This archived content originally appeared at, the predecessor site to, and does not carry the sponsorship of the Cato Institute.

Start from money, or credit?

The Austrian theory of money begins with Carl Menger. In his Principles of Economics he develops a theory, based on economic logic and compared to historical and anthropological evidence, of how money originated.

Menger’s emphasis is on commodity money, which Ludwig von Mises later classified as a form of “money in the narrow sense” and which textbooks today call a form of the monetary base. Menger has little to say about credit.

Later Austrians, building on Menger, developed a “monetary view of credit.” That is, they started from money as the basis of the analysis and discussed credit as deriving from money. This is the textbook view among economists generally, not just among the Austrians. One of the consequences of the monetary view of credit is that it tends to see money as the solid but narrow base on which is constructed a huge superstructure of credit. Diagrams often represent the monetary system as an inverted pyramid that looks likely to topple over at any time, leaving a pile of rubble.

The opposite approach is to take a “credit view of money,” as I think Henry Thornton implicitly did in his Paper Credit of Great Britain and as Joseph Schumpeter discussed in his History of Economic Analysis (see in particular part III, chapter 7, section 4). A credit view looks as money as a kind of residual for settling net credit balances where sufficient trust is lacking to roll over the credit. In the credit view, money is the capstone at the top of the settlement pyramid, which is not inverted and is therefore not in constant danger of toppling over. The credit view also implies less suspicion than the monetary view that banks are “creating money” and that fractional reserve banking is somehow inherently fraudulent.

Since Menger there has of course been a huge amount of further historical and anthropological research into the origins of money. The left-wing anarchist anthropologist David Graeber gives an interesting though tendentious account of some of it in his recent book Debt: The First 5,000 Years, using it to criticize Menger.

My sympathies are with the credit view, because credit is both logically and historically prior to money. Logically prior, because it is easier to imagine a society without money (where outstanding credits are extended, transferred, cleared, or written off) than to imagine a society without credit (where every transaction has to involve barter or cash payment at the moment of consummation). Historically prior, because even hunter-gatherer tribes functioned on a kind of credit, where people performed services for one another today expecting something in return in the future. It was not the exact calculation in terms of a monetary unit we now think of in connection with credit, but it did involve exchange of goods over time and was not simply a gift: those who were able to contribute to the tribe’s well-being but refused to do so tended to get expelled.

(An interesting aside: Menger had high regard for John Law’s ideas on the origins of money: see Appendix I of the Principles, page 318 in my book.)


  1. When Austrians talk about credit in the framework of monetary analysis I think they're referring to fiduciary media specifically, rather than credit more widely considered (i.e. debt). But, I don't see how the fact that debt precedes money shows how money came from this. The benefit of money is to lower transaction costs in general, not just for debt. It makes more sense, to me at least, to say that credit/debt is natural to exchange, and as such is independent of the existence of money.

    1. Good point. I always find the use of the word "credit" confusing without a qualifier. Sometimes it simply means "faith" in the government issuing/regulating a currency, and sometimes it means "debt." So, I now often use the terms "credit/faith" and "credit/debt." In effect, all money is "credit/faith" or it wouldn't serve as money, but it's money as "credit/debt" that's a big problem for many of us.

  2. I wonder what D. Graeber knows about C. Menger: in 5000 years, he cites Karl (sic) Menger's Origin of money as the only reference, and the single mention in the entire book is to say that "Menger and Stanley Jevons later improved on the details of the story, most of all by adding various mathematical equations" (sic).

  3. I can see how credit-based money could evolve without barter ever existing

    – Credit begins with informal "gifts" for which reciprocation is required
    – This becomes formalized into loans for which a specific repayment in the future is required
    – These future-commitments can then be used for spot-trades
    – These future commitments become standardized into credit notes and are used as money

    At some point in this process a universal way of evaluating different credit notes would be needed. In theory this could be done in terms of a commodity that was itself never used directly as money (though such a commodity would acquire a special status as the unit of account and ultimate settlement once adopted).

    It seems to strain credibility though to hold that spot-trades didn't exists before credit-money evolved. Assuming barter did exists then Menger's theory about how money evolves seems compelling. It seems very likely to me that both theories could be correct. As human society evolved the benefits of both spot-trades and future-trades became apparent and a monetary system based upon both requirements emerged. Both kinds of trades drove the adoption of a commodity unit of account and both credit notes and the commodity-money became medium of exchange.

  4. The title is misleading. It should be "Start from spot or credit". Even a credit transaction must be logically classified either as direct or indirect exchange, thus it is either barter or uses a medium of exchange. Menger and Mises do not claim that trade must emerge through spot transactions. That is why people like Graeber are wrong. All the examples of "credit" he mentions in the book are highly valued commodities, with the minor exception of tally sticks, but he misses that even the tally stick was a type of money substitute piggybacking on the pound (see Wikipedia article on tally sticks). He thinks that media of exchange magically gain a market price and liquidity without a background catallactic process behind this evolution.

    Menger was very thorough in analysing the evolution of money. He explains how transaction costs and liquidity affect the preferences for certain media of exchange. Even if historically it was true that credit matured before spot, the only thing it would indicate is that Menger's argument regarding transaction costs was correct, as typically, credit instruments have lower transaction costs than specie. I analyse the question of the effect of transaction costs on the choice of medium of exchange in my master's thesis "Economics of Bitcoin".

    Even if spot transactions didn't ever occur before money emerged, that still does not disprove Menger. It shows that Menger is misunderstood.

    1. Can you clarify, when you say "spot" or "spot transactions", do you mean any transaction that occurs at time = 0 (?)

      Thank you

      1. I think what cjdparis is hinting at here is that in a number of financial markets, a spot transaction is one that is agreed to today but not settled until typically two days later, whereas Peter Surda seems to have in mind what would in financial markets would be termed a "real-time settlement" transaction. And Peter, the way Menger offers his account of the origin of money it is unclear to me whether he thinks it is the only account, or whether, as I think, it is one possible account. If Menger meant it to be the only account, then yes, he would be disproved. I prefer to read him more charitably as offering one possible account, but not going into the detail of a treatise on money because he was writing a general book on economics.

        1. I see how my use of "spot transactions" is probably not the same one as used on financial markets. Let me clarify.

          What I mean by "spot" is an exchange of a present good for another present good. By "credit", I mean exchange of a present good for a future good. A gold nugget, for example, is a present good, while a credit instrument promising to pay gold in the future is a future good. The criticism of Menger implies that he saw barter as the former. But there is no requirement for this.

          I think that nowadays, it is quite common to view credit instruments as a present good (an immaterial medium of exchange) even though the gold standard branch of the Austrian school will disagree with me here. But I think that credit *as such* cannot originate as a present good, rather it must evolve through a future good, and the origin must thus be consistent with Menger's account. There are credit instruments which act as present goods but were never future goods (e.g. mutual credit and various complementary currencies), but I think that this is only possible because people are already accustomed to using credit as a present good, and a unit of account already exists. I don't think that it is possible for a credit instrument to "evolve" from a future good to a present good without an underlying unit of account, which implies that a medium of exchange (or even money) already must exist prior to that. And indeed, the examples presented by Graeber all indicate that that was the case. Whether "spot" transactions or "credit" transactions occurred first is irrelevant.

          So to put this into a more formal way, I see three stages of evolution of money:

          1. trade of present goods against present goods (what I call "spot")
          2. trade of present goods against future goods (what I call "credit")
          3. the evolution of a future good into a present good (i.e. credit becoming a medium of exchange accepted as a final settlement of debt)

          Now, people like Graeber claim that 2 came before 1, and this is supposed to disprove Menger. But that is not true. Menger's account does not require that 2 comes before 1 or that 1 actually ever occurs. What is important is that, as I claim, 2 must occur before 3.

  5. Hi,

    I got here from the Facebook page where it was requested that I not speak of fractional reserve lending, and instead to come here with any questions.

    First, though, I'd like to take a moment to get a handle on the use of the word "credit" here. Credit in the sense of faith and trust is all fine and well, but credit as it pertains to the loan of scarce goods is an entirely different animal, and credit as it pertains to lending something expecting even more of that exact same thing back in trade is intrinsically dangerous in my view. Gold, and to a lesser extent most metals, have an intriguing quality that made them useful as an intermediate exchange good in the past in that, for their size and weight, they maintained a good bit of value. I feel that the free silver movement has more or less laid precious metals as a useful base for money to rest though.

    Fractional reserve lending, just in general, is not lending. It is, at the very least, re-lending, and at worst a way to give some central authority somewhere the power to control the supply of money by manipulating the lending and money markets, which are inextricably linked. This is why when using a gold standard, over and over in history, there have been bank runs, or more accurately gold runs. At some point, someone gets enough money together to begin demanding your reserves of gold faster than you can liquidate your securities, because you simply have far too many securities out in relation to the amount of gold that there is.

    I would like to hear someone explain why we keep avoiding the seemingly obvious. Money as a concept is really simply credit in the trust sense. That is, we agree to allow someone to do something in exchange for a credit to their name that is respected centrally such that they can later call on that credit to get other goods and services later. This is not the institution "loaning" to an individual. Rather, it is an individual "loaning" to an institution. One has to trust the institution enough that they will respect the fact that the individual has already done work that the institution stated would be respected in trade for future goods or services.

    Whether the money is controlled by a governmental or private central authority is irrelevant. The problem is in putting this faith in any central authority. The mechanism for doing away with this central authority exists, and indeed has always existed, and that is to record and track the relationship between goods and services and to trade in those relationships.

    Thus I see the movement towards Local Exchange Systems or Community Exchange Systems the one that best reflects the natural state of trade, and I am not seeing why they are not given more attention in all of the various proposals folks put forward to resolve the issue of the Federal Reserve and other central banking authorities around the world.

    Thanks in advance for any information anyone can share. I'm going to now follow the advice I was given to search for articles relating to "fractional reserve", as it seems many of you are convinced there is nothing wrong with it intrinsically, and I'd be interested to know why.

    1. I will probably write a post within the next month addressing some of the points you raised. But money is not simply credit. Trade using commodity money need not involve trust. Fiat money is often based not on trust but on coercion.

      1. I look forward to that. Any thoughts at all about LETS or CES? It seems these would allow for multiple commodities to serve as money, and from these relationships and trade in services, the value of those services relative to many commodities would be established. Plus they are capable of decentralization and privatization, with the government's role firmly rooted in its tradition of regulation rather than operation.

        1. Not sure about LETS or CES, but I've found that this group is united mainly in its belief that there's something terribly wrong with central banking, but otherwise "free banking" means lots of things to lots of different people.

          My personal view on how to restore a pre-central bank era, but improved, version of "free banking" is for the government to simply allow (and proactively support) depositors to object to Federal Reserve notes, and claim their right to exclusively use Treasury-direct current-coin-only bank accounts, though I don't advocate taking physical possession of current coin. In my view this is the only legal or Constitutional remedy, the least disruptive, and it retains both the Federal Reserve (though only in its benign role as Congress's "fiscal agent") and some general improvements to the banking system that have occurred since 1913, such as FDIC insurance.

          As author Thomas Wilson reminds us in "The Power 'To Coin' Money" (1992):

          "Congress, of course, was given the authority to coin money, meaning precisely that: Congress was granted the power to regulate the quality of circulating media, i.e., to define the metallic content of U.S. coins and to assign legal tender status to those coins and comparable foreign coins. This does not mean that the regulation of coinage related to the quantity of money as such. The money supply was left to market determination." (pg. 4)

          In short, banking was generally meant to be let "free" to determine the money supply.

          Wilson continues: "Note-issuing banks existed in 1787, and their notes, devoid of legal tender standing, were presumably compatible with the Constitution–banking was viewed as a right of individuals or corporations to form business enterprises under the common law tradition." (pg. 5)

          1. I see. Thanks.

            I have heard the role of the Fed explained as Constitutional because it is also allowed to regulate the value of coined money. The Fed, then, is an instrument for regulating the value. "Coin" is interpreted in the broad sense of "create".

            But I see what you mean, and thanks!

          2. ShaneCRoach states: "I have heard the role of the Fed explained as Constitutional because it is also allowed to regulate the value of coined money. The Fed, then, is an instrument for regulating the value. "Coin" is interpreted in the broad sense of "create".

            Glad you brought this up because unfortunately it's a prevalent misconception held by the vast majority of U.S. lawyers, assuming they even care about this area of law.

            The Fed is in no way authorized to "regulate the value of coined money" under Article 1, Section 8, Clause 5. The coining power is exclusive to Congress, and "to coin" does not mean "to print" or "to create."

            The power to issue paper/electronic money, which, unlike the coining power, is a non-exclusive power that Congress can share with certain banks it charters, is derived from the Borrowing Clause (Article 1, Section 8, Clause 2) and the Legal Tender Cases (1869-1884).

  6. I think there's a conflation of two concepts. First the concept of bilateral exchange vs. triangular exchanges and second the concept of immediate exchange vs. exchange over time. Money enables triangular exchange and credit enables exchange over time.

    1. To me, LETS and CES also have the advantage of being multilateral rather than even triangular. They do require credit though in the trust sense that in order to fully replace a single monetary unit you have to trust the members of the organization to honor the work done through the system to return services at an exchange rate agreed upon at the time of rendering the service.

      If there's just absolutely, positively no interest in the community here at in LETS and CES, that's fine. I just am curious as to why it does not get more airtime in the emerging discussions about monetary policy reforms. I knew have had free banking in this country before, and that the accepted wisdom has been that it did not work well at all, so I am not entirely sure from which direction the community that forms the core of this blog is approaching the issue.

      1. I should have said multilateral instead of triangular, thanks. I think LETS systems have been discussed here before. There's even a new Larry White article at Reason that favorably mentions a LETS system. As to whether free banking worked in the USA, there was never exactly a national free banking system, since individual states heavily restricted banking with charters only in exchange for buying state bonds, and unit banking laws.

        1. Regarding free banking in the U.S. I poked around a bit after asking here and I see now what you mean about it never being truly free. I have read mixed reviews on Australia's "free banking" history. At least now I know what you're on about!

  7. Banking was more or less never considered a right of the individual in the United States or the area that would become them. Prior to the Revolution, what banking existed was state-run land banking. After the Revolution, the newly independent states quickly started passing restraining acts prohibiting private banking activity without a legislative charter. By 1800, not a single major commercial state had free entry into banking.

    By the way, bank notes are not a common law instrument, but part of equitable jurisdiction.

    1. MichaelM, that's interesting, but the point of my post is that under the Constitution Congress is not supposed to be in the banking business, nor have too close a relationship with a privately-owned central bank like the Fed, nor use the Fed to play shell games.

      Rather, Congress is only supposed to provide coined base money to private banks and depositors, and then let the banks issue notes and make loans based on their coin holdings, even if the coins are made of cheap metals.

      But for this version of free banking to happen U.S. lawyers need to recognize that depositors have a right to object to the use of the Fed's currency and to demand Treasury-direct coin-based accounts at any U.S. incorporated bank (under the 1819 McCulloch v. Maryland case).

  8. That post is supposed to be in reply to Rick, although it doesn't seem to have shown up like that .

  9. If the person running the "Ripple" network had backing in some commodity and facilitated trades rather than merely handing IOU's back and forth, it would be what I was looking for. Note that the article mentions banks have in house people to do the book keeping. That is precisely why banks are superior. The centralization and willingness to back the IOU is key to its universal acceptance. As the article says, albeit in a roundabout way, the utility of Ripple is limited to a closed group of people who trust each other enough to accept one another's IOU's.

  10. If its credit on money, it´s because there is a thing that constitutes "money". The fact that credit on money (credit money) is able to circulate does not transforms the credit on money in money itself. In the end money must exists in order for credit money to take place. We can say that it os a kind of money, yes,but performing less than money itself because there is a credit risk.

Comments are closed.