【作者】 张战仁；

【导师】 杜德斌；

【作者基本信息】 华东师范大学 ， 人文地理学， 2011， 博士

【摘要】 自我国统筹区域发展战略提出以来,区域增长差异是关于中国区域研究的持续热点,国内外学者围绕区域经济差异的推动因素展开了大量研究。但是,对于区域创新差异却缺少系统研究。然而,经济发展水平差异只是一种表象,经济发展的内在动力,具体说就是不同区域的技术资源,特别是工业技术资源及其创新活动的活跃程度上存在的某种程度的发散可以认为是更为深层的内在原因。在区域发展上,落后地区尽管可以通过市场购买或技术扩散效应提高社会创新活动对经济发展的贡献率。但是,受到外部环境和内部技术发展路径依赖两方面的强力制约,区域创新的空间异质性可能会长期存在。在这种情况下,要缩小我国区域经济发展差距,形成推动区域协调发展的政策安排,学界必须对区域创新差异问题加以关注。目前,尽管随着区域差异理论的日益完善、特别是新经济地理学理论诞生之后,学者在区域经济差异研究领域实现了空间革命,不仅注意到了区域产业互动联系、技术溢出、空间报酬递增对区域经济差异形成的贡献作用,而且,还以区际经济增长的可溢出性为着力点,探讨了增长地理溢出与区域经济差异形成的相关关系。但是,在区域创新研究中,尚未见有学者以地理空间为视角研究区域创新差异问题。而事实上,与经济活动一样,作为一个投入-产出过程,由于创新生产活动也涉及到主体间的互动联系及相互知识外溢,使得地理空间因素对区域创新差异的形成可能具有同样的功效。譬如,区域创新互动联系、知识外溢也有可能产生空间层面上的创新递增报酬,从而将引起创新地理集聚和形成区际差异；而创新知识的区际溢出则会促使区域创新行为产生空间交互作用,使得区际创新产生极化扩散效应,并形成超越行政边界的创新热点区域群或冷点区域群,等等。创新行为在少数区域地理集聚或在邻近区域成堆分布正是区域创新差异形成的重要空间表现形式。因此,不仅区域经济差异研究必须以地理空间为研究视角,以知识生产为目的的区域创新差异研究也需如此。本研究基于区域创新活动数据,理论探讨与实证检验相结合,采用地理空间视角,建立空间或行业实证检验模型,以我国的区域创新差异为研究对象,在时空两维角度及区域内和区域间两个空间层次,深入研究地理空间因素与我国区域创新差异非均衡发展的时空模式。本研究的焦点议题有三,一是区域知识外溢、创新互动联系与区域创新差异分析；二是区际知识外溢、创新地理溢出与区域创新差异分析；三是在动态变化角度分析创新空间交互作用与区域创新差异空间俱乐部趋同。本研究首先在系统梳理国内外有关区域差异研究的基础上,主要基于专利产出数据,首先直观分析了我国区域创新的地理集聚及空间集群格局。我国区域创新活动分布的地理集聚现象明显。与一般的经济活动相比,创新活动地理集聚倾向明显高于经济活动,且多空间尺度下的创新地理集聚都具有累积性特点；而在空间集群方面,我国区域创新差异不仅表现在创新活动的地理集聚上,其还表现在区域创新布局的成堆分布上。首先,省级空间尺度上,我国各省市创新活动分布的一个显著特点是相邻省份的创新发展基本处于同一分位区,其次,在地级空间尺度,区域创新成堆集群分布的态势更加明显。总体上,作为一个投入产出过程,创新活动地理集聚的累积性特点及比经济行为具有更高地理集聚度的客观事实等,可能意味着区域创新生产的互动联系、知识溢出与我国区域创新差异的产生具有一定的相关关系。而考虑到外在关联环境对一区域创新发展的重要性,区域创新成堆分布的这种规律性的空间关联结构显示,创新空间交互作用即创新地理溢出可能是我国区域创新差异产生的另一原因。进一步地在前文理论文献综述及稍后的直观分析基础上,本研究以地理空间为视角,区分了两种导致区域创新差异的创新互动溢出行为：一是区域知识外溢、创新互动溢出；二是区际知识外溢、创新地理溢出。与传统研究仅仅关注知识外溢的距离衰减性对创新地理集聚的影响不同,本研究将区域知识外溢赋予了一个新的内涵,认为与经济空间递增报酬一样,区域知识外溢、创新互动联系也会产生空间层面上的创新报酬递增,加之要素流动的集聚效应,这会从多方面引起区域创新差异。与传统研究相比,本文将知识外溢与区域创新差异的认识上升到了一个新的理论高度。此外,在区际知识外溢、创新地理溢出与区域差异研究方面,本研究主要关注的不是受区际知识外溢等影响下,区域创新地理溢出的整体溢出效果,而是试图论证多样性的创新空间交互作用对我国区域创新差异形成的影响作用。首先,对区域知识外溢、创新互动联系与区域创新差异的分析表明：(1)传统的供求及政策类因素对我国区域创新差异的形成仍有影响作用,在人力或资本投入规模上的差异可引起我国区域创新的非均衡发展,此外,政府的公共财政支持及区域不同的产业结构等政策或供求类因素也都与我国区域创新差异的产生关系密切；(2)从区域创新互动溢出因素来看,如果只强调以上这些传统的供求及政策类因素,则在区域创新差异的形成原因时,这明显有失全面性。事实上,在所有创新差异形成因素的实证检验模型中,创新互动溢出因素同样都具有较大的估计系数,而且,如果以最能代表区域创新发展状况的发明专利为解释变量,则创新互动溢出类因素的解释力度更强。考虑到发明专利指标对区域创新发展状况的更好的代表性,这显示了以地理空间为视角,从创新互动溢出角度系统论述区域差异产生原因的重要性。(3)从动态集聚角度来看,通过创新要素的跨区域流动等,我国区域内部不同的创新互动溢出环境,对我国区域创新差异的形成具有累积循环影响。在创新互动联系及相互溢出因素的影响下,循环累积作用是形成我国区域创新差异的一项重要机制。总体上,这说明了区域创新互动溢出对区际创新差异的形成具有综合影响。(4)从产业层面来看,区域知识外溢、创新互动联系与各产业创新活动的地理集聚关系密切。若一产业创新活动对地方知识外溢及创新互动环境要求越高,则该产业创新活动的地理集聚度通常也越高,两者的相关关系具有很高的统计显著性。鉴于各创新活动对地方互动溢出环境普遍较高的要求,从产业创新活动层面,这再次证实了区域不同的创新互动溢出环境对区际创新差异产生的重要影响。其次,至于区际知识外溢、创新地理溢出与区域创新差异分析,本论文将创新差异的形成因素从区域内拓展到区域外,从区域创新所处的外在关联环境入手,运用探索性空间数据分析方法,主要分析在地理距离、创新发展阶段及区域知识吸收能力等因素的影响下,多样性的创新空间交互作用对其区域差异形成的理论与实际作用,结果表明：(1)从我国区域创新的整体空间关联溢出效果看,我国区域创新整体是可溢出的,但是,这种溢出效果明显受到地理距离因素的影响。鉴于离创新增长极远近不同的区域,它们在创新发展上的可能不同表现,随距离增加而不断减小的创新溢出,在一定意义上能有效证明受地理距离等因素的影响,多样性的创新地理溢出对区际创新差异的形成具有贡献作用。(2)分地带(或分省区)从区域创新空间关联溢出的局部特征看,无论是地带内部的省级空间层面,还是省区内部的地级空间层面,我国区域创新的空间关联溢出结构都有不同的变化特征。在创新发达的东部地区,由于已跨越极化发展阶段,加之知识吸收能力较强,其内部各省区的创新一体化程度较高,创新正向的空间关联溢出明显,与此相反,在创新欠发达地中西部地区,其内部省区的创新协同溢出能力则较差,创新正向的空间关联溢出不明显,结果在创新发展的外在关联环境方面,创新欠发达省份远差于创新发达省份。此外,在省区内部的地级空间层面,创新的关联溢出变化与我国区域创新能力的非均衡空间布局也高度一致。总体上,这一结果说明,不仅受地理距离因素的影响,多样性的区际创新溢出对区域差异的形成具有理论上的贡献作用,在区域发展阶段及知识吸收能力等因素的影响下,创新空间关联溢出随区域不同表现出的规律性变化也显示,区域间多样性的创新空间关联溢出与我国区域创新差异的形成具有必然联系。(3)从地带交接处省市间的创新溢出情况看,东部省区与中部或西部省区间的创新发展仍基本处于发展的极化阶段,与中部相邻的东部省份通过人员、信息的流通获得了正向溢出,而与东部相邻的中部省份未能从东部省份的创新溢出中受益,与东部省份接壤的部分中西部省区还处于不利的逆向溢出环境中。概括而言,地带交接处的省市创新关联溢出仍然是有利于发达省市创新发展的这样一种空间关联格局。最后,在时间维度上,对于地理空间因素于区域创新差异动态变化的影响,本研究将区位空间依赖因素作为俱乐部趋同研究的区域分组标准,通过建立可测度地理空间效应对区域差异趋同影响的空间计量模型,对创新地理溢出与区域创新差异空间俱乐部趋同进行了实证分析,分析结果显示：(1)从全国总体考量,尽管我国区域创新差异并不存在动态的单一稳态趋同特征,相反,有悖于落后地区较之发达地区长期能保持更高增长速度的经典假设,我国区域创新的整体差异水平长期趋于发散。但是,从内部空间依赖模式相近的两个分组区域看,在我国区域创新差异整体水平不断扩大的同时,这两个分组区域内部的创新发展却逐渐趋同,形成了两个有特色的创新空间趋同俱乐部。(2)从创新地理溢出的贡献作用考量,我国区域创新差异的这种趋同特征不仅说明,与区域经济差异一样,我国区域创新差异并不存在单一稳态的趋同,而只会在具有某些相似条件的各分组区域内出现多稳态的相对趋同；同时,在空间依赖模式相近区域组内出现俱乐部趋同的实证结果还表明,地理溢出因素对我国区域创新差异的多稳态趋同有重要贡献作用。可以说,我国区域创新这种整体分异与局部多稳态趋同特征的形成是与创新的地理溢出分不开的。(3)从各分组区域的比较角度考量,地理溢出类因素对区域创新差异动态空间趋同的实际贡献作用随分组区域不同也有不同。高高组内区域更大的创新溢出系数λ,说明由于它们都具有相对较高的知识吸收能力等,创新地理溢出对平抑它们内部区域间的创新差异更具决定意义。更多还原

【Abstract】 Since the co-ordination of regional development strategy has been put forward, the differences in regional growth has been the core problem of China’s regional research for a very long time. Scholars, both domestic and abroad, have made abundant of research on the driving factors of regional economic differences. However, few systematic studies has been made on the regional differences of innovation. The spread of technological resources and relative innovation activity are the ultimate causes of regional differences in economy. Therefore, this thesis is going to deal with these questions:How to identify the factors causing regional difference in innovation, how to accelerate the development of innovation ability in undeveloped areas, and how to promote the coordinating development of regional economy and find its interior driving forces in China?Nowadays, with increasingly perfection of regional differences theory, especially after the birth of new economic geography theory, scholars not only took note of the interactive links of production, technological spillover, spatial increasing returns, i.e. capital/technology externalities contributing to the formation of regional economic differences, inside region, but also discussed the relative relation of geographical spillover and regional economic differences based on the spillover of regional economic growth. But in the existing research on regional innovation, none paid attention to regional innovation differences in the viewpoint of geographical space. Just like economic activity, innovation activities are essentially procedures of input and output. The innovation links and mutual spillover among regions caused a non-equivalent development of regional innovatioa For example, innovation links and knowledge spillover within the region will generate increasing returns of innovation in the spatial level, then cause geographical agglomeration of innovation and regional differences. The interregional spillover of innovative knowledge will lead to spatial interaction created by regional innovative behavior, make polarization diffusion effects, and form hot or cold regional clusters of innovation across administrative boundaries. Innovation behavior gathering in a few regions or forming clusters across regions is an important spatial performance formulated by regional innovation differences. Therefore, not only the research on regional economic disparities, but also the research on regional innovation differences whose main purposes are producing knowledge, must take geographical space as a basic starting point.Based on the data of regional innovation activities, combining theoretical discussion with case study, using relative theories of new economic geography and spatial measurement method, this research made a thorough study on time-space model of non-equivalent development of regional innovation and knowledge spillover and geographical spillover in the two-dimension viewpoint of time-space and intra-and inter-regional spatial levels. There are three core topics in this study:first, analyzing regional knowledge spillover, innovation interactive links and regional innovation differences; Second, discussing inter-regional knowledge spillover, innovation geographic spillover and regional innovation differences; third, explaining spatial club convergence of innovation spatial interaction and regional innovation disparities in a dynamic viewpoint.The paper systematically sorted out the relative research of regional difference in home and abroad, pointed out the next research directions of regional innovation differences, and analyzed geographical agglomeration and spatial clusters of regional innovation in China. It is evident of geographical agglomeration of regional distribution of innovation activities in China. Compared with the general economic activities, innovation activities tend to higher agglomeration than economic ones, and geographical agglomeration of innovation has cumulative features under multiple spatial scales. As for the spatial clusters, the differences in regional innovation are shown not only in the geographical concentration of innovation activities, but also in clustering distribution of regional innovation layout. First, one significant feature of innovation activity distribution of every provinces and cities is innovation development of adjacent provinces in the same sub-bit area; second, the pile distribution of regional innovation clusters is still more apparent in district-level spatial scales. In general, as an input-output process, the cumulative characteristics of geographical concentration of innovation activities, which have higher geographical agglomeration degree than economic activities, may mean there are certain relative relations between the interaction links of inter-regional innovation production, knowledge spillover and regional innovation disparity in China. While considering the importance of external environment to regional innovation development, this spatial correlation structure of pile distribution of regional innovation shows innovation spatial interaction may be another causes of regional innovation development in China.On the basis of previous theoretical literature review and later analysis, this study, taking geographic space as research perspective, sort out two spillover behaviors of innovation interaction which lead to regional innovation disparity:one is knowledge spillovers within the region and innovation overflow; another is inter-regional knowledge spillovers and innovation geographic overflow. Different from traditional research only paying attention to the different effect of distance decay of knowledge spillover on innovation geographic agglomeration, in this study, knowledge spillover within regions is given a new meaning, which believed that just like increasing returns of economic space, knowledge spillover within regions and innovation interaction would create innovation increasing returns. In addition to the agglomerating effect of factor mobility, all these will lead to regional innovation in many different ways. The main concern of this study is not overall spillover effect of regional innovation geographic spillover influenced by inter-regional knowledge spillover, but analysis on the effects of various spatial interaction of innovation on regional innovation disparity in China.At first, the analysis on regional knowledge spillover, innovation interaction and regional innovation disparity shows that:Firstly, traditional factors of supply-demand and policy still make effect on regional innovation disparity, and the differences of human or capital investment scale can cause non-balanced development of regional innovation.Secondly, using the patent, the most representative element of showing the condition of regional innovation development, as explanatory variables, the spillover factors of innovation interaction are more convincing. This also explains the significance of systematically discussing the causes of regional disparity from the angle of innovation interaction spillovers.Thirdly, from the perspective of dynamic agglomeration, the different spillover environment of innovation interaction within a region has a cumulative cycling effect on regional innovation disparities in China.Fourthly, from the industrial perspective, there is close relationship between regional knowledge spillovers, innovation interaction and various industrial innovation activities.As far as analyzing the relation between the regional knowledge spillovers, geographical spillover of innovation and the regional differences of innovation was concerned, the elements of innovation differences has been expanded from inner-region to outside in this research. Starting from the outside relative environment of regional innovation and using exploratory spatial data analysis methods, the author explained the theoretical and practical contribution to regional disparity of multiple spatial interactions of innovation under the influences of geographic distance, innovation development phase and knowledge absorption capacity of regions, etc. The conclusion has been made as follows:Firstly, the regional innovation is spilling from the overall spillover effects of spatial relevance of regional innovation in China, but this spillover effect is obviously influenced by geographical distance. Because of the different distance of regions from innovation growth pole, there may be different performance in innovation development:with the increasing of distance, innovation spillover is decreasing. In another words, it can be proved that diversifying geographic spillover of innovation has made a contributive effect on the regional innovation disparity under the influence of some elements, such as geographic distance.Concerning the local characters of regional innovation spillovers, the structure of China’s regional innovation spillovers, no matter in the province scale or the prefecture-level city scale, presents different features. Due to having realized the stage of polarized development as well as being apt at absorbing new knowledge, innovation of the provinces within the eastern region integrated closely and spills over positively with each other. In contrast, among provinces of the central and western regions, which are considered less innovative, innovation coordination performances are weak, and few positive and spatial innovation spillovers are identified. As results, provinces whose innovation level is underdeveloped are inferior to the eastern innovative province in the external innovation environment. In addition, the variation of innovation spillovers highly is consistent with the uneven spatial pattern of innovation capability in China among the prefecture-level cities. Overall, our findings demonstrate that the regional inequality is not only affected by geographic distance, but also by the diverse regional innovation spillovers. Influenced by the stage of regional economic development and knowledge absorption capability, the regular variation of spatial innovation spillovers shows that the inequality of regional innovation in China inevitably relates to the diversity of spatial innovation spillovers.Concerning the spillovers among the adjacent provinces, the development level of innovation is primarily at the stage of polarizing between the eastern regions and central regions or western regions. Eastern provinces adjacent to the central regions acquire positive spillover from the flow of personnel and information, while the central provinces were in the disadvantage of diverse spillovers. In a word, the spatial patterns of innovation spillovers between neighboring provinces are still in favor of the development of the economy prosper provinces。Finally, in from the temporal dimension, the study adopts the locational dependent factors as the regional grouping standards of the club convergence research, and empirically studies the regional innovation convergence process by spatial econometrical model of measurable geographic spatial effects. The findings are as follow:In China mainland, the regional innovation inequality doesn’t present dynamic single steady-state convergent characters. In contrary, different from the classical hypothesis that the backward regions can keep higher rate of development in the long term, the regional innovation inequality diverges overall. However, in view of groups with similar internal spatial dependent modes, although the inequality of regional innovation is widening, these groups converge gradually and forma convergent club with characteristic innovation space.As to the contribution of innovation geographic spillovers, the convergent features of regional innovation difference in China imply that like the inequality of the economic level, single steady-state convergence of regional innovation does exist, but the areas in different groups with similar spatial dependent patterns achieve relative convergence in multiple of steady-states. Meanwhile, the empirical results of club convergence among regions with similar spatial dependent patterns show that geographic spillover factors contribute to the multiple steady-state convergence of regional innovation inequality in China. It can be said that the features of overall differentiation and local multi-steady-state of regional innovation in China are inseparated from the geographic spillovers of innovation.Given the different groups, factors of geographic spillovers contribute differently to the dynamic spatial convergence of regional innovation unevenness. A larger lambda of the areas in high-high groups shows that due to more capable of absorbing knowledge and technologies, geographic spillovers of innovation play more significant roles in reduce inequality in these groups.更多还原