• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.



Page history last edited by Michael J 12 years, 8 months ago

Erstelldatum 16.04.2007 09:17:00

Heiner Ganssmann

April 2007

Doing Money

The social construction of monetary systems.

Introduction: What this project is about, how it differs from existing literature (general theory of money, no respect for disciplinary boundaries, attempt to use a precise and consistent vocabulary). Architecture à la Simmel: conditions vs consequences of monetary systems.


1.The basics

a) it does not make sense to talk about markets without talking about money,

b) that it does not make sense to talk about money without talking about the state (in whatever rudimentary form).

“Monetary system” (MS)

refers to a social machine, not to be mixed up with traditional frequent answers to the “What is money”-questions, such as pointing out the functions of money or the properties of “money objects”.


Minimum properties of MS.

How many players do you need - You need at least three M users

What do they need to know - to be able to count, add and subtract, and keep a record

What are the rules of the game?

  • a time horizon that allows for open-ended playing
  • at least two kinds of players, those who create and those who use money
  • money objects (always a creature of authority?) must be produced as social artifacts
  • MS inputs are bids and offers backed by money and goods/services
  • To be reproducible, the system has to produce a sufficient share of its inputs as outputs
  • Prices are pairs of specified quantities of goods/services and numbers of monetary units. Prices signal conditions for property change.



There are some preliminary answers to such questions.

  • You need at least three M users who are able to count, add and subtract before any of the problems typically associated with the needs for money can arise (for example, you need more than two players to solve typical barter problems, to go beyond “yes/no” answers to transaction opportunities that are governed by the “double coincidence of wants”-condition; only starting with three players the question of consistency of exchange relations emerges).


  • You also need a time horizon that allows for open-ended playing (or, in other words, if you construct a system that is to end in equilibrium, the need for money disappears).
  • You need at least two kinds of players, those who create and those who use money.


Money objects (always a creature of authority?) must be produced as social artifacts before they can be used in markets. So a first set of questions concerns the production/creation of money objects. Not only who produces what but why? As to the “why”, we run into typical problems of functionalist explanations. Standard answers to “why money?” point to the functions of money: measure of value, means of exchange, store of value, etc. That makes sense if we consider functions as answers to needs. However, for the reference to needs to have explanatory value, they have to be “pre-monetary” needs. So one big question is: How do agents in a pre-monetary setting discover that something they do not know yet (and that we will call “money” in retrospect) is an answer to their problems? (Answers can be grouped according to big bang or gradualist solutions, etc.)........


Given money and money users,

  • MS inputs are bids and offers backed by money and goods/services;
  • MS outputs are prices and transactions.

Transactions result in new combinations of goods/services and money held by participants. In time, goods and services disappear in use and consumption, whereas money may be lost, burnt, buried, but in general it stays in circulation.


To be reproducible, the system has to produce a sufficient share of its inputs1 as outputs, so at least part of the MS transactions have to fulfill corresponding requirements. Prices are not just the result of transactions, but also govern them.


What are prices?

Pairs of specified quantities of goods/services and numbers of monetary units. What do prices do? Prices signal conditions for property change (they may be expected, demanded, offered or effective). Price signals in turn can be used to coordinate actions without a central coordinating institutions (=>Hayek´s story)



2 Second Part


Social conditions, presuppositions of monetary systems: How are MS constructed?

Basic answer: they are built out of social conventions answering needs for action coordination.


The emergence of common rules of action coordination (=conventions) can be taken to follow the evolutionary scheme of variation, selection and retention according to what “works”.


Conventions lead to some objects being recognized as M. They amount to rules of using M. Searle´s distinction of regulative and constitutive rules and the formula: “X counts as Y in C” capture the gist of the general problem. How do social institutions develop? Iterative system building à la Searle (--X1 counts as Y1 in C, leading to: X2(=Y1) counts as Y2 in C...). Iteration can only be reconstructed in a diachronic perspective.


Convention approach accounts for the formal aspect of building social institutions, the reliance on some kind of agreement, or, in Searle´s terms, on collective intentions.


MS can be seen as a social machine for coping with uncertainty in economic matters (“economic” being understood as having to do with material provisioning).


There is a background of uncertainty rooted in general properties of social systems (double contingency, indeterminacy (Hardin)), but there is additional uncertainty resulting from the structure of market economies, the way in which they combine a division of labor with private independent production (the DoL&PiP economy).


DoL&PiP Economy

The decisive characteristic here is the extent to which such an economy is regulated by ex-post coordination. Insofar as uncertainty results from the social world, it can be partially reduced by forms of action coordination. In general, M systems work as systems of action coordination, by establishing a metric that links vast networks of participants unknown to each other, bound by (formally) voluntary agreements to give or take M at each link, with normal M users not able to change the quantity of M objects in circulation (pair-wise transaction governed by a constant sum condition).



2.“Thin” and “thick” theories of money.

In contrast to normal economic theory with its concept of market frictions to be overcome with the help of money, Marx has emphasized the need for social recognition of private individual efforts in the DoL&PiP-economy and the role of M as the instrument of such recognition.


Thin theory: In neoclassical theories of M, it only has to carry a light burden. Characteristic assumptions and presumptions: Equilibrium vs. M. General idea: Markets have a spontaneous “natural” tendency to generate equilibrium and Pareto optimality. Money has no place in equilibrium. It is only a veil. It is neutral. To create a place for money, it has to be built in by assuming market frictions in a world constructed by previously assuming perfect information and no uncertainty, rational expectations, etc.


Outcome: Older arguments are about why it is socially preferable to use money (overcoming restriction in barter, saving transaction costs). Newer arguments (search theories, OLG models) are about why it is individually preferable to use M given that everybody else is using M (surprise!). Alternative proposition:

Markets and money are products of a co-evolution.

Markets cannot perform wonders, but whatever wonders they do, they only work with money as an uncertainty absorption device. Thick theories: Marx: M as a mechanism of social recognition. M as “universal equivalent”. Other functions of M. Commodity M and the emergence of credit. Weber: M and market struggle resulting in effective prices. Effective prices as the underpinning of economic calculation; economic calculation as the underpinning of the general process of (formal) rationalization.



For Marx,

M mediation was a sign of reification plus alienation (connected to and at the root of capitalist domination/exploitation, s. part 3). The questions that follow here are

a) whether modern societies and the corresponding economies –market or not-- can work without such mediation. Unlikely.

b) If not, under what conditions could money mediation be understood as in modern sociology: as merely a useful and normatively unobtrusive extension of action coordination, as a means to go beyond the action coordination achieved by language use?


Has M lost its bite as a hierarchical instrument of social control?

Both Marx and Weber would argue against this. Money has excess properties compared to language. M as an effective, mostly irresistible lever in social relations and transactions; M objects as things endowing their holder with social power because everybody needs money and it always flows away when used to buy what is needed.


Other questions

(À propos M object as things: what happens when cash disappears? When paperless credit is used: Trade-off between physical presence of M as cash and increasing institutional back-up required for more “abstract” money. Peculiarities of cash: two-party agreements are sufficient to accomplish trade; cash leaves no paper trail.)


4.The metric of money. How is it constructed? (Two proposals in theories: Micro first step: define the M unit, work outward from there. Macro: define the M aggregate, move to micro units, transactions from there). Because the metric is needed, the “working fiction of a monetary invariant” is a requirement for markets to be sustainable (Phillip Mirowski). How can it be established? Money and metric in anthropology (Helen Codere: cognitive prerequisites for developed M system: counting, measuring, record keeping=systems of numbers, amounts and measures, writing...). How can monetary invariant (stability of purchasing power) be accomplished? Combination of top-down and bottom-up forces results in effective prices (in Weber´s sense) that can then be used for cognitive monetary operations: accounting, calculation, strategic decision-making. Top down: Authority and money. Authority –with a view to the whole-- can declare something (X counts as Y in C) to be unit of account, legal tender, fiat currency. Monetary policy: regulating the money supply, reacting to crises (part 3). Further top-down interventions: Taxation, accounting regulation, default and bankruptcy laws. Monetary system participants are not a mass of atomistic individuals on equal footing. Rather, even abstracting from the issue of money creation, there is always (at least) one big player, the state (on national level) opposed to many smaller, private players. In any case, authority can fix standards, turn customs and conventions into laws, and has unique possibilities of control (fiscal and monetary policy) exercised by legislation, regulation and material intervention, taxing and spending. Bottom-up: Two sets of forces. Bids/offers in a competitive setting lead to fixing prices and ensuing trades. Observed effective prices become spontaneously emerging standards for all market participants. The “law” of one price. Arbitrage operations react to price inconsistencies and tend to eliminate them. (see appendix) The joint effect of top-down and bottom-up forces, if all goes well: the MS works, even if the monetary invariant is a fiction.



5.Nonetheless, the monetary invariant remains a fiction. How can M users take the fiction as real? Their own contribution consists of observing a “local” constant sum condition, or a conservation principle. If player i holds Mi and player j holds Mj, after their transaction, the sum of their money holdings will still be Mi + Mj, although the M will be distributed differently between them. The systemic contribution traditionally consisted of using M stuff that was in recognizably short supply by nature, like kauri shells or gold, etc. After commodity M: Social construction of M loses its constraining “anchors” in the natural world. Today, the degree of stability of M depends unambiguously on behavior of M authorities (but also on the absence of extreme market reactions, s. part 3). Authorities normally work against in- or deflation to maintain confidence in stability of MS. But nation states with their currencies operate in world markets. External forces: Exchange rate mechanisms, arbitrage and competition on currency markets as additional factors. Internally: Market regulation is used to prevent monopolistic/­oligopolistic pricing and ensure exposure to competition. But overall, we are talking about an increasingly complex, intransparent system, so how can people risk to use M? Money and trust. Shubik tries to explain evolution of money as a combination of minimizing both transaction costs and trust requirements. There seems to be a trade off: full value commodity M (if it ever existed) in spot markets has minimum trust requirements, to accept cashless credit as a means of payment requires maximum trust. However, with cash, we do not have to trust persons, rather we trust their money. Second, with paperless credit, we do not trust M in the form of physically present M objects, but trust in the working of the institutions constituting the credit system. So trust is not something between persons, it is trust in institutional arrangements. Given monetary history, maybe it is better to speak of “suspended distrust”. How far does it imply relying on extra market resources in action coordination (s. for example Inglehart´s world value survey on the correlation between (generalized) trust and GDP p.c.)?


3. Third Part: Consequences of money use (restricted to macro level).



Sharp inequality as a result of using M, as demonstrated by econophysics. Is there a trade-off between inequality and monetary stability?


Money and crisis: Stabilization and institutional learning via crises. Dynamics of monetary innovations.


Money as power: Capitalism. Start from Benetti/Cartelier, Shubik´s emphasis on the necessity of default/bankruptcy rules: What happens if players default? One answer: proletarization—without M, players lose the ability to initiate economic activity. Construction of a society-wide single exit situation. Only “money buys membership in industrial society” (Rainwater). Everybody has to get M. (How is the “silent force of economic relations” (Marx) emanating from this single exit situation related to trust? Perhaps most of us cannot help but trust in money...)






Aglietta, M., Cartelier, J. (1998), Ordre monétaire des économies de marché, in: Aglietta, M., Orléan, A., La Monnaie Souveraine, Paris, Odile Jacob. 129-157.


Benetti, C., Cartelier, J. (1980), Marchands, salariat et capitalistes, Paris: Maspero.


Codere, H. (1968), Money-Exchange Systems and a Theory of Money, in: Man, N.S., vol. 3, No. 4, 557-577.


Hellwig, M. F. (1993), The Challenge of Monetary Theory, in: European Economic Review 37, 215-242.


Krause, U. (1979), Geld und abstrakte Arbeit, Frankfurt a.M.


Lewis, D.K. (1969), Convention, Cambridge, Mass. (Harvard UP)


Mirowski, P. (1991), Postmodernism and the social theory of value, in: Journal of Post Keynesian Economics, 13, 4, 565-581.


Searle, J. (1997), Die Konstruktion der gesellschaftlichen Wirklichkeit, Reinbeck: Rowohlt.


Shubik, M. (1999), The theory of money and financial institutions, 2 vls., Cambridge: MIT press.








On arbitrage


In terms of my general framework, arbitrage introduces a new type of player into the market game. Arbitrageurs will observe price relations. If they see an opportunity of gains from pure trade, they will engage in trade. For example in the 18th century, they can buy glass beads in Europe, ship them to Africa, have them exchanged there for gold, let the gold be shipped back to Europe and sell it for more money then they spent for glass beads plus transaction costs. Normally, the attraction of high gains from pure trade will draw more players into this part of the game, leading to shifts in supply and demand. Most likely, if none of the players can control the respective market (in contrast to the monopolies in colonial trade), increased competition will lead to a decrease in the gains to be made. In that sense, "The suicidal stimulus of this gain would always make it disappear", as Schumpeter (1952:61) put it.


A price system would tend towards a system of equivalence relations2 if all possible gains from arbitrage tend to be eliminated. The argument is theoretically interesting, not least because it relies on the unintended consequences of actions, but apparently it is hard to prove formally (see Ellerman, D. 2000).


The self-elimination property of arbitrage could help to explain why the working fiction of a monetary invariant can be maintained: If markets have a built-in tendency towards consistent prices that do no longer allow for gains from trade, market players will have an easier way to “read” prices. There may be all sorts of factors and events triggering price changes, but at least there is only a low risk that endogenous market forces will be responsible for surprising price changes all the time. In other words, there is less uncertainty involved in taking prices as parametric—which is what agents will have to do with most prices when calculating their strategies.


With regard to money the arbitrage argument is, in sum: Money makes it easier/possible to recognize arbitrage possibilities. But because arbitrage possibilities are self-eliminating, the resulting price system3 will approach consistency properties that in turn make it more feasible to reach some rationality in the construction of market strategies. Such strategies will normally be built from calculations in which an important subset of money prices is taken as given.






David Ellerman, Towards an arbitrage interpretation of optimization theory (World Bank 2000)


Ulrich Krause, Geld und abstrakte Arbeit, Frankfurt a.M., Campus 1979.




1 The share that it cannot draw out of its environment.


2 For a careful discussion of the formal properties of equivalenc relations, cf. Krause 1979.


3 "Hence, if one wished to leave arbitrage operations aside and at the same time to generalize the equilibrium established for pairs of commodities in the market, it would be necessary to introduce the condition that the price of either one of any two commodities (chosen at random) expressed in terms of the other be equal to the ratio of the prices of each of these two commodities in terms of any third commodity."(Walras 1954:161)

Comments (0)

You don't have permission to comment on this page.