过滤的动态面板收敛模型。结果发现,制造业劳动生产率在1988-1997年以年均大约3%的速度扩散,而在1998-2005年以年均大约4.5%的速度收敛,意味着地区差异在前期趋于扩大,而在后期趋于缩小。这不同于现有基于整体层面的结论,支持了在更微观层面上研究地区差异问题的必要性。采用收入分布法的分析验证了上述结论的稳健性。其次,在放宽规模报酬不变的假定和重新定义技术集空间后,本文利用改进的增长分解方法,对制造业劳动生产率的增长进行了分解,从增长绩效角度剖析了地区差异演变成因。本文发现,增长的主要来源在1988-1997年和1998-2005年两个时期相异,前期为资本深化,后期为技术进步;但是,增长的主导因素并非驱动地区差异演变的关键因素,表现为技术进步驱动了前期差异的扩大,而资本深化则主导了后期差异的缩小。再次,本文考察了结构因素在制造业劳动生产率地区差异演变过程中的作用。利用改造后的偏离份额分析法和方差的自然分解法,得到了结构差异在地区差异中贡献较为次要的结论。然后,从增长的结构效应出发,发现结构转变对西部地区缩小与东部地区的差异发挥了主要作用,但对中部地区缩小与东部地区的差异却没有起到主要作用。最后,本文分析了集聚效应对制造业劳动生产率地区差异的影响。对集聚效应的估计采用了面板模型,区分了静态和动态效应,消除了变量的空间依赖性。在采用空间俱乐部模型区别了集聚效应的地区差别后,结合集聚结构的实际变动,结果发现专业化效应在整个研究期间起到了缩小中西部与东部差异的作用,而规模效应只在1998-2005年对差异缩小起到了促进作用。综上,制造业劳动生产率地区差异的演变趋势在1988-1997年和1998-2005年分别表现为扩大和缩小。在前期差异扩大过程中,技术进步和行业增长效应起到了推动作用。在后期差异缩小过程中,发挥积极效应的是资本深化,而非技术追赶或技术进步,表明在收敛过程中产生作用的是新古典的资本收敛机制而非技术扩散理论的技术收敛机制;结构效应主要作用于西部与东部之间的差异缩小,意味着结构转变的收敛机制在西部发挥了作用;专业化效应和规模效应也对缩小差异起到了正面作用,这源自集聚效应的地区差别与各地区集聚结构变动差异的共同作用。以上结论验证了新古典和新增长等理论对解释我国地区增长差异和地区差异演变问题的有效性,提供了相应的经验证据。本文结论包含的政策启示是,提高中西部的创新能力、经济效率和技术吸收能力,加快结构转变,强化专业化发展,以及促进东部的产业转移等。更多还原

【Abstract】 The research papers in existence on regional disparity in China have been quite abundant. However, there is no consistent recognition. Because the papers have some localization on hierarchy of research object,research methods and standpoints, farther work on the question is rather needed. This paper researchs on the evolution of labor productivity of China’s manufacturing and makes explainations with some new views to growth performance, structural factor and agglormertaional effect. An empirical evidence form relatively microcosmic hierarchy has been afforded to understanding the regional disparity. It gains some significative results and provides several revelations on policy making.Firstly, the growth convergence of labor productivtity has been estimated to find out the evolution of regional disparity. Based on constructing dynamic economic spatial weight matrix, the marked spatial dependence in manufacturing’s labor productivity has been finded. To control the effect of spatial dependence, several spatial econometric convergence models and dynamic panel convergence model using spatial fitering data have been used. The results show that the labor productivity has disconverged by 3% in 1988-1997, but converged by 4.5% in 1998-2005, which means that the disparity has been enlarged in the first period and reduced in the second period. The results are different from that on the macro-economy and so it is need to research the regional disparity on the more micro hiberarchy.To make sure the correctness of above conclution, the income distribution method has been used, and the conclution has been proved robust.Secondly, to relax the hypothesis of constant return to scale and redefine the space of technique sets, the decomposition method of the growth of labor productivity has been improved. Using the corrected decomposition method of the labor productivity’s growth, the paper analyzes the drivers of evolvements of regional disparity in labor productivity of manufacturing, perspective on the performance of growth. The results are as follows. The main forces of the growth of manufacturing’s labor productivity are variational in the two periods that are 1988 to 1997 and 1998 to 2005. Capital deepening is the whole source of growth in the first period and technology progress is dominant in the second period. However, the drivers of the evolvements of the regional disparity are not as same as the forces of growth. Technology progress is the driving factor enlarging the regional disparity in the first period and capital deepening is the dominant factor reducing the regional disparity in the second period.Thirdly, the effects of structural factors to the evolution of reginonal dispariatiy of manufacturing’s labor productivity have been discussed. Using the rebuilt shift-share and natural decomposition of variance, the result shows that the effect of structural differences to disparity is minor. Then, after analysising the structural effect of growth, some results have been uncoverd, which structural change is a key factor to the gap reducing between the West and the East, but not a key factor to that between the Middle and the East.Lastly, the impace of agglomerational effect on the evolution of regional disparity of manufacturing’s labor producitivty has been analyzed. The agglomerational effect has been estimated. It is different from the existent researchs that panel data model has been adopted, and static effect and dynamic effect distinguished, and spatial dependence of variable eliminated. Based on using spatial regime model to disginguish the agglomerational effects between the East and the Midwest and calculating the real changes of agglomeration structure, the agglomerational effect to disparity has been estimated. The result shows that the effect of specialization has a positive impact on the gap reducing between the Midwest and the East in the whole research period, and the effect of scale impacts positively on the gap reducing only in 1998 to 2005.In conclusion, in 1988-1997 and 1998-2005, the labor productivity of regional disparity in manufacturing evolved differently, which enlarged in the first period and reduced in the second period. In the first period, the effects of technology progress and industry-growth increased disparity. In the second period, capital deepening played actively to disparity reducing, but technique’s catch up and technology progress don’t do it. It shown that the mechanism of the neoclassical capital convergence, not that of technology convergence in technology diffusion theory, worked in the process of disparity reducing. Structural effect mainly acted on the disparity deflating between the West and the the East, implying that the convergence mechanism of structural change worked in the West. The effects of specialization and scale impacted positively on the disparity reducing, which root in the conjunct impacts of the agglomerational effect and the agglomeration structure’s change. The above conclutions validate the validity of some growth theories, including neoclassical and new growth theories, on explaining regional growth dispatity and the evolution of regional disparity in China, providing some empirical facts.Based on the above conclusion, some policies and measures could be apopted: enhancing innovational ability, economic efficency and technology absorbency in the Midwest, and speeding up structural change, intensifying specialization, and promoting the industries transferation from the East to the Midwest.更多还原