【作者】 徐磊；

【导师】 冉茂盛； 黄凌云；

【作者基本信息】 重庆大学 ， 技术经济及管理， 2009， 博士

【摘要】 技术进步是经济长期增长的源泉,而在开放的经济系统中,技术的进步主要依赖于国内研发和国际技术溢出。在目前中国研发能力及科技水平仍旧有限的情况下,如何有效利用国际技术溢出促进我国的技术进步并进一步促进经济增长显得尤其重要。尽管外商直接投资(FDI)已经成为国际技术溢出的重要渠道,但其溢出效应尚未得到一致的经验证据支撑,其深层次的原因在于吸收能力对FDI技术溢出效果的影响。目前,关于吸收能力与FDI技术溢出之间关系的研究大多集中于实证研究,理论研究较为欠缺;另外,由于现有文献尚未形成对吸收能力的统一认识,往往强调技术水平、人力资本水平或金融深化程度等单一因素,因此得出迥异的实证研究结论。虽然也有学者尝试从“门槛效应”角度解释吸收能力与技术溢出效果间的关系,受限于计量研究手段,尚未对“门槛效应”的具体模式进行深入的分析,且大多停留于静态分析框架。鉴于此,本文以吸收能力这一重要研究视角为出发点,围绕吸收能力与FDI技术溢出的关系这一研究主题,系统地分析中国FDI技术溢出“门槛效应”的具体门槛模式。旨在明晰吸收能力对FDI技术溢出的具体作用机理,客观评价中国现阶段的引资政策及各地区的吸收能力,为各地区结合本地吸收能力情况制定未来的引资政策提供实证依据。主要研究内容包括:首先,对发展中国家吸收能力与FDI技术溢出的关系进行理论分析,为文章的实证部分提供严密的理论基础;其次,基于宏观层面吸收能力层次性的考虑,本文分别选择对外开放程度、区域创新能力及金融发展程度作为三个层次吸收能力的代理指标,分层次地分析基于这三个层次吸收能力的FDI技术溢出门槛模式;然后,进一步从广义的角度对吸收能力进行重新的界定,构建指标体系对中国各省市的广义吸收能力在时间维度和空间维度的相对变化趋势进行定量测度,并将时滞性引入实证模型,从总体上动态地分析FDI技术溢出的广义吸收能力门槛模式。与现有研究相比,本论文的创新工作主要体现在以下几个方面:①以发展中国家为研究对象,构建两阶段古诺(Cournot)寡头竞争模型,从理论上刻画引入技术溢出效应后,内、外资企业为了获取或预防溢出效应而进行的费用决策和最终的产量决策过程,进而对内资企业的吸收能力与FDI技术溢出的一般作用机制进行理论分析。研究表明:当内资企业的吸收能力低于门槛值时,均衡技术溢出率会随着内资企业吸收能力的升高而递增,此时,内资企业愿意为获取技术溢出效应支付更多的费用,而外资企业却减少为预防技术溢出效应而支付的费用;当内资企业的吸收能力高于门槛值时,随着内资企业吸收能力的提升,无法确定均衡技术溢出率的变化,获取技术溢出及预防技术溢出的费用变化方向也无法确定。研究发现了发展中国家的FDI技术溢出存在“吸收门槛”效应,充分解释了吸收能力与FDI技术溢出关系实证研究差异的本质原因,为后文的实证研究奠定了良好的理论基础。②基于宏观层面吸收能力层次性的考虑,利用中国1985-2007年各省市的面板数据构建“门槛模型”,分别从对外开放程度、区域创新能力及金融发展程度三个不同层面分析FDI技术溢出的吸收能力“门槛效应”,并进行了“门槛效应”的检验、实际门槛值的测算及相应门槛模式的分析。实证研究表明:中国各省市的FDI技术溢出效应对于三个不同层次的吸收能力都存在显著的“双门槛效应”,各自的门槛模式具有不同形式,基于对外开放程度的门槛模式是类似的“倒U”型模式;基于区域创新能力及金融发展程度的门槛模式都为“三级梯度”型模式。扩展分析显示:目前FDI对我国大多数省市都具有显著的技术溢出效应,然而这些省市对技术溢出效应的吸收还不够充分,对FDI溢出效应的吸收效果还有提升的空间。③基于各层次吸收能力之间协同性的考虑,从广义的角度对吸收能力进行重新的界定,通过构建指标体系对1985-2007年中国各省市的广义吸收能力进行定量测度,很好地刻画了样本时间范围内我国各省市广义吸收能力的时间变化趋势及地区差异。在此基础上,文章将FDI技术溢出的时滞性引入实证模型,利用动态门槛回归技术动态地分析广义吸收能力与FDI当期溢出效应之间的关系,深入分析基于广义吸收能力的动态门槛模式。广义吸收能力定量测度的结果表明:1985-2007年,我国各省市吸收能力基本呈现稳定的缓慢上升趋势;东部沿海省市普遍具有较高的吸收能力,而西部地区吸收能力普遍处于较低水平。动态门槛模型的GMM估计结果验证了FDI技术溢出滞后性的存在性,在考虑技术溢出的时滞性后我国各地区当期的FDI技术溢出效应存在基于广义吸收能力的“三门槛效应”,该门槛效应呈现“四级梯度”型门槛模式。截至2007年底,由于我国大部分省市广义吸收能力未跨越最高门槛值,当期FDI技术溢出效应显著但未达到最高值,说明这些省市对当期FDI溢出效应的吸收效果具有一定的提升空间。更多还原

【Abstract】 Technological progress is the source of long-term economic growth, and it mainly depends on domestic R&D and international technology spillovers, in an open economic system. Since the current R&D capability and technology level are still limited in China, how to take advantage of the international technology spillovers to promote technological progress and economic growth is particularly important. Although FDI has become the main channel of international technology spillovers, but its technology spillovers effect has not been supported by empirical evidence consistently, and its deep-seated reason is that the absorptive capability can affect FDI technology spillovers. At present, most of the studies on the relationship between absorptive capability and FDI technology spillovers are empirical studies, the existing literature is lack of theoretical study. In addition, the different empirical conclusions came out, since the inconsistent cognition of absorptive capability and the emphasizing the single factor, such as technology level, human capital or financial development and so on. Although some scholars have tried to explain the relationship between absorptive capability and FDI spillovers from the view of“threshold effect”, but due to the deficiencies of the methodology, the specific mode of“threshold effect”has not been analyzed, and most of these studies remain in a static analysis framework.In view of this, this dissertation focused on the relationship between absorptive capability and FDI technology spillovers, and did in-depth research on the specific mode of the FDI technology spillovers“threshold effect”based on the absorptive capability perspective. The main purposes of the study were: learning the mechanism between the absorptive capability and FDI technology spillovers; evaluating the current policy of attracting foreign investment and absorptive capability of each province; providing empirical evidence for making future policy of attracting foreign investment according to absorptive capability in each province. Main research contents include: firstly, the theoretical study on the relationship between absorptive capability and FDI technology spillovers in developing country was done in order to provide theoretical basis for the later empirical studies; secondly, considering the hierarchy of the macro absorptive capability, opening degree, regional innovation capability and financial development were chosen as the proxy variables of three dimensions of absorptive capability, and the threshold modes of FDI technology spillovers based on each dimension of absorptive capability were analyzed; thirdly, based on the re-definition of absorption capability from the generalized perspective, a index system was constructed in order to measure the generalized absorption capability of each province, and then the time-lag characteristic of FDI technology spillovers was introduced into the empirical model in order to study the dynamic threshold mode of FDI technology spillovers based on generalized absorption capability. Compared with the existing research, the innovative work of this dissertation is mainly reflected in the following aspects:①Focusing on the developing country, a two-stage Cournot oligopoly model was constructed in order to illustrate the fees decisions for gaining or preventing the spillovers effect and the output decisions of domestic-funded and foreign-funded enterprise, and then the theoretical analysis on the general mechanism between absorption capability and FDI technology spillovers was done. The research indicates that: When the absorptive capacity of domestic enterprise belows the threshold value, the equilibrium rate of technology spillovers increases as the absorptive capacity of domestic-funded enterprise increases. At the same time, domestic-funded enterprise is willing to pay more for gaining the technology spillovers, while the foreign-funded enterprise is willing to decrease the pay for preventing the technology spillovers. When the absorptive capacity of domestic enterprise is higher than the threshold value, the change of equilibrium rate of the technology spillovers and the cost of gaining or preventing the technology spillovers can not be determined. The existence of absorptive threshold effect in developing country was demonstrated, the different results of all the empirical researches on the relationship between absorptive capability and FDI technology spillovers was fully explained, and this lays a good theoretical basis for the later empirical study.②Considering the hierarchy of macro absorptive capability, the inter-province panel data of China during 1985-2007 was used to construct a threshold model, in order to analyze the threshold effect of FDI technology spillovers from opening degree, regional innovation capability and financial development perspective, respectively. And then the empirical test of threshold effect, the calculation of the true threshold value and the analysis of specific threshold mode were done. The empirical results indicate that the FDI technology spillovers has the significant“dual-threshold effect”based on opening degree, regional innovation capability and financial development in China. However, the mode of each threshold effect is different; the threshold mode based on the opening degree looks like“inverted U”type, the threshold modes based on the regional innovation capability and the financial development are both“three gradient”type. The extended analysis shows: At present, the FDI technology spillovers are significant in most provinces, but the FDI technology spillovers in these provinces are absorbed insufficiently.③Considering the cooperativity of each dimension of absorptive capability, a re-definition of absorption capability based on the generalized perspective was done, a index system was constructed in order to measure the time-varying trends and regional differences of generalized absorption capability in each province. On this basis, the time-lag characteristic of FDI technology spillovers was introduced into the empirical model, and the dynamic threshold regression technology was used to analyze the relationship between generalized absorptive capability and current technology spillovers of FDI, the in-depth analysis of dynamic threshold mode based on generalized absorptive capability was done. The measurement of generalized absorptive capability shows: During 1985-2007, generalized absorptive capability of each province increases with a low speed; the eastern coastal provinces almost got a higher absorptive capability, while the western provinces got a lower absorptive capability. The GMM estimation result of dynamic threshold model verifies the existence of time-lag characteristic of FDI technology spillovers, and it shows that: After considering the time-lag characteristic of FDI technology spillovers, the current technology spillovers of FDI has significant“tri-threshold effect”, the mode of it is the“four gradient”type. By the end of 2007, since the generalized absorptive capability in each province has not been cross the max threshold value, the current technology spillovers of FDI became significant but did not reach the highest value, and it indicates that the absorption of current technology spillovers of FDI has some room for improvement.更多还原