企业的企业家——契约理论(txt+pdf+epub+mobi电子书下载)

作者：张维迎

出版社：上海人民出版社

格式: AZW3, DOCX, EPUB, MOBI, PDF, TXT

企业的企业家——契约理论试读：

我首先要感谢我的导师Donald Hay和Jim Mirrlees教授的指导。一个隐含契约是，Donald负责对论文的总体指导，Jim负责模型的技术性问题。在过去三年里的学习期间，我基本上是每两周分别见他们一面，讨论与论文有关的各种问题，从基本思想到分析方法无所不包。我确信，我是牛津大学最奢侈最幸运的学生，能分享导师如此多时间的学生的确是不多的。Donald和Jim的鼓励和学术指导对我完成这篇论文是非常关键的。好几次，当我被一些技术性问题或理论问题搞得感到无路可走时，是他们帮我找到了解决问题的办法。Donald为我所做的远远多于一个导师假定要做的，我对他的感激之情是无法用语言表达的。

我还要感谢Meg Meyer博士。在1992年秋季Donald休假期间，她承担了对我论文的指导工作。特别是，她对我M. Phil论文的建设性评论，对我把M. Phil论文发展成目前的这篇博士论文帮助极大。Meg对我的帮助并没有随她的官方责任的中止而中止。她总是我的一位编外导师。

在我在牛津大学做访问学生和读学位的前后五年时间里，我从牛津大学的许多经济学家的授课中获益匪浅。他们当中，我要特别感谢Christopher Bliss，John Vickers和E. Eshag。我还要特别感谢Cyril Lin博士和A. K.Cairncross爵士，若没有他们的帮助，我是不大可能到牛津读书的。

我要感谢我在牛津的中国朋友的帮助。他们当中，我要特别提到金立佐、刘楚俊、宋丽娜、秦朵、谢多、王大鸿、程原、魏刚、秦才功、刘芍佳。

我要感谢世界银行提供的奖学金，感谢伦敦经济学院和Robbins家族提供的Lionel Robbins纪念奖学金，感谢英国政府的海外学生奖学金，感谢Nuffield学院提供我第三学年全额奖学金，尽管我实际上并没有使用这笔钱。

最后，我感谢我的妻子马晋红陪伴我在牛津读书，感谢她的理解和支持。在我学习的最后阶段，我们的儿子亚声出生，使我在繁忙的论文写作中得到无穷的欢乐。我将这篇博士论文献给他，表达我对他的爱。我要感谢我岳母，她来牛津照顾我的妻子和儿子，把我从家务中解放出来，使我的论文得以最后完成。我要感谢我父母的养育之恩和无私的爱。张维迎1994年4月于牛津大学ABSTRACT

A BRIEF DESCRIPTION OF THE THESIS

The firm is the typical organizational form of the market economy. The most significant characteristics of the firm are the asymmetric contractual arrangements between different participants（factor-owners）in both distribution of returns and control rights. Within the firm, some participants are called“employers”, while others are called“employees”. Employers hold“authority”over employees and are entitled to claim the residual returns, while employees are obliged to obey the authority of employers within certain limits and are entitled to fixed wages. In the terminology of principal-agent theory, employers are principals and employees are agents. This“micro”asymmetry between employers and employees directly determines a“macro”asymmetry. In society, employers belong to an upper-class, while employees belong to a lower-class. For this reason, this topic about the firm attracts attention not just from economists but also from sociologists, political scientists, politicians and, in particular, social reformers.

The employment relationship takes place between capital and labour. An important question which has puzzled economists as well as others for long time is: why does capital hire labour rather than labour hire capital? This question is specially relevant today for two reasons. First, almost all socialist countries have experienced the failure of the socialist planned economy and have now begun a market-oriented reform program. Although Yugoslavia’s experiment has shown that a labour-managed economy cannot be an efficient option, there is no guarantee that other socialist countries will not be attracted by the labour-hiring-capital system when they begin to deviate from the traditional planned economy. In particular, for ideological reasons, the labour-hiring-capital economy may be thought to be the only“acceptable”choice for some socialist countries. Secondly, in the joint-stock company,“ownership”is separated from management and the traditional conception of the employer is no longer as relevant as in the owner-managed firm. Instead, shareholders hire the management who in turn hire workers. That is, the traditional single agency relationship between a capitalist-entrepreneur and the workers has been replaced by an agency-chain between capitalists and management, and management and workers. Many economists have focused their attentions on how capitalists as the principal make an optimal incentive scheme to induce the management（agents）to act in their best interests, or how the managerial behaviour deviates from shareholders’ interests; but the most fundamental question is why the principalship should be assigned to capitalists rather than management in the first place. The logic behind this question is, if the firm’s output does not directly depend on the actions taken by capitalists, why could not the incentive problem associated with the separation of ownership and management be solved by assigning the principalship to the management and let the management work for themselves? Or more generally, why do we need capitalists?

This thesis is intended to explore the elements determining the assignment of principalship within the firm: Why does capital hire labour rather than labour hire capital? Why does the entrepreneur monitor workers rather than workers monitor the entrepreneur? Why do capitalists rather than workers select the management of the firm? What factors determine who will be the entrepreneur in equilibrium? We are concerned with an economy in which all economic actions fall into two types: marketing and producing. By“marketing”we mean the activities of“discovering the relevant prices”[Coase（1937）, p. 390] including speculating about profitable opportunities, forecasting market demands and making“judgemental decisions”[Casson（1982）] of“what to do, and how to do it”[Knight（1921）], in Schumpeter's words, setting up a production function. By“producing”we mean all the activities of transforming inputs into outputs“physically”under the given production function（technology）and according to marketing decisions.

Individuals in the economy are assumed to differ in（1）their marketing ability（entrepreneurial ability）, denoted by θ;（2）personal assets , denoted by WO; and（3）risk-attitudes, denoted by R. Because individuals differ in their marketing ability, it may be profitable for them to cooperate by setting up a“firm”through which individuals who have advantages in marketing specialize in making marketing decisions, while those who are not good at marketing specialize in producing（note that we assume that individuals are identical in their producing ability）. Because of“uncertainty”[Knight（1921）] and“team production”[Alchian and Demsetz（1972）], the firm involves an agency problem-some member may take actions（e. g. , shirking）which benefit himself but cost others. The key organizational issue is to design a contractual arrangement between different participants of the firm so as to make each member as responsible for his own actions as possible. We will argue that the member who does marketing should be assigned to be the principal to claim the residual return and to monitor others, not just because he is the major“risk-maker”but mainly because his actions are the most difficult to monitor. Thus he becomes the entrepreneur while those who do producing become the workers.

Under the assumption that personal assets WO are costlessly observable for all individuals while marketing ability θ is private information（or observable only at some cost）, we will demonstrate that capitalists with high marketing ability will be the winners of the competition for being the entrepreneurs because their costlessly observable capital stocks can work as a device to signal information about marketing ability of the would-be entrepreneur, and the arrangement therefore saves transaction costs. In other words, when information of ability is asymmetric between the insider and outsiders, only those would-be entrepreneurs who possess enough personal assets can be trusted as qualified entrepreneurs. Capitalists are more likely to be honest, credible, responsible and industrious when they choose to be entrepreneurs. They have less incentive to overstate their entrepreneurial ability, or to overinvest. A capitalist can earn“pure”profit, because his capital economizes on transaction costs by signaling information. In short, we show that capital-hiring-labour is a mechanism which guarantees that only qualified people will be chosen to be entrepreneurs（/managers）; in contrast, if labour hires capital, the market for entrepreneurs（/managers）would be full of lemons（i. e. , too many unqualified people would choose to do marketing）.

Finally, we set up a general equilibrium entrepreneurial model of the firm, in which marketing ability θ, personal assets WO and risk-attitudes R are identified as the three key factors determining the choices of being an entrepreneur or a worker or a manager or a pure capitalist. We will show that there is a general equilibrium in which all individuals have their（constrained）optimal choices and different forms of the firm are chosen so that both the labour market and the capital market stay in equilibrium（goods market equilibrium can be understood as a by-product of labour market equilibrium and capital market equilibrium）. In particular, we will show that the equilibrium relationships（both in pecuniary and non-pecuniary terms）between the firms’members depends on the joint distribution of marketing ability, personal wealth and risk attitudes in the population. Given asymmetry of distribution of entrepreneurial ability and distribution of personal wealth, the joint-stock company as a cooperation between capital and ability occurs if the costs of searching for high ability people are not prohibitively high. And it is also socially optimal to allow capitalists to select the management because the more personal assets a person holds, the more incentives he has in searching for high ability people. We argue that the major function of shareholders is to select a high ability manager rather than to monitor an incumbent manager.

The basic ideas of this thesis can be described by using a two-way classification of individuals（we omit the risk-attitude dimension）.

In Table 1.1, for simplicity, we assume that both marketing ability and personal wealth have two-point distributions: for any individual, θ takes one of two values,θH（high）or θL（low）; similarly, WO takes one of two values, WOH（rich）or WOL（poor）. Thus there exist four types of individuals, denoted by E, C, M and Z respectively. Type E individiuals are rich in both marketing ability and personal wealth:（θH, WOH）; Type C are rich in personal wealth but short of marketing ability:（θL, WOH）; Type M have high marketing ability but low personal wealth:（θH, WOL）; Type Z are poor both in marketing ability and personal assets:（θL, WOL）. If all transactions between individuals take place through spot-markets, each individual has to work as an individual businessman in deal ing with both marketing and producing. Taking Type Z as the yard-stick, C has an advantage in capital factor, M’s advantage is in markeing ability, and E has advantages in both capital and marketing ability. Obviously it may be profitable for different types of individuals to cooperate by forming a“firm”in which some individuals are specialized in marketing activities while others are specialized in producing activities. Two problems associated with the firm are: first, how to allocate the marketing function and producing function to different members; second, how to resolve the agency problem by assigning principalship. What we are going to demonstrate is that, as the firm substiutes for the spot-market,（i）Type E become entrepreneurs by doing marketing, monitoring producing-members, and claiming the residual;（ii）Type M become managers by doing marketing, monitoring producing-members but being monitored by Type C, and sharing some risk with Type C;（iii）Type C become capitalists by selecting and monitoring managers, and bearing risks; Type C and Type M together become joint-entrepreneurs; and（iv）Type Z become workers by specializing in producing, and receiving a fixed return. We also show that bargaining power of each type depends on the joint distribution of θ and WO in population. For instance, an increase in the proportion of Type C but will disadvantage Type C but advantage Type M（managers）and probably Type Z（workers）.

The thesis is organized as follows. In Chapter one, we give a brief description of the main ideas of the thesis and present a critical review of the existing theories of the firm（including the contractual theory, the entrepreneurial theory, and the managerial theory）. Chapters 2-4 are the body of the thesis addressing how principalship（residual claim and authority）is assigned among the different members of the firm. Our demonstration consists of three major steps. In Chapter 2（the first step）, we demonstrate why principalship should be assigned to the marketing member rather than to the producing member to maximize total welfare（equivalently, to minimize agency costs）. In so doing, we argue that differences of marketing ability between individuals are the original rationale for the occurrence of the firm; we identify marketing with Coase’s“discovering the relevant prices”but focus on aspects ignored by Coase. We make a distinction between the self-monitored incentive and the being monitored incentive. We argue that there is a trade-off between the self-monitored incentive and the being-monitored incentive associated with assignment of principalship, and it is optimal for the marketing member to be the principal because such a contractual arrangement can guarantee that total welfare are maximized. In Chapter 3（the second step）, a hidden information model is used to show why priority in being entrepreneurs is given to capitalists. In so doing, we focus on how the capital endowment of a would-be entrepreneur can function as a signal of his marketing ability. Specifically we show that the individual critical ability for being an entrepreneur is increasing with his personal wealth, unless the individual’s personal wealth exceeds a certain level. Under the assumption that marketing ability is not observable（or not costlessly observable）, it is shown that priority in being an entrepreneur（marketing member）and / or the right of selecting the person to undertake marketing should be given to capitalists because such a contractual arrangement can ensure that only qualified candidates win the competition for being entrepreneurs（or managers）. This conclusion implies that imperfect capital markets may be socially optimal. In Chapter 4 （the third step）, based on the arguments given in Chapters 2 and 3, a general equilibrium entrepreneurial model of the firm is set up; the main properties of the equilibrium will be derived; and the partition of the population into entrepreneurs, manager, pure capitalists and workers will be identified. We show that in equilibrium,（a）individuals with high ability, high personal wealth and low risk-aversion become entrepreneurs,（b）individuals with low ability, low personal wealth and high risk aversion become workers,（c）individuals with high ability but low personal wealth become managers hired by capitalists, and（d）individuals with low ability but high personal wealth become“pure”capitalists to hire managers. Chapter 5 concludes the thesis and directs attention to some promising aspects for our future research.

One of the implications of our hypothesis is that because advantages of capital over labour are associated with information-cost saving, we may predict that these advantages will be diminishing as other signals become available. Education is one such signal, which may reveal some information on marketing ability and therefore help some MBA-holders to become managers. In the extreme case, if information is perfect, capital would become a pure production factor and would lose all its advantages over labour. In fact, in this case nobody has any advantages over others in marketing, and thereafter the firm becomes redundant in Coase’s sense. However, if we believe that marketing is some kind of innate ability which is not entirely educable, capital will still enjoy advantages over labour in signaling information about a person's marketing ability.第二版序言

本书的书名《企业的企业家—契约理论》读起来有一些拗口，但它准确地表达了本书的主题。

新古典经济学中讲的企业理论，实际上是生产决策理论，即一个生产单位（“厂商”）如何选择投入要素的组合和产出，以最大化利润。它强调的是要素价格、生产技术和需求函数如何决定最优的生产选择。这一理论对理解市场均衡和资源配置的效率是有意义的，但掩盖了企业内部复杂的组织关系和激励问题，所以又被称为“黑箱理论”（black box theory）。最早对新古典企业理论发起批判的是罗纳德·科斯，他在1937年发表的《企业的性质》一文中认为，新古典企业理论甚至没有办法解释企业为什么存在，因为它假定交易成本是零，而在没有交易成本的情况下，所有的生产都可以通过个人之间的市场交易完成，根本无须借助于企业这样一种以“权威”为特征的组织形式。由科斯开创、威廉姆森等人发展的现代企业理论被称为“契约理论”，这一理论把交易成本放在中心地位，认为企业是市场交易的替代方式。

企业的契约理论解释了企业的存在以及企业内部所有权和激励的重要性，加深了人们对企业作为一种制度的理解。契约理论的最大缺陷是没有对企业家在企业形成中的作用予以足够的重视，契约理论中的企业仍然是没有企业家的企业。由于这个原因，契约理论虽然能够解释企业为什么存在以及企业所有权为什么重要这样的问题，但不能解释为什么企业决策者应该拥有剩余索取权以及应该由谁来选择决策者的问题。简单地说，契约理论关注的是激励问题，而不是经营者的选择问题。但在我看来，经营者的选择比激励机制更为重要。一个没有企业家的企业不是真正的企业。

在本书中，我把企业家放在中心地位。我试图建立一个以企业家为中心、契约关系为本质的企业理论，故称之为“企业的企业家—契约理论”。我认为，在市场经济中，制定经营决策（生产什么、如何生产）是最重要的功能，所谓的“企业家”就是最善于制定决策的人，而人口中具有高超的决策能力的企业家并不多，企业的价值就是让企业家制定决策，而为了激励企业家制定正确的决策，企业家必须承担决策的风险，也就是拿剩余收入而不是合同收入。进一步，由于有关企业家能力的信息是不对称的，如何保证企业决策者是真正具有企业家素质的人就成为企业制度要解决的一个重要问题。我们所观察到的资本主义企业制度就是自由市场为解决这个问题而做出的制度安排，它不是法律的选择，而是竞争演化的结果。简单地说，“资本雇佣劳动”是一种能够保证只有合格的人才会被选做企业家（/经营者）的机制。

从1994年开始，我将自己的企业理论应用于国有企业的分析，撰写了一系列的文章，这些论文收集于我的《企业理论与中国企业改革》一书中（北京大学出版社，1999年，2006年）。我的一个基本结论是，国有企业虽然可以通过利润分成和奖金制度对经营者提供短期激励，但没有办法解决经营者的长期激励问题，更没有办法解决经营者的选择问题。原因在于，代表国家行使所有权的政府官员并不是真正的资本所有者，他们虽然有选择经营者的权力，但并不对选择的后果承担财务风险，他们重视的只是控制权收益，而不是货币形态的收益，因而不可能有真正的积极性选择最具企业家素质的人担任经营者。由于这个原因，国有企业与其说是经济组织，不如说是政治组织，严重的权力斗争是国有企业的重要特点。解决国有企业的经营者选择问题和长期激励问题的唯一出路是对国有企业实行民营化，创造出真正的所有者。

企业的企业家—契约理论也是我分析公司治理结构的基本理论框架。近十几年来，无论在理论界还是实务经济部门，公司治理结构都是一个热门话题，但在我看来，流行的公司治理理论可以被称为“经理人中心模型”（manager-centered model），这一模型表现出对市场机制的不信任和对企业家精神的忽视。如果我们想对市场和公司治理有一个正确的理解并提供正确的政策建议，我们必须从“经理人中心模型”转向“企业家中心模型”（entrepreneur-centered model）。在《产权、激励与公司治理》一书（经济科学出版社，2007年）中，我试图提供的就是一个“企业家中心的公司治理结构理论”。

我有关企业理论的进一步发展包括在《市场的逻辑》一书（上海人民出版社，2010年, 2012年）的第一部分。与科斯的企业理论不同，我认为，企业不是市场的替代物，而是市场的运行方式。简单地说，企业是一个连带责任的组织。市场的有效运行依赖于交易者之间的信任，正是通过所有者（/企业家）对企业团队成员的连带责任和品牌企业对供应商的连带责任，消费者才有可能监督生产者、信任生产者，市场的交易成本才能降得足够低，市场才能有效、有序运行。如果没有企业这样的组织，根本不可能有大规模的市场交易发生，更不可能有全球化市场的出现。从这个意义上讲，科斯错了。进一步，对消费者来说，大企业类似一个“总承包商”，这个“总承包商”为消费者承担着监督众多中小企业的责任，节约了消费者的监督成本，如果没有大企业，大量复杂的产品根本不可能得到消费者的信赖。从这个意义上讲，两百多年前亚当·斯密对大企业的批评也错了。由此，我们也有必要重新思考反垄断法的经济学基础。反垄断法的经济学假定企业只是一个生产单位，而没有把企业当作声誉的载体和创新组织。传统经济学所谓的“完全竞争”实际上是没有竞争，“完全竞争”的市场根本不可能是一个有效、有序的市场。

本书初版于1995年，自出版后，就成为中国经济学界和管理学界引用最多的著作之一，这对作者是最大的慰藉。本次再版时，正文部分没有改动，但我对附录做了调整，将汪丁丁、张春霖和张曙光三位经济学家对本书写的长篇评论以及我对他们评论的《答复》收入附录（这四篇原来收入《企业理论与中国企业改革》一书的附录中），同时删去了原附录中的四篇文章。这三篇书评是邓正来先生组织的，曾发表于他主编的《中国书评》1995年5月号（总第10期）。邓正来已于2013年元月离开了我们，借此机会，我也想表达我对他深深的谢意和思念之情。他是一位杰出的学者，也乐于助人，他对我在学术上的帮助让我终生难忘。

最后，我要感谢本书第一版责任编辑何元龙先生和本版策划编辑姚映然女士，何元龙先生为本书第一版的成功做出了重要贡献，姚映然女士对这次再版从内容调整到版面设计都尽心尽力。出版我的系列作品并把本书作为其中之一本是施宏俊先生的动议，在此一并感谢。张维迎2014年1月24日前　言

本书是我在牛津大学的博士论文的中文译本。这篇论文要探讨的问题是什么因素决定市场经济中企业委托权（所有权）的安排：为什么资本雇佣劳动而不是劳动雇佣资本？为什么企业家监督工人而不是工人监督企业家？为什么资本所有者选择经营者而不是工人选择经营者？什么因素决定在均衡中什么人将成为企业家？我试图找出能解释上述问题的最基本因素。尽管论文本身是非常理论性的，但我选择研究这个题目恰有很强的现实背景。20世纪80年代中后期，我在国家体改委中国经济体制改革研究所从事经济理论和改革政策研究工作。当时经济改革过程中出现的许多问题使得企业家问题成为经济理论界的热门话题。经济学家几乎一致认为，造就企业家队伍对保证改革成功和体制的有效运行具有关键的作用，但在有关如何造就企业家，特别是企业家的形成与所有制的关系上，经济学家之间很有分歧。当时的主流观点是，企业家是重要的，但所有制是不重要的；造就企业家队伍的关键是公平的竞争环境和充分的经营自主权，而不是所有制。有些经济学家甚至引证市场经济中“所有权与经营权分离”的“事实”特别是日本的例子说明，正是由于所有者不起作用了，企业家才真正有了用武之地。我当时的观点与这种主流观点相反，我认为，企业家是特定的财产关系的产物；没有真正的财产所有者，就不可能有真正的企业家；因此，造就企业家队伍的关键是所有制改革（见张维迎《企业家与所有制》一文，中国经济体制改革研究所《经济体制改革研究报告》1986年第30期；张维迎：《造就真正的企业家》，《人民日报》1986年9月19日）。但是，我当时还缺乏一种理论对那些流行的浮浅观点予以有力反驳。1987年10月，我到牛津大学进修，开始接触近二三十年来发展起来的现代企业理论文献。但是，尽管我从这些文献得到很大启发，现存的文献并没有提供我一个现成的武器，比如说，科斯等人研究了为什么存在企业，但没有回答为什么是资本家而不是工人成为企业的老板。70年代中后期发展起来的委托—代理理论将股东作为委托人，经理作为代理人，研究委托人如何设计最优激励合同诱使代理人努力工作；但是，在这种理论中，委托—代理关系本身是给定的，而在我来看，更为基本的问题是，究竟谁应该是委托人谁应该是代理人？为什么资本所有者成为委托人？特别是，既然企业的收益不直接依赖资本所有者的行动（如理论所假定的），为什么经理人员的积极性问题不能通过把他们直接变成“委托人”来解决？1990年9月我回到牛津读博士学位，将博士论文选题定为“为什么资本雇佣劳动”这样一个题目。我认为，回答这个问题的关键是剖析古典资本主义企业中经营者—企业家—资本家三位一体的现象：为什么从事经营决策的人索取剩余成为企业家？为什么资本家拥有成为企业家的优先权？只有解释了这些问题，才能真正理解现代企业制度。得益于信息经济学的发展，我的研究工作进行得非常顺利。到1991年底，论文的基本思想和模型化工作已经完成，论文的初稿作为硕士论文获得l992年牛津大学经济学研究生最佳论文奖（the George Webb Medley Prize for the best thesis），使得我有理由相信，我的研究为上述问题提供了一个一家之言的答案。

1994年8月，我毕业回国，开始向国内同行介绍自己的理论观点并运用这些观点分析中国国有企业的改革。余晖先生和刘世锦博士将我论文的第1章的一部分翻译在《经济研究》1994年第11期上发表；我应樊纲博士之邀在中国社科院经济研究所双周学术研讨会上作了一次演讲；在为北大和社科院研究生开的《产业组织理论》课上，我将自己的企业理论作了比较系统完整的介绍；此外，我回国后写的两篇有关中国国有企业改革的论文先后在《经济学消息报》、《改革》杂志和《中华工商时报》上发表。所有这些使我的理论在读者中引起了兴趣，不少读者写信或打电话告诉我说，他们觉得我的理论很新颖，希望能读到论文的全文。我的一些学生则建议我将博士论文译成中文出版。这可以说是出版这本书的来由。

准确地讲，出版这本书有两个目的，一是传播理论思想，二是介绍研究方法。就后一目的而言，我心目中的主要读者对象是正在读经济学的博士研究生、硕士研究生及高年级大学生。自到北大教书以来，我与经济学研究生有着频繁的交往，我发现，国内研究生对国外研究生如何做博士论文并不了解。作为一名教师，我想我有责任向学生介绍一些这方面的情况，为中国经济学与世界经济学接轨做点贡献。大致来说，国外经济学博士论文可以分为两类，一类是经验研究（empirical studies），另一类是理论研究（theoretical studies）。经验研究主要是使用统计数据和计量模型检验一种假说（hypothesis），这种假说可能是自己的，也可能是他人的；理论研究主要是用严格的数学方法证明一种假说，这种假说应该是自己的（一个例外是用一种新的方法证明别人已经证明过的东西）。经验研究因为涉及大量收集和处理数据的工作，耗费的时间要长一些，但成功的把握性大，因为你总可以收集到一些新的资料进行分析，比如说，别人做了美国的消费函数，你可以做中国的消费函数；有人做了中国的消费函数，你还可以做北京市的消费函数，如此等等，你总可以做出一些东西来。理论研究是在书斋里做文章，一般来说要快些，但难度大风险也大，因为你必须做别人不曾做过的东西，创造一种新理论模型，比如说，对某种观测到的现象提供一种新的解释，证明一个新的定理，数学推导是绝对马虎不得的。这是为什么绝大多数研究生选择做经验研究而不是理论研究的主要原因。当然，无论是经验研究还是理论研究，你必须在你所选择的领域里作出自己的贡献（尽管对什么叫“贡献”并没有严格的定义），所以，你必须对自己研究领域的现有文献非常熟悉。事实上，熟悉文献是每个博士生要过的第一关。如果连别人干了些什么都不熟悉，你怎么能知道自己是否在做贡献呢？当然，写出别人已经做了些什么（或许夹杂些自己的评论）仅仅是研究的起点，你必须做出别人没有做过的东西。经验研究的一个例子是，普林斯顿大学的一位研究生的博士论文研究曼哈顿鱼市上是否存在种族之间的价格歧视。我的博士论文属于理论一类。但是，与大部分理论性论文不同的是，我论文的选题比较“大”（这确实是不寻常的）。另外，我的论文的不同部分之间具有逻辑上的连贯性（coherence），论文的结构类似一本书的结构，而大部分博士论文更像论文集。（顺便说一下，在大部分学校，几篇独立的文章[essays]就可以构成一篇博士论文。）但是，就研究方法而言，我的论文是一篇典型的博士论文。读了这篇论文，读者大致可以了解，国外经济学的理论性博士论文是怎样做的。我的导师曾告诉我，这篇论文将是他们未来的研究生的一个范本。

如大多数理论文章一样，这篇论文也不得不使用一些数学。但是，坦率地讲，这里用到的数学知识是非常一般的，基本上没有超出微积分和概率论，比许多学术杂志上的文章用到的数学要简单得多。对受过中级以上微观经济学训练的读者来说，读这本书是不应该有什么困难的。或许，读者可以用这本书来检查一下自己的微观经济学基础。如果你觉得读这本书还有些困难，那么，我建议你再回炉一下自己的微观经济学。我并不期待这本书受到所有经济学学生的喜爱，但我相信，对那些有志于经济学研究，特别是企业理论研究的读者来说，认真读一读这本书是值得的。现在在校的经济学学生是10年后中国经济学界的主力军。就目前中国经济学界的情况而言，一个中等以上水平的研究生要发表几篇文章也许并不难，但要使自己在10年以后也能发表有一定学术价值的文章，现在进行一些基础性投资是非常必要的。当然，对那些只对观点感兴趣的读者来说，读一读本书的文字也就够了。

与本书的主体部分不同，本书中的四个附录是有关中国经济的。它们可以看作是现代企业理论（包括我本人的理论观点）在分析中国经济中的应用，已分别在杂志上发表。我将它们收集在此，是想为理论的应用提供一个范例。[1]

我要感谢余晖、刘世锦、张春霖、李仁贵、周慈敖、吴有昌、马捷、王中华、阎伟等为本书做的翻译工作，感谢我在北大和社科院的学生以及其他读者对我的研究成果的兴趣，感谢上海人民出版社陈昕

试读结束[说明：试读内容隐藏了图片]

下载完整电子书