消费波动、消费增长和中国经济波动

【作者】 李凌；

【导师】 权衡；

【作者基本信息】 上海社会科学院 ， 政治经济学， 2009， 博士

【摘要】 自改革开放以来,中国的经济增长方式一直是以投资和出口驱动的,消费则处于非主导的地位,“高增长、高投资、高出口依存度、高储蓄和低消费”成了中国宏观经济典型的特征写照。然而在总需求的三大组成部分中,投资具有“引致性”和“派生性”的特点,(净)出口贸易则易受外部市场环境和本国在国际贸易中的地位等不确定因素的影响,投资和(净)出口贸易的出发点与落脚点都是为了消费,唯有消费才具有决定性的意义。如果缺乏消费的有力支撑,新增的投资产能就难以及时被市场消化,大规模投资引致的结构性通胀将被产能过剩和库存化所取代,随之而来的通货紧缩又将取代结构性通胀,从而会引发一轮又一轮通胀与通缩交替出现的经济波动。但国内需求不足又迫使中国经济提高对外依存度,更加依赖于投资和出口驱动的增长模式,从而产生一定程度的路径依赖。因此,以投资和出口拉动的经济增长方式是不可持续的,消费率过低既不利于社会成员生活质量的提高,也难以实现共同分享经济发展的成果,这同经济发展的初衷相背离。2008年下半年,一场突如其来的全球范围内的金融危机在很大程度上抑制了中国对外贸易的增长,东部沿海地区的制造业和加工贸易业最先受到冲击,随着时间的推移,衰退通过产业间的传导机制向国内服务业和其他地区的工业产业逐渐蔓延,使得投资消费失衡的矛盾更加显露无疑。在理论界,对现行增长模式的否定已然达成共识,但问题在于如何实现经济增长模式从以投资和出口为主导向以消费为主导的转化,如何实现从转变经济增长方式向转变经济发展方式的转化,如何实现社会经济的可持续和科学发展,让更多的社会成员享受到改革开放和经济发展的成果,这就需要对中国的消费和消费者行为作一个比较全面的考察。消费波动和消费增长可能是一个比较适宜的切入点,它不仅是一个关系社会成员生活水平和经济福祉的微观现实问题,也是一个涉及宏观经济健康运行的重大理论问题。本文以中国居民的消费波动和消费增长为研究主线,重点解决四个问题:一是中国居民消费波动的经验事实有哪些;二是中国居民消费增长的决定因素有哪些;三是为什么改革开放以后中国的消费波动大于产出波动;四是消费波动和消费增长哪个更为重要。值得强调的是,在消费波动和消费增长的关系上,本文尤其强调消费波动的重要性,并将其置于经济波动的一般理论中进行阐述和分析,这是因为单纯从消费增长的视角研究消费的做法,往往忽视了消费过程的随机性,在转型经济国家,这种随机性与经济波动的关系尤为显著,要理解消费增长的源泉所在必须先充分认识消费波动的随机性。除导论、结论和政策建议外,全文共分六章。第一章是基础理论部分,第二章是方法论,第三章整理和研究了中国居民消费波动的经验事实,第四章分析和检验了中国居民消费增长的决定因素,第五章揭示和阐述了中国消费波动大于产出波动的原因,第六章比较和讨论了消费波动和消费增长的经济福祉与社会福祉。具体而言,第一章通过对五种经济波动的理论解释作简要梳理和评介,旨在为中国经济波动和消费波动研究奠定理论基础,这五种理论分别是传统的凯恩斯经济周期理论、货币主义经济周期理论、新古典经济周期理论、真实经济周期理论和新凯恩斯经济周期理论。第二章通过对不同的宏观结构计量方法和时间序列滤波方法的比较,阐明本文所使用的RBC方法和CF滤波法的特点及其合理性。由于RBC模型的数值求解是本文的一个难点,为此,第二章还给出了一个从数据到GUASS程序的实例。第三章通过对中国消费波动经验事实的比较静态分析,发现改革开放以后消费波动大于产出波动的经验事实,并从消费分层的角度,印证了消费波动大于产出波动的判断在总量和个量数据上的统一。消费波动大于产出波动大多产生于发展中国家,是转型经济区别于成熟市场经济的重要标志之一。从消费结构的角度看,1991年以来城镇和农村居民在消费支出构成方面的不同步波动特征正在加强,这主要是由满足更高层次的异质性需求引起的,城镇居民的消费支出对农村居民而言具有“示范效应”;从不同收入等级消费群体的消费结构角度看,随着收入等级递增,城镇居民消费波动先下降再上升,农村居民消费波动递减。考虑到农村居民消费的绝对水平相较城镇居民而言,仍属于城镇的中等偏下水平,因此,农村居民收入等级和消费波动之间的变化规律恰好体现了城镇居民“U”型曲线的前半支,而且随着收入的增加,农村的消费波动和消费增长存在同步上升的趋势。从消费波动和经济波动的因果联系看,Granger因果分析揭示了消费波动和经济波动之间的联动性,以及城乡人均消费品支出和按收入等级分类消费群体消费波动的内在传导机制。相较而言,城镇消费波动的传导机制比农村更丰富,而农村消费波动的传导机制则过于单调;城镇消费波动具有收入维度的向下传导性,而农村消费波动则具有收入维度的向上传导性。第四章在对消费波动经验事实整理的基础上,根据中国居民消费行为的三个观察现状,建立含流动性约束和内部消费习惯的跨期最优模型,从消费函数的角度分析影响消费增长的因素,如收入、消费习惯、利率、不确定性和各类宏观经济政策,其中着重剖析消费变动与收入变动之间的(非)对称关系,并运用1991~2006年省际动态面板数据测算中国城乡居民消费的过度敏感性,分别对流动性约束、短视行为、风险厌恶情绪、预防性储蓄及λ假说进行检验。实证结果表明:城镇居民和农村居民都表现出一定的预防性储蓄动机,城镇居民面临的收入不确定性大于农村居民。城镇居民和农村居民的消费变动都呈现出对预期收入变动的过度敏感性,城镇居民消费变动的收入敏感性高于农村居民,城镇居民非耐用消费支出变动的收入敏感性低于农村居民。城镇居民消费变动的收入敏感性表现出收入变动的对称模式,参数估计支持“短视行为”假说;农村居民的收入敏感性表现出收入变动的非对称形式,参数估计支持远期流动性约束假说和“损失厌恶情绪”假说,“损失厌恶”情绪在收入增长衰减伊始达到最大值,然后逐渐平复,即使陷入危机,消费变化也不会像刚开始那样剧烈。此外,中国居民消费的过度敏感性大于美国居民,这主要是由于中国居民面临更紧的流动性约束、更弱的向上的收入流动性,以及拥有更低的财富水平所导致的。第五章在对消费过度敏感性测算的基础上,在劳动可分RBC模型中,加入流动性约束、政府消费冲击和居民消费习惯冲击和等需求因素,将RBC模型的微观基础从一元情景拓展到二元情景,在异质性代表性行为人的动态随机一般均衡(RBC)框架内,分别检验以上三种需求因素对消费波动性和相对波动性的影响,解析中国“消费易动之谜”。研究同时还对“就业变动之谜”、政府消费与居民消费和投资的关系、以及收入差距等经济结构方面的因素对经济波动的影响加以讨论。研究指出,中国的“消费易动性”源于居民的消费习惯冲击,习惯强度越大,当期消费对下一期消费而言,习惯存量就越高,消费结构也就越不稳定,消费波动就越剧烈。相反如果习惯强度较小,那么当期消费所形成的习惯对于下一期而言无需消耗更多的消费量去补偿,因而消费结构也就越稳定,消费波动就越平滑。中国的经济改革与社会转型带来的消费意识觉醒和产业结构升级,使得消费结构、消费偏好和消费习惯都处于不断的变化之中,这才是形成中国等一些发展中国家的消费易动现象的真正原因。研究同时也否定了政府消费和流动性约束是造成中国1978年以后消费波动大于产出波动的原因。其中,政府消费冲击只在一定“量的积累”上才会引起消费波动大于产出波动,“量的积累”主要体现在两个方面,一是政府消费支出与GDP的占比,二是政府消费本身的波动性;流动性约束只有在几乎全部人口受到流动性约束的情况下,才会增大消费与产出的相对波动性。但含有流动性约束的二元情景RBC模型可就经济结构方面的变量——如流动性约束、收入差距和消费差距等——对经济波动的影响进行量化分析。第六章通过建立含消费习惯养成的消费波动和消费增长的经济福祉成本模型,运用1991~2006年城乡居民人均按收入等级分组消费数据,计算不同消费群体消费波动和消费增长的经济福祉成本。研究认为,当行为人具有较大的风险厌恶情绪时,平抑消费波动所收获的经济福祉的提升可能数十倍乃至上百倍于消费增长1%所引致的经济福祉的提升;习惯强度的引入一定程度上使得具有不同风险态度的消费者,在看待增长和波动孰重孰轻的问题上,具有更为一致的评价。但经济福祉成本并不能涵盖社会发展的全部要义,通过对经济福祉成本比较模型的反思和批判,研究把消费的经济福祉分析延伸到社会福祉分析,阐述消费结构与产业结构、消费信贷与金融创新、消费增长与社会保障、以及科学消费与科学发展的关系,并结合当前金融危机背景下,就如何把握中国经济在外部冲击下的运行规律和前进方向、拉动内需和转变经济发展方式,提出了相应的政策建议。更多还原

【Abstract】 It is not consumption but investment and export that have dominated the economic growth mode since China has been opening to the outside world, which can be featured as high growth, high investment, high export-GDP ratio, high saving and low consumption. Among the three parts of aggregate demand, however, investment and international trade may serve as a starting point of consumption. From such specific view, investment is characterized by inducing and derivative while international trade is full of uncertainty such as inconstant international market environment and the changeful competitive position in the international trade. If the domestic market is lack of forceful support from consumption, the unique ultimate factor of aggregate demand, large scale of investment could not be absorbed regularly and quickly, pushing price up and resulting in structure inflation. With the time goes by, excess production capacity and abundant inventory will appear, followed with deflation. That inflation and deflation come forth by turns demolish the economic stability. At the same time, domestic demand shortage compels Chinese economy to increase export-GDP ratio, which rely more on the growth mode driven by the investment and export. Such phenomenon is called path-dependence. So the economic growth driven by investment and export is not sustainable. Furthermore, low consumption rate is also not propitious for the social members to raise their living standards and to share the fruits of economic development since 1978.A sudden and global finance crisis has greatly restrained Chinese trade growth since the second half of 2008. Manufacturing and processing trade in the east coastal region have suffered from the shock first. With the time goes by, recession spreads out from the east to the inland and from manufacturing and processing trade to service and the other part of industry step by step through the inter-industry transmission mechanism. Conflict between consumption and investment is aggravating. In theory, it has already reached an agreement on negation of the going economic growth mode. The question is how to realize the transform from the going economic growth mode to the mode driven by the consumption, the transform from economic growth to economic development and the transform from non-sustainable development to sustainable and scientific development. What we need now may be an all-round review on the consumption and consumer behavior. It is found that consumption volatility and growth is a comparative suitable breakthrough point for both individual welfare and healthy macroeconomics.The essay is tightly around the topic of consumption volatility and growth of Chinese household, trying to answer the four questions specific to Chinese household as follows.⑴What are the style facts of consumption volatility?⑵What determines the consumption growth?⑶Why consumption fluctuates more than output does? And⑷which is more important, consumption volatility or consumption growth? Moreover, the essay pays more importance or attention to the consumption volatility than consumption growth and put the analysis into the general theory of business cycles. This is not for the reason that consumption growth is less important but for paying only attention to growth is inclined to neglect the randomicity of the consumption process. Such randomicity is highly correlated to economy stability especially in transition countries like China. It is never too late to learn clearly about the randomicity before understanding how to increase the consumption.Besides introduction and conclusion, the essay is composed of six chapters as follows. Chapter 1 is about the basic theory of business cycles. Chapter 2 contains methodology and calculation methods. Chapter 3 compiles the stylized facts of consumption volatility. Chapter 4 tests the determined factors of consumption growth. Chapter 5 reveals the real reason that consumption fluctuates more than output does. Chapter 6 compares the economic and social welfares of consumption volatility and growth.More specifically, chapter 1 is the theoretical base of the essay and gives an overview of main theoretical explanation of business cycles including traditional Keynesian business cycles theory, monetarism business cycles theory, new classical business cycles theory, the real business cycles theory and new Keynesian business cycles theory.Chapter 2 demonstrates why and how the essay uses the methods of RBC and CF filter by comparing different structure macro-econometrics and filters respectively. For the solving procedures of a RBC model is a quite difficult point in the context, chapter 2 also gives an example from data to GUASS code.Chapter 3 checks the fact that consumption volatility is about one and a half times as much as output volatility since 1978 by utilizing CF filter on both macroeconomic and microeconomic data in China. Such economic phenomenon, one of the obvious identification marks of transitional economy to differ from developed economy, always occurs in developing countries. From the perspective of consumption structure, some part of consumption between urban household and rural household has getting more and more non-synchronous since 1991, which results from the heterogeneous demand in the higher level. Urban consumption has somewhat demonstrating effect on the rural household. From the perspective of the consumption structure divided by income group, with the income increasing, urban consumption fluctuates down and up while rural consumption fluctuates down straight. Accounting for that the absolute value of rural consumption is much lower than or as much as the middle level of the urban consumption, the relationship between income and consumption about rural household just shows the first half of U-curve which reflects the relationship between income and consumption about urban household. It is likely that the consumption volatility is going to keep pace with the consumption growth when the income goes up in rural. From the perspective of Granger causality between consumption volatility and output volatility, they are mutual Granger causal and Granger effect. Comparatively speaking, the volatility transmit mechanisms in urban are much more abundant than those in rural, and urban consumption volatility transmits from individuals with higher income to those with lower income but rural consumption volatility is on the other way round.Chapter 4 presents a theoretical and dynamic panel study on Chinese urban-rural household consumption about inter-temporal decision-making individual with inner habit formation and liquidity constraint based on the stylized facts supported by previous chapter and three main behavior observations on Chinese consumer, considering a series of variables as income, consumption habit, interest rate, uncertainty, macroeconomic policies and so on so forth. What is emphasized is the asymmetric relationship between the shift of different part of consumption and income growth. The essay using provincial level panel data for the period from 1991 to 2006, it is found that both urban and rural household consumptions exhibit precaution saving intentions and excess sensitivity to personal predicted income. Comparatively speaking, urban household faces more risks than rural household and the extent of the response of total (nondurable goods) consumption to the changes in predicted income is larger (weaker) for urban household than for rural household. More specifically, the mode of urban household consumption is near to the hypothesis of“myopia”where consumption changes in proportion to the predicted income. Evidence from rural household is in favor of the“loss aversion”hypothesis where consumption changes in no proportion to the predicted income. Rural consumption reduces much when one predicts that his income is going to be low. International comparison proclaims that excess sensitivity of Chinese household is much bigger than that of American household owing to more tensive liquidity constraint, weaker upper income mobility and lower wealth for Chinese household.Chapter 5 releases one of the basic hypotheses of the traditional RBC model developed by Prescott and Kydland (1982) and Hansen (1985) by allowing two different individual behavior in one RBC model, called heterogeneous agents RBC model. The first type agent is immune to liquidity constraint and the second type agent faces a complete credit restriction preventing them to inter-temporarily transfer their income. Another two kinds of demand shocks are also learnt as government purchase shock and habit formation shock. The essay tests three main hypotheses to explain why consumption fluctuates more than output does according to the latest documents in the extended RBC model. Several related topics are also discussed including employment variability puzzle, relationship between government purchase and inhabitant consumption and investment, and some economic structure factors as income gap and consumption gap. It is found that the real reason for the problem that consumption volatility is bigger than output volatility comes from the consumption habit formation shock. To the next period consumption, the more the habit strength is, the higher the habit stock will be. If one wants to keep inter-temporal utility stable, consumption structure may be changeable and consumption volatility will be high. On the contrary, if the habit strength is small, one needs not to compensate for the consumption habit formed in the previous period. One will feel more satisfactory with the analogue consumption. Under the circumstance, consumption will be stable and smooth. Consumption consciousness awareness and industry upgrading, brought by the Chinese economic reform and social transition, make consumption change in structure, preference and habit quickly and fluctuate more than output does. The essay also rejects the hypotheses of government purchase shock and liquidity constraint. It is showed that government purchase shock will not increase consumption relative volatility to more than one until its proportion to GDP and the volatility of itself reach to a very high extent. Although liquidity constraint is not the main reason for raising consumption relative volatility, a lot of economic structure variables which may not express anything in the homogenous agent RBC mode can be learnt meaningfully by the heterogeneous agents RBC model.Chapter 6 discusses about the judgment that consumption volatility and growth are the two aspects of the same process. How can the agent weigh which one is more important? The essay uses welfare cost model of consumption volatility and growth with habit formation to compare one to the other quantitatively by means of the urban-rural microeconomic data divided by income group from 1991 to 2006. It is presented that when the agent is risk-averse, the cost of consumption volatility may be tens or hundreds times as much as that of 1% decreasing in consumption growth. Those who have different risk attitudes but nearly the same level of habit strength are probably to be in accordance with each other on accessing the importance of consumption volatility or growth. But these economic analyses are far from the needs of social development and individual goodness. What we need is social welfare analyses but not only economic analyses, which contains the relationship between consumption structure and industry structure, between consumption credit and finance innovation, between consumption growth and social security, between scientific consumption and scientific development. The essay concludes that several essentials should be attached great importance to such as to establish healthy consumption concept, to develop credit markets, to open up the rural consumption markets, to reduce to duties and fees, to increase up mobility of income, to adjust economic structure and so on.更多还原