Home > The Economics Profession > Post-Keynesian free banking – or towards a string theory of money

Post-Keynesian free banking – or towards a string theory of money

from Merijn Knibbe

Arjo Klamer has long been ridiculed. Fellow economists did not take him serious – as he was not only against the Euro back in the nineties but also predicted Euro doom and Euro gloom. One year ago, the Euro was still hailed as a spectacular succes. But the world changes.  Klamer was right. But what’s his alternative? As I read in the newspaper, he proposes a whole variety of moneys: special money for transactions between certain companies, special kinds of money for specific regions – just imagine. Should we ridicule him for this? Probably not – the very thing is already happening. We just don’t see it – as we use the wrong, orthodox, mythical definition of money. This definition can be found in, for instance, the Abel, Bernanke (that’s Ben Bernanke) and Croushore textbook (p237-p239) – but it’s all over the place.

Herewith goes a modest proposal for a more specific, demystified definition of money.  The background of this proposal is, in the end, economic history: any economic historian who looks at money will find that the thing changes and that modern definitions of money do capture some aspects of historical money (including present day money) but that other aspects are left aside. The petty debts at the grocers or the crooked debit and credit at the company store are not counted as money – while they should.  That’s a definition of money past. A good definition should also cover present money and future money.  The present: look at food stamps: 64 billion of (limited) spending power in 2010. For the future of money: just look at http://www.groupon.com/.  That’s money: a means of exchange, a store of value and a unit of acount – but it’s not seen and, worse, counted as money by many economists.  A definition of real world money should actively cover such kinds of money – as people do use those kinds of means of payment.  Any definition of money should, therefore, be specific enough to describe the present state of affairs but general enough capture past and future changes, from indeed the days of the Greek to the future development of Groupons. It should also enable a more thorough analysis of the use of money be real people and real companies. My modest proposal is as follows:

 First, money is often defined as a unit of account, a store of value and a means of exchange. I accept this. I do not acccept the idea that only banks and comparable institutions are money-creating entities.  Many companies and even households can, and do, create money – not euro’ or renminbi’s but surely some kind of means of exchange, store of value and unit of account. But different kinds of money do not have the same propensities as store of value, unit of account or means of exchange – which gives rise to large problems when you do not acknowledge this but still try to estimate the amont of money (Burgess and Janssen, 2007).  Food stamps (Food debit cards, I should write) are, for instance, only a limited means of exchange. I call this: local money). The guilder of 20 stuivers was, for a very long time in a very large region in northern Europe, the preferred unit of account. The actual means of exchange were however all kinds of petty debts and a bewildering array of coins (petty debts as a kind of money: that’s 100% consistent with the ideas of the classicals: Nijstad, 2005). During the first decades of Kaapstad (Cape Town in South Africa), in the seventeenth century, the indigenous San people were not paid with ‘a general means of exchange’ as they did not live in an exchange economy and had not use for a general means of exchange – they were paid with, of course, tobacco and gin (soon, the Dutch and Flemish settlers started to use slaves from Mozambique or Madagaskar as they needed more labor and there was no real labor market).

Second, many money’s have strings attached: you can’t buy groceries with a mortgage – we might call this the M-string. Ireland will get a Euro loan with a German string attached: the lenders have to take a hit too (thanks, many thanks, Mrs. Merkel). And even the petty debts mentioned above (a kind of money superseded by credit cards, which unlike the petty debts do charge interest) had strings attached – the grocer or the carpenter knew you, and there was only so much debt you could acquire (oops, what went wrong in the USA mortgage market?). This leads to the next table: money as a historical entity – it gives bit of order to at least my understanding of the shape shifting we call money. I mean, we do need to have a definition of money which can be used by for instance economic historians like me, Post Keynesians like Klamer and Austrians inspired economists  alike. I hope this modest proposal will be a small step towards this (and yes, I do exclude neo classicals as their General Equilibrium models are non-monetary models of trade – that’s just beyond me. All evidence we have on non-monetary societies shows that barter is a side show at best. Without money, (almost) no trade.

  Universal means of Store of Unit of Strings
  exchange Value account attached
Euro’s, cash ++++ ++++ ++++ 0
Euro’s on a checking account ++++ ++++ ++++ +
Debit on a food debit card ++ +++ +++ ++++
Mortgage money debit + + ++++ ++++
Tobacco on a Cape town farm, 1673 + ++ ++++ +++
18-th century petty debts debit +++ +++ +++ +++
Debit at the company shop ++ ++ +++ ++++
Groupons + ++ + +++

Addendum: debit means: rights to pay, based upon reputation or an account of money. Means of exchange: how many different items can be paid with a specific kind of money. Store of value: how long does a certain kind of money last (Groupons: 6 months). Unit of account: is it difficult to estimate the value in a balance sheet? Strings attached: is there a person or institution which limits transactions? For instance: I can’t draw more than 500,– per day in cash from my regular checking account, the bank (in fact: the government) forbids larger amounts.

Summarazing: Abel, Bernanke and Croushore define money as: ‘assets that are widely used and accepted as payment’. That’s so wrong. Money ‘stringed’ to a mortgage is only accepted as payment for a specific house. Money consists of assets that are accepted as payment, widely or local. There’s not just one kind of money. Acceptance rests on all kinds of cultural, local, political and economic specifics.

P.S. – on the orthodox definition of money: money is supposed to solve the ‘mutual coincidence of wants’ problem. However, as money enables markets to grow, it’s money which gives rise to this problem. Non-monetary economies do not know this problem, as they do not know trade and have other means to regulate exchange. See the Abel, Bernanke and Croushore textbook, p. 239 on this fable.  There is of course some mathematics to this madness. Brunner and Meltzer (1971) have provided a mathematical version of this fable – but a fable mathematical told still is a fable.

Abel, A.B., B. Bernanke and D. Croushore (2011), Macroeconomics, Boston.

Brunner, K. and A.H. Meltzer, ‘The uses of money: money in the theory of an exchange economy’ in: The American Economic Review 61-5 pp. 784-805. 

Burgess, S. and N. Janssen (2007), ‘Proposals to modify the measurement of broad money in the United Kingdom: a user cosultation’ in: Bank of England Quarterly Bulletin 2007-III, pp. 402-416. 

Nijboer, H. (2005), ‘Fashion and the early modern consumer revolution. A theoretical explortation and some evidence from seventeenth century Leeuwarden” in: Blonde, B e.a. (eds.) Retailers and consumer changes in early modern Europe, England, France, Italy and the low countries (Tour), pp 21-36.

The Cape town information is obtained from:http://www.tanap.net/content/activities/documents/Orphan_Chamber-Cape_of_Good_Hope/index.htm

The earliest accounts and probate inventories in this database show some glimpses how the most advanced market economy of the time – The Dutch and flemish farmers in combination with the Dutch East Indian company settled in a stone age economy (cause of death: ‘slain by a Lion’, ‘killed by the San’).  Some of the ninenteenth century ones are in english, all of them show how much material culture as well as much of economic life resembled life resembled life around the Norht Sea (except for the slaves and the large amount of stock per farm). The slaves more or less occupied the place of boarding labor on the farms around the Norht Sea (including Scotland), as there was no labor market, money wages could not buy labor.  The settlers had to find a non-market, non monetary solution (which, alas, consisted of slavery).

  1. Peter T
    December 30, 2010 at 5:29 am

    I had a look at one account of a barter economy and it occurred to me that there was in fact money involved – it was just not mentioned. And this is inevitable unless all transactions are complete and instantaneous. Otherwise one or both parties have incurred a debt – which is just a very narrow form of money (Tom owes me 50 kg of apples – I want bacon, so I transfer Tom’s debt to the pig farmer – who wants nails, so transfers it to the smith, who….).

    For an illustration see the transaction recorded under “The Price of a Slave” at http://www.reshafim.org.il/ad/egypt/timelines/topics/slavery.htm. A definite price is given in units of silver, but no silver changes hands. Is this a monetary transaction? I would say yes.

  2. merijnknibbe
    December 30, 2010 at 10:24 am

    thanks for this reference. An interesting footnote in the text you mention:

    ‘Twenty pieces of silver: ‘Essrim kessef’ (i.e. 20 silver in the Hebrew bible), there was no coined money during the latter half of the second millennium’

    i.e. the price for which Joseph was sold by his brothers, with the accompanying bible text:

    ‘Then there passed by Midianite merchantmen; and they drew and lifted up Joseph out of the pit, and sold Joseph to the Ishmeelites for twenty pieces of silver [2]; and they brought Joseph into Egypt’

    There were merchantmen, there were markets, ‘deben’ (almost a typo, I was inclined to write ‘debet’) and ‘kit’ were a units of account for amounts of gold and silver, gold and silver were means of exchange and hoards of value – this was clearly not a barter economy. A ‘unit’ of silver clearly was the unit of account, though coins of silver were not yet the actual means of exchange – pieces of silver (or copper?) might have done the trick. Coins are a kind of ‘commodification’ (I hate the phrase as its so ugly, but it serves its purpose) of precious metals: guaranteed (the stamp of the king or city) amounts of guaranteed metal which saves on transaction costs, this great enemy of market behavior and, especially, barter.

    It might well be that some kind of coin was first used to pay (some kind of) a standing army, every soldier getting the same guaranteed (the stamp of the city) amount of metal – but experts seem to differ on this one. Once money existed, it served the needs of merchants (and the tax man) – while it also enabled the rise of a labor market of at least day labor and a class of day laborers (the bible contains clear evidence on this as well as on price setting).

    But Egypt and the surrounding countries were rather advanced societies, with cities, agriculture, tax systems, written records and the like. Hunter-gatherer societies are the kind of societies I meant. (Almost) no trade there – though the little trade there was (flint!, shells, precious gems) might have been highly important. As in todays family households, the fast majority of the (division of) labor was however ‘traditional’ in these societies, based upon gender, family position, age, skills and social position – not upon market behavior and monetary budget constraints. No money needed to pay the stay at home mum/dad – though you do of course need money for shoppings and rent (I do, by the way, not believe in ‘implicit’ markets. To me, markets are almost by definition based upon explicit contracts/exchange agreements and specific amounts of money).

    What we see nowadays is a shift of childcare from the family household to the market – but such a process takes time. A whole new sector of the economy has to grow and to develop before such ‘market behavior’ can exist, while people have to change their attitudes – just like a labor market has to develop before wage labor can become important to, for instance, farming.

    So, I stick to the point that non-monetary economies are different from barter economies – non-monentary barter economies are not efficient ad effective enough to survive. Trade needs money (taxes can do with taxes in kind, by the way).

    I also fully agree with a man like Hayek on the idea that ‘traditional’ behavior (INCLUDING MARKET BEHAVIOR ITSELF) has often been shaped by historical developments. The cape town settlers are a case in point – their (very market oriented) households and farms do look, according to the probate inventories so much like those ‘at home’ in Flandria and the Netherlands. They took their ‘bourgeois’ attitudes with them. But there was one problem. Labor. The day labor and boarding labor needed for this style of farming was not available. There were people (the San) but no laborers, as the San did not know a labor class, or small farmers who could provide labor, or the European system of young adults providing boarding labor in other households. The San economy did not yet know ‘commodified’ labor and did not yet need money. After some time, the settlers solved this problem with slavery.

    Sideline – when you read Desert Flower of Waris Dirie, it is amazing how much the nomadic lifestyle of Somalian goatherds anno 1980 still resembled the lifestyle of Abraham, Isaac and Jacob – up to the quarrels at the well.

  3. December 31, 2010 at 6:15 am

    Why did you not include gold and silver in your example table? Certainly a form of money with a long history across many cultures. Air miles also would be a very common form of “money”.

    I would also take issue with your definition of “store of value”. I agree that “how long does a certain kind of money last” in the sense that does the money “expire” is a factor, but surely what is more important or at least commonly understood by “store of value” in reference to money is its ability to maintain purchasing power.

    How useful is to know/classify money as lasting 6 month or 60 years if at the end of that period it has been so inflated that it purchases very little?

  4. merijnknibbe
    December 31, 2010 at 2:00 pm

    Good idea, including gold (copper, some time from now?) as well as a volatility index, indicating changes in value (indeed, almost always down for fiat money).

    Groupons, which is essentially commodity backed money issued by a store (or cinema of whatever), have a limited lifetime (often: six months) by contract, to enhance the propensity to spend it – it might therefore be usefull to use ‘lastability’ as an indicator.

  5. Peter T
    January 1, 2011 at 2:44 am


    Take your four attributes:

    1. Means of exchange
    2. Store of value
    3. Unit of account
    4. Strings attached.

    I. Does not have to be universal – just widespread enough to cover ordinary transactions (even now I cannot use my Visa card everywhere, not all banknotes are accepted globally and so on)

    2. Just denotes some reasonable degree of longevity

    3. Some easily assessed, not easily faked unit, available in enough quantity to settle residual amounts, persistent, widely recognised.

    4. Qualifies the above.

    What’s the minimum needed to meet these requirements? Here history helps.

    Hunter-gatherers (and modern households and politicians, and …) track social rather than material debts. These are not readily transferable, often short-lived, and not easily assessed.

    The first “money” were transferable written debts, expressed first in units of grain or other commodities, and then in silver. They took the form of clay bulla, with a record on the outside duplicating a record on the inside. In effect, a sophisticated tally-stick (there are some examples in the British Museum). These circulated for 2000 odd years before coins, and still circulate in the form of computer digits. The actual silver was for settling residual amounts, after all the goods and services had been netted out (there never was, and never could be, enough silver or gold to cover all transactions).

    Note that the very large and sophisticated Chinese economy did without actual coins other than copper cash for small change until quite recently (the unit of account was the silver tael, but this was not a coin – just a weight of silver).

    The point is that any system of exchange where the actual good or services (rather than the social act) is the focus needs to record transactions unless they are instant and complete, and this record is proto-money. As soon as you start recording in some standard unit, you have actual money. Coins are, historically, a convenient afterthought.

    Since the laws of this universe do not allow acts to be instant and complete, a barter economy of things cannot and does not happen. All such barter involves either proto-money or actual money – something that meets the first three tests above.

    • Alice
      January 1, 2011 at 7:56 am

      Take your four attributes:

      1. Means of exchange
      2. Store of value
      3. Unit of account
      4. Strings attached.

      Item 4 should read “commission attached to items 1 to 3.”

  6. merijnknibbe
    January 2, 2011 at 11:56 am


    thanks for your highly interesting comments, which teach me quite a lot.

    The ‘transferable written debts’ remind me of what happened in Ireland during the bankstrikes around 1970, when the banks closed down for about one year. Yes, that did happen – and Ireland did not sink into the waves. People started to issue such transferable debts in Ireland too:


    I do not want to understate the importance of ‘too big to fail’ banks, but such examples show that they might be slightly less pivotal to the economy than sometimes assumed, in the end, everybody whose reputation is trusted can issue some kind of money. But I guess that in those days, Ireland was much closer to the ‘tally stick’ economy than it’s nowadays and that the Irish could fall back upon not yet entirely forgotten habits as well as pub-centerd networks of acquitance and commercial trust. Or am I getting into romatic wishfull thinking here?

    I know next to nothing of Assyrian money, but a Palin (not Sarah, but Michael) television series on the development of arithmetic showed that the ‘records on the outside’ of the clay bulla’s were the origin of written, abstract arithmetic. These accountants were the most influential economists ever! The discovery of ‘0’ also enhanced financial accounting and all the historical accounts I’ve seen are ample evidence to the idea that for a monetary system to work, a large part of the people have to know some arithmetic – money is a cultural thing, too. Unlke the Irish and the Assyrian debts, the debts and credits in these eighteenth century accounts were not very transferable: they could be inherited by children and spouses but a creditor could not sell them to somebody else (I do at least not know any examples of this). That’s were my (far from perfect) system kicks in: not all debt is created equal and it might be usefull to have a yard stick to measure the tally stick – in this case it is possible to distinguish the Irish and Assyrian debts with the help of the ‘means of exchange’ attribute (as well as the ‘strings attached’ – but than I’ve got to become boring about eighteenth century inheritance law, custody of orphans and the like).

    You’re obviously working on this. Have you published anything?

  7. Peter T
    January 5, 2011 at 7:44 am


    Just thinking about this, as the ordinary explanations do not seem to quite add up.

    I think we have a progression of increasing abstraction here, from:

    A owes B some defined goods/services (or B is entitled to these from A) – the “tallystick” stage; to

    A owes some person some defined goods/services – the “bulla” stage; to

    A owes some person goods/services equivalent to n units of account – the “letter of credit” stage; to

    Any vendor acknowledging the unit owes some person goods/services equivalent to n units of account – the “coined money” stage

    From my reading, units of account and coins are quite distinct. Many medieval accounts are in livres tournois, although there was never a coin equivalent. Also it seems rare for coins to contain metal to their face value, although a premium attached to coins which did not vary over time, such as the Venetian sequin.

    It also seems that settlement in actual metal was not usual. It was more an occasional test of solvency.

    I think this attested historical progression focuses attention on some neglected aspects of money – such as the role of debt in creating it (and so the role of debtors in maintaining the standard) and its role as a metaphor rather than a direct counterpart to actual goods/services. There may be a useful parallel with language here, but I’m still thinking about that.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.