【作者】 兰宏；

【导师】 聂鸣；

【作者基本信息】 华中科技大学 ， 技术经济及管理， 2013， 博士

【摘要】 在经济全球化的背景下，企业往往以产业集群的形式聚集发展，其中许多发展中国家的产业集群或集群企业以嵌入全球价值链的方式参与全球竞争并寻求升级机会。在“市场换技术”政策的引导下，全球价值链下的我国各产业部门也试图通过国际合作和技术引进的方式来快速提升技术水平和产业竞争力。然而，市场难以换来核心技术和前沿技术，许多产业集群及企业被锁定在价值链的低端环节，不仅导致企业的低利润水平，而且容易引起本地经济的“贫困化增长”。因此，低端锁定问题是全球价值链下产业升级的一个关键问题。本文试图从集群企业知识积累和学习机制的角度，利用学习障碍这一概念对全球价值链下的低端锁定现象进行解读。本文首先在国内外相关文献研究的基础上，对产业集群本地学习和全球化学习的研究现状进行了评述。接着，引入知识流动理论分析了集群知识流动和价值链知识流动的特点，在此基础上构建了基于知识流动的全球价值链下集群企业的学习机制，并结合理论分析和基于CES函数的数理推导阐述了低端企业遭受学习障碍并导致低端锁定的机理。然后，提出将政府视作突破学习障碍的外生变量，并利用我国高技术产业的面板数据进行了实证分析。最后，围绕学习障碍这一核心概念，结合大飞机产业这一典型案例的分析，提出了政府主导下突破全球价值链下低端锁定的策略。基于以上研究工作，本文得出的主要结论如下：（1）集群企业主要通过基于自主创新的内生性途径、基于集群本地学习和价值链学习的外生性途径进行学习和知识积累。其中，产业集群具备的地理临近性和文化根植性特点有利于本地学习的进行和隐性知识的传播，而价值链学习则主要传输显性知识，较远的空间距离使价值链高端企业对知识流出具有较强的控制力。（2）嵌入性依赖和知识壁垒两大学习障碍是导致全球价值链下低端锁定现象的根本原因。前者致使企业学习的外部途径依赖，容易受领导企业牵制，而基于自主创新的内生性知识积累途径逐渐弱化甚至消失，后者则妨碍了低端企业获取领导企业把控的产业关键技术。在嵌入性依赖和知识壁垒的共同作用下，企业无法获取升级所需的知识积累且难以摆脱对现有技术发展路径的依赖，从而陷入低端锁定。（3）政府是突破低端锁定的外生力量，解决嵌入性依赖和知识壁垒两大学习障碍是产业升级的必要条件。政府可以承担路径突破中的转换成本，可以利用大型战略项目倒逼企业技术升级，主要目的是引导企业开展自主创新活动。利用自主创新和跨价值链学习可以重构主体的知识基础而摆脱嵌入性依赖，利用同类技术发展和知识获取型逆向投资可以绕开知识壁垒获取企业升级的必备知识。（4）大飞机产业的案例研究为我国集群企业突破学习障碍和低端锁定提供了一条可行性路径：依托自主创新和价值链重构，在突破学习障碍不断获取知识的同时占据价值链战略环节、夺取链条治理权。更多还原

【Abstract】 In the context of economic globalization, enterprises tend to centralizing develop inthe form of industrial cluster. In many developing countries the industrial clusters orcluster enterprises participate in global competition and seek upgrade opportunities byway of embedding themselves in the global value chain. Under the guidance of the policyof "market for technology", China’s industrial sectors in the global value chain are alsotrying to quickly upgrade their technological level and industrial competitiveness throughinternational cooperation and technology introduction. However, it is difficult to exchangemarket for the core technology and cutting-edge technology, so many industrial clustersand enterprises are locked up in the low-end part of the value chain, which not only leadsto the low level of enterprise profits, but also easily leads to "immiserizing growth oflocal economy. Therefore, the low-locked is a key problem for industrial upgrading in thecontext of global value chain.From the perspective of knowledge accumulation and learning mechanism of clusterenterprises, this paper attempts to interpret the low-locked phenomenon in the global valuechain using the concept of learning disabilities. Firstly, on the basis of review of relevantliteratures at home and abroad this paper observes the status quo of research on industrialclusters’ local learning and global learning. Secondly, this paper introducesknowledge-flow theory to analyze the characteristics of cluster knowledge flow and valuechain knowledge flow, and on this basis builds a knowledge flow-based learningmechanism for cluster enterprises in the global value chain, and combined with theoreticalanalysis and CES function-based mathematical derivation explains the mechanism howlow-end enterprises suffering from learning disabilities are locked up in the low-end partof value chain. Thirdly, this paper proposes government as an exogenous variable to breakthrough learning disabilities, and makes empirical analysis using the panel data of China’shigh-tech industries. Finally, around the core concept of learning disabilities, combinedwith analysis of a typical case of large aircraft industry, this paper puts forward agovernment-led strategy to break through Low-locked in the global value chain. Based on the above research, this paper draws the main conclusions as follows:(1) Cluster enterprises conduct learning and knowledge accumulation mainly byindependent innovation-based endogenous ways and clusters’ local learning and valuechain learning-based exogenous ways. Industrial clusters’ characteristics of geographicalproximity and cultural rooting are conducive to the conduct of local learning and thedissemination of tacit knowledge, while the learning of value chain is mainly to transmitexplicit knowledge, in which the greater spatial distance allows high-end enterprises in thevalue chain to have strong control over knowledge outflow.(2) Two learning disabilities of embedded dependence and knowledge barrier are theroot causes of the low-locked phenomenon in the global value chain. The former causescorporate learning dependent on exogenous ways, vulnerable to pin-down by leadingenterprises, while the independent innovation-based endogenous ways to knowledgeaccumulation gradually weaken or even disappear. The latter prejudices low-endenterprises from access to control of key industrial technology of leading enterprises.Under the dual impact of embedded dependence and knowledge barrier, enterprises cannotrealize knowledge accumulation required for upgrade, nor can they get rid of dependenceon the existing path of technological development, and thus fall into the low-lockedposition.(3) Government is an exogenous force to break through low-locked and a necessarycondition for industrial upgrading is solving the two learning disabilities of embeddeddependence and knowledge barrier. Government can take on the conversion cost incurredin path breakthrough and use large strategic projects to help enterprises on the verge ofcollapse achieve technology upgrades. The government’s main purpose is to guideenterprises carrying out activities of independent innovation. The use of independentinnovation and learning across the value chain allows enterprises to reconstruct knowledgebase and get rid of the embedded dependence, so that enterprises take advantage of similartechnology development and knowledge acquisition-oriented reverse investment to bypassknowledge barriers and obtain the necessary knowledge for upgrade.(4) The case study of large aircraft industry provides a feasible path for China’scluster enterprises to breakthroughs of learning disabilities and low-locked: relying on independent innovation and value chain restructuring, enterprises while breaking throughlearning disabilities to continuously acquire knowledge occupy the strategic links of valuechain and seize chain governance.更多还原