【作者】 杨杨；

【导师】 吴次芳；

【作者基本信息】 浙江大学 ， 土地资源管理， 2008， 博士

【摘要】 改革开放以来,中国经济维持了长达20多年的高速增长,与其相伴随的是,社会经济发展对土地利用空间和土地产品的需求不断扩大。由于土地资源的自然供给是有限的,土地供给的紧缺性与社会经济需求的增长性之间失衡发展的态势日益显化,有限的土地资源已经成为制约中国经济长期增长的障碍。根据Romer(2001)的假说,每个国家都不可避免会受到土地资源约束对经济的“增长阻尼”作用。如何定量测度符合中国土地资源特点的“增长阻尼”,对于土地资源宏观调控目标的制定具有重大战略意义,同时,强化土地资源约束问题的基础理论研究对推动中国经济走出土地资源瓶颈具有重要的现实意义。通过对已有研究的梳理发现,将土地纳入完整生产函数分析框架的代表性文献十分有限。度量土地资源约束对经济增长的影响,可以借鉴Romer(2001)“增长阻尼”的研究方法。但是,已有研究主要是基于C-D生产函数的模型框架,为了克服C-D生产函数替代弹性为1、忽略技术进步以及规模报酬不变的缺陷,本文构建了改进的二级CES生产函数模型作为基础模型度量土地资源约束对经济增长的影响程度。同时,考虑到不同研究区域土地资源状况的差异,因此,拓展了已有研究关于“存在土地资源约束”的假设,设置了“土地资源总量固定”、“土地资源紧缺状况更加严峻”和“土地资源紧缺状况有所缓解”三种情景假设。基于改进的二级CES生产函数,推导出三种情景假设下,土地资源对经济的“增长阻尼”测算公式。界定了研究的时间范围为1985-2005年,空间范围分为全国层面和省级层面两个层次。然后对数据进行整理,关键是界定了土地资源总量,将除了未利用土地以外的土地类型加总作为土地资源总量。全国层次的实证研究中,严格按照计量经济学的参数估计方法以及检验方法进行参数估计,以保证“增长阻尼”测算公式中所用参数的数据质量。在对数据进行平稳性和协整检验之后,首先采用最小二乘法进行参数估计,估计结果发现解释变量之间存在较为严重的多重共线性,违背了应用最小二乘法进行参数估计的基本假设。为了克服多重共线性,采用岭回归的方法进行参数估计。最后,根据1985-2005年间中国土地资源总量年均增长率0.638%的特点,选择“土地资源紧缺状况有所缓解”的情景假设3作为“存在土地资源约束”的假设条件,计量结果表明,土地资源约束对中国经济增长有着较大的影响,中国每年的经济增长速度比没有土地资源约束的情形下降低了0.75%。省级层面的实证研究中,按照与全国层面相同的逻辑,测算出了各省土地资源对经济的“增长阻尼”。重点是运用全局和局部空间自相关指数(Moran’s I和Moran’s I_i)分析省级“增长阻尼”的全局空间关联关系与局部空间关联关系,同时也分析了省级“增长阻尼”的空间集聚中心与空间孤立点。研究结果表明,总体上,省级“增长阻尼”在全国范围内具有较好的空间结构性和空间关联性,呈现出显著的空间集聚模式,从局部来看,省级“增长阻尼”表现出十分明显的局部空间差异,表现出高-高关联类型的沿海指向性与低.低关联类型的内陆指向性,说明沿海地区土地紧约束的状况更为显著。最后,基于土地资源“增长阻尼”的理论分析与实证研究,提出缓解土地资源约束对经济增长影响的路径选择,即依靠土地开发、整理、复垦和存量建设用地整理,“开源”;利用资本和技术对土地资源的替代作用,节约与集约利用土地资源,“节流”,通过“重点省份调控”,缓解沿海地区土地紧约束的状况。这样的路径选择表明,未来中国经济发展仍将处于稳态的增长路径之上。更多还原

【Abstract】 Since reform and opening-up, the Chinese economy has maintained more than 20 years fast growth. Following with the socio-economic development, people have more and more demand to land utilization space and land product. Because the natural supplies of land resources are limited, the unbalanced situation between the scarce of land supplement and the growth of social-economy demand, the limited land resources have already become a restriction factor of long-term economic growth in China. According to the assumption of Romer model (2001), owing to the limitation of land resources, no country can avoid the influence of "growth drag". So it is strategic significance to measure the "growth drag" according to the character of land resources of China. Meanwhile, strengthen the theory research of land resources-restriction is practical significance to overcome land resources bottleneck.After combing the literature, we discovered that the representative literature integrating land into the analysis frame of a complete production function is very limited. We can use the "growth drag" of Romer (2001) as the research method to measure the impaction of land resources limitation to economic growth. To keep the analysis manageable, the former researches started with the case of Cobb-Douglas production function, in which the elasticity of substitution among these factors is 1. But the hypothese is not true in the real economy world. So we choose the modified two-level CES production function as the basic model to study the "growth drag" of land resources. Meanwhile, the thesis modify the hypothese of former research and gives three hypotheses of "land resources restriction", which are "the quantity of land resources is fixed"、"the scarce of land resources is more stern "、"the scarce of land resources is more alleviation". Based on the modified two-level CES production function, we get the equations of "growth drag".Following with the determination of study range, we take 1985-2005 as the time series, and study on the country level and province level. Then we begin to deal with the data, in this part, the thesis make some statistics and economic definition about the variable of Y, K, L, T, and define the data of land resources as the whole land except the unused land.In the empirical study of country level, the thesis carries on parameter estimation according to the econometrics methods absolutely. After testing the stationary and co-integration, the thesis use the OLS to estimate the parameters first, but the serious multicollinearity between variables violates the basic hypothese. then we use the ridge regression method to overcome the multicollinearity. Finally, based on the hypothese 3, we get the growlh drag. The results indicated that the growth drag of China was 0.007512 annually.In the empirical study of province level, according to the same logic as country level, the thesis calculates the "growth drag" of each province. Methods of spatial autocorrelation indicators were employed to analysis the spatial correlates of "growth drag" of provinces. The results indicate that "the impaction degree of land resources limitation to economic growth" showed obvious spatial cluster pattern; the spatial clusters and outliers showed definite rules; "the impaction degree of land resources limitation to economic growth" showed obvious local spatial disparity pattern.Finally, based on the theory analysis and empirical study, the thesis offer some choice to alleviation the impaction of land resources limitation to economic growth, which is land readjustment, substitute of capital and technique to the land resources and important province control. Those policy choices indicate that if we insist the land resources protection policy, China will be on a balanced growth path in the future.更多还原