【作者】 盛垒；

【导师】 杜德斌；

【作者基本信息】 华东师范大学 ， 人文地理学， 2009， 博士

【摘要】 当今时代,知识和技术已成为经济社会发展的最关键因素,国家的兴旺发达、民族的繁荣进步,都高度依赖于知识和技术的持续创新。而研发活动作为生产知识并将知识转化为生产力的核心动力,就显得尤为重要。尽管研发与创新极为重要,但在经济全球化的背景下,任何一个国家都难以仅仅依靠本国力量实现技术的可持续创新,还需要充分利用外部技术资源,通过技术引进与技术扩散来提高本国科技水平和加速经济发展。尤其是发展中国家因自身技术创新能力的限制,更需要重视利用技术扩散来促进本国技术水平的提高和经济的快速发展。拥有、掌握与生产了世界绝大多数关键技术的跨国公司研发的日益全球化为发展中国家利用国际科技资源提升本国技术水平提供了机遇和可能。当前,世界许多国家和地区都纷纷出台相关政策,争相吸引外商到本国开展和设立研发活动,其主要目的就在于期望外资研发活动能给本国带来知识溢出效应,使跨国公司的先进知识、技术和经验在向子公司转移的过程中扩散到当地企业。于是,外资研发的知识溢出成为了近年来社会各界关注的焦点和国际投资领域研究的新方向。国内外已有的相关研究主要集中在经济学领域。经典经济学主要研究的是一个“点”上的经济活动规律,然而,任何经济活动都离不开一定的地理空间载体。知识溢出属于经济活动,但更是一种空间行为,在知识溢出的理论与实证研究中,如果忽视了经济行为的空间特性,假定经济行为是在一个过滤了空间的“点”上进行的,那么“知识”就无法“溢出”。因此,传统的外资知识溢出研究仅仅基于某一特定的“点”来考虑知识溢出,忽略了外资的空间知识溢出行为,导致外资知识溢出的检验结果存在明显的缺口,难以准确把握外资的知识溢出规律。另外,既有的研究大多还停留在理论分析层面,许多研究结论尚缺乏较有说服力的实际证据。本文尝试从“地理空间”的角度研究外资研发的知识溢出效应及其空间规律,试图弥补以上缺陷,为国内外相关研究领域的拓展提供一些有益补充。本研究基于翔实的统计数据,理论分析与实证研究相结合,采用新经济地理学相关理论和随机前沿分析方法,建立空间面板数据模型和行业面板数据模型,从区域内和区域间两个层次入手,以外资在华研发活动为研究对象,深入探究外资在华研发活动的空间集聚及知识溢出的客观规律。本文的核心议题有二,一是外资在华研发活动的空间格局和空间集聚机制,二是外资在华研发的知识溢出效应及其空间规律。本研究在梳理和评价国内外有关跨国投资知识溢出研究的基础上,利用最新的统计调查数据,首先系统分析了外资在华研发活动的空间分布格局。我国外资研发活动主要集中在东部沿海少数省份和区域,且高度聚集在部分全国性和区域性的经济中心城市,区域间差异十分明显。外商研发投资的空间集聚度不仅高于外商直接投资,也高于一般生产制造性活动的地理集中程度。然后,从新经济地理学视角重新审视外资研发活动的空间集聚机制,结果表明,传统的跨国投资区位理论已不能有效解释我国外资研发的空间集聚成因,而引入新经济地理变量后的检验结果为认识外资研发的空间集聚行为提供了更充分的启示。相比劳动力成本、优惠政策等传统区位因素,交易成本、技术外部性、资金外部性及区位自我强化效应等新经济地理因素是导致外资在华研发空间集聚格局的更重要原因。进一步从经济地理的角度,本文区分了两种不同的外资研发知识溢出行为:一是外资研发的区域内知识溢出,即区域外资研发活动对本区域产生的知识溢出;二是外资研发的区域间知识溢出,即其他区域外资研发活动对该区域带来的知识溢出。与传统研究仅注重区域内的知识溢出所不同,本文既分析了外资研发的区域内知识溢出效应,也考察了外资研发的区域间知识溢出效应,从而更真实地模拟了外资研发的知识溢出行为。在研究方法上,本文摒弃了传统的确定性前沿分析方法,选择更符合客观实际的随机性前沿分析方法和技术非效率模型对外资研发的区域知识溢出效应进行实证检验和估计。对外资在华研发区域内的知识溢出效应检验表明:(1)外资研发活动对我国省区技术效率的提升具有一定的促进作用,但这种溢出效应只有在外资研发进入中国较长一段时期后才会发生;外资研发进入当期不仅无法给中国省区带来技术效率进步,反而会造成技术非效率。因此,本研究证实了,虽然短期内外资研发难以对我国产生有效的知识溢出效应,甚至其负面影响更为突出,但从长期来看,外资研发对中国省区技术效率增长的积极作用是客观存在的。(2)在作用类型上,外资研发的进入对中国省区技术效率的影响主要反映为竞争效应,通过技术、工艺和管理上的示范作用直接带来中国省区技术效率的提升相对有限。(3)在作用途径上,外资研发主要是通过加剧市场竞争促使中国省区增加资本投入提高对外资研发知识和技术的吸收能力,而通过人力资本对外资研发知识的吸收效应未能带来省区技术效率进步,外资研发知识溢出效应的发挥显著地受到人力资本门槛的制约。(4)区域样本的分析结果则说明,外资研发的知识溢出效应呈现区域差异性特征,外资研发显著提升了东部省区的技术效率,但对中西部省区技术效率影响十分有限。(5)进一步对全部37个行业以及外资研发进入水平前10个行业的实证研究证明,外资研发对中国内资企业的自主研发能力确实产生了较为明显的促进作用,而不是相反。而且,行业中外资研发的进入程度越高,则该行业内资企业的自主研发能力也提高得越快。(6)模型中引入空间集聚变量后的分析发现,空间集聚确实更有利于外资研发知识溢出效应的发挥,如果在未考虑集聚变量的情况下外资研发对内资企业的知识溢出程度为λ,那么在考虑集聚变量的情况下外资研发对内资企业专利申请的溢出程度为2λ,对内资企业科技经费投入的溢出与促进作用更大为9λ,对内资企业科技人员投入的影响为6λ。对于外资研发的区域间知识溢出效应,本研究从“受溢方”的视角出发,建立了有别于以往文献的空间溢出分析框架和空间知识溢出模型,对地域邻近与外资研发知识溢出的空间关系以及外资研发知识溢出的空间规律进行实证分析,分析结果显示:(1)正如本文所假设的那样,一省外资研发活动不仅对本省存在知识溢出效应,对周边省份技术效率的提升同样具有溢出作用。从全国总体来看,与一省邻近省份的外资研发规模越大,那么该省份技术效率提升速度也越快;某一省份受惠于本省外资研发的知识溢出效应要大于来自与之相邻省份的外资研发知识溢出效应;并且,某一省份来自相邻省份的外资研发知识溢出主要是与之相邻且外资研发规模更大的省份,外资研发规模更小的相邻省份对其不存在显著溢出效应,因此区际外资研发知识溢出需要跨越一定的外资研发规模门槛。(2)从各区域来看,东部地区省份除了受惠于本省外资研发的知识溢出效应,其受惠于相邻省份的外资研发溢出效应更为可观;其中,东部地区省份受惠于同属一个区域相邻省份的外资研发知识溢出效应要大于与之不属同一区域相邻省份的外资研发知识溢出效应。本省外资研发对中西部省份不存在显著知识溢出效应,但中西部省份将受惠于与之相邻省份的外资研发知识溢出效应;而且,中西部地区省份仅受益于与之相邻且同属一个区域省份的外资研发知识溢出效应,与之相邻却不属同一区域省份的外资研发活动不会对其产生知识溢出效应。(3)从区域之间相互作用来看,与东部相邻的中部省份外资研发活动对东部地区技术效率提升产生了正向溢出效应,而与中部相邻的东部省份外资研发活动却不会对中部地区技术效率进步带来显著溢出效应;与中部相邻的西部省份外资研发活动不会对中部地区技术效率提升带来积极溢出效应,而与西部相邻的中部省份外资研发活动也同样不会对西部地区技术效率增长带来明显溢出效应。由此,本研究认为,经济发展程度、人力资本水平、科研技术实力等方面较为接近的区域之间更容易发生外资研发知识溢出现象并形成外资研发区际溢出的双向互动格局;区域之间发展水平差异越小,资源要素的区域流动越频繁,区域间相互作用强度可能越大,也越可能形成外资研发的区际知识溢出。(4)而进一步利用扩展后的外资研发空间知识溢出模型分析发现,外资研发的空间知识溢出不仅仅发生在相邻区域,在非相邻区域之间也同样存在,而且是有规律可循的。外资研发的空间知识溢出具有明显的地理边界,我国某一省份的外资研发活动仅对距离600km以内省份的技术效率提升产生积极作用,因此,外资研发知识溢出只在有限的空间范围内发生。与一般地理现象的扩散相似,外资研发知识溢出强度存在明显的空间衰减规律,并且,随着距离的增加,溢出衰减也越明显。与溢出源距离增加300km,溢出强度将减少50%;溢出距离再增加300km,溢出强度将减少70%;更远距离的外资研发溢出强度衰减为零。更多还原

【Abstract】 Nowadays, knowledge and technology innovation are the key factors to economy development. As the core driver of knowledge production and translate knowledge into productivity, R&D activity become even more significant. Although R&D and innovation are very important, under the background of globalization, it is difficult for any countries to innovate continuously only rely on its own resources. Therefore, introduce of new technology and technology diffuse become an effective way to improve science and technology of host country, which is especially significant for developing countries. Along with the R&D globalization of multinational companies, developing countries have the opportunity to improve their technology power by utilizing international R&D resources. In recently, most of countries all over the world have competed to attract foreign R&D activity. They expect advanced knowledge, technology or experience of foreign R&D can diffuse to local firms. Therefore, knowledge spillovers have become a hot point of researchers.Most research about knowledge spillovers of foreign investment come from economics. Classical economics mainly concentrate on "point" to study economical activities, but any economical activities can’t break away from the geographic space. As an economical activity, spillover also belongs to spatial behavior. So, if we neglect the spatial identity of knowledge spillover, knowledge is impossible to spillover. From this point of view, the traditional spillover research ignored the spatial knowledge spillover behavior of foreign investment, and the test of knowledge spillovers exist visible gap. Beside, lots of recent papers still rested on theory analysis, the research results lack of persuasive evidences. This paper attempts to study the knowledge spillover effects of foreign R&D to make up these limitations.Based on detailed data, combined with theory analysis and empirical study with new economical geography and stochastic frontier method, the paper established panel data model to research agglomeration and knowledge spillovers of foreign R&D in China.At first, this paper analysis the spatial distribution structure of foreign R&D in China. China’s foreign R&D concentrates in eastern regions, and mainly agglomerates in few countrywide or regional economic centers. The geography agglomeration degree of foreign R&D is higher than FDI and manufacture activities, the regional difference of foreign R&D distribution in China is visible. Then the paper studies the agglomeration mechanism of foreign R&D with new economic geography theory. According to the empirical research, traditional location variables can’t explain the reason of foreign R&D agglomeration, and the result of new economic geography variables offer more inspirations to understand agglomeration behavior of foreign R&D. Comparing to traditional location factors as labor cost and preferential policy, trade cost, technological externality, pecuniary externality and self-reinforcement effect are the more important factors which lead to geography agglomeration of foreign R&D.Secondly, the paper distinguishes intra-regional knowledge spillovers and inter-regional knowledge spillovers of foreign R&D, utilizes stochastic frontier analysis method and technical inefficiency model to estimate these two kinds of regional knowledge spillover effects. As to the intra-regional knowledge spillovers, we found that: (1) Foreign R&D improves technical efficiency of Chinese provinces, but this effect happens only after foreign R&D stay in China for a period of time. There is no knowledge spillover in the first year foreign R&D enter China, what’s worse, foreign R&D will cut down Chinese provinces’ technical efficiency in the current year. (2) Foreign R&D affects Chinese provinces’ technical efficiency mainly through competing effects, demonstrating effect is still limited. (3) Through intensifying the market competence, foreign R&D urged all the regions to augment capital input to improve the absorbency to knowledge and technology spillover from foreign R&D. But Chinese provinces can’t upgrade their technical efficiency through absorbing advanced knowledge of foreign R&D by labors. Therefore, knowledge spillovers of foreign R&D are limited by labor capital threshold. (4) According to the regional sample, there exist visible regional difference of knowledge spillover effects of foreign R&D. Foreign R&D investment advanced eastern provinces’ technical efficiency, but this effect is limited for center and western provinces. (5) Evidence from the whole 37 industries and the top 10 industries which have the highest foreign R&D entering intensity indicates that, foreign R&D do promote indigenous R&D capability of Chinese firms of these industries. Those industries which have higher foreign R&D entering intensity always have more powerful independent R&D capability. (6) Agglomeration is more effective for foreign R&D knowledge spillovers. While the spillover isλwhich don’t take into account of agglomeration variable, the spillover of foreign R&D to patent of native firms is 2λwhile take into account of agglomeration, spillover to technology input is 9λ, and spillover to technology labor input is 5λ.As for the inter-regional knowledge spillovers of foreign R&D, this research point out that:(1) The foreign R&D of a province not only spillover knowledge to itself, but also have spillover effects to neighbor provinces. From the point of the whole China, the more of the neighbor province’s foreign R&D scale, and the upgrade of this province’s technical efficiency is more quickly. The spillovers from own province’s foreign R&D is larger than that from its neighbor provinces’. What’s more, spillovers from foreign R&D of neighborhood is mainly from the near province whose scale of foreign R&D is larger than this province.(2) From the point of regions, eastern provinces not only accept spillovers from native foreign R&D, but also benefit more from neighbor provinces. Eastern provinces gain more spillovers from foreign R&D of neighbor eastern provinces than from foreign R&D of neighbor but not.eastern provinces. Native foreign R&D doesn’t spillover to central and western provinces, but these provinces can benefit from foreign R&D of its neighbor provinces. Be similar to eastern provinces, central and western provinces can get spillovers from foreign R&D of neighbor and the same central or western region provinces, but foreign R&D of neighbor provinces and not belong to the same region will not have spillover effect.(3) From the point of spatial interaction, foreign R&D of center provinces which neighbored to eastern provinces exist knowledge spillover effects to eastern region, but foreign R&D of eastern provinces which neighbored to central provinces have no spillover effects to central provinces. Foreign R&D of western provinces which neighbored to central provinces has no spillover effects to central provinces, and the same as foreign R&D of central provinces which neighbored to western provinces. Therefore, it is easier for foreign R&D of regions which with the similar development level to produce spillover effects.(4) Spatial knowledge spillover of foreign R&D not only exists in neighbored regions, but also happens between not-neighbored regions. There is visible geography boundary of inter-regional spillover of foreign R&D, the foreign R&D of a province in China can only spillover to the provinces within 600km far away from this province, so the spatial spillover of foreign R&D in China is limited in a definite distant range. Besides, the spillover intensity decays distinctly with distance. When distance far away from spillover center adds 300km, spillover intensity will be fading away 50%. When distance adds another 300km, spillover intensity will be fading away 70%. While the distance is even farther, the spillover intensity will be fading away completely.更多还原