From Peter Blau, Exchange and Power in Social Life.New York: Wiley, 1964, pp. 88-97.


Social Exchange

The moral type on the other hand is not based onstated terms, but the gift or other service is given as to a friend,although the giver expects to receive an equivalent or greaterreturn, as though it had not been a free gift but a loan; and as heends the relationship in a different spirit from that in which hebegan it, he will complain. The reason of this is that all men, ormost men, wish what is noble but choose what is profitable; and whileit is noble to render a service not with an eye to receiving one inreturn, it is profitable to receive one. One ought, therefore, if onecan, to return the equivalent of services received, and to do sowillingly. . . . Aristotle, The Nicomachean Ethics

Processes of social association can be conceptualized, followingHomans' lead, "as an exchange of activity, tangible or intangible,and more or less rewarding or costly, between at least two persons.''[l] Social exchange can be observed everywhere once we are sensitizedby this conception to it, not only in market relations but also infriendship and even in love, as we have seen, as well as in manysocial relations between these extremes in intimacy. Neighborsexchange favors; children, toys; colleagues, assistance;acquaintances, courtesies; politicians, concessions; discussants,ideas; housewives, recipes. The pervasiveness of social exchangemakes it tempting to consider all social conduct in terms ofexchange, but this would deprive the concept of its distinctivemeaning. People do things for fear of other men or for fear of God orfor fear of their conscience, and nothing is gained by trying toforce such action into a conceptual framework of exchange.

Mauss and other anthropologists have called attention to thesignificance and prevalence of the exchange of gifts and services insimpler societies. "In theory such gifts are voluntary but in factthey are given and repaid under obligation. . . . Further, what theyexchange is not exclusively goods and wealth, real and personalproperty, and things of economic value. They exchange rathercourtesies, entertainments, ritual, military assistance, women,children, dances, and feasts; and fairs in which the market is butone element and the circulation of wealth but one part of a wide andenduring contact." [2]

The institutionalized form the exchange of gifts frequentlyassumes in simpler societies highlights the two general functions ofsocial, as distinct from strictly economic, exchange, namely, toestablish bonds of friendship and to establish superordination overothers. The creation of friendship bonds is typified by theceremonial Kula exchange in the Western Pacific, where "the Kulapartnership provides every man within its ring with a few friendsnear at hand, and with some friendly allies in the far-away,dangerous, foreign districts." [3] A polar example of theestablishment of superordination over others is the potlatch in theAmerican Northwest, in which "status in associations and clans, andrank of every kind, are determined by the war of property. . . . "[4] What is most interesting, however, is that the exchanges in thesame institution serve sometimes to cement peer relations andsometimes to produce differentiation of status, contradictory asthese two consequences appear to be.

The basic principles underlying the conception of exchange may bebriefly summarized. An individual who supplies rewarding services toanother obligates him. To discharge this obligation, the second mustfurnish benefits to the first in turn. Concern here is with extrinsicbenefits, not primarily with the rewards intrinsic to the associationitself, although the significance of the social "commodities"exchanged is never perfectly independent of the interpersonalrelation between the exchange partners. If both individuals valuewhat they receive from the other, both are prone to supply more oftheir own services to provide incentives for the other to increasehis supply and to avoid becoming indebted to him. As both receiveincreasing amounts of the assistance they originally needed ratherbadly, however, their need for still further assistance typicallydeclines.

"The profits from exchange decrease with the number of exchanges;"[5] in technical terms, the marginal utility of increasing amounts ofbenefits eventually diminishes. If we need help in our work, forexample, five minutes of an expert's assistance are worth much to us,and another five minutes are perhaps just as valuable, but once hehas aided us for half an hour another five minutes of his time areundoubtedly less significant than were the first five. Ultimately,the declining marginal utility of additional benefits is no longerworth the cost of obtaining them, and the point at which this happensfor both partners, often after some adjustment in the ratio at whichthey exchange services, governs the level of transactions mostadvantageous for both at which the volume of exchange between thempresumably becomes stabilized. Although personal considerations forinstance, the desire not to antagonize a colleague modify theserational decisions, such factors also can be taken into account inmore complex versions of the basic model, at least in principle.

Take the association of a new member of a profession with arespected senior colleague as an illustration of these processes. Thejunior is rewarded by the senior's stimulating expert discussions ofprofessional matters and by the senior's willingness to treat him asa colleague, which symbolizes acceptance as a full-fledgedprofessional. He reciprocates by his deferential admiration, which isrewarding for the senior. The gratification the senior derives frombeing listened to with great respect prompts him to devote some ofhis limited time to the association, but his gratification is notproportionately increased if he extends the period in which the otheradmires his expert opinions from half an hour every few days toseveral hours daily. Moreover, the more time the senior devotes tothe association, the costlier it becomes for him to further restrictthe time available to him for other activities. Hence, he will beinclined to limit the time he spends in discussions with the juniorto the level at which the support he receives from his admirationstill outweighs in significance the advantages foregone by takingtime from other pursuits. At this point, however, the junior maystill profit from further association with the senior. Since hisadmiration does not suffice to increase the association time, thejunior must endeavor to furnish supplementary rewards, for example,by doing odd jobs for his senior colleague, thereby obligating him toreciprocate by devoting more time to the association than heotherwise would. Eventually, the marginal advantages for the juniorof associating still more with the senior will no longer outweigh themarginal cost of providing more services for him, and the exchangewill tend to level off. The assumption is not that individuals makethese calculations explicitly but that such implicit calculationsunderlie the feelings of boredom or pressure from other work thatprompt their decisions to spend only a certain amount of timetogether.

Unspecified Obligations and Trust

The concept of exchange can be circumscribed by indicating twolimiting cases. An individual may give another money because theother stands in front of him with a gun in a holdup. While this couldbe conceptualized as an exchange of his money for his life, it seemspreferable to exclude the result of physical coercion from the rangeof social conduct encompassed by the term "exchange." An individualmay also give away money because his conscience demands that he helpsupport the underprivileged and without expecting any form ofgratitude from them. While this could be conceptualized as anexchange of his money for the internal approval of his superego, hereagain it seems preferable to exclude conformity with internalizednorms from the purview of the concept of social exchange. [6] Asocial exchange is involved if an individual gives money to a poorman be cause he wants to receive the man's expressions of gratitudeand deference and if he ceases to give alms to beggars who withholdsuch expressions.

"Social exchange," as the term is used here, refers to voluntaryactions of individuals that are motivated by the returns they areexpected to bring and typically to id fact bring from others. Actioncompelled by physical coercion is not voluntary, although compliancewith other forms of power can be considered a voluntary servicerendered in exchange for the benefits such compliance produces, asalready indicated. Whereas conformity with internalized standardsdoes not fall under the definition of exchange presented, conformityto social pressures tends to entail indirect exchanges. Men makecharitable donations, not to earn the gratitude of the recipients,whom they never see, but to earn the approval of their peers whoparticipate in the philanthropic campaign. Donations are exchangedfor social approval, though the recipients of the donations and thesuppliers of the approval are not identical, and the clarification ofthe connection between the two requires an analysis of the complexstructures of indirect exchange, which is reserved for chapters eightand ten. Our concern now is with the simpler direct exchanges.

The need to reciprocate for benefits received in order to continuereceiving them serves as a "starting mechanism" of social interactionand group structure, as Gouldner has pointed out. [7] When people arethrown together, and before common norms or goals or roleexpectations have crystallized among them, the advantages to begained from entering into exchange relations furnish incentives forsocial interaction, and the exchange processes serve as mechanismsfor regulating social interaction, thus fostering the development ofa network of social relations and a rudimentary group structure.Eventually, group norms to regulate and limit the exchangetransactions emerge, including the fundamental and ubiquitous norm ofreciprocity, which makes failure to discharge obligations subject togroup sanctions. In contrast to Gouldner, however, it is held herethat the norm of reciprocity merely reinforces and stabilizestendencies inherent in the character of social exchange itself andthat the fundamental starting mechanism of patterned socialintercourse is found in the existential conditions of exchange, notin the norm of reciprocity. It is a necessary condition of exchangethat individuals, in the interest of continuing to receive neededservices, discharge their obligations for having received them in thepast. Exchange processes utilize, as it were, the self-interests ofindividuals to produce a differentiated social structure within whichnorms tend to develop that require individuals to set aside some oftheir personal interests for the sake of those of the collectivity.Not all social constraints are normative constraints, and thoseimposed by the nature of social exchange are not, at least, notoriginally.

Social exchange differs in important ways from strictly economicexchange. The basic and most crucial distinction is that socialexchange entails unspecified obligations. The prototype of aneconomic transaction rests on a formal contract that stipulates theexact quantities to be exchanged. [8] The buyer pays $30,000 for aspecific house, or he signs a contract to pay that sum plus interestover a period of years. Whether the entire transaction is consummatedat a given time, in which case the contract may never be written, ornot, all the transfers to be made now or in the future are agreedupon at the time of sale. Social exchange, in contrast, involves theprinciple that one person does another a favor, and while there is ageneral expectation of some future return, its exact nature isdefinitely not stipulated in advance. The distinctiveimplications of such unspecified obligations are brought into highrelief by the institutionalized form they assume in the Kuladiscussed by Malinowski:

The main principle underlying the regulations ofactual exchange is that the Kula consists in the bestowing of aceremonial gift, which has to be repaid by an equivalent counter-giftafter a lapse of time. . . . But it can never be exchanged from handto hand, with the equivalence between the two objects beingdiscussed, bargained about and computed. . . . The second veryimportant principle is that the equivalence of the counter-gift isleft to the giver, and it cannot be enforced by any kind of coercion.. . . If the article given as a counter-gift is not equivalent, therecipient will be disappointed and angry, but he has no direct meansof redress, no means of coercing his partner. . . . [9]

Social exchange, whether it is in this ceremonial form or not,involves favors that create diffuse future obligations, not preciselyspecified ones, and the nature of the return cannot be bargainedabout but must be left to the discretion of the one who makes it.Thus, if a person gives a dinner party, he expects his guests toreciprocate at some future date. But he can hardly bargain with themabout the kind of party to which they should invite him, although heexpects them not simply to ask him for a quick lunch if he hadinvited them to a formal dinner. Similarly, if a person goes to sometrouble in behalf of an acquaintance, he expects someexpression of gratitude, but he can neither bargain with the otherover how to reciprocate nor force him to reciprocate at all.

Since there is no way to assure an appropriate return for a favor,social exchange requires trusting others to discharge theirobligations. While the banker who makes a loan to a man who buys ahouse does not have to trust him, although he hopes he will not haveto foreclose the mortgage, the individual who give another anexpensive gift must trust him to reciprocate in proper fashion.Typically, however, exchange relations evolve in a slow process,starting with minor transactions in which little trust is requiredbecause little risk is involved. A worker may help a colleague a fewtimes. If the colleague fails to reciprocate, the worker has lostlittle and can easily protect himself against further loss by ceasingto furnish assistance. If the colleague does reciprocate, perhapsexcessively so out of gratitude for the volunteered help and in thehope of receiving more, he proves himself trustworthy of continuedand extended favors. (Excessive reciprocation may be embarrassing,because it is a bid for a more extensive exchange relation than onemay be willing to enter.) By discharging their obligations forservices rendered, if only to provide inducements for the supply ofmore assistance, individuals demonstrate their trustworthiness, andthe gradual expansion of mutual service is accompanied by a parallelgrowth of mutual trust. Hence, processes of social exchange, whichmay originate in pure self-interest, generate trust in socialrelations through their recurrent and gradually expanding character.

Only social exchange tends to engender feelings of personalobligation, gratitude, and trust; purely economic exchange as suchdoes not. An individual is obligated to the banker who gives him amortgage on his house merely in the technical sense of owing himmoney, but he does not feel personally obligated in the sense ofexperiencing a debt of gratitude to the banker, because all thebanker's services, all costs and risks, are duly taken into accountin and fully repaid by the interest on the loan he receives. A bankerwho grants a loan without adequate collateral, however, does make therecipient personally obligated for this favorable treatment,precisely because this act of trust entails a social exchange that issuperimposed upon the strictly economic transaction.

In contrast to economic commodities, the benefits involved insocial exchange do not have an exact price in terms of a singlequantitative medium of exchange, which is another reason why socialobligations are unspecific. It is essential to realize that this is asubstantive fact, not simply a methodological problem. It is not justthe social scientist who cannot exactly measure how much approval agiven helpful action is worth; the actors themselves cannot preciselyspecify the worth of approval or of help in the absence of a moneyprice. The obligations individuals incur in social exchange,therefore, are defined only in general, somewhat diffuse terms.Furthermore, the specific benefits exchanged are sometimes primarilyvalued as symbols of the supportiveness and friendliness theyexpress, and it is the exchange of the underlying mutual support thatis the main concern of the participants. Occasionally, atime-consuming service of great material benefit to the recipientmight be properly repaid by mere verbal expressions of deepappreciation, since these are taken to signify as much supportivenessas the material benefits. [10] In the long run, however, the explicitefforts the associates in a peer relation make in one another'sbehalf tend to be in balance, if only because a persistent imbalancein these manifestations of good will raise questions about thereciprocity in the underlying orientations of support andcongeniality.

Extrinsic benefits are, in principle, detachable from the sourcethat supplies them, but their detachability is a matter of degree. Atone extreme are economic commodities, the significance of which isquite independent of the firm that supplies them. The value of ashare in a corporation is not affected by the broker from whom we buyit. At the other extreme is the diffuse social support we derive in alove relationship, the significance of which depends entirely on theindividual who supplies it. The typical extrinsic benefits sociallyexchanged, such as advice, invitations, assistance, or compliance,have a distinctive significance of their own that is independent oftheir supplier, yet an individual's preferences for them are alsoaffected by his interpersonal relations with the supplier. Althoughthe quality of advice determines its basic value for an individual,regardless of who furnishes it, he tends to prefer to consult acolleague whose friendly relations with him make it easy for him todo so rather than a more expert consultant whom he hardly knows. [11]The ease with which he can approach a colleague, the jokes andconviviality that surround the consultation, and other rewards heobtains from the association combine with the quality of the adviceitself to determine the value of the total transaction for him.Indeed, the exchange of instrumental assistance may sometimes largelyserve the function for participants of providing opportunities forexchanging these other more salient rewards. Going over and helping afellow worker with his task might simply be an excuse for chattingwith him and exchanging social support.

Since social benefits have no exact price, and since the utilityof a given benefit cannot be clearly separated from that of otherrewards derived from a social association, it seems difficult toapply the economic principles of maximizing utilities to socialexchange. [12] The impersonal economic market is designed to stripspecific commodities of these entangling alliances with otherbenefits, so to speak, and thus to make possible rational choicesbetween distinct alternatives with a fixed price. Even in economicexchange, however, the significance of each alternative is rarelyconfined to a single factor, which confounds rationaldecision-making; people's job choices are affected by workingconditions as well as salaries, and their choices of merchants, bythe atmosphere in a store as well as the quality of the merchandise.Although the systematic study of social exchange poses distinctiveproblems, the assumptions it makes about the maximization ofutilities implicit in choice behavior are little different from thosemade by the economist in the study of consumption.

In production and in marketing, profit is at a maximum "when themarginal cost and the marginal revenue are equal," [13] and sinceboth these quantities are defined in dollars, an unequivocalcriterion for maximizing exists. But in consumption maximizinginvolves equating the marginal utilities obtained from expending anextra dollar in alternative ways, and this comparison of utilitiesposes, in principle, the same problem the study of social exchangedoes. Indeed, economists typically do not attempt to measureutilities directly to ascertain whether they are equated but simplyinfer that they are from the distribution of consumer expenditures orfrom other economic decisions. Thus, if a scientist accepts anacademic job at a lower salary than he could command in industry, theso-called psychic income he obtains from his university position isassumed to equal or exceed in utilities the difference in salary.Similar inferences can be made from the observable conduct in socialexchange. Moreover, these inferences about the value the socialrewards exchanged have for individuals can be used to derive testablehypotheses concerning the group structures that will emerge amongthem and the structural changes that will occur under variousconditions, as will be exemplified in chapter seven.

Impressionistic observation suggests that people usually dischargetheir social obligations, even though there is no binding contractthat can be enforced, in contrast to the contractual obligations ineconomic exchange, which can be enforced through legal sanctions. Thereason is that failure to discharge obligations has a number ofdisadvantageous consequences, several of which do not depend on theexistence of a norm of reciprocity. Suppose an individual to whom aneighbor has repeatedly lent some tools fails to reciprocate by doinghis neighbor a favor when an opportunity arises. He can hardly borrowthe tools again next time he needs them, and should he be brashenough to ask for them the neighbor may be reluctant to lend them tohim. The neighbor is also likely to become less friendly toward anindividual who refuses to do favors after he accepts some. Besides,the neighbor will probably distrust him in the future and, forexample, be disinclined to trust him to repay for having their commonfence painted but ask him for payment in advance. Chances are,moreover, that the neighbor will tell other neighbors about theingratitude of this individual, with the result that this person'sgeneral reputation in the community suffers. Specifically, theneighbor's complaints will prompt numerous others to think less wellof him, to hesitate to do him favors, and generally to distrust him.The first neighbor and the others have reason to act in this mannereven if they were only concerned with protecting their ownself-interest. The existence of a norm of reciprocity among themfurther reinforces their disapproval of him and their disinclinationto do favors for him, now as a punitive reaction against a violatorof a moral standard as well as to protect their own interest.Finally, an internalized norm of reciprocity would make him feelguilty if he fails to discharge his obligations, subjecting him tosanctions that are independent of any actions of others. The multiplepenalties that failure to discharge social obligations evokesconstitute pressures to discharge them.


1. George C. Homans, Social Behavior, New York Harcourt,Brace and World, 1961, p. 13. This entire book has been an importantsource of inspiration for the analysis presented here, as noted inthe preface. For Homans' first statement on the subject, see his"Social Behavior as Exchange," American Journal of Sociology,63 (1958), 597-606.

2. Marcel Mauss, The Gift, Glencoe: Free Press, 1954, pp.1, 3.

3. Bronislaw Malinowski, Argonauts of the Western Pacific,New York: Dutton, 1961, p. 92.

4. Mauss, op. cit., p. 35.

5. Homans, Social Behavior, p. 70.

6. Ludwig von Mises refers to this type as autistic exchange."Making one-sided presents without the aim of being rewarded by anyconduct on the part of the receiver or of a third person is autisticexchange. The donor acquires the satisfaction which the bettercondition of the receiver gives to him. The receiver gets the presentas a God-sent gift. But if presents are given in order to influencesome people's conduct, they are no longer one-sided, but a variety ofinterpersonal exchange between the donor and the man whose conductthey are designed to influence." Human Action, New Haven: YaleUniversity Press, 1949, p. 196.

7. Alvin W. Gouldner, "The Norm of Reciprocity," AmericanSociological Review, 25 (1960), 161-178, esp. p. 176.

8. This is not completely correct for an employment contract orfor the purchase of professional services, since the precise servicesthe employee or professional will be obligated to perform are notspecified in detail in advance. Economic transactions that involveservices generally are somewhat closer to social exchange than thepure type of economic exchange of commodities or products ofservices.

9. Malinowski, op. cit., pp. 95-96.

10. See Erving Goffman, Asylums, Chicago: Aldine, 1962, pp.274-286.

11. See Blau, The Dynamics of Bureaucracy (2d ed.),University of Chicago Press, 1963, pp. 129-131.

12. See Homans, op. cit., p. 72.

13. Kenneth E. Boulding, Economic Analysis (3d ed.), NewYork: Harper 1955, p. 552.


Back tothe Syllabus