【作者】 高远东；

【导师】 陈迅；

【作者基本信息】 重庆大学 ， 数量经济学， 2010， 博士

【摘要】 空间计量经济学作为一门新兴的学科,正被广泛应用到增长经济学、城市经济学、区域经济学等多个领域。其中,空间计量主要应用领域之一——经济增长,正成为国内外空间计量学者研究的热点。国内外既有的空间计量实证研究,通过和非空间模型回归结果比较,验证了将空间计量学应用在区域经济增长研究的有效性和对区域经济增长,以及区域经济发展差异更合理的解释。然而,国内目前对经济增长的空间计量研究仍较为粗浅,大部分是在原经济增长模型中直接加入空间变量,改造方法粗糙,而且大部分也是基于空间截面数据进行的研究,对空间效应估计有所夸大。此外,从投入要素角度进行的研究,也未能充分考虑实物资本的异质性和人力资本的空间效应。为真实、准确的测度国内资本、外商直接投资(FDI)和人力资本对经济增长的贡献,及其空间外溢显著性和外溢程度,以及在我国东、西部不同区域具有的显著差异,本文基于省域技术水平间存在空间依赖性的理论假设,并且充分考虑了资本空间外溢的异质性,构建了空间计量模型实证研究。首先,本文就国内资本、FDI和人力资本对中国区域经济增长的作用进行了概述;对我国经济增长及东、中、西部不同区域经济发展存在的差异,从投入要素角度出发,进行了统计分析。其次,基于省域技术水平的空间依赖性和资本外溢性的假设,并假设FDI区别于国内资本具有空间外溢异质性,将FDI从国内资本分离,基于索罗模型,构建了空间扩展索罗模型,对中国省域及东部地区进行了固定效应空间面板估计,验证了假设的正确性,并得出了有力的结论。而且,进一步运用地理加权回归测度了国内资本和FDI及其外溢效应在我国不同省域的异质性。再次,本文将资本扩展至广义资本,基于上述同样假设,从MRW模型出发,构建了空间扩展MRW模型,对中国及东、西部地区进行了固定效应面板估计,测量了人力资本及其空间外溢效应的显著性和外溢强度,以及在东、西部区域的不同作用。同时,与MRW模型进行对比,证实了空间扩展MRW模型更优。此外,为进一步研究人力资本作用,本文从Benhabib-Spiegel模型出发,构建了空间Benhabib-Spiegel模型,将人力资本作用分解为自主创新能力和对先进技术的吸收能力,运用同样的空间计量方法进行了研究。得出了中国全域及东、西部地区人力资本的自主创新能力均不显著等一系列有价值的结论。最后,对前述实证结果进行归纳、综合分析,提出主要结论如下。①省域技术水平间存在显著的空间依赖性。②空间扩展索罗模型、空间扩展MRW模型和空间Benhabib-Spiegel的计量结果显著优于索罗模型、MRW模型和Benhabib-Spiegel模型。③国内资本空间外溢性显著,呈现出从东部到西部递减的显著特征。国内资本及其空间外溢性的总和对经济增长的贡献从东到西递减。西部地区,国内资本对本区域经济增长贡献最大,即西部地区经济增长对国内资本依赖性最大。④区别于国内资本,FDI空间外溢具有显著的异质性,其对省域人均收入的贡献和溢出强度均小于国内资本。而且,FDI的外溢性,除在我国东部地区显著外,其余均不显著。⑤全国、东部和西部地区,人力资本及其空间外溢效应均不显著。⑥东部地区人力资本及空间外溢性对区域经济增长的贡献均大于西部地区。⑦在全国,人力资本自主创新能力不显著,对外来先进技术的吸收能力显著。分区域而言,东部地区人力资本对外来先进技术的吸收能力显著,西部地区则不显著;东部和西部地区人力资本的自主创新能力均不显著。而且,以促进生产要素在不同区域的有效投入,FDI外溢作用的最大程度发挥,区域人力资本水平提高,以及省域间或区域间经济协调增长为目的,进一步提出了针对性的政策建议。更多还原

【Abstract】 The Spatial Econometrics as a rising subject is applying in lots of economic fields of Economic Growth, Regional Economics and Urban Economics. Moreover, the Economic Growth has gradually become one of the focuses what the spatial econometric scholars study. The abundant literatures have proved it is necessary and robust that the spatial variables account for regional economic growth and per income difference among regions or countries by comparing the empirical results between the spatial economic growth model and non spatial economic growth model.But, the spatial econometrics study about economic growth in China is superficial. The majority of spatial econometric model is rough, which is presented by adding spatial terms of variables into non-spatial economic growth model. In addition, the majority of existing literatures run the spatial econometric regression using the cross-sectional data. It leads to that the spatial effect is exaggerated. The correlative study from production factor aspect has ignored the the spatial effect of human capital, and the heterogeneity of FDI different from the domestic capital.Based on the physical capital externalities, the heterogeneity of the capital’s spatial spillover and the technological spatial interdependence between provinces, the paper develops spatial econometric models and runs an empirical regression in order to estimate truly the contribution of domestic investiment, FDI and human capital on regional economic growth,and the spatial effect of domestic investiment, FDI and human capital which is significantly different in Chinese different region.Firstly, this paper analyses the effect of domestic investment, FDI and human capital on regional economic growth, and has a statistical analysis about regional economic growth and the difference of different regions in China. Secondly, the paper supposes that the spatial interdependence of the change of the aggregate level of technology exists between economies, and the physical capital externalities exist. Based on the Solow economic growth models, the paper develops a Spatial Solow Model and conduct Fixed-effect Spatial Panel Estimation using in the sample of China and China’s eastern region. The empirical results approve the correctness of hypothesis and receive lots of vigorous conclusion. Again, the methos of GWR is used to estimate the contribution of domestic investment and FDI on regional economic growth and their spatial effects, which is heterogeneous in Chinese different provinces. Thirdly, based on the same hypothesis, the paper expands physical capital to generalizede capital, and presents a Spatial MRW Model on the basis of MRW model. The paper has Fixed-effect Spatial Panel Estimation using in the sample of China in order to estimate the spatial effect of human capital and different effection in different regions in China. The results display the Spatial MRW Model is superior to MRW model.Fourthly, the paper develops a Spatial Benhabib—Spiegel model on the basis of the economic growth models proposed in Benhabib and Spiegel(1994) in order to study the effect of human capital. The effect of human capital is divided into the ability of absorbing the advanced technology from techchnology leader and the ability of innovation in Sptial Benhabib—Spiegel. The results including the spatial effect of human capital is not sidnificant, is gained.Finally, the paper summarizes the results which are in the following.①The technological spatial interdependence between provinces exist significantly.②Compared with Solow Model, MRW Model and Benhabib-Spiegel,The regression results of Sptial Solow Model, Spaital MRW Model and Spatial Benhabib-Spiegel are better.③The spatial spillover effect of domestic capital which is decreasing from east to west is significant. The contribution of domestic capital and its spatial spillover effects to economic growth is decreasing from east to west. The contribution of domestic capital to economic growth is the largest in China’s western region.In other words.④Different from domestic capital, the spatial spillover effect of FDI has the heterogeneity. The contribution of FDI to economic growth and the degree of spillover is less than domestic capital. But, the spatial spillover effect of FDI is insignificant except in China’s eastern region.⑤The spatial spillover effect of human capital is insignificant in the whole country, China’s eastern region and China’s western region.⑥The contribution of human capital and its spatial spillover effects to economic growth in China’s eastern region is not more than in China’s western region.⑦The ability of innovation of human capital is insignificant while the ability of absorbing the advanced technology from techchnology leader is significant in China. Specifically, the ability of absorbing the advanced technology from techchnology leader is significant in China’s eastern region while is insignificant in China’s western region. The ability of innovation of human capital is insignificant in China’s eastern and western regions.Moreover, the paper put forward some suggestion in order that production factors is invested effectively, the spill over effection of FDI is exerted as great as possible, and the economic level is increased harmoniously between different regions in China.更多还原