A critique of price theory with special reference to decentralized exchange

Abstract

Progress in the theoryof exchange has been conceived as a refinenent of equilibrium states, and the question of existence of equilibrium as a set of prices at which purely national trading plans are consistent has been confused with the different question of how plans based on nationally predetermined equilibrium prices can actually be carried out. Along the way, formal theory has been emptied of concern with potentially observable activities of transactors; and theoretical accounts of markets and exchange arrangements have been emptied of empirical content. The quenstions I raise are, first, how thw neowalrasian approach evolved and led to the omission of trading and markets, and second, how it constrained understanding of the execution of trades. The argument in this dissertation is as follows. In chapter II, it is shown that the program set forth by walras of providing an explanation for the convergence to equilibrium in competetive exchange is still today an unresolved question. Walras was unable to carry forward to the general model his view of the workings of free competition in markets where bargaining and trading take place in the course of the adjustment process. But if Walras settled matters upon the assumption of no-arbitrage, the neowalrasian formalization entirely lost sight of trading and pursued tâtonnement as if it described real adjustment in a competitive economy. Chapter III begins with a comparison of adjustments to equilibrium in Walras and Marshall, to conclude that both authors are basically in agreement, and that the separation to be established is with Arrow-Debreu. Next , a line of evoltuion in price theory from Marshall to Chamberlin and on to triffin is described; when the general equilibrium edifice was topped by the proofs of existence (and stability) by Arrow Debreu in the fifties, the Marshallian tradition in price theory had become engulfed in a purification by means of which general inderdependence does without markets. Decentralized trading arrangements decome thereby an oversight in theoretical analysis and this is why chapter IV opens with a discussion of whether a theory of markets and intermediation is available, and how the consideration of transaction costs by coase sidestepped the issue. Moreover, transaction costs have both been called to explain the emetgence and the impediment to the formation of markets. 'Markets' are a misnomer for 'prices' in the theory of general interdependence, and an unnticed confusion between exchange and markets ensures. A broad distinction between brokered and non-brokered markets is attempted and, next, decentralization is characterized in its dimensions, which regard information, the determination of the price signal, and logistics. In this perspective the Arrow-Debreu model of general exchange fails in all three aspects: fisrt, because convergence is not computable; second, decause for consistency of trading plans exchange rates are required to be common, given, and freely know; third, because only with centrally coordinated multilateral barter can the theory preclude bargaining by traders in attempts to transform planned trades into feasible executions. The consistency of informationally decentralized trading with predetermined equilibrium prices has often been questioned (e.g. Ostroy and starr[1990]), and analyses of execution have been attempted with unsatisfactory results. Direct barter at equilibrium prices hardly meets full execution under decentralized information, i.e.,it is not feasible. The introduction of intermediation or a medium of exchange (monetary exchange being a special case of indirect barter, and one of central interest here because it is potentially desentrslized) is shown to enable feasibility of execution. In spite of hardly fitting a world of costless exchange, the consideration of these facilitating devices helps expose the contradictions carried by attempts to dichotomize pricing and trading. Given these negative findings, the doubt arises whether and how the Walrasian program can be furthered to include and treat logistics of exchange. Were we to stay within the Walrasian program, twob possible solutions are discussed, but both face obstacles. One approach is to formalize trading in the process of bargaining, but the consideration of feasibility of desentralized execution raizes doubts about the adequacy of the estabilized notion of equilibrium ( as a notional state) as a basis for understanding trading. The other possible approach is to start with logistics of transactions in an informationally desentralized setting, without postulating equilibrium prices as preexistent to trade. Finally, chapter V offers concluding comments on Arrow's [1959] contribution to the theory of price adjustment that reinforce my earlier arguments showing that existing theories of price and quantity adjustment which attempt to dispense with 'the auctioneer' have been uncuccessful in defining alternative theoretical trading schemes that would satisfactorily describe trading phenomena in the decentralized nonbrokered markets that are a ubiquitous feature of every modern economy.