The Bazaar Economy: 
Information and Search in Peasant Marketing




There have been a number of points at which anthropology and economics have come to confront one another over the last several decades-development theory; preindustrial history; colonial domination. Here I want to discuss another where the interchange between the two disciplines may grow even more intimate; one where they may come actually to contribute to each other rather than, as has often been the case, skimming off the other's more generalized ideas and misapplying them. This is the study of peasant market systems, or what I will call bazaar economies.


There has been by now a long tradition of peasant market studies in anthropology. Much of it has been merely descriptiveinductivism gone berserk. That part which has had analytical interests has tended to divide itself into two approaches. Either the bazaar is seen as the nearest real world institution to the purely competitive market of neoclassical economics- "penny capitalism"; or it is regarded as an institution so embedded in its sociocultural context as to escape the reach of modern economic analysis altogether. These contrasting approaches have formed the poles of an extended debate between economic anthropologists designated "formalists" and those designated "substantivists," a debate that has now rather staled for all but the most persevering.


Some recent developments in economic theory having to do with the role of information, communication, and knowledge in exchange processes (see Michael Spence; George Stigler; Kenneth Arrow; George Akerlof; Albert Rees) promise to mute this formalism-substantivism contrast. Not only do they provide us with an analytic framework more suitable to understanding how bazaars work than do models of pure competition; they also allow the incorporation of sociocultural factors into the body of discussion rather than relegating them to the status of boundary matters. In addition, their actual use on empirical cases outside the modern "developed" context may serve to demonstrate that they have more serious implications for standard economic theory and are less easily assimilable to received paradigms than at least some of their proponents might imagine. If this is so, then the interaction of anthropology and economics may come for once to be more than an exchange of exotic facts for parochial concepts and develop into a reciprocally seditious endeavor useful to both.





The bazaar economy upon which my discussion is based is that of a town and countryside region at the foot of the Middle Atlas in Morocco I have been studying since the mid-1960's. (During the 1950's, I studied similar economies in Indonesia. See the author, 1963.) Walled, ethnically heterogeneous, and quite traditional, the town is called Sefrou, as is the region, and it has been there for a millenium. Once an important caravan stop on the route south from Fez to the Sahara, it has been, for about a century, a thriving market center of 15,000-30,000 people.


There are two sorts of bazaar there: 1) a permanent one, consisting of the trading quarters of the old town; 2) a periodic one, which meets at various spots--here for rugs, there for grain--outside the walls on Thursdays, as part of a very complex regional cycle involving various other market places and the other days of the week. The two sorts of bazaar are distinct but their boundaries are quite permeable, so that individuals move freely between them, and they operate on broadly the same principles. The empirical situation is extremely complex--there are more than 600 shops representing about forty distinct commercial trades and nearly 300 workshops representing about thirty crafts--and on Thursdays the town population probably doubles. That the bazaar is an important local institution is beyond doubt: two-thirds of the town's labor force is employed there.


Empirical detail aside (a full-scale study by the author is in press). the bazaar is more than another demonstration of the truth that, under whatever skies, men prefer to buy cheap and sell dear. It is a distinctive system of social relationships centering around the production and consumption of goods and services--that is. a particular kind of economy, and it deserves analysis as such. Like an "industrial economy" or a "primitive economy," from both of which it markedly differs, a "bazaar economy" manifests its general processes in particular forms, and in so doing reveals aspects of those processes which alter our conception of their nature. Bazaar. that Persian word of uncertain origin which has come to stand in English for the oriental market, becomes. like the word market itself. as much an analytic idea as the name of an institution, and the study of it, like that of the market, as much a theoretical as a descriptive enterprise.





Considered as a variety of economic system, the bazaar shows a number of distinctive characteristics. Its distinction lies less in the processes which operate and more in the way those processes are shaped into a coherent form. The usual maxims apply here as elsewhere: sellers seek maximum profit. consumers maximum utility; price relates supply and demand; factor proportions reflect factor costs. However, the principles governing the organization of commercial life are less derivative from such truisms than one might imagine from reading standard economic textbooks, where the passage from axioms to actualities tends to be rather nonchalantly traversed. It is those principlesmatters less of utility balances than of information flows-that give the bazaar its particular character and general interest.


To start with a dictum: in the bazaar information is poor, scarce. maldistributed. inefficiently communicated. and intensely valued. Neither the rich concreteness or reliable knowledge that the ritualized character of non market economies makes possible, nor the elaborate mechanisms for information generation and transfer upon which industrial ones depend. are found in the bazaar: neither ceremonial distribution nor advertising: neither prescribed exchange partners nor product standardization. The level of ignorance about everything from product quality and going prices to market possibilities and production costs is very high. and much of the way in which the bazaar functions can be interpreted as an attempt to reduce such ignorance for someone, increase it for someone, or defend someone against it.





These ignorances mentioned above are kiloit'll ignorances, not simply matters concerning which information is lacking. Bazaar participants realize the difficulty in knowing if a cow is sound or its price right, and they realize also that it is impossible to prosper without knowing. The search for information one lacks and the protection of information one has is the name of the game. Capital. skill, and industriousness play. along with luck and privilege, as important a role in the bazaar as they do in any economic system. They do so less by increasing efficiency or improving products than by securing for their possessor an advantaged place in an enormously complicated, poorly articulated, and extremely noisy communication network.


The institutional peculiarities of the bazaar thus seem less like mere accidents of custom and more like connected elements of a system. An extreme division of labor and localization of markets. heterogeneity of products and intensive price bargaining, fractionalization of transactions and stable c1ientship ties between buyers and sellers, itinerant trading and extensive traditionalization of occupation in ascriptive terms--these things do not just co-occur, they imply one another.


The search for information--laborious, uncertain, complex, and irregular--is the central experience of life in the bazaar. Every aspect of the bazaar economy reflects the fact that the primary problem facing its participants (that is, "bazaaris") is not balancing options but finding out what they are.





Information search, thus, is the really advanced art in the bazaar. a matter upon which everything turns. The main energies of the bazaari are directed toward combing the bazaar for usable signs, clues as to how particular matters at the immediate moment specifically stand. The matters explored may comprise everything from the industriousness of a prospective coworker to the supply situation in agricultural products. But the most persistent concerns are with price and quality of goods. The centrality of exchange skills (rather than production or managerial ones) puts a tremendous emphasis on knowing what particular things are actually selling for and what sorts of things they precisely are.


The elements of bazaar institutional structure can be seen in terms of the degree to which they either render search a difficult and costly enterprise, or facilitate it and bring its costs within practical limits. Not that all those elements line up neatly on one or another side of the ledger. The bulk have effects in both directions, for bazaaris are as interested in making search fruitless for others as they are in making it effectual for themselves. The desire to know what is really occurring is matched with the desire to deal with people who don't but imagine that they do. The structures enabling search and those casting obstructions in its path are thoroughly intertwined.


Let me turn, then, to the two most import ant search procedures as such: clientelization and bargaining.





Clientelization is the tendency, marked in Sefrou, for repetitive purchasers of particular goods and services to establish continuing relationships with particular purveyors of them, rather than search widely through the market at each occasion of need. The apparent Brownian motion of randomly colliding bazaaris conceals a resilient pattern of informal personal connections. Whether or not "buyers and sellers, blindfolded by a lack of knowledge simply grop[ing] about until they bump into one another" (S. Cohen, quoted in Rees, p. 110), is, as has been proposed, a reasonable description of modern labor markets, it certainly is not of the bazaar. Its buyers and sellers, moving along the grooved channels c1ientelization lays down, find their way again and again to the same adversaries.


"Adversaries" is the word, for clientship relations are not dependency relations, but competitive ones. Clientship is symmetrical, egalitarian, and oppositional. There are no "patrons" in the master and man sense here. Whatever the relative power, wealth, knowledge, skill, or status of the participants--and it can be markedly uneven--c1ientship is a reciprocal matter, and the butcher or wool seller is tied to his regular customer in the same terms as he to them. By partitioning the bazaar crowd into those who are genuine candidates for his attention and those who are merely theoretically such, c1ientelization reduces search to manageable proportions and transforms a diffuse mob into a stable collection of familiar antagonists. The use of repetitive exchange between acquainted partners to limit the costs of search is a practical consequence of the overall institutional structure of the bazaar and an element within that structure.


First, there is a high degree of spatial localization and "ethnic" specialization of trade in the bazaar which simplifies the process of finding clients considerably and stabilizes its achievements. If one wants a kaftan or a mule pack made, one knows where, how, and for what sort of person to look. And, since individuals do not move easily from one line of work or one place to another, once you have found a particular bazaari in whom you have faith and who has faith in you. he is going to be there for awhile. One is not constantly faced with the necessity to seek out new clients. Search is made accumulative.


Second. clientelization itself lends form to the bazaar for it further partitions it, and does so in directly informational terms. dividing it into overlapping subpopulations within which more rational estimates of the quality of information. and thus of the appropriate amount and type of search, can be made. Bazaaris are not projected. as for example tourists are, into foreign settings where everything from the degree of price dispersion and the provenance of goods to the stature of participants and the etiquette of contact are unknown. They operate in settings where they are very much at home.


Clientalization represents an actor-level attempt to counteract, and profit from. the system-level deficiencies of the bazaar as a communication network--its structural intricacy and irregularity, the absence of certain sorts of signaling systems and the undeveloped state of others. and the imprecision, scattering, and uneven distribution of knowledge concerning economic matters of fact--by improving the richness and reliability of information carried over elementary links within it.





The rationality of this effort. rendering the clientship relation dependable as a communication channel while its functional context remains unimproved. rests in turn on the presence within that relation of the sort of effective mechanism for information transfer that seems so lacking elsewhere. And as that relation is adversary. so is the mechanism: multidimensional intensive bargaining. The central paradox of bazaar exchange is that advantage stems from surrounding oneself with relatively superior communication links, links themselves forged in sharply antagonistic interaction in which information imbalances are the driving force and their exploitation the end.


Bazaar bargaining is an understudied topic (but see Ralph Cassady). a fact to which the undeveloped state of bargaining theory in economics contributes. Here I touch briefly on two points: the multidimensionality of such bargaining and its intensive nature.


First. multidimensionality: Though price setting is the most conspicuous aspect of bargaining. the bargaining spirit penetrates the whole of the confrontation. Quantity and/or quality may be manipulated while money price is held constant. credit arrangements can be adjusted. bulking or bulk breaking may conceal adjustments. and so on, to an astonishing range and level of detail. In a system where little is packaged or regulated. and everything is approximative. the possibilities for bargaining along non-monetary dimensions are enormous.


Second, intensiveness: I use "intensive" in the way introduced by Rees, where it signifies the exploration in depth of an offer already received, a search along the intensive margin, as contrasted to seeking additional offers. a search along the extensive. Rees describes the used car market as one in which intensive search is prominent as a result of the high heterogeneity of products (cars driven by little old ladies vs. taxicabs. etc.) as against the new car market. where products are considered homogeneous, and extensive search (getting new quotations from other dealers) predominates.


The prominence of intensive bargaining in the bazaar is thus a measure of the degree to which it is more like a used car market than a new car one: one in which the important information problems have to do with determining the realities of the particular case rather than the general distribution of comparable cases. Further. it is an expression of the fact that such a market rewards a "clinical" form of search (one which focuses on the diverging interests of concrete economic actors) more than it does a "survey" form (one which focuses on the general interplay of functionally defined economic categories). Search is primarily intensive because the sort of information one needs most cannot be acquired by asking a handful of index questions of a large number of people, but only by asking a large number of diagnostic questions of a handful of people. It is this kind of questioning, exploring nuances rather than canvassing populations, that bazaar bargaining represents.


This is not to say that extensive search plays no role in the bazaar; merely that it is ancillary to intensive. Sefrou bazaaris make a terminological distinction between bargaining to test the waters and bargaining to conclude an exchange, and tend to conduct the two in different places: the first with people with whom they have weak clientship ties, the second with people with whom they have firm ones. Extensive search tends to be desultory and to be considered an activity not worth large investments of time. (Fred Khuri reports that in the Rabat bazaar, bazaaris with shops located at the edge of the bazaar complain that such shops are "rich in bargaining but poor in selling," i.e. people survey as they pass, but do their real bargaining elsewhere.) From the point of view of search, the productive type of bargaining is that of the firmly clientelized buyer and seller exploring the dimensions of a particular, likely to be consummated transaction. Here, as elsewhere in the bazaar, everything rests finally on a personal confrontation between intimate antagonists.


The whole structure of bargaining is determined by this fact: that it is a communication channel evolved to serve the needs

of men at once coupled and opposed. The rules governing it are a response to a situation in which two persons on opposite sides of some exchange possibility are struggling both to make that possibility actual and to gain a slight advantage within it. Most bazaar "price negotiation" takes place to the right of the decimal point. But it is no less keen for that.





G. A. Akerlof, "The Market for 'Lemons': Quality, Uncertainty and the Market Mechanism," Quart. J. Econ., Aug. 1970, 84, 488-500.

Kenneth J. Arrow, The Limits of Organization, New Yark 1974.

R. Cassady, Jr., "Negotiated Price Making in Mexican Traditional Markets," Amer. Indigena, 1968,38,51-79.

Clifford Geertz, Peddlers and Princes, Chicago 1963.

----, "Suq: The Bazaar Economy in Sefrou," in Lawrence Rosen et aI., eds., Meaning and Order in Contemporary Morocco: Three Essays in Cultural Analysis, New York forthcoming.

F. Khuri, "The Etiquette of Bargaining in the Middle East," Amer. Anthropologist, July 1968, 70, 698-706.

A. Rees, "Information Networks in Labor Markets," in David M. Lamberton, ed., Economics of Information and Knowledge, Hammondsworth 1971, 109-18.

M. Spence, "Time and Communication in Economic and Social Interaction," Quart. J. Econ., Nov. 1973,87,651-60.

G. Stigler, "The Economics of Information," in David M. Lamberton, ed., Economics of Information and Knowledge, Hammondsworth 1971, 61-82.



(*) The Institute for Advanced Study.



The Bazaar Economy, in: The American Economic Review, Vol. 68, No. 2 (Suppl.: Papers and Proceedings of the Ninetieth Annual Meeting of the American Economic Association) (May, 1978), pp. 28-32.


online source:


JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

The American Economic Review  is currently published by American Economic Association. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at


Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

Using this text is also subject to the general HyperGeertz-Copyright-regulations based on Austrian copyright-law (2001), which - in short - allow a personal, nonprofit & educational (all must apply) use of material stored in data bases, including a restricted redistribution of such material, if this is also for nonprofit purposes and restricted to the scientific community (both must apply), and if full and accurate attribution to the author, original source and date of publication, web location(s) or originating list(s) is given ("fair-use-restriction"). Any other use transgressing this restriction is subject to a direct agreement between a subsequent user and the holder of the original copyright(s) as indicated by the source(s). HyperGeertz@WorldCatalogue cannot be held responsible for any neglection of these regulations and will impose such a responsibility on any unlawful user.

Each copy of any part of a  transmission of a HyperGeertz-Text must therefore contain this same copyright notice as it appears on the screen or printed page of such transmission, including any specific copyright notice as indicated above by the original copyright holder and/ or the previous online source(s).