【作者】 柴洪辉；

【导师】 顾海英；

【作者基本信息】 上海交通大学 ， 农业经济管理， 2009， 博士

【摘要】 从长期的视角来看,只有少数发达国家跨入了持续的经济增长阶段,大多数国家仍在长期的经济停滞中挣扎。经济增长并非均衡而普遍地展开,世界贫富差距在不断扩大。是什么原因造成了长期经济增长的国家差异是发展理论和增长理论都在探讨的问题,越来越多的学者开始从理论上关注人类社会从停滞到增长的内在机制。正如Lucas(1988)所说的那样:“在这类问题中所涉及的人类福利是如此惊人:一旦你开始思考这类问题,就很难再思考任何其他问题。”传统的研究通常把长期经济增长过程划分为马尔萨斯陷阱和持续增长两个阶段,从而对于长期增长问题的研究主要集中在从停滞阶段到增长阶段的转型。这样的阶段划分在某种程度上把工业革命视作技术革命而非渐进的变革,从而使得构建起来的理论模型在处理逃出马尔萨斯陷阱时依赖于外生的技术冲击。然而已有的证据表明,无论从马尔萨斯陷阱到现代持续增长转型的速度看上去有多迅猛,历史证据都表明这种转型本身都是一个长期渐进的内生过程。新增长理论在解释现代平衡增长上很成功,但是对于处理停滞阶段的表现差强人意,发展理论在讨论从停滞到增长阶段的结构转型上有着自己的理论构建,但是无法提供长期的一致的理论视角。正是在这个意义上,如Galor and Weil(1999,2000)提出了三阶段的划分方法并且构建了统一增长模型(United Growth Model)来解释人类历史上所经历的长期增长过程。Galor(2005)指出,构建一个统一的增长框架是一项十分具有挑战性的工作,需要进行方法论上的革新。依照经济史数据,我们可以将长期的经济增长过程划分为停滞、缓慢增长、高速增长和平衡增长四个阶段。以此为基础,可以把Galor and Weil的构建统一增长模型的思想进一步扩展,构建起能够解释世界经济在两千年的发展中经历的停滞、缓慢增长、高速增长、平衡增长全过程的理论模型。基于四个阶段的划分,本文构建了一个涵盖结构转型和人口转型的内生长期经济增长模型。该模型经济依次演化,经历了停滞、缓慢增长、高速增长跨入了平衡增长,并且在缓慢增长和高速增长阶段内生的实现了结构转型和人口转型,从而将发展理论和增长理论的核心问题整合在了一个内生长期增长框架之下。本文在Galor and Weil(2000)的基础上,以人力资本和资源为生产要素构建起一个具有农业生产部门和非农业生产部门的经济模型,以家庭的理性选择为基础,从分析家庭在养育孩子上的教育投入为出发点,研究家庭的教育投入的起点和教育投入的演化,从而分析教育投入水平和技术进步率的相互影响,进一步讨论了人口、教育投入水平、技术进步率的相互影响和联合演化问题。在联合演化的基础上,以教育投入水平变化为主线讨论了人口转型问题,以技术进步率的变化为主线讨论了结构转型问题,最后利用上述的结论讨论了长期增长过程和驼峰形事实。遵循上述思路,本文的结构安排如下。第一章为导论。以长期的经济增长数据为基础,本章主要讨论了长期经济增长的过程,描述了马尔萨斯陷阱、库兹涅茨事实、卡尔多事实、人口转型和驼峰形事实,提出了从长期的视角来看典型事实统一于长期增长过程的观点。以这一观点为核心,本章讨论了增长理论和发展理论通常认可的传统的增长两阶段划分模式以及新近研究中的三阶段划分模式的缺陷,指出正式由于对于马尔萨斯陷阱含混不清的界定导致了增长阶段划分的不明晰,提出了以长期经济增长数据为基础的停滞、缓慢增长、高速增长和平衡增长四阶段划分方法。从长期增长的四阶段划分出发,指出增长理论和发展理论不相容的根本原因在于没有从长期增长过程整体视角的角度考察增长问题,从而从融合增长理论与发展理论的角度提出了本文立论的问题。第二章为文献综述。本章以新增长理论的兴起为基础,考察了Romer(1986)和Lucas(1988)所开启的新增长理论对于增长理论和发展理论融合方面的研究文献,从长期增长与发展理论融合研究、长期增长中技术与人口关系研究、长期增长与人口转型关系研究和长期增长与结构转型关系研究四个方面梳理了这些文献的思想渊源和基本贡献,评述了当前研究中的缺憾和不足。在此基础上,文章对于Galor and Weil提出的构建统一理论的思想和其理论框架(G-W模型)做了详细的探讨。分析表明,G-W模型实质上是在新古典框架之下汲取了演化经济学的思想,对新增长理论中的人力资本模型和研发模型进行有机整合的结果G-W模型的理论贡献在于构建了当前最为复杂的融合增长理论与发展理论的增长模型,在构建统一增长模型(United Growth Model)方面做出了奠基性工作;G-W模型不足之处在于做为单部门模型只能讨论从马尔萨斯陷阱到现代增长的演化问题,而无法讨论长期增长过程中的结构转型和驼峰形增长。第三章构建了两部门内生统一增长模型的基本框架。本章以Galor and Weil关于构建统一增长理论的思想内核为基础,针对G-W模型的缺陷和可能的扩张方向构建了一个两部门长期增长模型。不同于G-W模型的地方在于,本文放弃了G-W模型的演化分析方法而沿用了新古典的分析框架,通过两部门之间交换价格的设定从人均收入的角度构建起了一个内生长期增长框架,从而使得本模型可以在长期增长框架下分析长期增长过程中的结构转型、人口转型、驼峰增长。在本模型中,经济系统中有着农业生产部门和非农业生产部门两个部门,进入两个部门的生产要素为人力资本和资源,价格由两个部门之间的产品交换量来确定。经济系统中家庭的效用函数为C-D函数,每个家庭通过选择自己的消费水平和孩子的数量和质量来最大化自己的效用。由于家庭期望的孩子的收入水平进入了家庭的效用函数,所以每个家庭当期所作出的抉择都受到此前各期家庭的抉择的影响,同时又会影响到以后各期家庭所作出的选择,这样就把家庭当期的选择问题和家庭长期的延续问题关联起来,从而把一个短期的家庭优化问题转化成了一个长期的经济增长问题。在本模型中,除了进入生产方程的资源系外生给定之外,人力资本、教育投入水平、技术进步、人口增长都实现了内生化,搭建了后文分析人口转型、结构转型、长期增长的基本分析框架。第四章讨论了模型经济的教育演化问题。本章通过对于教育的价值的讨论,运用日本和英国的数据,探讨了教育对于长期经济增长的促进作用。以此为基础,本章对模型经济中的教育问题进行了讨论,指出本模型在教育问题上与G-W模型的不同之处在于家庭的教育投入问题与收入密切相关。对于家庭优化结果的分析表明,在理性决策的支配下,家庭在相当长的时期之内会选择零教育投入水平。当且仅当家庭选择教育投入的门槛条件突破之后,家庭才会选择对自己所养育的孩子进行教育投资。本章的分析表明,技术进步率的高低对于教育投入水平的选择有着关键的影响。尽管教育投入水平会随着技术进步率的提高而提高,但是对于技术对教育投入水平影响的讨论和教育投入水平,对于教育投入水平和技术进步率的演化关系上我们选择了Logistic曲线来描述二者的关系。这样的选择既是基于现实的观察,也是技术发展轨迹的要求。至此,我们对模型经济中的教育问题进行了完整的讨论,为下一章讨论技术的演化提供了支撑。第五章主要讨论了模型经济中的技术演化问题。本章我们从技术变迁的历史出发,讨论了以工业革命为代表的技术进步是一个连续的缓慢的技术变迁过程而非突变性的技术革命,指出技术革命的称谓割裂了技术发展的历史从而忽略了技术变迁的连续性。以技术的连续演化为主线,本章讨论了模型经济的技术进步问题。文章指出,从科学发现的逻辑来看,教育驱动的技术进步率的演化模式遵从logistic曲线,人口规模驱动的技术进步率也遵从Logistic曲线。以此为基础,我们讨论了模型经济的技术变迁的动力和演化,分析了人口规模、教育投入水平和技术进步率三者联合演化过程,给出了模型经济中教育和技术进步率的演化轨迹。研究表明,在教育投入的门槛条件没有达到之前,经济系统中的教育投入持续为零,技术进步率会缓慢上升,从而人均收入水平会上升,经济缓慢增长指出人力资本的演化路径表现为在( e, g )平面中沿着et+1曲线向右上方移动, et+1曲线上的( et +1 , gt+1 )、( et + 2 , gt+ 2)等点所代表的人力资本水平即为家庭理性选择下的人力资本存量。第六章主要讨论了人口增长及转型问题。本章从人口增长的历史出发,回顾了有关人口转型的研究,讨论了人口、技术和增长研究的方法论问题,借鉴蔡昉(2006)的观点,对于模型经济中的人口增长问题进行了分析。研究表明,模型经济中的人口增长存在着门槛条件,在门槛条件未能达到之前,经济将长期处于停滞之中,而且这种停滞是稳定的,经济自身不具备突破这种停滞的机制。在停滞阶段,模型经济中的生产保持稳定,非农产业产品投入农业产业的比例稳定,产业间的劳动力分配稳定。如果某一冲击使得人口增长的门槛条件突破了,经济中的人口会实现持续增长。在这个增长过程中,人口的增长率先上升后下降表现出了明显的转型过程。模型经济的人口转型过程与历史上的长期人口增长数据的趋势相吻合,从而使得本文所构建的长期增长模型可以较好地解释现实经济长期增长中的人口转型现象。第七章主要讨论了结构转型问题。本章以库兹涅茨的产业结构变迁的研究为引言,讨论了产业结构变迁和转型的思想渊源以及最新的研究进展,探讨了当前在新增长框架下讨论结构变迁问题的理论模型的贡献与缺陷。基于此,本文运用教育演化、技术演化和人口转型的相关结论研究了模型经济的结构转型问题。结果表明,模型经济在人口开始增长之后,农业和非农产业之间的劳动力分配就开始演化,农业部门的实际劳动力份额逐渐降低并趋于稳定,非农产业部门(不包含教育)的实际劳动力份额逐渐上升并趋于稳定,教育产业的劳动力份额在非农产业(不包含教育)的劳动力份额稳定之后开始上升并逐渐趋于稳定。与劳动力演化相一致的是农业产业的收入份额逐渐下降,而非农产业的收入份额逐渐上升后稳定,教育产业的收入(产值)份额在非农产业收入稳定之后开始上升。因此,模型经济中产业间劳动力份额的演化和收入份额的演化与现实经济在长期增长过程中的结构转型趋势相一致。作为一个引申的讨论,本章在结尾部分指出,模型经济中农业的发展需要工业的支撑。尽管随着经济增长,投入农业的工业品相对份额在降低,但其绝对值上升,这是农业自身发展和经济持续增长的需要;而工农业产品剪刀差的存在将会导致农业部门的人口增长率较高发展缓慢以及工农业收入差距不断扩大,而且工农业产品剪刀差对于工农业收入差距后人口增长率差距的扩张具有自我加速的扩张效应。第八章讨论了长期增长和驼峰问题。本章以发展理论和增长理论各自面临的困境出发,讨论了新增长理论的新古典特性。新增长理论的新古典特性对于讨论长期增长问题无法提供有效的理论平台,正是在这个意义上Galor and Weil提出了构建统一的增长模型的必要性。以此为基础,本章对于模型经济增长率的研究表明,只要人口增长的门槛突破了,经济就会持续增长,经济增长率先上升后下降,最终会趋向于一个固定的增长速度的平衡增长路径,表现为一个明显的驼峰增长过程,这也进一步验证了我们关于经济增长阶段的划分。因此,现实中各个国家或区域的经济增长率横向比较的驼峰现象源自于各自处于不同的经济增长阶段上。第九章为结论性评述。本章从四个方面讨论本文所构建的统一增长模型的基本结论并简要评述了本文构建的长期增长模型,指出本文的主要创新之处在于构建了一个可以较为完整地解释长期增长过程的统一增长模型。本文研究的不足之处主要在于无法将增长的原动力内生化,后续的研究方向可能在内生化原动力、讨论模型中资源边界的扩展和汲取演化经济学方法几个层面上展开。本文的主要创新之处在于四个方面:(1)本文构建了涵盖停滞、缓慢增长、高速增长和平衡增长四个增长阶段的内生两部门长期增长模型,将长期增长过程中的人口转型、结构转型和驼峰形增长整合在了一个内生长期增长框架中。(2)本文构建起来的长期增长模型对于人口转型的讨论整合了人口转型和工业化的关系,可以清楚的判明模型经济中工业化和人口转型的时间相继关系,实现了模型的逻辑时间与现实经济的历史时间的相互吻合。(3)本文构建的模型经济的人口增长率随着经济增长趋向于零的确定性结论解决了G-W模型的最终人口增长率无法判别的问题(4)本文构建的长期增长模型存在一个渐进的平衡增长路径,解决了G-W模型本质上不存在平衡增长路径的问题。本文的不足之处在于模型无法内生化增长原动力问题,因为本模型依然依靠一次性的外生的冲击来跨越初始增长的启动问题。另外一个不足之处在于模型中我们只能描述出一个较弱的驼峰增长趋势,但是我们无法讨论经济增长率在何时达到最高点并开始下降。更多还原

【Abstract】 From a long run view, only a few countries have been stepping on the sustainable growth path while the others have still strugled for escaoping stagnation. That means, however, economic growth isn’t unrolling in the world and it has witnessed the expanding of distance between the poorest and richest in the world. What caused this arosed more scholars’interest in peering the mechanisim of long-run growth which are discussed in the growth theory field as well as development theory field. As Lucas(1988) argued,“so many welfare are involed in such questions so that you can never moved out when you begin to think of it.”Traditional research divided the long-run growth into two phases: Malthusian stagnation and modern growth and thus take more attention to the escaping from Malthusian stagnation to modern growth. Such a division has somehow taken the industrial revolution to a technological breakthrough rather than a progressive process and therefore its theoretical model need exogenous shock to modeling the escaping process.However, existing historical data showed that process gradually and endogenously prograssed in long term no matter how rapidly was the escaping from Malthusian stagnation to modern growth. New growth theory can successfully explained the modern growth stage while has no significancs to stagnation stage; those theoretical model constructed with stangdard growth theory and integrated some significance from development in it can provides some insights to the structural transition but have provided no united view to long-run growth theory. So, Galor and Weil proposed division of 3 stages and tried to constructed united model to explain the long-term growth economic growth, which is a challenging work and needed the innovation of methodology.According to historical economic data, we found the long-term economic growth process of the world can be divided into stagnation stage, slow growth stage, high growth stage and balanced growth stage. This paper, with the enlightenment from Galor and Weil’s works done, modeled those four stages endogenously in a long-term run growth framework which could finished structural transition and demographic transition in it, i.e. such a model has combined the core issue of growth theory and development theory into one long-term endogenouse growth framework.This dissertation, based on the G-W model, constructed a economic model with agricultural sector and non- agricultural sector. It firstly analyzed the education level chosen by parents and then come to evolution of education level and its threshold. Then it turned to the united evolution of population scale, education level, and technological progress rate. With the conclusion obtained, the dissertation discussed the demographic change mainly against education level, the structural change mainly against technology. Furthermore, it discussed the long-run growth and the hump-shaped fact. Along with this technological routine, the dissertation was organized as follow.Chapter 1 offered an introduction to this research. Based on the long-run historical data of word economy’growth, we discussed the Malthusian Stagnation, Kuznets’Facts, Kaldor’s Facts, Demographic Transition and Hump-Shaped fact and pointed out that the traditional 2-stage view of long-run growth process was somehow not exact and so the 4-stage would be more useful. Based on these discussions, it came to conclude the research issue of this dissertation.Chapter 2 offered a literature review. In this chapter, we checked the existing research with the mainstream of New Growth Theory and emphasized on the merging of development theory and growth theory with comments and critics. At last, we discussed the G-W model especially in its theoretical contribution and its imperfection.Chapter 3 constructed a united growth model. In this chapter, we take over the phenomenon of G-W model to construct an over-lapping endogenous economic growth model with agricultural sector and non-agricultural sector and thus constructed the framework for further analysis.Chapter 4 takes great deal to education evolution of the model congstructed. This chapter firstly dealt with the value of education with the historical data of Japan and the Great British to show the role of education in economic growth. Then it came to discuss the relation between education and income, the threshold condition of education investment and the influence of technological progress. At the end, the upper bound of education level was discussed.Chapter 5 mainly discussed the evolution of technology in model economy. Beginning with the history of technology change, this chapter argued the technological change emerging in so called industrial revolution was such a slowly changed process rather than a revolution in its natural meaning. In the view of slow long-lasting change of technology, it talked about the momentum of technology change and evolution and then analyzed the united evolution of population, education and technological progress rate.Chapter 6 mainly discussed the demographic growth and transition of the model economy. With long-term historical data of population, This chapter argued the methodology of demographic research and discussed the stagnation of population in model economy and its stability. Then it found the threshold condition of population growth and analyzed the process of it. Based on the growth rate, this chapter attaches more importance to the sustainable growth and come to the conclusion of endogenously demographic change in model economy.Chapter 7 mainly discussed the structural change of the model economy. With the introduction of the research works by Kuznets, this chapter recalled the source of structural change and reviewed the existing modeling works of three paths. Then it turned to discuss structural change of this model economy based on chapter 3 and chapter 4 and come to the relation between agricultural sector and non-agricultural sector. As a result, it found the support of non-agricultural sector to agricultural sector played an important role in the growth of model economy. At the end, it discussed the issue of the price between agricultural goods and non-agricultural goods when it lost balance.Chapter 8 discussed the long-term growth and the hump-shaped fact question. Beginning with the dilemma faced respectively by growth theory and development theory, this chapter explored the neo-classical character of New Growth Theory and thus argued the necessity of constructing the United Growth Model. Then it turned to the discussion on the long-term growth and hump-shaped fact from a view of growth rate.Chapter 9 briefly reviewed the long-term growth model constructed in this paper and pointed out that the main innovation of this paper referred to the long-run growth model constructed which could endogenously discuss the structural change, demographic change and long-run growth as well and therefore integrated the development theory into growth theory to some extent. Based on this, end of this chapter were argued the deficiency of the model constructed and the following research from three directions.The main innovation of this paper referred to the long-run growth model constructed which could endogenously discuss the structural change, demographic change and long-run growth as well and therefore integrated the development theory into growth theory to some extent. The second innovation lied in the first discussion of the support of industrial sector to agricultural sector in a long-term growth model. The discussion of the price difference between industrial sector to agricultural sector and thus the conclusion obtained should be mentioned as another one innovation in this paper which would never discussed in existing model.更多还原