社会互动、不确定性与我国居民消费行为研究

发布时间：2018-04-13 01:29

本文选题：社会互动 + 社会性学习　；参考：《西南财经大学》2010年博士论文

【摘要】：在传统的经济学中,价格理论是整个经济学理论的核心。价格的形成过程,通常是不同的相关行为人独立地对市场信号起反应,人们通过市场价格而间接地形成联系,相应地,仅仅对市场信号起反应的行为人成为经济学理论中的行动主体。然而在现实中,人们往往感受到文化、规范、社会结构等在价格以外,或者称之为“非市场因素”对经济活动,尤其是微观消费行为的重要影响,并由此在不同的国家或地区形成差异化的消费结构。例如金融危机后,奢侈品厂商发现奢侈品消费群在亚洲更趋向年轻化,大量中等收入阶层的年轻人成为消费主体并有着旺盛的需求,于是纷纷将亚洲,尤其是中国作为主体市场并以前所未有的速度快速扩张。除了中国高速发展的经济成为消费量持续增长的重要保障,不可忽视的,是中国普遍的“攀比”和“送礼”文化带动了奢侈品消费的爆发式增长。以此为契机,我们关注特有的文化、制度、事件和政策背景下,我国居民的微观消费行为特征,探索非市场因素对消费活动的作用规律,并借此评价目前部分经济政策的效果。论文研究的出发点主要有两个：一是以社会互动(social interaction)为媒介,观察个体的消费决策与参照群体的消费行为之间的联系,以及作用的机理。这里提到的社会互动,最初源于社会学的概念,指人与人之间在特定社会情境中相互作用,彼此影响,从经济学上讲,这种影响“并不直接在市场上通过价格变化而产生”(Baker and Murphy,2000)。参照群体因研究背景而异,可以是个体的家庭、朋友、邻居或同事等有一定往来的群体(Scheinkman,2005),但他们都必须源自基于个体感受的共同的社会距离空间(Akerlof,1997)。其意义在于,社会互动作为信息传播的有效途径之一,能够使互动群体获得信息而免于或支付较少的成本,另一方面,群体的行为会对个体起到示范作用并相互影响,例如一个新产品进入市场,随着消费的人数逐渐增多,潜在消费者会通过更多的人群互动获得信息,或受到影响而进行消费,这就是消费理论中提出的溢出效应,或者叫外部性。如果是正向的溢出,对消费会形成促进,却不需要改变商品价格或其他外部属性。在政策层面,公共政策可以通过社会互动而放大其效果,这就是所谓的“社会乘数”(social multiplier),社会乘数的存在将有助于提高公共政策的有效性(Becker and Murphy,2000)。例如,自2008年底以来,我国相继制定了家电、建材下乡、新型节能电器消费补贴等一系列针对耐用品消费,拉动内需的政策,在短期内已显示其效果,如果消费活动中存在溢出效应,补贴不但会增大当期的购买,对潜在的消费者也将形成间接的促进作用。比如,对存在社会乘数效应的消费品提供一定补贴,即使补贴终止后,正的外部性也将对带动长期消费形成一定的持续作用。我们要证实的,就是消费活动中是否确实存在溢出效应,所选择的媒介便是人与人的社会互动(当然,还可能有广告、媒体等作为其他的渠道),并探讨溢出效应产生的内在机制,以及出现社会乘数效应的商品特征。二是借助突发事件形成的自然实验,来量化人们面对不确定性事件时的反应,进而对消费决策和市场的影响。这里所涉及的不确定性和传统文献有所不同,以往关于不确定性与消费的研究以预防性储蓄或流动性约束为主,侧重未来收入不确定性与当前预防性储蓄的关系,我们所关注的,是以不可预测的自然灾害为主体的突发事件对人们消费决策的影响。这在我国目前有较为普遍的意义。本世纪以来,国内各地自然灾害频繁发生,灾害发生的时间、规模基本不可预测,并常常持续相当的时间和频率,危害当地居民人身安全,也增大了财产损失的可能。突发事件导致的不确定性增长,可能改变个人的消费决策,这种消费支出和预防性储蓄的本质一样,是为了降低不确定性冲击而加以保险。这里涉及到的不仅是未来收入降低(损失),还包括伤亡风险,是以往文献未考虑的。出于对风险的规避,部分人群会采取一定的自我保护措施,如购买保险(目前自然灾害险种极少且门槛较高),或更换更安全的房屋居住,甚至进行区域迁移。我们想要证实的,是突发事件的不确定性和常见的收入不确定性相比,二者体现出的消费者行为变化是否一致?如果不一致,又有着怎样的规律?另一方面,通过研究不同人群的自我保护策略,分析其行为决策对受灾城市经济波动、消费市场,以及人力资本流动的影响,对制定适宜的灾后重建政策,促进受灾城市的经济恢复和社会稳定,具有积极的现实意义。论文的内容分为六章：第一章主要对本文的研究背景、问题和所采用的方法进行说明,并总结论文的创新点和不足。在方法部分,侧重于对第三至五章的实证部分采用的方法和主要结论进行初步说明。第二章对国内外消费理论发展及相关研究回顾并作出评述。提出本文的理论基础,与相关理论的传承和区别,并提出本文的创新和不足。第三章基于社会互动理论,验证消费活动中存在的溢出效应,并探讨溢出效应的形成机制。本章建立了社会互动过程的理论模型,并采用由中国国家统计局农调队进行的,覆盖全国范围的城郊、乡镇和农村家庭的耐用品消费调查数据,构建线性概率模型和风险函数(Harzard function)分解,并运用代理变量和工具变量的两阶段最小二乘法回归,得出较为稳健的结果。第四章以自我保护理论为基础,研究自然灾害的不确定性对居民消费行为的影响以及人力资本流动的冲击。利用汶川地震后,对主要受灾的大中城市房地产市场进行调查得到的微观数据,对自我保护理论和预防储蓄理论进行检验并得出结论。第五章,在第四部分研究的基础上进一步拓展,构建离散选择连续内生解释变量模型(Discret Choice Model with Continous Edougenous Explanatory Variables),侧重讨论自我保护和自我保险理论的区别及不同的决定因素,并研究在支付成本可预期与不可完全预期的情况下消费者行为的差异。第六章,对前几章的相关理论和实证结果进行总结,提出研究不可避免的局限性及后续研究的拓展,以及相关的政策建议。本文的主要结论如下： 1、以中国农村耐用品消费为例,研究发现,农村家庭的彩电、洗衣机等耐用品消费会显著受到村庄以内其他村民消费行为的影响,1997-1999年,样本中彩电消费50.5%的购买率源于他人消费的带动,也就是内生的社会互动的作用,尤其是当新产品推出或家庭的首次购买时效果最大。证实了消费行为中存在溢出效应,而形成的机制,源于消费者在社会互动中的社会性学习与信息共享。90年代末,中国农村家庭购买耐用品的目的侧重商品使用性能,收入的差别对溢出的作用并不显著,受到更好教育的户主进行社会性学习的能力越强,消费的溢出效应越大,并排除了消费的攀比或羊群效应。我们相信,溢出效应在消费活动中是普遍的,但溢出的机制会有一定差异。比如奢侈品消费会呈现更强的攀比或财富显示的特征,选择的样本数据和时间跨度不同,会反映出不同的规律。 2、预防性储蓄理论最主要的结论在于：不确定性同财富积累之间有正相关关系,不确定性越高,财富的预防性积累越多(Fisher,1956:Skinner, 1988；Guiso,Terlizzese,1992)；个人对当前收入或财富更加敏感,同时对预期的将来收入有较低反应(Zeldes,1989).我们通过汶川地震的自然实验,发现消费者在灾害之后会实施自我保护的措施,主要通过增大预防性消费来实现,例如,地震后偏好较低的楼层或者建筑质量更好的房屋,甚至进行区域迁移。由于同时存在伤亡风险和财产风险,我们得到与以往研究不同的结论：不确定性与预防性消费不完全是正相关关系,会由收入差异而不同。低收入者会根据不确定性的大小选择预防性消费,二者是正相关的；然而,高收入者的预防性消费不受风险大小的影响,过去发生的突发事件成为显示信号,无论未来风险如何,高收入者会进行更多的预防性消费。地震后,消费者更青睐较低的居住楼层;高收入群体有更强的风险厌恶程度,将在震后选择质量更好的房屋；区域性的购房优惠补贴政策对大中城市商品房市场将起到一定的稳定作用。 3.采用汶川震区周边大中城市的问卷调查数据,利用地震前后人们住房选择变化来检验消费者对不确定事件及其信息的反应,从自我保护和自我保险理论出发,论证二者不同的决定因素,并研究在预防性消费支出可预期与不可完全预期的情况下消费者行为的差异。从实证的角度验证了自我保险行为受收入分布的影响,自我保护行为更多受个人风险偏好的影响。同时表明,对受灾区域大中城市的房地产市场而言,地震以后,收入越高的群体将更注重房屋建筑质量,高质量房屋的相对需求会有一定增长。在长期,女性、有小孩的家庭以及受损严重市县的消费者将更注重房屋的抗震强度,而震后及时实行的购房补贴优惠政策对商品房市场的稳定起到了一定的促进作用。从长期来看,地震对商品房市场的整体影响并不太大,尤其是地震导致的心理冲击对消费者并不构成显著影响,针对商品房市场的相关政策,可以短期为主,旨在减少房地产价格的波动；在长期,尤其是以居住需求为主的家庭消费更注重房屋质量,复核和调整工程建设的抗震设防标准、注重建筑物的抗震性,提高城市住房的安全性和保障性,在灾后重建中更显得尤为重要。本文的主要创新点包括：第一,论文选题的新颖性。目前国内外关于非市场因素对消费行为的实证研究较少,本文从社会互动、不确定性等不同的方面对消费行为理论进行丰富与拓展,证实社会互动、信息共享与知识溢出对消费有正的外部性并促进消费的可持续增长。同时,采用“自然实验”的实证思想刻画不确定性对住房消费支出的影响。在现今国内外的有关房地产的研究当中,采用自然实验的方法进行深入的研究相对较少,尤其是在国内的现有文献来看更是如此。采用“自然实验”的方法在很大程度上由于它的完全随机性可以解决很多经济问题的内生性,因此在理论的证实或证伪以及对现实问题的研究过程中占有重要地位。第二,学术观点的创新性。论文提出了以下较有创新性的结论：(1)消费的外部性产生的途径之一源于于人与人之间思想、信息与创意的交流与共享形成的知识外溢效应,从另一个角度验证了知识溢出已成为经济增长与社会和谐发展的核心要素。(2)内生的社会互动的作用,尤其是当新产品推出或家庭的首次购买时效果最大。受到更好教育的户主进行社会性学习的能力越强,消费的溢出效应越大。(3)不确定性与预防性消费不完全是正相关关系,会由收入差异而不同。低收入者会根据不确定性的大小选择预防性消费,二者是正相关的；然而,高收入者的预防性消费不受风险大小的影响,过去发生的突发事件成为显示信号,无论未来风险如何,高收入者会进行更多的预防性消费。第三、微观数据的独特性。第三部分的研究中,我们使用覆盖全国各省的城郊、乡镇和农村居民样本进行实证研究,结果具有一定的代表性和普遍性；第四、第五部分,笔者设计了受灾城市居民微观消费行为及个人特征的调查问卷,在实地调研、广泛论证的基础上,累计发放调查问卷一千五百多份,收回有效问卷一千余份,获得了及时的第一手研究数据。这些研究数据,有力地支持了本文的研究结论,具有可再利用性和共享性,对后续研究具有积极的参考价值。本文还存在着许多不足之处,希望得到各位专家的批评和指正。 (一)虽然本文对居民的消费需求的影响进行了系统的实证分析,但由于影响因素的复杂性、多样性以及资料不足,尚有部分结果不能更加严谨的论证,比如耐用品消费中品牌效应的论证不能区分溢出效应的类型。 (二)数据的特征变量,尤其是第二问卷的家庭特征不够充足,难以获得更稳健的检验。由于是横截面数据,目前无法在时间跨度上对预防性消费的跨期分配进行检验,如果有地震较长时间后再一次调查的数据,将会得到一些更有趣的结论,例如收入和风险主观判断对市场影响时间的长短,高收入人群与低收入人群的预防性消费是否会随着时间而变化等等,能够更好地完善理论和实证假说。
[Abstract]:In the traditional economics , the price theory is the core of the whole economic theory . The formation process of price , usually the different relevant actors react independently to the market signal , and the people who respond indirectly to the market signal form the main body of action in the economics theory .

The starting point of the research is two : one is the social interaction as the medium , the relationship between the consumption decision of the individual and the consumption behavior of the reference group is observed , and the mechanism of the action is observed . The social interaction mentioned here is originally derived from the concept of sociology , the interaction between the person and the person in a specific social situation and the influence on each other . From the economics , the influence is not directly generated in the market through price change ( Baker and Murphy , 2000 ) . Reference groups may vary depending on the background of the study ( Scheinkman , 2005 ) , such as families , friends , neighbors or colleagues of an individual , but they all have to come from a common social distance space based on individual feelings ( Akerlof , 1997 ) . The significance lies in that social interaction is one of the effective ways of information transmission , which can make the interactive group obtain information without or pay less cost . On the other hand , the behavior of the group can play a role in the individual and interact with each other , such as the overflow effect proposed in the consumption theory , or the external property . For example , since the end of 2008 , our country has formulated a series of policies for consumer goods consumption and pulling domestic demand , such as household appliances , building materials , new energy - saving electric appliance consumption subsidies , etc .

Second , by means of the natural experiment formed by the emergency , we quantify the response of people in the face of uncertain events , and then influence the consumption decision - making and the market .

The contents of the thesis are divided into six chapters :

In the first chapter , the background , problems and methods used in this paper are explained , and the innovation points and shortcomings of the thesis are summarized . In the method part , the methods and main conclusions adopted in the empirical part of the third to fifth chapters are mainly described .

The second chapter reviews and comments on the development of consumption theory at home and abroad and the related research . The paper puts forward the theoretical basis , the inheritance and the difference of the relevant theories , and puts forward the innovation and deficiency of this paper .

In chapter 3 , based on the theory of social interaction , the spillover effect exists in the consumption activity , and the formation mechanism of spillover effect is discussed . In this chapter , the theoretical model of social interaction process is established , and the survey data of the consumption of durable goods in the suburbs , towns and rural families covering the whole country , which is carried out by the National Bureau of Statistics of China , is used to construct the linear probability model and the risk function decomposition , and the two - stage least square regression of the agent variable and the tool variable is used to obtain a more robust result .

In chapter 4 , based on the theory of self - protection , the influence of the uncertainty of natural disasters on the behavior of residents and the impact of human capital flow are studied . After the Wenchuan earthquake , the micro data obtained from the investigation of the real estate market of the major affected cities is analyzed , and the self - protection theory and the prevention deposit theory are tested and concluded .

In chapter 5 , on the basis of the research of the fourth part , we further expand the discrete choice continuous endogenetic interpretation variable model ( Discret Choice Model with Theory ous Edoused Variable Variables ) , focus on the differences between self - protection and self - insurance theory and different determinants , and study the difference of consumer behavior when the payment cost is expected and cannot be fully expected .

Chapter 6 summarizes the relevant theories and empirical results of the previous chapters , and puts forward the inevitable limitation of the study and the expansion of the follow - up study , as well as the relevant policy recommendations .

The main conclusions are as follows :

1 . As an example , the study shows that the consumption of durable goods such as color TV and washing machine in rural households is significantly affected by other villagers ' consumption behavior within the village . In 1997 - 1999 , the purchasing rate of 50.5 % of color TV consumption in rural households is due to the consumption of others , which is the effect of the consumer ' s social interaction . It is believed that the spillover effect is common in the consumption activity , but the spillover effect is different . For example , the consumption of luxury goods will show stronger climbing or wealth display characteristics , and different patterns are reflected in the selected sample data and time span .

2 . The most important conclusion of the theory of preventive savings lies in the positive correlation between uncertainty and wealth accumulation , the higher the uncertainty , the more preventive accumulation of wealth ( Fisher , 1956 : Skinner , 1988 ;
Guiso , Terlizzese , 1992 ) ;
Individuals are more sensitive to current income or wealth while at the same time have a lower response to the expected future income ( Zeldes , 1989 ) . Through the natural experiment of Wenchuan earthquake , we find that consumers will implement self - protection measures after the disaster , mainly by increasing the preventive consumption . For example , after the earthquake , we get better houses and even carry out regional migration . As a result of the risk of casualty and property , we get different conclusions from previous studies : uncertainty is positively related to preventive consumption , which is different from the income difference .
However , the preventive consumption of the high - income people is not influenced by the risk size . In the past , the emergency becomes the display signal , and no matter how the future risks , the high - income people will carry out more preventive consumption . After the earthquake , consumers prefer lower residential floors ; higher income groups have stronger risk aversion levels , and will choose better quality houses after the earthquake ;
The regional purchasing preferential subsidy policy will play a certain stabilizing role in the commercial housing market in China .

3 . Based on the survey data of the large and medium - sized cities around Wenchuan earthquake area , the change of housing selection before and after the earthquake is used to verify the consumers ' response to uncertain events and their information . In the long term , the higher the income will pay more attention to the housing ' s seismic intensity . In the long term , the higher the income of housing subsidy will pay more attention to the seismic intensity of the housing market . In the long term , the impact of the earthquake on the housing market is not too big , especially the impact on the housing market , which can be short - term , and aims at reducing the fluctuation of real estate price .
In the long run , especially the household consumption based on the living needs pays more attention to the quality of the house , check and adjust the seismic fortification standards of the project construction , pay attention to the earthquake resistance of the building , improve the safety and security of the urban housing , and become more important in the post - disaster reconstruction .

The main innovations of this article include :

At the same time , in the domestic and foreign research of real estate , this paper studies the influence of uncertainty on the consumption of housing consumption . At the same time , the method of " natural experiment " is used to study the influence of uncertainty on the consumption of housing . At the same time , the method of " natural experiment " is used to solve many economic problems .

Second , the innovation of academic viewpoint . The paper puts forward the following innovative conclusion : ( 1 ) One of the ways of the external creation of consumption derives from the knowledge spillover effect formed by the exchange and sharing of thought , information and creativity between man and man .
However , the preventive consumption of high - income people is not influenced by the risk size , and in the past , unexpected emergencies have become the display signals , and the high - income people will carry out more preventive consumption regardless of the future risks .

Thirdly , in the study of the third part , we study the sample of villages , towns and rural residents covering all provinces of the country . The results are representative and universal ;
The fourth and fifth part , the author designed the questionnaire of the micro - consumption behavior and personal characteristics of the residents in the affected city . On the basis of field investigation and extensive demonstration , more than 1,500 questionnaires were distributed , more than 1,000 questionnaires were collected , and the first - hand study data were obtained .

There are many shortcomings in this paper , and it is hoped that all experts will be criticized and pointed out .

( 1 ) Although this paper makes a systematic empirical analysis on the effect of the consumption demand of the residents , some results cannot be more rigorous proof because of the complexity , diversity and insufficient information of the influencing factors , such as the demonstration of the brand effect in the consumption of durable goods cannot distinguish the types of spillover effects .

( 2 ) The characteristic variables of the data , especially the second questionnaire , are not enough to obtain more robust test . Because of the cross - sectional data , it is impossible to test the cross - period distribution of the preventive consumption on the time span . If there is more data after the earthquake for a long time , it will get some more interesting conclusions , such as the subjective judgment of income and risk , the length of the influence time of the market , whether the preventive consumption of the high - income population and the low - income population can change over time , etc . , and can better perfect the theory and the demonstration hypothesis .

【学位授予单位】：西南财经大学
【学位级别】：博士
【学位授予年份】：2010
【分类号】：F126

【参考文献】