George Selgin versus David Graeber on the Origin of Money

The blogosphere has recently seen a new debate on the origin of money between George Selgin and David Graeber:

Ilana E. Strauss, “The Myth of the Barter Economy,” The Atlantic, February 26, 2016
http://www.theatlantic.com/business/archive/2016/02/barter-society-myth/471051/
Though not by Selgin or Graeber, I gather that this article set off the debate.

George Selgin, “The Myth of the Myth of Barter,” Alt-M Ideas for an Alternative Monetary Future, March 15, 2016.
http://www.alt-m.org/2016/03/15/myth-myth-barter/
This is Selgin’s original critique of Graeber.

George Selgin, “Graeber, Once More,” Alt-M Ideas for an Alternative Monetary Future, March 24, 2016
http://www.alt-m.org/2016/03/24/graeber-once-more/
After some debate on Twitter, Selgin posted this reply.

For those of you who don’t know, George Selgin is a monetary economist (whose research includes banking theory, monetary history, and free banking, among other things), and David Graeber is an anthropologist and author of the fascinating book Debt: The First 5,000 Years (2011).

To some extent, we are simply revisiting the debate that David Graeber had with Robert Murphy had in 2012.

I have intellectual respect for both George Selgin and David Graeber. For example, Selgin’s paper “Those Dishonest Goldsmiths” is an excellent refutation of the Rothbardian cult on their history of the origin of fractional reserve banking in England.

But, on this issue, I get the impression that people here are talking past one another. I offer this in the spirit of constructive and friendly criticism of both George Selgin and David Graeber, and hope it is taken in that way.

Let me begin with a positive assessment of Selgin’s analysis.

Selgin argues the following:

(1) that just because there is no modern anthropological evidence for pure barter economies, it does not necessarily follow that these things did not exist in the distant past before money was invented.

(2) Selgin complains:

“What I do deny, and vigorously, is anthropologist David Graeber’s claim that the existence of gift economies undermines ... ‘the entire discourse of economics.’”

(3) there is an irrational hatred of capitalism from some left-wing people, who want to blame slavery and imperialism on the use of money, and even paint “Adam Smith … as an enabler of slavery and imperialism.”

(4) that human societies can have quantified debts in terms of commodities owed without money.

(5) that, if we define value as subjective value, then money is not an objective measure of value, as the Austrians argue. It is also true that, generally speaking, during exchanges one person subjectively values the good he receives more highly than the good he parts with. Exchange is not an exchange of equivalents in that sense when we consider subjective value.

I can grant George Selgin all these points. In fact, I do. He is essentially correct on these points.

Selgin is a highly intelligent man, and he has scored hits against left-wing people here.

On (2), of course the refutation of simplistic models of money emerging from pure barter in neoclassical textbooks doesn’t refute all of neoclassical economics. Such a claim is absurd.

On (3), yes, it is the worst sort of left-wing irrationality to blame slavery and imperialism on money or capitalism.

On (5), I also appreciate Selgin’s complaints about the Aristotelian view of exchange:

“The idea that money is a ‘measure of value,’ like the related idea that exchanges are necessarily exchanges of equivalents, is among the hoariest of economic fallacies. It plays a prominent part in Aristotle’s economics — and, not coincidentally, in Aristotle’s condemnation of all sorts of ‘capitalist’ activity. Smith himself, in subscribing to a modified labor theory of value, was unable to break free of it. It is more than a little ironic that Graeber, in flinging all sorts of undeserved criticism at Smith, cleaves to him when it comes to his one indisputable mistake.

The notion that money is a ‘measure of value’ is but a particular instance — albeit one that has managed to linger on in some economics textbooks — of the mistaken belief that economic exchanges are exchanges of equivalents.”
George Selgin, “The Myth of the Myth of Barter,” Alt-M Ideas for an Alternative Monetary Future, March 15, 2016.

One could also add: Marx took over that same idea of exchange as necessarily exchanges of equivalents from the Classical economists and Aristotle, and it is a major reason for the train wreck of illogic and flawed argument in Chapter 1 of volume 1 of Capital.

We get to the crux of Selgin’s counterargument here:

“My concern, though, isn’t with Graeber’s sweeping condemnation of modern economics, or of the economic arrangements for which modern economists are supposedly to blame. It’s with his particular claim that there’s no merit in Smith’s account of the origin of money, or in the later accounts of other economists, including Carl Menger. Despite what these economists have argued, money couldn’t have grown out of barter, Graeber insists, because the ‘fabled land of barter’ that these accounts posit never existed.”
George Selgin, “The Myth of the Myth of Barter,” Alt-M Ideas for an Alternative Monetary Future, March 15, 2016.

Unfortunately, it seems to me that this misunderstands Graeber’s theory.

Graeber, as far as I am aware (I could be wrong), does not say that there is absolutely no merit of any kind to the views of Adam Smith or Carl Menger.

Graeber does not deny that money in some historical circumstances can emerge from barter, especially in long distance trade.

Right on p. 75 of Debt: The First 5,000 Years (2011), Graeber says:

“Throughout most of history, even where we do find elaborate markets, we also find a complex jumble of different sorts of currency. Some of these may have originally emerged from barter between foreigners: the cacao money of Mesoamerica and the salt money of Ethiopia are frequently cited examples. Other arose from credit systems, or from arguments over what sort of goods should be acceptable to pay taxes or other debts. Such questions were often matters of endless contestation.” (Graeber 2011: 75)

This statement requires that, yes, there is some merit to the theories of Smith and Menger, but suitably qualified.

The point is, however, that the Smith, Menger and neoclassical “barter origin” theory of money is not universally true and cannot be in its classical form a correct theory of the origin of money. It’s flawed.

What is needed is an eclectic theory that takes careful account of history and anthropology.

Moreover, even Selgin himself concludes that “notwithstanding the fact that credit is older than barter, Smith’s theory is, after all, not all that far removed from the truth.”

So if Selgin has asserted that

(1) the empirical evidence strongly suggests that gift exchange and debt–credit exchanges without money long preceded the invention of money in human history, and

(2) the classical and orthodox barter spot trade theory of the origin of money needs revision.

then it follows directly that Selgin has conceded something significant to his opponents.

Selgin states:

“So, how true is Graeber’s account, and just how fatal is it to the "fable" that economists like to tell? For answers, we need look no further than the evidence Graeber himself supplies. For on close inspection, that evidence itself suffices to show that, notwithstanding the fact that credit is older than barter, Smith's theory is, after all, not all that far removed from the truth.

A paradox? Nothing of the sort. The simple explanation is that, while subtle forms of credit or outright gift giving may suffice for affecting exchanges within tightly-knit communities, exchange within such communities hardly begins to take advantage of opportunities for specialization and division of labor that arise once one allows for trade, not just within such communities, but between them, that is, for trade between or among strangers. One need only recognize this simple truth to resuscitate Smith's theory from Graeber’s seemingly fatal blow. Simple forms of credit may come first; but such credit only goes so far, because it depends on a repeated interaction, and the trust that such interaction both allows and sustains.”
George Selgin, “The Myth of the Myth of Barter,” Alt-M Ideas for an Alternative Monetary Future, March 15, 2016.

At this point, Selgin cites with approval Graeber’s own views that barter trade, historically speaking, was probably far more important between ethnic groups and different communities, particularly in long distance trade.

Selgin continues:

“The question is, what did Smith really ‘imagine’? His story of the butcher and the baker notwithstanding, his reference to pastoral societies makes it perfectly evident that he understood the difference between conduct among ‘villagers’ and conduct among strangers. His theory of the origins of money ought to be understood accordingly. It is a theory of how, when opportunities for trade arise among strangers, bringing with them further scope for the division of labor, trade will be ‘choked and embarrassed’ if it must occur by means of barter, but will cease to be so once barter gives way to the employment of money. In portraying such cases as exceptions to the rule that ‘credit’ proceeds barter, Graeber simply fails to understand that such ‘exceptions’ are all that matters in assessing Smith’s theory.” ….

In short, a generous reading of Smith, far from making him out to be a right bungler when it comes to matters ethnographic, yields a relatively sophisticated view, according to which kinship and ‘credit’ first predominate, but then give way, as strangers meet, first to barter, but eventually to monetary exchange, which in turn allows for the growth of commerce, which ends up reducing the role of kinship and kin-based credit relationships.”
George Selgin, “The Myth of the Myth of Barter,” Alt-M Ideas for an Alternative Monetary Future, March 15, 2016.

But, in my view, this “generous reading of Smith” is really just reading into Smith things that aren’t there. Selgin has actually re-interpreted Smith from ideas in the Theory of Moral Sentiments, and – at the very least – one would have to admit that the explicit theory as presented in The Wealth of Nations needs revision.

Selgin then goes on to argue that Menger, in particular, was well aware of the importance of gift exchange and debt–credit exchanges in premodern human societies before money. Here there are good points. I agree Graeber is unfair to Menger. For example, Menger’s 1892 article “On the Origin of Money” (at least in its English translation) is mercifully free from mathematics, and clearly not filled with “mathematical equations.”

It is true that Menger on this issue and certainly on his other important economic and political views was far from the dogmatic Misesians and Rothbardian anarcho-capitalists who claim his legacy. Menger was a subtle thinker and made real contributions to economics, and was no extremist libertarian.

Here Selgin notes that “Menger understood perfectly well that ‘credit,’ in Graeber’s loose sense of the term, is older than either monetary exchange or barter,” and cites a passage in Menger’s article “Geld” in the Handwörterbuch der Staatswissenschaften (vol. 3), 1892. pp. 730–757. But Selgin cites the 2002 translation of L. B. Yeager and M. Streissler of that article which is actually based on the 3rd edition of 1909, not the original article of 1892.

I have a minor quibble here. I know it is a pedantic point, but did Menger’s views on this evolve over time and move away from his 1890s theory? (see Appendix 1 below).

At any rate, let us put this aside, and move to a much more important point.

If we read Menger’s classical article of 1892 in its English translation by C. A. Foley published in the Economic Journal, we find an interesting qualification that Menger makes to his theory:

“It is not impossible for media of exchange, serving as they do the commonweal in the most emphatic sense of the word, to be instituted also by way of legislation, like other social institutions. But this is neither the only, nor the primary mode in which money has taken its origin.” (Menger 1892: 250).

That leaves open the possibility that money can be instituted by a government, though Menger, as far as I can see, clearly did not think that this was the primary and earliest manner by which money had emerged.

But, at the same time, it is also very difficult to see how this isn’t a substantive concession to chartalism by Menger.

That passage by Menger also shows us the divide between Menger’s nuanced view of the origins of money and the stridency of Rothbard:

“[sc. Mises’s] Regression Theorem also shows that money, in any society, can only become established by a market process emerging from barter. Money cannot be established by a social contract, by government imposition, or by artificial schemes proposed by economists.” (Rothbard 2009: 61).

However, Menger remained an advocate of the barter spot trade theory of money’s origins in ancient times, although he was willing to concede what later Austrians have emphatically denied.

But let us return to Selgin’s defence of Menger.

Selgin is essentially arguing the following:

(1) that Menger understood that gift exchange and debt–credit exchanges existed in premodern human societies before money.

(2) that the pure barter economy model from which money emerges is therefore an oversimplification.

(3) that the standard Austrian and neoclassical view on the origins of money can be easily rescued by arguing that it was trade and barter exchange between different communities that was the fundamental origin of money: money emerged in inter-communal barter spot trade as the most saleable good (or goods) became the medium of exchange (as in Menger 1892: 249, though Menger seems to conceive of this happening within communities), and over time money then spread more and more to exchanges within tribes or communities, and internal economies became monetised.

However, I find profound problems with this defence.

For one thing, this is not the mainstream neoclassical view, which does indeed proclaim the naïve theory of the emergence of money from barter spot trade as a universal theory and takes insufficient account of the importance of debt–credit relationships.

Secondly, it is a rather curious state of affairs to see that Selgin is actually proposing a theory that seems very much like the one proposed by Karl Marx in Chapter 2 of volume 1 of Capital:

“Objects in themselves are external to man, and consequently alienable by him. In order that this alienation may be reciprocal, it is only necessary for men, by a tacit understanding, to treat each other as private owners of those alienable objects, and by implication as independent individuals. But such a state of reciprocal independence has no existence in a primitive society based on property in common, whether such a society takes the form of a patriarchal family, an ancient Indian community, or a Peruvian Inca State.

The exchange of commodities, therefore, first begins on the boundaries of such communities, at their points of contact with other similar communities, or with members of the latter. So soon, however, as products once become commodities in the external relations of a community, they also, by reaction, become so in its internal intercourse. The proportions in which they are exchangeable are at first quite a matter of chance. What makes them exchangeable is the mutual desire of their owners to alienate them. Meantime the need for foreign objects of utility gradually establishes itself. The constant repetition of exchange makes it a normal social act. In the course of time, therefore, some portion at least of the products of labour must be produced with a special view to exchange. ....

The necessity for a value-form grows with the increasing number and variety of the commodities exchanged. The problem and the means of solution arise simultaneously. Commodity-owners never equate their own commodities to those of others, and exchange them on a large scale, without different kinds of commodities belonging to different owners being exchangeable for, and equated as values to, one and the same special article. Such last-mentioned article, by becoming the equivalent of various other commodities, acquires at once, though within narrow limits, the character of a general social equivalent. This character comes and goes with the momentary social acts that called it into life. In turns and transiently it attaches itself first to this and then to that commodity. But with the development of exchange it fixes itself firmly and exclusively to particular sorts of commodities, and becomes crystallised by assuming the money-form. The particular kind of commodity to which it sticks is at first a matter of accident. Nevertheless there are two circumstances whose influence is decisive. The money-form attaches itself either to the most important articles of exchange from outside, and these in fact are primitive and natural forms in which the exchange-value of home products finds expression; or else it attaches itself to the object of utility that forms, like cattle, the chief portion of indigenous alienable wealth. Nomad races are the first to develop the money-form, because all their worldly goods consist of movable objects and are therefore directly alienable; and because their mode of life, by continually bringing them into contact with foreign communities, solicits the exchange of products.” (Marx 1906: 99–101).

“An adequate form of manifestation of value, a fit embodiment of abstract, undifferentiated, and therefore equal human labour, that material alone can be whose every sample exhibits the same uniform qualities. On the other hand, since the difference between the magnitudes of value is purely quantitative, the money commodity must be susceptible of merely quantitative differences, must therefore be divisible at will, and equally capable of being re-united. Gold and silver possess these properties by nature.” (Marx 1906: 102).

Of course, it is not exactly the same. But, if we just ditch Marx’s mystical labour theory of value nonsense, we actually have a theory here not far different from the one Selgin proposes.

According to Marx, money emerges by necessity from barter exchange of commodities between different communities (Marx 1990: 181–182). Money must be a uniform, portable, fungible and divisible commodity. It is implied that money then spread into communities and it monetised exchanges within tribes or communities.

Something like this was probably also the view proposed by the German Historical School economists of the 19th and early 20th centuries such as Max Weber (1978: 673–674) and Karl Bücher (1901), who argued that money emerged from barter between different societies, not within societies (Karl Polanyi may also have held a position close to this).

But, once again, even if we adopt this revised theory of money’s origins, modern anthropology and history suggest that it has serious problems and is in need of revision.

Our starting point is not – repeat not – that there is nothing of value in Smith’s or Menger’s theories, but (1) that it is not a universally applicable theory, and (2) that there are other important ways by which money can emerge.

Even Graeber does not deny that money in some historical circumstances can emerge from barter between strangers, especially in long distance trade (Graeber 2011: 75).

Now Selgin already admits that primitive money-less societies are frequently dominated by debt/credit transactions, or “gift exchange,” not by barter spot trades.

I assume he also seems to accept that even in cases where goods exchange for goods in spot trades, social relations can complicate matters considerably, and historically barter seems to have been prevalent between one community and another, or, that is to say, between people who were strangers and where relationships were implicitly or explicitly hostile (as in Graeber 2011: 29–30).

While a non-enumerated system of debts/credits or gift exchange might not give rise to money, there is clearly a role for debt in the history of money (Graeber 2011: 40). In the real world, gift exchange and debt/credit arrangements existed long before money, and societies could develop an effective system of exchange in which debt/credit or gift exchange transactions were the predominant system (Graeber 2011: 40).

But, under the theory that Selgin proposes, doesn’t this mean that societies that had no interest in large-scale trade with foreigners had little reason to develop money? Wouldn’t it also imply that the double coincidence of wants problem would be largely overcome in such an early tribal or isolated society?

More likely, such communities would sometimes develop what anthropologists call “non-commercial money” or “ceremonial money,” which is non-commercial in the sense that it is not used for everyday purchases of goods and services, or only rarely for such ordinary goods. It is thus non-commercial in the sense that it is not a universal medium of exchange. The purpose of such non-commercial money is social (see also here).

Graeber calls non-commercial money “primitive monies” and gives examples such as the shell money in the Americas or Papua New Guinea, cattle money in Africa, bead money, feather money, and so on. These are rarely used to buy everyday items in the societies that use them. Instead, they are employed in social relations like marriages and to settle disputes (Graeber 2011: 60).

At this point, we should consider the views of older anthropologists. In 1949, A. H. Quiggin published A Survey of Primitive Money: The Beginnings of Currency (London).

The summary of the author’s views on the origin of money is worth quoting:

“Writers on the origin of the use of money often start with a consideration of barter and its inconveniences. From ‘silent trade’ (a primitive though abnormal form of barter) they trace the evolution of trading and money side by side, relying mainly on literary evidence for probing into the past.

This study relies mainly on the tangible evidence of the actual types of primitive money or money-substitutes used by ‘unrisen’ people and others all over the world, and is concerned with the purposes, when discoverable, for which they were used. The evidence suggests that barter in its usual sense of exchange of commodities was not the main factor in the evolution of money. The objects commonly exchanged in barter do not develop naturally into money and the more important objects used as money seldom appear in ordinary everyday barter. Moreover, the inconveniences of barter do not disturb simple societies. The variety of material and the complexities of uncivilized attitudes towards money preclude generalizations, but the evidence appears to support the following line of argument.

In the beginning Man lived in self-supporting and self-contained groups. Except in an area where provisions are unlimited, a society depending on hunting and food-gathering for its subsistence is necessarily unsociable and ‘has no truck’ with its neighbours. Early exchanges were in the way of present-giving, and were expressions of friendship with no ulterior economic purpose, although the latter – an expectation of an adequate or even improved return – cannot be excluded from human dealings. Present-giving or gift-exchange, seen in simple forms in the Andamans, Torres Straits or New Zealand, may develop into elaborate ceremonial as in Fiji or the North-West of America, but remain distinct from trading with money.

Barter develops between areas of contrasted produce, such as coastal and inland, forested and open country. We see the barter of fish or shells for vegetables, game for bananas, &c., in Melanesia or the Congo, and the establishment of regular markets. Trading voyages such as those of Torres Straits and New Guinea take us a stage further by the introduction of conventional presents. But so far there is no need for any medium of exchange such as is commonly described as money.

This is the state of affairs over about half the world at the present day. Barter suffices for most of the natives of Australia, New Zealand and the islands of the Pacific, and for the less-advanced peoples of Africa, Asia and the Americas, where native economy is not upset by the trader and the missionary.

The use of a conventional medium of exchange, originally ‘full-bodied’ but developing into ‘token’ money, is first noted in the almost universal customs of ‘bride-price’ and wergeld. When sister-exchange is not practicable, some other value must be substituted; where life for life is not demanded, some equivalent must be found. The history of ‘bride-price’ and wergeld (which has yet to be written) shows how formal the customary gifts become, fitted to definite scales of value. It is not without significance that in any collection of primitive currency the majority of the items are described as ‘used in bride-price’.

When once a system of conventional gifts or payments with a definite scale of values has been established (and this is necessary for ‘bride-price’ and for wergeld) the first steps are taken in the evolution of money. It develops thereafter in response to human needs into the accepted medium of exchange. Nutzgeld [viz. “useful money” or useful objects in exchange – LK] still remains Nutzgeld. Cattle may constitute wealth and form a standard of value. They cannot, strictly speaking, be called money. Money, to be generally acceptable, needs more convenient material and finds the four essential qualities (portable, divisible, durable, recognizable) in shells, beads or metals. Two further qualities have been shown to be necessary, one geographical and one more difficult to define.

The objects that come to be used as money are mainly non-local, or if local are the product of a special area or a special class; and they have prestige or essential virtue, religious or magical. Cowries and beads, most universal of all forms of primitive money, have magical as well as monetary value and still hold their own over a large part of the world, though everywhere disappearing now with the advent of the trader and trade tobacco. Metals best illustrate the transition from ‘full-bodied’ to ‘token-’ money. The spears and hoes of Africa, the knives and spades of China, and the spits of Argos are familiar examples. The tools may become amorphous and valued according to their weight in metal, or survive as attenuated imitations of their former selves. Metal, whether gold, silver, copper, iron or tin, is everywhere useful and everywhere valued, and estimated by size, shape or weight. Ingots are preliminary stepping-stones to coins. Ingots, as lumps or bars, develop in response to local needs or whims in special forms, such as manillas, Katanga crosses and Kissi pennies, Malay hats and Siamese bullets, or our own currency bars and ‘ring-money’.

To us, looking backward, the next step appears obvious and inevitable, but it was only in rare spots (possibly only in one rare spot) in the Old World that the final stage was reached, and definite weights of metal, rounded, flattened and stamped, can be called coins. Here the study of primitive money comes to an end.” (Quiggin 1949: 321–322).

It is depressing really how good empirical research like this is often forgotten and ignored in neoclassical economics, and there are two fundamental points to be taken from Quiggin’s analysis:

(1) many primitive people were able to exist without the modern developed form of money, and even where it arose in trade between foreigners it seems to have been largely limited to that sphere.

(2) the origin of money within a society can also be linked to social customs like bride-price and wergild (“man-money” or compensation for murder).

These conclusions seem to have been widely held by early 20th century anthropologists. Paul Einzig, for example, using the anthropological work of his day, argued a long time ago that “money first developed to serve matrimonial, political or religious payments was only later adopted gradually for commercial purposes” (Einzig 1948: 984).

So let us now focus on (2). I rely on Philip Grierson’s The Origins of Money (1977) (see here on that book).

As we have seen, in primitive societies, there often arises non-commercial money or ceremonial money. It is mainly used in social interactions, often formal social events such as marriage, wergild and bloodwealth payments, political relations (e.g., potlatch, moka), and fines and compensations (compensation for adultery, or for things lost), and may only be rarely used, if at all, for everyday purchases or commercial transactions (Grierson 1977: 15–16).

Sometimes this non-commercial money develops into money as a more general medium of exchange in some societies (though often enough it remains in its traditional role in others where no universal medium of exchange arises). So where does the origin of non-commercial money as a general medium of exchange and unit of account come from?

Grierson (1977: 19) proposes that the social custom of wergeld and wergeld-like customs are the answer. Wergeld (literally, “man-money”) is the paying of compensation for murder or other injuries and even theft of personal property. The object of wergild is to stop blood feuds and violence in revenge, and to provide an adequate measure of the things lost, as well as compensation. The objects that arose as standards of value as non-commercial money in tribal wergild payments did not necessarily arise by barter spot trade of the most saleable commodity (Grierson 1977: 21; 28–29).

Often in tribal societies objects of high social status or conferring “prestige” or even thought to have magical power will function as non-commercial money. Thus such non-commercial money doesn’t necessarily arise by barter spot trade as the most saleable commodity.

Most interesting is the linguistic evidence from many societies which shows how the word for money arose etymologically from concepts related to wergild and debt. The English word “pay” comes via French payer from the Latin word pacare, meaning “to pacify,” “make peace with.” In certain societies, non-commercial money arose as a standard for measuring value related to wergild-like customs and possibly even things like bride-wealth, but did not necessarily develop into general purpose money/commercial money.

Compensation payments are made in various goods, such as cattle, bondmaids, and precious metal, but it is likely a common unit of account was developed to simplify calculation of payments, which later spread to the wider community in economic transactions.

Where commercial money arose from non-commercial money, Grierson makes the following argument:

“… where societies have developed the notion of money as a general measure of value, it will, I believe, most often be found that a system of legal compensation for personal injuries, at once inviting mutual comparison and affecting every member of the community, lay behind them.” (Grierson 1977: 29).

So it is possible that in some societies commercial money arose from its previous role in systems of legal compensation.

But this clearly isn’t the end of the story either. What about ancient advanced civilisations?

The origin of money in ancient Mesopotamia appears to be in the development of an abstract money of account in the temple and palace institutions. These temples and palaces were institutions with large internal centrally planned economies, with complex weights and measurements for internal accounting of the products produced, received and distributed, and rent and interest owed. Many prices were set and administered in the money of account which developed from weight units. The two units of account were (1) the shekel of silver (which was equal to the monthly grain ration) and (2) barley (Hudson 2004). Silver money of account spread to the private economy mostly as a means of reckoning debts to temples and palaces (Hudson 2004: 115). But many ordinary people could pay in commodities, and the administered pricing system in terms of silver/grain that was developed in the temples was to assist in calculation of payments in kind.

In ancient Egypt, money appears as the most important unit of account called the deben (or uten), which was a unit of weight, originally equated to 92 (or 91) grams (Henry 2004: 92; there was also the unit called the khar for measuring wheat or barley, and 1 khar was equivalent to 2 deben of bronze). The measure of value for various goods was thus fixed weight units and historically no doubt these units arose from copper, silver, grain and gold. The unit of account system appears to have been developed by complex palace, government and temple institutions for internal accounting. While goods came to be denominated in terms of deben, in early times during the period of the Old Kingdom there were no physical deben changing hands in the private economy. That is to say, the deben did not function as a physical means of payment, and did not emerge by barter spot transactions as the most saleable medium of exchange. Even though goods and services were measured in a deben unit of account, payment was made in goods.

An important element in both these historical processes was the institution (or institutions) where surplus products were stored from taxation, tribute and gifts. These institutions dealt with complex flows, in and out, of goods: they were palace and temple complexes. Accounting systems, weight measures and writing are connected with just such institutions, and, importantly, some abstract unit of account arose by which to measure relative values of stored goods and inflows or outflows of goods. Since loans were also no doubt made from surplus products stored, repayment of loans in kind was facilitated by a unit of account.

In primitive human societies by the end of the Stone Age (c. 2.9 million years to 4500 BC), before the literate urban civilizations, agricultural communities developed where surpluses were stored, most probably held as a communal resource. The question of how the origins of a unit of account or measure of value could be related to the emergence of stored communal wealth is an interesting research question that deserves further study.

It is quite possible that silver and gold were used as non-commercial money or ceremonial money in ancient societies too before they became abstract units of account.

So even in developed ancient civilisations with cities we have huge collectivist temples and governments (sometimes presided over by god-kings) with large-scale economic and central planning.

So where are they in neoclassical theory and Selgin’s theory?

Given its very scarcity, it is unlikely that silver would have arisen as a unit of account and medium of exchange in, say, Mesopotamia from internal barter trade as the most saleable good precisely because there wasn’t enough of it.

More likely, it was imposed as a unit of account from above, by the collectivist temples and governments. The ancient government–temple complex may well have adopted it because it was important in foreign and international trade, but even here that requires yet more massive revision to Selgin’s theory, taking account of ancient institutions.

But it may well have been adopted merely because it was high-prestige object with a religious significance and from its role as a weight unit. That is, it arose partly because as a weight unit it was easily convertible into an abstract unit of account.

We could also examine the origin of money in ancient Greece, which I have done in detail in this post.

In short, before coins the state of affairs was this:

(1) cattle or oxen functioned as a largely abstract unit of account (but not a common medium of exchange) and

(2) iron spits might have been a very limited or weak medium of exchange.

In Homer’s epics which reflect the social reality of Dark Age and early Archaic period Greece, cattle or oxen are a type of unit of account, but the actual means of payment tend to be many other types of goods, not just cattle (Peacock 2011: 49–54).

Now one might argue that cattle did become the most common medium of exchange but then receded in importance to become a mere unit of account, but there are serious problems with this view. First, the emergence of a “cattle/ox” unit of account in Greece appears to be related to religion and cult offerings, not emergence of cattle as the most saleable good.

While cattle no doubt had value in market trades, they were an important sacrificial animal and offering to the gods. The Greeks appear to have developed a cattle or ox unit of account derived from the value these animals had in ritual and sacrifice (Seaford 2004: 61; the view was originally proposed by Laum 1924). Religious rituals and then temples had a preeminent place in ancient Greek society, and the city government’s major responsibility was to honour, appease and placate the gods by offering sacrifices. In this sense, the ancient Greek temple and city are not separate entities, but really one and the same. So the emergence of an ox unit of account can be seen as another state-based, institutional process affecting economic life.

Metal coinage in Lydia and Greece was an invention of the state, and the first Lydian coinage was struck in electrum and used to pay soldiers and mercenaries. This was most probably a high prestige object and perhaps even non-commercial money. At most, it was simply one of many goods used in conventional barter trades: there is no convincing evidence that it was the reigning medium of exchange (money) that had already emerged as the most saleable good in spot barter trades before it was adopted by the Lydian state.

Instead, electrum was a high prestige commodity selected by the state, standardised and used as a form of payment as wages. Its subsequent rise in market trades on a significant scale as a common medium of exchange was then induced by the exchange of these coins for goods by soldiers, and the need to acquire the coins themselves to pay taxes.

Again, for full analysis of the emergence of money in Greece, see my post here.

So, finally, we have a vast, vast body of empirical evidence that the orthodox Mengerian and neoclassical explanations of the origins of money are deeply flawed and not adequate to describe the full complexity of human history. I am afraid even Selgin’s revised Mengerian theory, curiously similar to that of Karl Marx, suffers from the same limitations.

Appendix 1: Menger’s Writings on the Nature and Origin of Money
The writings of the Austrian economist Carl Menger on money extend well behind his classic article of 1892, and include the following:

Menger, C. 2007. Principles of Economics (trans. Grundsätze der Volkswirtschaftslehre [1st edn. 1871] by J. Dingwall and B. F. Hoselitz), Ludwig von Mises Institute, Auburn, Alabama. pp. 257–285.

Menger, C. 1892. “Geld,” in Handwörterbuch der Staatswissenschaften (vol. 3). 730–757.

Menger, C. 1892. “On the Origin of Money” (trans. C. A. Foley), Economic Journal 2: 238–255.

Menger, C. 1909. “Geld,” in J. Conrad et al. (eds.), Handwörterbuch der Staatswissenschaften (vol. 4; 3rd edn.), Fischer, Jena. 555–610.

Menger, C. 1923. Grundsätze der Volkswirtschaftslehre (2nd rev. edn.), Hölder-Pichler-Tempsky, Vienna.

Menger, C. 2002 [1909]. “Money” (trans. L. B. Yeager and M. Streissler), in M. Latzer and S. W. Schmitz (eds.), Carl Menger and the Evolution of Payments Systems, Edward Elgar, Cheltenham, UK. 25–108. [N.B. this is a translation of Menger 1909.].

We must remember that Menger’s treatment of money developed in the course of his life.

Now L. B. Yeager and M. Streissler’s translation of Menger’s article on money from the Handwörterbuch der Staatswissenschaften is based on the 3rd edition of 1909, not the original article of 1892.

Was the passage that Selgin cites from Menger (2002 [1909]) actually in Menger’s 1892 article “Geld”? This is a pedantic point, but an interesting one, since Menger may have developed and modified this 1890s views as he got older.

Further Reading
“The Nature, Origin and History of Money 101.”

“A Note on Menger on the Nature and Origin of Money,” July 28, 2012.

“Menger on the Origin of Money,” January 5, 2012.

“Menger’s Nuanced View on the Origin of Money,” November 6, 2012.

“Observations on Non-Commercial Money,” February 18, 2012.

“Philip Grierson on the Origin of Money,” March 21, 2012.

“The Origin of Money and Coinage in Western Civilisation: The Case of Ancient Greece,” April 5, 2013.

BIBLIOGRAPHY
Bücher, K. 1901. Industrial Evolution (trans. S. Morley Wickett), H. Holt and Co., New York.

Einzig, Paul. 1948. “New Light on the Origin of Money,” Nature 162.4130 (25 December): 983–985.

Graeber, D. 2011. Debt: The First 5,000 Years. Melville House, Brooklyn, N.Y.

Grierson, P. 1977. The Origins of Money. Athlone Press and University of London, London.

Henry, J. F. 2004. “The Social Origins of Money: The Case of Egypt,” in L. R. Wray (ed.), Credit and State Theories of Money: The Contributions of A. Mitchell Innes, Edward Elgar, Cheltenham, UK. 79–98.

Hudson, M. 2004. “The Archaeology of Money: Debt Versus Barter Theories of Money’s Origins,” in L. R. Wray (ed.), Credit and State Theories of Money: the Contributions of A. Mitchell Innes, Edward Elgar, Cheltenham. 99–127.

Laum, B. 1924. Heiliges Geld: eine historische Untersuchung über den sakralen Ursprung des Geldes. Mohr, Tübingen.

Marx, Karl. 1906. Capital. A Critique of Political Economy (vol. 1; rev. trans. by Ernest Untermann from 4th German edn.). The Modern Library, New York.

Marx, Karl. 1990. Capital. A Critique of Political Economy. Volume One (trans. Ben Fowkes). Penguin Books, London.

Menger, C. 1892. “On the Origin of Money,” Economic Journal 2: 238–255.

Peacock, M. S. 2011. “The Political Economy of Homeric Society and the Origins of Money,” Contributions to Political Economy 30: 47–65.

Quiggin, A. H. 1949. A Survey of Primitive Money: The Beginnings of Currency. Methuen, London.

Rothbard, M. N. 2009. The Essential von Mises. Ludwig von Mises Institute, Auburn, Alabama.

Seaford, R. 2004. Money and the Early Greek Mind: Homer, Philosophy, Tragedy. Cambridge University Press, Cambridge.

Selgin, G. “Those Dishonest Goldsmiths,” revised January 20, 2011
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1589709

Weber, M. 1978. Economy and Society: An Outline of Interpretive Sociology (eds. G. Roth and C. Wittich; trans. E. Fischoff et al.), University of California Press, Berkeley and London.