Sunday, April 14, 2013

What is Money? Take 15,273


Defining money is so difficult and so contentious that it's almost not worth trying, but somehow I am still compelled to define and define again. (I feel the same about the concept of value). I've burned through quite a few definitions in the last year; the one I currently offer on the theory tab is "money is debt," which I no longer quite agree with, but don't want to replace until I'm fully acclimated to my current take. You would think I would be rushing to correct my error, but conveniently, my theory doesn't rely on the definition of money itself; it relies on the definition of the capacity to exchange, aka the aggregate limit to spending, or money times its potential velocity, in a given period of time. Because the capacity to exchange is an indivisible quantity, it doesn't particularly matter where I draw the line between money and velocity.

Still, the question haunts me, most importantly because my theory also assumes that price level expectations cause inflation, and price level expectations are partly determined by what real people think is "real money" and what people think is purely inflationary. There is also a very real difference between expansion in the capacity to exchange via increased velocity of money, and expansion in the capacity to exchange via an increase in money itself; whereas the former tends to occur because people have more real output (boats, shoes, and toys, oh my) to exchange, the correlation between real output and pure money expansion is usually much weaker, a point over-enthusiastically codified in the quantity of money theory. 


So what is money? Specifically, what is money in a fiat system, the type of system we live in today?  Before I offer my definition, let's look at some of the most popular definitions out there. Ultimately all of these definitions point to the same thing, the quantity we all know, love, and thirst after, and ultimately all of these definitions are vague or incomplete enough to be accurate. The question then becomes what is the most useful, or most direct definition of money?

I have sought to here categorize four popular definitions of money, and not to mimic any particular authors. Most economists actually choose some particular combination of these definitions, and the Chartalist and Credit Theories of Money overlap significantly.  I also have a feeling that if an MMT theorist ever reads this they'll come running after me with a pitchfork for reducing their definition to the inverse of MB. 

1) The Econ 101 definition

Money is.. 
  • A store of value 
  • A unit of account 
  • A medium of exchange 
The Econ 101 definition is pretty much a cop out, as it could explain a myriad of things, both fiat and commodity, and doesn't offer any insight into the fluctuation of prices and monetary aggregates. This IMF video is cute, but it most definitely doesn't cut it: 

In academic economics these characteristics are more often discussed as the functions of money, rather than as the components of a definition, but this is often the most advanced definition that the uninitiated get. 

2) The Central Bank definition 

Money is... 

The monetary base, also known as narrow money, central bank money, government money, or high powered money, is a liability of the central bank. Through open market operations, the central bank increases or decreases MB by increasing or decreasing the size of its balance sheet. M1, M2, M3 etc,  also known as commercial bank money or broad money, is MB multiplied by the fractional reserve banking system. 

The seemingly arbitrary distinctions between the various tranches of money in the central bank definition was one of the original inspirations for the capacity to exchange.

3) The Chartalist, or Modern Monetary Theory definition

Money is.. government deficit. Government spending creates money, and taxation destroys it. 

For someone familiar with the central bank definition of money, the MMT definition can be quite confusing, and seemingly off target, but when you think about how central bank creates money though open market operations it begins to make sense. Open market operations usually involve specifically the purchase of treasury bonds by the central bank; in other words one branch of government purchases debt from another branch of government, ie, borrows from no one. Thus any expansionary monetary policy can be seen as government deficit. Note, however, that expansionary monetary policy doesn't always correspond with a new issuance of debt, and that sometimes the central bank purchases other types of securities during open market operations, so obviously this flip definition isn't perfect. 

4) The Credit Theory of Money definition 

Money is.. Credit. 

Of the four definitions of money listed here I am most partial to this one, but it begs the question, if credit is money, can any average Joe who buys on credit create money? The answer to this, I think, was best expressed by Hyman Minsky in Stabilizing an Unstable Economy "everyone can create money; the problem is in getting it accepted." 

Wandering Statement of My Definition 

When anything is particularly confusing, it is generally useful to think about it in the simplest case, what I like to call the philospher's "state of nature." Imagine a world without money or organization of any sort, no governments, no central banks, no financial institutions at all, just a spattering of humans who can each make things, and who occasionally find it useful to swap the things they make for the things other people make.  They can swap things for things directly (barter) but this inconveniently requires both parties to have something the other desires at the same time. They can also swap things much more practically if one party issues an IOU in exchange for a real good or service, settles the debt by giving the holder of the IOU an equally valuable real good or service at some later date. 

In this economy I can imagine three basic transactions:
  1. Real good or service swapped for real good or service (barter) 
  2. IOU issued for real good or service 
  3. IOU swapped for a real good or service (For example, when the Jim, the holder of Anne's IOU, gives Diego Anne's IOU to pay for a real good or service. This may be easier than issuing a new IOU)
Now imagine that sometimes people default on their IOUs, and that some people default more than others. For instance, imagine that Timmy is a good-for-nothing loafer, and Tamara is diligent and hardworking. Everyone in the community will accept Tamara's IOU, but only a few people are willing to risk accepting Timmy's IOU.   To compensate for the greater risk, Timmy starts promising to settle his IOU with more real goods and services than he issued it for--ie, he promises to pay a higher rate of interest. 

Now I can imagine two more types of transactions: 
  1. IOU issued for an IOU (For example, if Tamara is willing to take Timmy's IOU, but Jim isn't, Timmy can issue an IOU for Tamara, and Tamara can issue an IOU for Jim, and Jim can give Timmy the real goods and services he wants. Provided that Timmy doesn't default, Tamara profits from the interest)*
  2. IOU swapped for an IOU (For example, if Anne is holding Timmy's IOU, and Diego is holding Tamara's IOU, but Anne decides she'd rather have a less risky and lower paying IOU, and Diego decides he'd rather have a more risky and higher paying IOU, they can swap)
*Note, I don't really distinguish between a new IOU issued for a issued new IOU, as in Timmy/Tamara's case, and a new IOU issued for a previously issued IOU, as would be the case if Diego was already holding Tamara's IOU, and agreed to accept a new IOU from Timmy so that he could profit from the interest. I see these two cases as essentially equivalent. 

So now we have five types of transactions:
  1. Real good or service swapped for real good or service (barter) 
  2. IOU issued for real good or service (credit creation)
  3. IOU swapped for a real good or service (purchase) 
  4. IOU issued for an IOU (arbitrage) 
  5. IOU swapped for an IOU (hedging) 
The first type of transaction is quite rare and is generally ignored by economics today; barter exchange does not factor into national accounts, is not tracked by financial markets, and for the most part does not enter the models of academic economists. 

The third type of transaction is a simple purchase or payment. This does not create money, but it does speed the velocity of money, or increase the number of times a single IOU is used to purchase value in a period of time. 

The fourth type of transaction is the issuance of a loan. This also does not create money, but does indirectly speeds the velocity of money by allowing the borrower to make purchases which, presumably the lender did not feel like making, and, presumably, the borrower would not have been able to make otherwise. This matching of purchasing power with purchasing desire is the basic value add of financial intermediation. 

The fifth type of transaction is the rest of modern finance; trading (already issued) stocks and bonds, securitizing and reselling consumer loans, diversifying portfolios, etc etc. This makes financial intermediation safer, by allowing lenders to hedge the risk of an IOU default. Because direct lending can be very dangerous for the lender and may not occur without the ability to hedge, this fifth type of transaction also indirectly speeds the velocity of money (but less directly than a loan). 

Only the second of these transactions, the issuance of an IOU in return for a real good or service, is the transaction that actually creates money. If you, for example, eat dinner at the local tub and put in on your tab, you just created money because you issued your IOU for a real good or service, a meal. However, like Timmy, your IOU isn't generally trusted. The bartender can't go around using your pub tab to buy furniture or fur coats, because no one would accept it.  (As Minsky said, "everyone can create money; the problem is in getting it accepted.") 

In a fiat system, the government's IOU is the only IOU trustworthy enough to be used as a general medium of exchange. This is probably a both consequence of the government's "monopoly on force," ie unique ability to tax, and an artifact of its regulation of commodity money before the fiat system.  Large and reputable institutions can also issue IOUs sufficiently trustworthy to be used as mediums of exchange, and have in the past, (ie, independent bank notes) but this rarely happens anymore. Today, most governments also issue their IOUs in a sneaky sort of way; instead of printing money flat out and giving it to, say, construction workers building government roads, or soldiers serving in the army, the government instead has the central bank issue a new liability to buy treasury bonds, uses the money made from those bonds to go out and buy real goods and services. 

Every IOU issued for a real good or service is multiplied by the particularl velocity of that IOU, a velocity directly increased by transactions of type 3 and indirectly increased by transactions of types 4 and 5. Government IOUs, dollars, euros, pounds, yen, often are multiplied hundreds of times in a year, whereas individual IOUs, like the pub tab, are rarely multiplied at all, and can be excluded from economic consideration without terribly distorting the overall calculation of the capacity to exchange. Thus my definition on the theory tab (Money is debt) is a bit too inclusive; it includes 4 as a type of money when really only 2 qualifies. 

Concise Statement of My Definition 

Money = an IOU offered in return for a real good or service, specifically a government IOU = a special type of credit = specially determined by the government's unique ability to tax = central bank liabilities = a quantity that acts as a medium of exchange, unit of account, and store of value. Or: 

In conclusion, my definition is not at all revolutionary, but I do think it's important to understand why, on a deeper level, I think it is true. 


  1. I wouldn't say "money is an IOU".
    Rather, in my opionion money is a voucher to buy whatever is on the market.
    It is issued by central banks (high powered money) and commercial banks (most of the money supply!) on behalf of the whole economy.

    With fiat money it makes no sense to say the bank owes the debtor anything. Actually, nobody owes the moneyholder anything, not even the economy as a whole.
    However, since the other economic agents themselves need these vouchers to purchase goods , everybody will accept them as payment.

    Fiat money has to be credit, because if it were'nt, the first receiver ("A") would live as a freeloader (like a money counterfeiter).
    Only the fact that he himself (at a later date) needs money to pay back his credit forces him to produce something himself to sell on the market.

    Since the following receiver of money ("B") exchanges the money from A for whatever he (B) has sold, he (and all the others after him) are under no obligation to pay back someone, or to spend the money.

    When banks recollect the credit, they obviously do that, on the micro level, in their own interest.
    But the "macro" logic behind it is that they force him to pay back "B" what he has received from him.

    B has sold goods to A and received a piece of paper with no intrinsic value.
    But then, after some time, A needs this paper back (plus some) to repay his credit.
    Now the exchange works in the other direction:
    The "worthless" Paper returns from B to A, and products go from A to B.
    And once the "piece of paper" has been returned to the bank, the money is automatically annihilated.

    So in effect we have seen a barter, facilitated by the help of a voucher. Which by being returned to the bank is automatically voided.

    The state makes no exception and does not have any privileges (under normal circumstances, not like Germany in 1923, or Zimbabwe in our days). The only exception are coins, which indeed the state creates. (Their emission is a form of taxation, because the state does not give anything in return - other than coins of mostly low value - for the purchases he makes with these coins.)

    The chartalist theories of money are therefore BS:
    - The state does not (under normal circumstances) create money (coins apart)
    - And most certainly money is not voided, when people pay their taxes. Instead, they (involuntarily) hand over (some of) their "vouchers" to the government, without receiving a direct exchange in goods. (But of course they do profit indirectly.) And now the state uses these vouchers the same way any citizen would: To purchase goods - or to redeem his debt. Only in this case (and only when the creditor is a bank) is the money annihilated. But that has nothing whatsoever to do with the state.

  2. Of course, you might give a different description of the transactions between A and B:

    When A buys something from B, he is giving him an "IOU" guaranteed by the bank which has issued the credit.
    When, later, B makes a purchase from A, he hands him back his "IOU" and (only) then A has effectively "paid" B.

    However, for A it's not really his own IOU, but the bank's. So he returns it to the bank and thereby redeems debt.
    So, with that model in mind you might call money an IOU.

    However, I still hold "voucher" to be the better term and the better concept. Because in reality of course it is not the same money ("piece of paper") that will be "returned" to A, and it won't normally come from B.

  3. Thanks for the comment!

    I think I follow. Yes, your absolutely right that fiat money in most societies is created by central banks, and that "voucher" is a more appropriate word to describe this creation than "IOU." I was interested when writing this post in thinking about how money might arise independent of financial institutions and the state. Without a bank, that first transaction would involve the transfer of A's IOU, not a voucher borrowed by A from the bank.

    If we are talking about banks, I think the point of interest is the fact that B accepts the bank voucher from A. For B to do that, I think he either has to really trust A... that A will make something worth purchasing, and that will accept the voucher back... in which case the voucher is really no different than an IOU... or he has to really trust the bank, in that the bank will be able to give him something he wants in return for the voucher. But why does he trust the bank?

    If there had never been a bank before. and I were B, I would trust the bank only if the bank was rich, was known for keeping his word, and has promised to vouch for A. I would be trusting the bank's reputation to make good on his IOU.

    Of course, today we trust the bank because everyone trusts the bank, and therefore everyone takes the bank's vouchers. The personal reputation of the bank no longer matters. That's the magical thing about money. If enough people put their faith in it, it doesn't really matter where it came from.

    Not sure that makes sense.. but hope so!

    1. Of course our present day monetary system has evolved over time. I guess any reconstruction would have to try and reconstruct the historical steps of this development.

      But today, as it is, you don't trust the bank in order to accept "their" money. Because the bank does not redeem it. This, of course, was different in the olden days when a banknote was a promissory bill for the delivery of gold. (Though in fact even then probably rather illusionary.)

      What you put your trust in when you accept a banknote is that later one someone else will accept it from you: its function as a voucher, which you can redeem anywhere on the market.

    2. Come to think of it: Yes, there is a certain amount of trust involved.
      In the central bank or, more precisely, in that the central bank won't just print money and give it away (to the state), thereby depreciating its voucher-value (because there are more vouchers than goods out there).

      This is, of course, what happened in Germany in and after WW I (hyperinflation of 1923) and II (when in 1948 the old Reichsmark were exchanged, for, I think, less than 1/10 th of their nominal value, in to DM).
      Between the end of WW II and the introduction of the Deutschmark many would not accept the Reichsmark any more. So cigarettes became a means of exchange. (Then being rare in Germany, because we grow very little tobacco and in those days for obvious reasons could not afford to import it either.)