When the history of money is described, very often the claim is made that people invented money because barter was impractical: a quick Google Search lists many places where this claim is made, e.g. (1), (2), (3), etc.

It sounds plausible, but not everything that sounds plausible is actually true. Is there any credible evidence for this claim?

As an example for an alternative explanation, David Graeber poses that money was invented to keep track of debt.

  • 12
    @matt_black - Depends how you want to define debt, I could very easily sell/trade you three cows in exchange for two horses and until you provide me those two horses you are debt to me.
    – rjzii
    Commented Nov 3, 2012 at 17:48
  • 3
    Also, the impracticality of barter is only obvious once money is invented. Money makes trade dramatically easier especially for people who are not neighbours. And trade usually makes all parties better off.
    – matt_black
    Commented Nov 4, 2012 at 1:32
  • 5
    Both "invented" and "deemed" make it sound like there was some central authority that made a decision. Evidence on the ground is that money developed as an organic process later codified by Authority. Commented Nov 4, 2012 at 14:13
  • 2
    I don't think the two stories are incompatible. No one is arguing that money did not start as promisory notes, ie. formalized dept. The mainstream is that this became a medium of exchange because it was more convenient than barter. This does not contradict Graeber, at least not the parts that wikipedia recounts. Graeber departs from mainstream by arguing that deferred payment, where one sides obligations are met at a later stage, where more common in early economies than barter.
    – Taemyr
    Commented Jun 19, 2014 at 9:29
  • 6
    Graeber claims that anthropological consensus is against there having ever been a "barter" stage before money existed, which is what would be required for money's raison d'etre to be to avoid barter. I suggest reading the first few chapters of Graeber's book and following up the references from there. I'm very sceptical of the idea that money was invented because barter is impractical and would want to see better evidence anyway. Commented Sep 15, 2014 at 17:10

4 Answers 4



Was money invented because barter was deemed impractical?

This is very difficult question to answer because in your question you are asking about whether barter was deemed impractical rather than whether it actually was objectively impractical, and money is literally as old as civilization itself perhaps even older as I will explain below, so we can only conjecture what was going through the heads of people who first used money, but impracticality of barter is very plausible, and by mainstream economists preferred explanation of origin of money, but the way how the question is posed we will likely never know for sure. Furthermore, while this is plausible and preferred explanation in economics, it is by no means the only explanation. Other explanations cannot be readily dismissed in this case, neither they are necessarily mutually exclusive (a phenomenon may have more than one cause).

Full Answer

Part I: What is Money and Barter

First, before going any further let us set our definitions for money straight. laymen often throw economic terms like money or barter without rigorous understanding of definitions (or lack of thereof) of the terms, and that might lead to fallacious assertions such as made in some of the other answers here. Furthermore, note since the sites that were referenced in opening post are economics sites (or purport to be), we need to be using definitions from economics not anthropology or history or other disciplines, otherwise we could easily commit an equivocation fallacy as definitions between these fields are not necessarily uniform.


Money in Economics by definition is something that can fulfill the following functions (See Mankiw 20181):

  1. medium of exchange - an item that buyers give to sellers when they want to purchase goods and services

  2. unit of account - the yardstick people use to post prices and record debts

  3. store of value - an item that people can use to transfer purchasing power from the present to the future

This is rigorous definition of money you can find in any economics textbook or generally academic work. It is also widely accepted definition. This is also consistent with the definition used by the sources linked to by original poster (one of the links does not work but the definition of money from Investopedia and owngoldandsilver.com match this textbook definition).

Important takeaway from this definition is to understand that money is not just what laymen think of when they think of money (e.g. coins, notes etc) it can take any form, shells, sacks of grain, small rocks etc as long as it is used as a medium of exchange, store of value and unit of account (oldest recorded money was plant and livestock based cca 9000- 6000BC see Glyn (2002)2).


Defining barter is slightly more problematic. In mainstream economics* barter was, and to some extent still is (except of some subfields) defined broadly as (Mankiw 2018)3:

the exchange of one good or service for another—to obtain the things they need.

Mankiw (2018) is undergraduate text, but in graduate texts such as Mas-Colell et al (1995)4 which is more rigorous, in models describing barter economies never makes any other qualification on barter beside requiring exchange be done by trading goods and services directly for each without declaring some good money.

This is also consistent with the definitions of barter mentioned on the site original poster linked. For example the owngoldandsilver.com states:

Most people understand barter. The premise is simple. I work and produce something and then I exchange my production, the fruit of my labour, for someone else's production.

Investopedia defines it as:

Barter is an act of trading goods or services between two or more parties without the use of money —or a monetary medium, such as a credit card. In essence, bartering involves the provision of one good or service by one party in return for another good or service from another party.

However, anthropologist define barter differently, because they contrast barter with gift giving in a way that it is not done in economics.

For example, in Mauss’s analysis of The Gift referenced in Avery's answer gifts are never rigorously defined but he constantly ties it to three features: to reciprocation, to an obligation to give and to the maintenance of social relations (Testart, 1998)5. Well if the gift giving is obligatory reciprocal then under general broad definition of barter in economics mentioned above, the definition that is demonstrably used by linked material by OP, the gift is barter (since there is reciprocal exchange of goods and services), and the so called 'gift-giving economy' is just barter economy as described by general economic models such as those discussed in Mankiw or Mass-Colell et al referenced above.

More generally, we have forensic evidence that exchange of goods and services and some rudimentary division of labor (which by definition requires reciprocal exchanges) was taking place even in prehistory, in upper Paleolithic** (see Villotte et al, 2019)6), even before cca 9000BC when we have evidence for money so it had to involve barter. Anthropologist might call exchanges in these societies gift giving, mainstream economists in general would call this barter. Saying one or the other is perpetuating myth is absurd, its just matter of semantics. Since sources, in OP question use economic semantics answer should use it too.***

To be also fully transparent, recently the anthropological definition influenced some subfields of economics, because in past gift-giving was neglected by economics profession, and if one wants to analyze gift giving broad definition of barter (such as the one used in general economics) can lead to problem as it leaves very little space for defining gift.

However, only very small portion of already niche work on gift giving in economics directly uses anthropological definition of gift. Actually, as discussed by Elder-Vas (2020)7, the gift is poorly defined in economics, (he tries to alleviate that by proposing rigorous definition), but under many including one proposed there, gift giving should not include obligation for reciprocity, and it is incompatible with anthropological work on gift giving economies, which would still be characterized as barter economies in large body of economic literature.

Consequently, in economics, hunter-gatherer societies of past would generally not be considered gift giving economies (perhaps there are some exceptions, but to an extent reciprocity in such societies was obligatory, and average member of such society could not just continue receiving "gifts" without some explicit or implicit obligation of contributing back these were barter economies). This does not mean that anthropology cannot have its own definition, but its wrong to assert that all social sciences have to follow definition set by anthropology, there is no need for such semantic imperialism, different fields have routinely different rigorous definitions. However, here the sources cited by OP use economic definitions, so to address them we have to use economic definitions as well.

So did Money Arise from Barter or Something Else?

Now that we settled definitions of money and barter we can finally attempt to answer the question.

First we have to establish some time-line since if money arisen to solve inefficiencies of barter, barter necessarily has to come first.

This is not difficult, again as mentioned in the previous section we have evidence dating to pre-history that division of labor was taking place in upper Paleolithic (see Villotte et al, 2019)6). When you divide labor, you have to make some exchanges, as that is fundamentally point of division of labor, and as long as these exchanges create even implicit obligation for reciprocity they are barter, under broad definition of barter in economics. As discussed earlier, according to Glyn commodity money was developed cca 9000- 6000BC. So money could arise from barter.

Furthermore, generally economic literature shows that barter is highly inefficient form of exchange (see Mankiw ibid pp 604), due to problem of double coincidence of wants (e.g. with barter you need to find someone else who wants exactly what you offer), so there would definitely exist a motive for society to arrive at some solution (populations that manage their resources more efficient would presumably have evolutionary advantage).

Consequently, it is plausible to describe origin of money was to select one or few items preferred barter items to all others which then became partly accepted because of their qualities in acting as media of exchange. These could be preferred barter items for a number of reasons, some because of convenient and easy storage, some because of high value densities and portability, others for durability. These barter items, being highly and widely desired, would be more easily to exchange, and at some point they would became money.

Nonetheless, as mentioned in the first paragraph we cannot be sure if people deemed barter to be impractical. We know with hindsight that barter is inefficient, but that is easy to say from the vantage point of modern economist (like it is easy to laugh now at ancients thinking earth is center of the universe). Money is older than written language so there is no way for us to know what people were thinking when money was invented, and it is not even clear that money was invented consciously as the explanation by Glyn shows.

However, this is not the only plausible story for origin of money. It is also possible that money arisen from debt. Here dating works as well as for barter since we have some evidence for debt existing in Upper Palaeolithic. This evidence is based on findings of "tally sticks", some of which can be dated as far back as to 44,200 and 43,000 years in the past (see Francesco d’Errico et al. 2012)8. We cannot be 100% sure what the tally sticks were used for (e.g. they could be used for religious purposes, math etc) but keeping track of debt is one plausible explanation for some of the tally sticks found, since we know that in other historical periods tally sticks, especially split tally sticks were commonly used for tracking debt. However, the earliest references to debt, where we can be 100% sure we are talking about debt date to ancient Sumer (see Ferguson 2008)9

Hence, a possible explanation for money arising is that people at some point decided to trade tally sticks, instead of barter.

This is what creates a difficulty here. Note that by definition money has to be medium of exchange, unit of account and store of value. You cannot just have tally sticks recording debt and call that money if you don't use them as medium of exchange you cannot say that we are dealing with money. If you trade some commodity but do not use it as an unit of account or store of value, again we are not talking about money.

Consequently, since money, by definition requires that the thing you call money is both unit of account and medium of exchange, the moment you "invent" money you solve both problems of barter and tracking debts simultaneously, and it is difficult to clearly pinpoint what exactly was the impetus that compelled society to develop money (possibly it could be even both reason at the same time).

What even worse, these are not the two only explanations that are there. As argued by Glyn:

Money originated very largely from non-economic causes: from tribute as well as from trade, from blood-money and bride-money as well as from barter, from ceremonial and religious rites as well as from commerce, from ostentatious ornamentation as well as from acting as the common drudge between economic men.

In fact Glyn argues that one of the main reason for development of money was neither to solve problem of barter, nor necessarily settle debts, but rather to settle tax obligations (which can be treated separately from debts, even though not all authors make that distinction).

To be perfectly honest, this is not fully settled question, so it is hard to say who is right, but most mainstream economists prefer the barter origin theory (or actually in literature this is known as the metalist view). This is due to heavy influence of Jevons10 who emphasized high transaction costs created by barter, which imply there is a large incentive to solve this issue as soon as possible, and other forerunners of present day mainstream economics such as Schumpeter11 or Samuelson (1967)12. Samuelson was especially influential since he is the 'father' of 21st century modern mainstream economics. This is also the explanation preferred by many standard and widely used economics textbooks such as above mentioned Mankiw's Principles, or Samuelson's Economics etc.

In economics, the main advocates for the debt origin of money (known as chartalist view) come from heterodox and often fringe subfields such as Marxist economists or Modern Monetary Theory advocates such as Randall Wray. As a whole these are very fringe and not widely accepted fields of economics (and I am still being very generous with my language here).

Because these fields have very bad reputations ideas coming from these subfields are sometimes too quickly dismissed. It does not help that these subfields often talk very ideologically about origins of money, to the point where one has an impression that they oppose the barter origins just because they argue its inherently connected to decentralized market exchanges.

Nonetheless, one has to admit that when it comes to the topic of the origin of money these strands of literature might be actually correct and should be taken more seriously. While, the anthropological research is not really applicable here as it simply uses different definition of barter/gift and not consistent definitions of barter (so its conclusions about (non)prevalence/existence of barter have very little to no bearing on the origins of money from barter as defined in economics), there is some evidence that barter might not be as extremely inefficient (and thus high transaction costs) as is often thought as argued by Svizzero and Tisdell (2019)13 who show that there is some archeological evidence for barter persisting in some flexible forms even in some ancient civilizations which shows that barter cannot as extremely inefficient as often thought, as otherwise such civilizations would likely not be able to bear the burden of high transaction costs created by barter.

This being said, this is not completely settled question, I do not think it will even be fully settled in foreseeable future. Since both money and debt likely predate written languages it is extremely difficult to discover what was original impetus for creation of money. Note, there is a consensus that money is both important as a solution to double coincidence of wants, and as a unit of account and ways how to settle debts. And I do not think that anyone doubts that as soon as money was "invented" it was quickly used for both, making this very difficult problem to solve.

* Mainstream economics is the economics taught at most universities and the most widely accepted set of economic theories.

** Paleolithic age was roughly 2.5 million years ago to 10,000 B.C, we are talking about upper Paleolithic so closer to 10,000 B.C, but that is still older than 9000 B.C. for which we have first evidence for money.

*** This is why the answer by the user Avery is highly misleading when the user calls prevalence of barter myth held by social sciences. That is plainly fallacious statement, that would be like if economist would tell to physicists inflation [meaning cosmic inflation] is a myth since we know that inflation is increase in aggregate price level. This is just an fallacy of equivocation.


  1. Mankiw, N. G. (2018). Principles of economics. 8th ed. Cengage Learning. pp 605.
  2. Davies, Glyn. A history of money from ancient times to the present day, 3rd ed. Cardiff: University of Wales Press, 2002. pp 41-44
  3. Mankiw, N. G. (2018). Principles of economics. 8th ed. Cengage Learning. pp 604.
  4. Mas-Colell, A., Whinston, M. D., & Green, J. R. (1995). Microeconomic theory (Vol. 1). New York: Oxford university press.
  5. Testart, A. (1998), ‘Uncertainties of the “Obligation to Reciprocate”: A Critique of Mauss’, in W. James and N. J. Allen (eds), Marcel Mauss: A Centenary Tribute, New York: Berghahn Books, pp. 97–110
  6. Villotte, S., Churchill, S. E., Dutour, O. J., & Henry-Gambier, D. (2010). Subsistence activities and the sexual division of labor in the European Upper Paleolithic and Mesolithic: evidence from upper limb enthesopathies. Journal of Human Evolution, 59(1), 35-43.
  7. Elder-Vass, D. (2020). Defining the gift. Journal of Institutional Economics, 16(5), 675-685.
  8. Francesco d’Errico et al. (2012) Early evidence of San material culture represented by organic artifacts from Border Cave, South Africa. Proceedings of the National Academy of Sciences 109(33)
  9. Ferguson, N. (2008). The ascent of money: A financial history of the world. Penguin. pp 29.
  10. Jevons, W. S. 1910. Money and the mechanism of exchange. London: Kegan Paul.
  11. Schumpeter, J. A. (1963). History of economic analysis (No. 330.153 S392). Oxford University Press, New York, NY (EUA).
  12. Samuelson, P.A. (1967). Economics: An Introductory Analysis, 5th Edn., McGraw-Hill, New York.
  13. Svizzero, S., & Tisdell, C. (2019). Barter and the Origin of Money and Some Insights from the Ancient Palatial Economies of Mesopotamia and Egypt.
  • 1
    "I do not think it will even be fully settled in foreseeable future." -- why the hedge about the foreseeable future? The history of the prehistoric will never be fully settled (barring the invention of time machines). Commented Jul 2, 2021 at 11:32
  • @JohnColeman you are right, I guess as an academic I am used to a bit more cautious way of writing
    – 1muflon1
    Commented Jul 2, 2021 at 12:01
  • This answer seems to promote the use of economics to solve this problem (definitions, economist's models, comparing schools of economic thought). This seem to completely the wrong discipline to ask. We are looking for historical evidence. We should be asking historians, or more likely, archaeologists. Sure, economists may be able to justify why the archaeologists find what they find, or suggest ideas for what to look for, but surely this requires the work of people with dirty fingernails.
    – Oddthinking
    Commented Jul 2, 2021 at 17:11
  • 1
    @Oddthinking note I am relying in the answer on historical and archeological evidence, I am only stating that for this answer one has to use correct terminology, since the OP asks about validity of the statement from sites that use economic terminology. Note I am not relying here on theoretical arguments. When it comes to the plausibility of "barter origin of money" I fully rely on archeology and historical evidence I just use 'correct terminology' in order to avoid equivocation fallacy. If OP asks about barter in economics context but some other field has completely different definition
    – 1muflon1
    Commented Jul 2, 2021 at 17:36
  • we have to make sure that empirical results are compared using the right definitions. For example, in economics (and the sites OP draws from are economic sites) money has specific definitions that can include various mediums. Maybe in a history book historian can consider only coins money, and lets hypothetically state that one historical source states that earliest discovery of coins was in year x. Well just citing that source to 'debunk' the economic statement about money, without adjusting definitions is equivocation fallacy, since in economics shells might be considered money and then we
    – 1muflon1
    Commented Jul 2, 2021 at 17:48

I think your linked sites, and numerous others, do provide significant evidence that what promissory notes, and then money gave us was a much easier way to carry the equivalent of a couple of cows, some gold and an acre of land.

Economist Ludwig von Mises, in his 1912 book The Theory of Money and Credit, had this to say on the Origin of Money (ellipses and bold indicating my edits):

I.1.4 Suppose that A and B exchange with each other a number of units of the commodities m and n. A acquires the commodity n because of the use-value that it has for him. He intends to consume it. The same is true of B, who acquires the commodity m for his immediate use. This is a case of direct exchange.

I.1.5 If there are more than two individuals and more than two kinds of commodity in the market, indirect exchange also is possible...

I.1.6 Let us take, for example, the simple case in which the commodity p is desired only by the holders of the commodity q, while the comodity q is not desired by the holders of the commodity p but by those, say, of a third commodity r, which in its turn is desired only by the possessors of p. No direct exchange between these persons can possibly take place. If exchanges occur at all, they must be indirect; as, for instance, if the possessors of the commodity p exchange it for the commodity q and then exchange this for the commodity r which is the one they desire for their own consumption. The case is not essentially different when supply and demand do not coincide quantitatively; for example, when one indivisible good has to be exchanged for various goods in the possession of several persons.

I.1.7 Indirect exchange becomes more necessary as division of labor increases ...

I.1.8 Thus along with the demand in a market for goods for direct consumption there is a demand for goods that the purchaser does not wish to consume but to dispose of by further exchange...

I.1.9 Now all goods are not equally marketable...

I.1.10 It was in this way that those goods that were originally the most marketable became common media of exchange...

I.1.11 Thus the requirements of the market have gradually led to the selection of certain commodities as common media of exchange...

I.1.12 This stage of development in the use of media of exchange...For hundreds, even thousands, of years the choice of mankind has wavered undecided between gold and silver. The chief cause of this remarkable phenomenon is to be found in the natural qualities of the two metals. Being physically and chemically very similar, they are almost equally serviceable for the satisfaction of human wants...

(Translated by H. E. Batson 1934, additions 1953, this edition published by Liberty Fund in 1980 and retrieved from The Library of Economics and Liberty.)

Have a read of the rest of that paper for much more in depth description!

  • 2
    The example is actually just bartering, but with gold/silver instead of directly. While coins were made their value was in the metal itself and it wasn't uncommon to break a coin into smaller pieces.
    – liftarn
    Commented Jun 12, 2014 at 11:20
  • 11
    But no reference to any actual historical period in which this is said to happen. I.e. it reads like armchair economics rather than actual history. Commented Sep 15, 2014 at 17:11
  • 2
    The natural qualities of gold and silver that made them useful as currency was that they were 1) pretty; 2) relatively rare and 3) pretty much useless for any other purpose. Commented Jul 5, 2020 at 15:00
  • 1
    @Shadur For gold it doesn't tarnish and has a unique color. It is easy to identify gold without much experience in identifying gold. Once metalurgy was more common, however, it became less easy in identifying pure gold. See why Archimedes yelled "Eureka!" to understand why. And though most have been proven not beneficial, ancient people's believed in health benefits attached to both gold and silver.
    – user11643
    Commented Jul 6, 2020 at 22:50
  • 2
    @FrancisDavey We can actually observe this in prison
    – user541396
    Commented Jul 9, 2020 at 0:41

A famous practical example was found in WWII prisoner of war camps where cigarettes became a de-facto currency. This happened without any central organisation or planning; prisoners started with barter, and over time found that cigarettes had become the medium of exchange because they were more-or-less fungible, small and common enough to buy small things with, and desirable in their own right by at least some of the prisoners.

  • This does not answer the question "Was money invented because barter was deemed impractical?"
    – Avery
    Commented Jul 6, 2020 at 15:08
  • 2
    @Avery As dmckee says in the comments on the OP, the word "deemed" is problematic. It makes it sound like some central authority decided "This barter system needs upgrading". What the example shows is how an initial barter system evolved into a money system without any central organisation. Whether this is because barter was "deemed" to be impractical or whether it was merely found that currency worked much better is a semantic question I have deliberately avoided. Commented Jul 6, 2020 at 15:50
  • 3
    That semantic question is probably important in an answer. Was this prison-cigarette currency or just a common barter item among other barter transactions? The semantics are hard to avoid. Some definitions of currency require a central plan, for example.
    – user11643
    Commented Jul 6, 2020 at 22:59
  • In this example a form of currency arose among people who already had a robust notion of money -- so it has no obvious relevance for the question of how the notion of money arose in the first place. Commented Jul 2, 2021 at 11:26

The answer beginning, "Suppose that A and B exchange with each other" is unreferenced and factually inaccurate. The story being told in that answer has been known to be false for 100 years.

Here is a constructive critic of Graeber affirming this:

As Graeber explains, however, pure barter as such never existed. The myth of barter is a handy story for explaining the origins of money in terms that warrant market theories of price and commodity exchange. But, historically and ethnographically, it does not hold water. None of this is news to anthropologists. It is, after all, the basis of Mauss’s analysis of The Gift.

Bill Maurer, “Debating Anthropology's Assumptions, Relevance and Future: David Graeber's Wunderkammer, Debt: The First 5,000 Years,” Anthropological Forum 23(1)

The book cited is The Gift by Marcel Mauss, a classic of anthropology published in 1925. Mauss' analysis of pre-monetary gifting is now widely used by historians in order to prevent the simplifications that the modern concepts of barter and currency can create. For example, I just read an excellent book, The City-State of Boston 1, which discusses to great effect how the early American gift material called wampum was neither money nor barter, and how its uses changed with the arrival of economically interested European settlers.

The critique of Graeber made in the article being quoted does not come from debunking the myth of barter, which is unquestioned among social scientists, but from the question of how the myth of barter arose.

State-credit theories and barter-commodity theories did, alternately and often cyclically, as Graeber demonstrates in the second half of the book, rise to prominence, capturing the imagination of noble and patriarch, king and parliament, authors and activists alike. In so doing, they conjured a world after their own image. The theories properly speaking do not describe, they perform. But that performing makes difficult the quest for origins with which Graeber begins his analysis: the origins of the ethical imperative to repay a debt.

ibid., emphasis added

This goes beyond the scope of this question, but basically, Graeber is correct to reject the myth of barter which appears on the "Library of Economics and Liberty" and other such untrustworthy websites. The major concern with his book is his account in how the concept of personal debt and the myths of barter and state credit arose, which complicate the issue.


  1. The City-State of Boston: The Rise and Fall of an Atlantic Power, 1630–1865 by Mark Peterson – April 23, 2019 - Amazon Link
  • As mentioned in a comment above, I find this discussion hard to follow without a definition of "money" and "barter". What is "pre-monetary gifting", and how does it differ from "pure barter"? What is "the modern concept of currency", and what is its alternative? Maybe an outline of the "unquestioned" alternative history of money would make this more accessible to non-experts like me?
    – IMSoP
    Commented Jul 6, 2020 at 20:56
  • 2
    In chat, I questioned whether Graeber's attack on a barter system is a straw man. The essence of that concern being: who says there were planned and understood barter systems that not only preceded currency, but altogether were ignorant of it until it was so devised? The claim is that money was invented because barter is impractical. Indeed, barter is impractical, so is it that Graeber only is highlighting that currency arose at the same time as civilization?
    – user11643
    Commented Jul 6, 2020 at 23:29
  • @user541396 this article seems to admit that there's no evidence that "money [was] invented because barter was deemed impractical". it doesn't present new evidence, it just claims that its "armchair reasoning," based on zero observation, is simply "logic" and that there's no other logical explanation possible
    – Avery
    Commented Jul 9, 2020 at 3:31
  • An answer based purely on logic is actually unacceptable on this website; see skeptics.meta.stackexchange.com/questions/1019/…
    – Avery
    Commented Jul 9, 2020 at 3:33

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .