【英文原文】Theperson in relation to others is what is missing in standard economic theory, whichappears not to see that what is relevant to people is not to be found only in peoplethemselves, but also in what happens among them. An economic science that assumedall agents to be asocial, and failed to consider that the person qua person matterswould be a poor science indeed, and ultimately of little use. Even the “new socialeconomics” of Durlauf and Young (2001) – certainly an interesting line of thought –offers late and quite often simplistic answers, because it posits an “economicman” who, like the mushrooms Hobbes talked about in De Cive, comes onstagealready full formed. The self comes before the social relation, so that the latterbecomes strictly instrumental. This is why this literature fails to account forthe importance of reciprocity, which is regularly interpreted as the “specialcase” of an exchange of equivalents relation, in which the agents pursueenlightened self-interest.

The culture of modernity is responsible forthis reductionist stance, whereby contracts and incentives (plus, of course, awell defined institutional arrangement) would be all that the economy needs.This means refusing to see that gift as gratuitousness always counterposes itslogic of overabundance to that of equivalence typical of contracts. It is knownthat there are moral traditions like the Puritan which considers actionscomplying with duty superior to those stemming from love. As Max Weber wrote inthe Protestant Ethics: “a higher ethical value is attached to the accomplishmentof duty without love than to sentimental philanthropy” (Weber1967: 202). Why?Because while the Kantian morality seeks universal principles independent of social ties, (and marketexchange seeks equivalence also independent of those ties), things tend, on thecontrary, to be embedded in social ties. They circulate embedded in a socialrelationship; not in the name of an abstract principle. The gift is aninvitation to a relationship, since when one receives a gift it induces animpulse to give. Economic theory needs to think of an agent who can combine freedomof choice with relationality, for if human relations alone would produce an equivocalcommunitarianism, freedom of choice by itself would resume all the shortcomingsof axiological individualism.

A phenomenon that has contributed greatlyto bringing the relationality principle back into economic discourse is thehappiness paradox, or “Easterlin’s paradox” (after the American scholar RichardEasterlin who proposed it in 1974). Pascal (Pensées, nol. 425) observed that“All men seek happiness. This is without exception. Whatever different meansthey employ, they all tend to this end. … This is the motive of every action ofevery man, even of those who hang themselves.” Now as long as economic doctrinewas able to convince that “to be” happy was the same thing as “to have”happiness, it succeeded in masquerading utility as happiness and persuadingpeople that maximizing utility was not just rational but reasonable, i.e. anact of wisdom.

The problems came to a head just when itwas discovered empirically, that the relation between per capita income –as an indicator, albeit rough and ready, of utility – and subjectivewell-being (happiness) can be graphed as an inverted “U” (a parabola concaveupwards). That is, above a certain level further increases in per capita incomeactually diminish the subjective perception of well-being. I do not intend todwell on the countless explanations suggested for this paradox. They run fromthe psychological (the treadmill effect) through the economic (positional externalities)to the sociological (based on the notion of relational goods). The literatureis vast and deep-rooted (Bruni and Zamagni 2016b).

In another work I have dealt with thepeculiar characteristics of relational goods and their meaning in today’sadvanced societies (Zamagni 2005b). Here, let me add that the main reason theindividualist paradigm can never treat relational goods adequately is that withthese goods it is the relation as such that constitutes the good; that is, theinterpersonal relation does not exist independently of the good, which is producedand consumed at the same time. This means that my knowledge of the identity ofthe other with whom I have a relation is indispensable for there to be a relationalgood at all. By contrast, the assumption underlying the exchange of equivalentsis that it is always possible to replace the person or persons on whom my well-beingdepends with other persons. (If I am not satisfied with my regular butcher, Ican always go to another. But I cannot replace the person who provides me witha service of care without altering my own index of happiness.) As Philip Wicksteed(1910) saw clearly, the primary foundation of the capitalistic market is notegotism but “non-tuism,” because business is better done with people whose personalidentity one does not know. From the relational perspective, however, the relationshipwith another person presupposes recognition and receptiveness: welcoming apresence that, in its humanity, is common to me and in its otherness, distinctfrom me. No easy task, certainly – “Hell is other people,” as Sartre saidin No Exit – but essential if we want to overcome the severe shortage ofrelational goods that typifies our society. Individualism is a useful guide forutility that depends on goods and services that can be enjoyed even inisolation. But it is a poor maestro for happiness, given that true happinessrequires being at least two in number, as Aristotle wrote. To quote ure,“The Lord God said, it is not good that the man should be alone” (Genesis 2:18,King James version).

The fact is that I need the other todiscover that it is worth preserving myself; in order to flourish, as in theAristotelian “eudaimonia”. But the other too needs me to recognize him/her assomeone whose flourishing is good. Since we need the same recognition, I willact towards the other as if before a mirror. Self-fulfilment is the result ofthat interaction. The original asset that I can put at the disposal of theperson in front of me is the capacity to recognize the value of that person’sexistence – a resource that cannot be produced unless it is shared. Whatmatters here is to see that this implies recognition of the other – notjust her right to exist but of the necessity that she exist as a condition formy own existence, in relation to him/her. Recognizing the other person as anend in himself and recognizing him as the means for my own fulfilment arereunified, which resolves the reductive dichotomy between Kantian morals, whichrequire that we see others strictly as an end in themselves, and a theory ofinstrumental rationality in which others are seen as the means to one’s own ends.The good of self-fulfilment is attained when there is reciprocal recognition. Notethat the fact that my recognition of the other person brings with it thereciprocal recognition that I myself need does not make this disposition merelyinstrumental. For the self is constituted in part by the recognition conferredby others.

3.5 ReciprocityVersus Exchange of Equivalents

经济学在相当长的一段时间内已经认识到，社会互动产生外部性，而后者往往会延缓社会有利平衡的运动。这源于这样一个事实：社会决定（影响他人效用的决定）和个人决定（其影响仅限于代理人的范围）有基本区别。澄清这一点：尽管我决定与朋友们的关系的质量和强度不受消费的影响，比如说香蕉，而不是苹果，但我决定结婚与否，获得学位，与其他人有显著的直接关系。事实上，这样的决定有助于定义我的身份，从而影响我与他人的关系。它可能发生，正如G. Akerlof（1997）指出，我的选择的主要决定因素是影响我想象我的决定会对我与他人的关系网络；而我的选择对我的效用–这是标准的理论本身限制考虑–可能是次要的水平直接效果评价。

近年来，社会互动理论明确地考虑到与社会决策有关的外部性，这是从互惠原则的大量实证研究中得出的结论。对kahnemann和特韦尔斯基对经济心理学的工作，劳动经济学，Bewley，sudgen对公共物品的自愿捐款，使Fehr和其他人对合同的执行情况，清楚地表明，往复的行为不能在自我兴趣偏好和行为导向的结果解释。这是现在某些互惠是不同于所谓的往复利他主义或直接的相互关系，由R.M. Axelrod针锋相对针锋相对的策略例证（2006）说。往复的利他主义者，事实上，准备回报只有一定的报酬预计未来从他的行动。这样的报酬可以在无限重复博弈中出现，也可以作为完全信息有限博弈中的均衡解。此外，这种行为与利己的偏好完全相容，因为未来的收益（或未来惩罚的可信威胁）可能会导致自私但有耐心的代理人在重复博弈的情况下做出短暂牺牲。

礼尚往来，另一方面，是一种行为回答“感知礼”为法尔克和Fischbacher（2001）把它–，礼貌用语是指无论是分配正义感和行为意向的比较。那些有往复偏好的人的动机不是基于未来收益的任何前景。这样的一个人倾向于作出物质上的牺牲，以支持那些同样被处分的人，并惩罚那些没有的人。例如，惩罚游戏中的不公平不公正的行为的最后通牒，是一种互惠。再次，在协调博弈中，大多数实验室实验表明，玩家首先倾向于一种劣势平衡，然后出人意料地改变策略，收敛到优势平衡点。

要进行这种反转，必须有几个人选择对预先发生的动作的结果最好的回复。换句话说，一定数量的玩家在短期内做出牺牲，教导其他玩家协调优势平衡的方法。

文献中存在着互惠原则的两个概念。一，这仍然是在因果的角度来看，是建立在对公平的感觉：它是厌恶不公平的分布，刺激往复响应。另一个基于礼貌原则的概念不仅考虑了行动的后果，而且考虑了行动背后的意图。（在最后通牒博弈，例如，某提供的钱是被其他玩家更经常如果提供帮助的人显然不是比如果同时提供信号的用意是好的–让我们说因为支持者发现在自己需要一个特定的情况下）。不用说，第二概念是对等理论中的一个有趣的发展，因为它允许代表代理人的效用作为一个功能不仅对原始游戏的好处，但也表达感知–一词，法尔克和Fischbacher的礼貌回应的期限（2001）正确所谓的“互惠”的效用。但从理论上说，这一理论的解释力远低于最初的预期。对我来说，这是互惠的标准理论都假定一个人版本的框架，也是对利他行为的理论基础。事实上，在这两种情况下，个人都具有相互依赖的偏好；换句话说，他们对其他人的回报感兴趣，也对他们自己有兴趣。归根结底，这两种方法之间的区别在于偏好的相互依赖性的不同规范，但其背后的概念结构仍然相同。

事实上，声明的目的这一文献是表明互惠偏好是一致的，尽管外表，与个人主义的法令。总之，想法，是在信息不完全–因此偏好不能被观察到–偏好，而不是个人主义是不可信的，因此最终由个人喜好为主，由于代理商，虽然他们知道在人群中的各种类型的相对频率，不知道的人他们发现自己的喜好相互作用。因此，结论是，在大群体中，在存在不完全信息的情况下，毫无疑问，进化倾向于在市场经济等社会背景下运作的代理人的个人偏好。

不难看出这类论点的基本弱点。关键的一点是，在这文献的长期互惠的意义被耍弄。反之，互惠假设是对他人身份的认识，最重要的是，代理人与其他人的关系准备就绪，主流做法减少了互惠，而这只是一种关系。因此，如果别人不知道我是谁，即他们不知道在阿克洛夫和Kranton的感觉我的身份（2000），他们是领导对我的表现个人主义，因为正是这种模式的出现作为一个进化稳定在所有那些由其他任何一种”突变”。事实是，身份意味着认识自己并被认可。自我认同的维度意味着自我认识，对自己和自己的生活经验的记忆；被认可的维回忆，需要每个人将插入一个关系网络促进自我知识由于提供的信息，那些与人互动。正因为身份属于存在，而不是拥有，它是不可谈判的。

鉴于此，我认为，S.Kolm所提出的概念的相互作用（2000）更相关。对他来说，互惠是一系列相互独立、相互联系的双向转移。独立意味着每一次转让本身都是自愿的，即自由；换言之，没有转移是把另一个付诸行动的先决条件，因为没有任何外在的义务可以作用于代理人的思想。正是这一特点，区别于基于等价物的交换关系，互惠关系的原则，也是由一组双向传输，但其自愿，所以说，全球在这个意义上，它是在一组传输应用，而不是单个转移了本身。换言之，等价交换关系所隐含的转让是另一方的前提条件，因此，第三方（例如，法官）可以随时干预，以履行合同义务。这样的事发生的相互关系，即使这是真的，对等互惠的交流，在一定程度上，假设自愿，反对命令关系。然而，与此同时，互惠的自由要比交换等价物的自由大得多，因为一个方向的转移在相反的方向上是强制性的。双向互惠转移的双向性是后者与纯粹利他主义的区别，后者包括单向转移。可以说，互惠在交换与纯粹利他之间处于中间地位。

鉴于上述情况，可以得出结论，互惠的关系需要某种形式的平衡之间的什么人给什么人期望得到的，平衡的，然而，这不是在一个确定的关系交流表达（即在一个相对价格），因为它可以根据与道德意向等同情、仁爱的强度不同，或团结的感觉是由参与关系的代理付诸实践。无论如何，互惠在某种程度上是指个人之间相互作用的战略层面：如果那些接受我调动的人没有给我一些回报的迹象，我就会被诱导与他们断绝关系。那么，在等价交换关系中发生的区别在哪里呢？答案是双重的。在交换中，交换比率（即均衡价格）的确定在逻辑上先于交换项目的转移，只有在买方和卖方商定了作为交易对象的房子的价格时，交换才发生。在互惠关系中，另一方面，在逻辑上和时间上，转移先于反转移，在这种情况下，开始关系的代理人不能要求任何权利，而只能是期望。其次，互惠关系可以修改游戏本身的结果，是否因为互惠的实践倾向于稳定剂的亲社会行为的人发现自己在囚徒困境式语境相互作用，或是因为互惠的文化倾向于修改内源性的代理人的偏好序。作为基本的信任博弈的结果，如果我发现自己需要别人的情况下，我不能可靠地约束自己，对未来有帮助的承诺，在理性选择理论的意义上的理性主体，虽然能帮助我，肯定不这么做的话，我知道一个自私自利的经纪人，他猜想，我不会有丝毫的兴趣收到往复的青睐。但这是不是如果我潜在的帮手，知道我是谁的做法互惠（Kolm 2000）。现在可以理解为什么，不像交换原则发生了什么（和强迫），互惠不能解释的自身利益：动机和对别人的态度是互惠的概念的一个基本组成部分。这就是为什么标准的经济理论，取决于理性选择的方案，未能充分解释互惠的概念：后者是关系不能与声誉是一种无形资产–困惑。

【英文原文】Economicshas known for quite some time that social interaction produces externalitiesand that the latter tend to delay the movement towards socially advantageousequilibria. This follows from the fact that there is a basic difference betweensocial decisions (those that influence the utility of others) and individualdecisions (those whose effects remain limited to the sphere of the agent). Toclarify the point: whereas the quality and the intensity of my relationshipswith friends are not affected by my decision to consume, say, bananas insteadof apples, my decision to get married or not, to obtain a degree, or not, hasdirect effects of marked relevance to others. In fact, decisions of this kindhelp to define my identity and thus influence my relationships with others. Itmay thus happen, as G. Akerlof (1997) notes, that the main determinant ofmy choice is the impact I imagine my decision will have on the network of myrelations with others; whereas the evaluation of the direct effects of mychoice on my level of utility – which is what the standard theoryrestricts itself to considering – could be of secondary importance.

In recent times, a theory of socialinteractions, explicitly taking into account the externalities associated withsocial decisions, is the one stemming from the huge amount of empiricalresearch on the principle of reciprocity. The work of Kahnemann and Tversky oneconomic psychology, that of Bewley on labor economics, that of Sudgen on thevoluntary contribution to public goods, that of Fehr and others on the enforceabilityof contracts, clearly shows that reciprocating behavior cannot be explained interms of self-interested preferences and behavior oriented to results. It is bynow certain that reciprocity is something different from the so-calledreciprocating altruism or from direct reciprocity, as exemplified by thetit-for-tat strategy R.M. Axelrod (2006) speaks of. The reciprocatingaltruist, in fact, is prepared to reciprocate only if some recompense isexpected in the future from his action. Such a recompense can emerge as anequilibrium solution in games infinitely repeated or else in finite games withcomplete information. Furthermore, this kind of behavior is perfectlycompatible with self-interested preferences, since the prospect of future gains(or the credible threat of future punishments) can induce selfish but patient agentsto make momentary sacrifices in contexts of repeated games.

Reciprocity, on the other hand, is abehavioral answer to “perceived courtesy” as Falk and Fischbacher (2001)put it – where the term courtesy is taken to mean both the sense ofdistributive justice and the intention to behave fairly. The motivation ofthose who have reciprocating preferences is not based, then, on any kind of prospectof future gain. An individual of this kind is disposed to make material sacrificesin favor of those who are similarly disposed, and to punish those who are not.For example, to punish unfair iniquitous behavior in a game like ultimatum, is aform of reciprocity. Again, in games of coordination, most of the laboratory experimentsshow that the players, first, tend towards an inferior equilibrium and then,unexpectedly, change their strategy and converge to the dominant equilibrium.

For this inversion to take place it isnecessary that several individuals choose actions that are not best replies tothe results of the moves that occurred immediately beforehand. In other words acertain number of players “make sacrifices” in the short run to teach the otherplayers the way to coordinate on the dominant equilibrium.

Two conceptions of the principle ofreciprocity are present in the literature. One, which remains within theconsequentialist perspective, is founded on aversion to the sense of injustice:it is aversion to unfair distributions that stimulates the reciprocatingresponse. The other notion, based on the principle of perceived courtesy, takesinto account not only the consequences of the action, but also the intentionsbehind it. (In ultimatum game, for example, a certain offer of money isrejected by the other player much more often if the person making the offer isevidently not wellintentioned than if the same offer signals a goodintention – let us say because the proponent finds herself in a particularsituation of need). Needless to say, the second notion represents aninteresting development within the theory of reciprocity, since it allows torepresent the utility of the agents as a function not only of the pay-offs ofthe original game, but also of the term expressing the response to the courtesyperceived – a term that Falk and Fischbacher (2001) rightly call “utilityof reciprocity”. But on reflection, the explanatory power of the theory is agood deal lower than the expectations it initially aroused. To me, the reasonis that both versions of the standard theory of reciprocity presuppose the sameindividualist framework that is also basic to the theory of altruism. In bothcases, in fact, individuals possess interdependent preferences; in other words,they are interested in some way in others’ pay-offs as well as their own. Thedifference between the two approaches, in the final analysis, lies in thedifferent specification of the interdependence of the preferences, but theconceptual structure behind it remains the same.

Indeed, the declared purpose of thisliterature is to show that the preferences for reciprocity are compatible,despite appearances, with the individualist statute. The idea, in short, isthat when information is incomplete – and therefore the preferences cannotbe observed – preferences that are not individualistic are not credibleand hence end up by being dominated by individualistic preferences, since theagents, though they know the relative frequency of the various types in thepopulation, do not know the preferences of those with whom they find themselvesinteracting. The conclusion is therefore that within big groups and in thepresence of incomplete information, there can be no doubt at all that evolutionfavors individualistic preferences in agents that operate in social contextslike a market economy.

It is not difficult to see the basicweakness of a line of argument of this kind. The point is that in thisliterature the meaning of the term reciprocity is being juggled with. Whereas,reciprocity postulates the knowledge of the identity of others, and above allthe readiness on the part of the agent to be in relation with others, the mainstreamapproach reduces reciprocity – which is a relation – to a merepreference. It is therefore obvious that if others do not know who I am, i.e.they do not know my identity in the sense of Akerlof and Kranton (2000), theyare led to behave individualistically towards me, since it is precisely thismode that emerges as the one evolutionarily stable among all those generated byany other kind of “mutant”. The fact is that identity means to recognizeoneself and be recognized. The selfrecognition dimension impliesself-knowledge, the memory of oneself and one’s own experience of life; thedimension of being recognized recalls the need of every person to be insertedin a network of relations facilitating self knowledge thanks to informationprovided by those with whom the person interacts. Precisely because identitybelongs to being, rather than having, it is not negotiable.

In view of this I believe theconceptualization of reciprocity proposed by S. Kolm (2000) to be morerelevant. To him, reciprocity is a series of bi-directional transfers, independentof each other and at the same time interconnected. Independence implies thateach transfer is voluntary in itself, i.e. free; in other words, no transfer constitutesa prerequisite for putting the other into action, since there is no external obligationcapable of acting on the mind of the agent. It is precisely this characteristicwhich differentiates reciprocity from relations based on the principle of the exchangeof equivalents, relations that are also constituted by a set of bi-directional transfers,but in which the voluntariness is, so to speak, global in the sense that it is appliedwithin a set of transfers, and not to each single transfer taken by itself. In otherwords, the transfers implied by the relation of exchange of equivalents are eachthe prerequisite of the other, so that a third party (for example, a judge) canalways intervene to render contractual obligations enforceable. Nothing likethis happens with reciprocity, even if it remains true that exchanges ofequivalents and reciprocity, to the extent that both postulate voluntariness,are opposed to command relation. At the same time, however, there is much morefreedom in reciprocity than in the exchange of equivalents, where the transferin one direction becomes obligatory by the transfer in the opposite direction.The bi-directionality of the transfers characterizing reciprocity is whatdifferentiates the latter from pure altruism, which consists in one-directionaltransfers. On the whole, it may be said that reciprocity occupies the intermediateposition between exchanges and pure altruism.

In view of the above, it may be concludedthat the relation of reciprocity requires some form of balancing between whatone gives and what one expects to obtain, a balancing, however, that is notexpressed in a defined relation of exchanges (i.e. In a relative price), sinceit may vary according to the intensity with which moral dispositions such assympathy, benevolence, or the feeling of solidarity are put into practice bythe agents involved in the relation. At any rate, it is true that reciprocity insome way refers to the strategic dimension of interaction between individuals:if those who receive my transfer do not give me some sign of reciprocation, Ishall be induced to break off relations with them. So where does the differencelie with what happens in the exchange of equivalents relations? The answer liesin a dual order of circumstances. In exchanges the determination of theexchange ratio (i.e. the equilibrium price) logically precedes the transfer ofthe item exchanged – only after the purchaser and seller have agreed onthe price of the house which is the object of the transaction can the exchangetake place. In the relation of reciprocity, on the other hand, the transferprecedes, both logically and temporally, the counter-transfer, about which theagent who begins the relation cannot claim any right, but only an expectation.Secondly, the reciprocity nexus can modify the results of the game itself,whether because the practice of reciprocity tends to stabilize pro-social behaviorin agents who find themselves interacting in contexts of the prisoner’s dilemmatype, or because the culture of reciprocity tends to modify endogenously thepreference orderings of the agents. As the results of the basic trust game, ifI find myself needing the help of others in circumstances in which I cannotcredibly bind myself to some commitment for the future, a rational agent in thesense of the theory of rational choice, though capable of helping me, willcertainly not do so if, knowing that I am also a self-interested agent, heconjectures that I will not have the slightest interest in reciprocating thefavor received. But this is not the case if my potential helper knows that I amsomeone who practices reciprocity (Kolm 2000). It is now possible to understandwhy, unlike what happens with the principle of exchange (and with coercion),reciprocity cannot be explained in terms of selfinterest only: motivations andthe attitude towards others are a fundamental component of the concept ofreciprocity. This is why standard economic theory, hinging on the scheme ofrational choice, fails to explain the notion of reciprocity adequately: thelatter being a relation cannot be confused with reputation – which is an intangibleasset.

【良知天下、良知企业家、制度企业家、阳民智库及其周刊原创文章统一投稿邮箱：Ayoungman@sina.com】返回搜狐，查看更多