【作者】 袁凯；

【导师】 陈建军；

【作者基本信息】 浙江大学 ， 产业经济学， 2013， 博士

【摘要】 本文旨在阐述和验证基于企业异质性的产业空间分布动力机制,并反思和提出适当的政策建议。围绕这一中心任务,本文沿着“理论梳理→现象总结→理论分析→实证验证→政策建议”的技术路径,从以下方面展开：第一,系统梳理和比较现有理论关于产业空间分布动力机制的研究和认识,总结其演进脉络,反思其可能存在缺陷,寻找理论分析切入点,为全文提供理论支撑。从大地理范围视角,传统区位理论和新经济地理学主要对产业空间分布机制提供两种动力解释：一是基于自然地理条件的禀赋优势,二是基于动态集聚的金融外部性。从小地理范围视角,转向后的传统经济地理对对产业空间分布机制提供的动力解释是基于制度关系等非经济因素的知识外溢效应或技术外部性。尽管二者存在一定差异,但有两个相同点：同样属于相对企业的外部动力、同样以企业同质性或代表性为基本假定。这显然与经济活动基本现实相悖：企业决策不仅依赖外因,内因才是最基本的源动力；多样性和差别化是企业的最基本特点。脱离这一认识,可能直接导致理论构建出现基础性偏误或认知缺陷。新近兴起的新新经济地理学为此提供了新注解,从微观层面的研究表明企业异质性将导致企业区位决策差异。第二,系统梳理和总结产业空间分布历史路径、现实轨迹与基本形态的特征,分析其演化规律,探寻理论认识和实践现实之间的背离或差距,为全文提供现实背景和实践支撑。从国内外历史视角对产业空间分布演变轨迹的总结发现我国产业空间分布有两个基本特点：非均衡性和动态演化性。进一步从对中国当前产业空间分布形态进行比较后提出,产业分布的空间非均衡性具有“二重性”特征,即“产业份额”与“生产效率”耦合的二重“中心-外围”结构。这一认识与传统理论下产业空间分布的产业份额“一维性”具有本质区别。第三,在总结理论和现实不足的基础上,提出本文基本的逻辑框架和全文理论路线,并计划从基于产业份额(企业同质性)和基于生产效率(企业异质性)两个视角展开,为全文立“骨架”。第四,分别通过理论和模型分析了二重空间的形成与演化机制,并进行了对比,为引出实证检验的假说做铺垫。首先,本节沿着产业空间分布演化的时间序列,分析了同质性企业视角下的产业空间分布动力和机制作为后文分析的背景和参照,发现：传统理论主要提供的三种动力分别是禀赋优势、金融外部性和制度与政策。三种动力作用路径都是“经济关联”,即通过基于市场的物质和服务交换实现的,属于企业外部动力。企业同质导致了企业内部行为差异缺位。其次,在企业异质条件下,先通过一个基本模型分析认为选择是产业空间分布的新动力,并利用一个Hotelling模型分析了选择效应作用基础——基于生产效率差异的竞争门槛效应机制,然后进一步通过对影响企业竞争行为的异质性两个维度进行拓展性的分解,验证了竞争门槛效应的作用机制。这一机制本质上是与“经济关联”既有区别又有联系的“知识关联”机制。最后在对上述两个不同动力机制和作用路径进行对比性分析,发现传统动力机制是二重性空间形成路径的做成部分之一,基于企业异质性的选择效应是形成路径的关键组成部分,二者分别都通过经济关联和知识关联作用于生产效率,进而影响企业区位决策和产业空间分布。第五,首先对二重空间形成机制或影响因素进行了总结,并提出了待检验的实证假说,然后通过计量和案例进行实证检验,结果支持了本文基本假说。本文核心假说是：产业空间分布与生产效率之间呈现“二次项”关系(倒U型结构)即随着“中心”生产效率提高,产业空间分布表现出先集聚、再扩散的空间轨迹。然后分别通过一个计量模型和一个实证模型进行检验。在面板模型中,引入滞后一期的劳动生产率平方项作为选择机制(倒U型关系)的工具变量,并利用我国数据进行检验,结果验证了理论假说。并在此基础上大致推算了我国产业空间分布转折的“门槛”,大致出现在2002年前后。尽管这与产业空间分布出现转折的观点存在滞后的情况,但可能是由于企业耐受度的差异和我国可能存在劳动生产率高估造成的。总体上,结论是符合我们预期的。同时也借鉴现有成熟方法对传统因素进行了简单面板回归,以便于前者进行比较。结果也是肯定的支持这些因素的作用。必须指出,不管是那种分布模式,其演化基本起点都是从最初的非均衡开始的。第六,基于理论和实证分析,总结全文核心观点、理论创新,展望未来研究方向,并反思两种条件下政策模式,提出了适当建议。在政策建议中,简单分析了两种作用机制下政策模式效果和预期,发现对现实存在的“污染避难所”和地区产业发展效率差距扩大的情况,传统机制不能提供有效解释,甚至可能是其中推手。因此,有必要从区域和产业政策进行深入反思,不仅要引导产业空间布局优化,更要从更高视角进行适度的制度安排,通过引导劳动合理分布实现资源整体配置更有效和区域发展更协调。第七,针对本文理论模型、实证模型和政策建议都比较粗糙的实际情况,本文认为对未来的研究方向应该是对理论机制和视野进行进一步拓展,对实证的计量方法(比如动态)、工具变量(比如选择更好变量)和数据(比如,时间周期更长、数据指标更完备或微观企业等方面数据)进行拓展。针对两种机制不同的政策方向,需要深入从企业异质性视角探索可行的政策模式,更理性和真实的反应产业空间分布规律,从而经济活动整体效率提高提供支撑。更多还原

【Abstract】 This paper aims at explain and verify the dynamics and mechanism of the spatial distribution of industries based on firm heterogeneity,then reconsider and propose proper policy strategy.For this purpose,it carries out as following technological path,"theory exploring→practice summarizing→theoretical→analyzing→empirical verifying→policy advising".(i).Systematically comparing and exploring current research on the dynamic mechanism of he spatial distribution of industries,summarizing its evolving paths and possible defection, thus to find proper breakthrough point and provide theocratical support.Traditional Location Theory and New Economic Geography has provided two explanations:Endowment advantages based on natural and geographical conditions and pecuniary externalities based on dynamic agglomeration. Direction-turned traditional Economic Geography has provided one explanation from rather small geographical sense,that is:knowledge spillover or technology externality. Despite the above two explanations differ in some sense,it has two commons:both are external drive force for firms and both take heterogeneous or representative firm as basic condition, which are apparently against with the practice:the inner factor should be the very fundamental fore rather than external ones;diversification and differentiation are the basis features of firms. Beyond above recognition may cause base bias or defections in theory structure.The newly emerged New New Economic Geography has provided new explanation for this,that is,research shows that firm heterogeneity will cause firm’s location choice difference.(ii).Systematically summed the historical path,current trace and basic pattern for the spatial distribution of industries,analyzed its evolving rules,exploring the difference between theocratical recognition and practical situation,thus to provide the practical background and support.It found that China’s spatial distribution of industries had two fundamental features,unbalancing and dynamic evolving. Feather work show that the "Core-Periphery" structure of spatial distribution of industries is dual-coupled by industry share and productivity.This is essentially different from one dimension of industry share by traditional theory.(iii).It proposed the logic framework and theoretical path for the whole dissertation on the basis of above theoretical and practical summarizing,and planned to extend from two view points in terms of firm homogeneity and heterogeneity. (iv).Analyzed and compared the dynamic mechanism in term of above two viewpoint by theoretical models,it put fro ward the hypothesis to-be-verified. It firstly analyzed the mechanism in homogeneous firm and found that the three traditional forces including endowment advantage,pecuniary externality and institution and policies,which work through economic linkages(E-linkages) in terms of market exchange of materials and services, were external ones.Firm homogeneity caused the lack of interaction inside the firm.It secondly analyzed the dynamic mechanism with heterogeneous,in which selection is the new driving force that is based on the competitive thresholds in terms of productivity difference. Then it extended to discompose heterogeneity in two dimension thus to verify how the "threshold" work through,which is both differed and linked by Knowledge linkages(K-linkages) with E-linkages. And finally,it compared the above dynamics and mechanism and proposed the hypothesis to be checked,that is,the relation between the industrial spatial distribution and productivity showed as quadratic or inverted U shape. That means with the increase of productivity in the center,the spatial trajectory of industrial distribution showed to be:agglomerating in the beginning and then dispersing.It should be pointed out that the starting point of evolving began from a imbalanced point in either models.(v).It then verified the core mechanism with econometric and case analysis and supported the above hypothesis.Using panel model,it firstly checked the factor of the dynamic with homogeneous firm including endowment,pecuniary eternality and policy,which turned out to be in accordance with our expectation. The it checked the case with heterogeneous firm with China’s panel data,in which the lagged quadratic term of productivity instrumented the selection mechanism(inverted U shape).The result valid supported the hypothesis,with which it calculated the threshold in China that turned up around2002.The lagging may be caused by the firm’s adapting capability and the overestimation of the labor productivity,(vi).Finally it summed the core ideas,theoretical innovation based on above analysis and expected the future research. In the policy suggestions,it analyzed the policy model and effects in term of the two different mechanisms and found the existing "pollution haven" effect and the extending of regional disparity,which is not properly explained by the traditional mechanism. Therefore,it is necessary to rethink the regional and industrial policy,which has to have proper industry allocation in regions as well as correction of institution arrangement,thus to guide labor allocate resource in more efficient way. (vii)Though it has achieved the goal we set in the beginning,it is still a primary work in term of theoretical model,empirical verification and policy research.Then the future direction will be further extending over above three dimension,eg:econometric methods(using dynamic panel),proper instrument variables(there might have more fit ones),data(in rather long time period or more complete date like firm level ones). According to the above two policy directions,it needs more policy research in heterogeneous firm condition,thus to reflect the rules of industrial distribution and to support the whole productivity activities.更多还原