Learning to choose a commodity-money: Carl Menger's theory of imitation and the search monetary framework

Abstract

This paper studies Carl Menger's theory of the emergence of a commodity money. We propose an interpretation of Menger's learning by imitation process based on the search theoretical formal framework. We show that there exists a tension between the importance of intrinsic properties of commodities and the pure conventional self-fulfilling expectations of agents. This confirms the role of imitation in the emergence of monetary equilibria in search theory. We conclude that Menger's approach may support the idea that the fundamental property of a commodity-money (namely its great liquidity) is the result of its emergence process and not necessarily of its original intrinsic properties.

Keywords:

• Carl Menger
• search
• money
• evolution
• imitation

Notes

* I would like to thank Carlo Benetti, Jean Cartelier, Jimena Hurtado and the participants to the Ph. D. Seminar held at the VIth Annual Conference of the European Society of History of Economic Thought, University of Crete, Rethymno, Crete, 14 March, 2002. Special thanks go to Professor Eric B. Streissler for his insightful remarks. The final version of this article greatly owes to the suggestions of the EJHET's two anonymous referees. It goes without saying that all remaining errors are mine. The author is research fellow at Paris X University, where he is pursuing a doctoral program in economics. He is a member of the PHARE research centre.

See Streissler (2002: 20 – 22).

The first article in this line is Trejos and Wright (1995). From this article on most scholars have adopted and improved this approach.

Menger's main work on money, Geld has been recently translated and published in English by Leland B. Yeager and Monika Streissler (in Latzer and Schmitz 2002). Other important works by Menger on money are: Menger (1892a) many references to money in Menger (1963 [1883]); the chapter VIII on money in Menger (1976 [1871]) and Menger (1892b) an article in French concerning money as a measure of value.

Even if Menger constantly refers to historical facts, most of his arguments are purely theoretical. In fact, he affirms that his theory of the origin of money does not preclude the possibility of a different process of emergence of a general medium of exchange, in particular by the action of a state. He is mainly interested in establishing a logical theory of a spontaneous emergence of money rather than an empirically founded explanation. See in particular Menger (2002 [1909]: 33) (pages 16 to 17 in the original text in German) and footnote 12 pages 91 to 92.

‘To be sure, the difficulty often emphasized, namely, that with the prevalence of barter the person offering a commodity is unlikely to find those persons whose commodity he needs and to be found by the persons who need his commodity, had already been overcome by the emergence of markets wherever trade has reached a considerable extent and importance and could count on the requisite legal security. As a rule, the commodities offered on barter markets, …, tend to be arranged so expediently that every market participant will just as easily find those who are offering the commodities that he wants as he can easily be sought out and found by those who want the commodities he is offering: therefore, the essential difficulty of barter is not the meeting of the contracting parties' (Menger 2002 [1909]: 27; pages 5 – 6 in the original text in German).

‘The theory of money necessarily presupposes a theory of the saleableness of goods. If we grasp this, we shall be able to understand how the almost unlimited saleableness of money is only a special case, – presenting only a difference of degree – of a generic phenomenon of economic life – namely, the difference in the saleableness of commodities in general’ (Menger 1892a: 243).

Klein and Selgin formalize the medium of exchange's selection process as a Polya urn computer simulation. Saleability of goods is represented by the number of beads of a particular colour in an urn. Agents draw randomly, one after another, a bead from the urn and they returned the ‘selected’ bead plus another of the same colour. After a complete tour, all initial beads are removed and another identic ‘selection’ turn begins. The entire process finished when all beads are the same colour. This is a steady-state situation. Within this framework the probability to choose a particular object among a set of different objects is not affected by the lower or greater success of using this medium of exchange in the market. The actual network externality of money is thus not captured by this simulation because there is no social interaction considered here. Everything happens as if agents imitate the most frequent strategy without evaluating its relative success in exchange.

The (implicit) description of market exchange in general equilibrium theory exhibits a substantial amount of as-if centralization, certainly too much to permit a role for money. Alternatively put, the Walrasian model of exchange is not much concerned with how commodities are exchanged (Ostroy 1989: 191). Indeed, the absence of a decentralized exchange theory constitutes the major difference between modern competitive general equilibrium models and Menger. The centralized coordination of transactions, in charge of either an auctioneer or any other ‘fiction’ such as a ‘clearing house’, rules out any possibility of analysing the role of a general medium of exchange: No theory of money is offered here, and it is assumed that the economy works without the help of a good serving as medium of exchange (Debreu 1959: 28).

Ostroy and Starr (1974) is a classical article on the description and treatment of these problems.

Rupert et al. (2001) contains a very complete review of the monetary search literature. This article proposes also a generalized framework for the search theoretical approach to monetary economics.

A fiat money is an object, differentiated from commodities, which nobody wants to consume.

A classifier system is a list of condition-action statements and an associated real number for each statement. This real number is called the strength of the classifier (i.e. each condition-action statement). This system allows to select winner strategies between simulation periods. The special characteristic of these systems is that they modify both consumption and exchange strategies, while consumption conditions the initial state of agent's holdings in the next period and so the possible exchange strategies. (See Marimon et al. 1990: 334 – 42.)

See Mailath (1992) for a detailed presentation of the application of replicator dynamics in economics.

According to bounded rationality hypothesis and the type of evolutionary dynamics of the model, equilibria are characterized as fixed points of a strategy selection dynamics. Whenever a particular strategy is adopted by the entire population, the selection dynamics finds a rest point. A possible equilibrium situation is thus a set of fixed points for every population that represent a set of pure strategies for the underlying game.

In terms of evolutionary game theory, when this propensity to imitate ‘successful’ strategies is positive for each type, it can be said that the imitation dynamics follows a Darwinian evolution.

Additional information

Notes on contributors

Andrés Alvarez

* I would like to thank Carlo Benetti, Jean Cartelier, Jimena Hurtado and the participants to the Ph. D. Seminar held at the VIth Annual Conference of the European Society of History of Economic Thought, University of Crete, Rethymno, Crete, 14 March, 2002. Special thanks go to Professor Eric B. Streissler for his insightful remarks. The final version of this article greatly owes to the suggestions of the EJHET's two anonymous referees. It goes without saying that all remaining errors are mine. The author is research fellow at Paris X University, where he is pursuing a doctoral program in economics. He is a member of the PHARE research centre.