It’s the Endogeneity, [Redacted]

A few weeks ago, just when I was trying to sort out my ideas on whether, and, if so, how, the Chinese engage in currency manipulation (here, here, and here), Scott Sumner started another one of his periodic internet dustups (continued here, here, and here) this one about whether the medium of account or the medium of exchange is the essential characteristic of money, and whether monetary disequilibrium is the result of a shock to the medium of account or to the medium exchange? Here’s how Scott put it (here):

Money is also that thing we put in monetary models of the price level and the business cycle.  That . . . raises the question of whether the price level is determined by shocks to the medium of exchange, or shocks to the medium of account.  Once we answer that question, the business cycle problem will also be solved, as we all agree that unanticipated price level shocks can trigger business cycles.

Scott answers the question unequivocally in favor of the medium of account. When we say that money matters, Scott thinks that what we mean is that the medium of account (and only the medium of account) matters. The medium of exchange is just an epiphenomenon (or something of that ilk), because often the medium of exchange just happens to be the medium of account as well. However, Scott maintains, the price level depends on the medium of account, and because the price level (in a world of sticky prices and wages) has real effects on output and employment, it is the medium-of-account characteristic of money that  is analytically crucial.  (I don’t like “sticky price” talk, as I have observed from time to time on this blog. As Scott, himself, might put it: you can’t reason from a price (non-)change, at least not without specifying what it is that is causing prices to be sticky and without explaining what would characterize a non-sticky price. But that’s a topic for a future post, maybe).

And while I am on a digression, let me also say a word or two about the terminology. A medium of account refers to the ultimate standard of value; it could be gold or silver or copper or dollars or pounds. All prices for monetary exchange are quoted in terms of the medium of account. In the US, the standard of value has at various times been silver, gold, and dollars. When the dollar is defined in terms of some commodity (e.g., gold or silver), dollars may or may not be the medium of account, depending on whether the definition is tied to an operational method of implementing the definition. That’s why, under the Bretton Woods system, the nominal definition of the dollar — one-35th of an ounce of gold — was a notional definition with no operational means of implementation, inasmuch as American citizens (with a small number of approved exceptions) were prohibited from owning gold, so that only foreign central banks had a quasi-legal right to convert dollars into gold, but, with the exception of those pesky, gold-obsessed, French, no foreign central bank was brazen enough to actually try to exercise its right to convert dollars into gold, at least not whenever doing so might incur the displeasure of the American government. A unit of account refers to a particular amount of gold that defines a standard, e.g., a dollar or a pound. If the dollar is the ultimate medium of account, then medium of account and the unit of account are identical. But if the dollar is defined in terms of gold, then gold is the medium account while the dollar is a unit of account (i.e., the name assigned to a specific quantity of gold).

Scott provoked the ire of blogging heavyweights Nick Rowe and Bill Woolsey (not to mention some heated comments on his own blog) who insist that the any monetary disturbance must be associated with an excess supply of, or an excess demand for, the medium of exchange. Now the truth is that I am basically in agreement with Scott in all of this, but, as usual, when I agree with Scott (about 97% of the time, at least about monetary theory and policy), there is something that I can find to disagree with him about. This time is no different, so let me explain why I think Scott is pretty much on target, but also where Scott may also have gone off track.

Rather than work through the analysis in terms of a medium of account and a medium of exchange, I prefer to talk about outside money and inside money. Outside money is either a real commodity like gold, also functioning as a medium of exchange and thus combining both the medium-of-exchange and the medium-of-account functions, or it is a fiat money that can only be issued by the state. (For the latter proposition I am relying on the proposition (theorem?) that only the state, but not private creators of money, can impart value to an inconvertible money.) The value of an outside money is determined by the total stock in existence (whether devoted to real or monetary uses) and the total demand (real and monetary) for it. Since, by definition, all prices are quoted in terms of the medium of account and the price of something in terms of itself must be unity, changes in the value of the medium of account must correspond to changes in the money prices of everything else, which are quoted in terms of the medium of account. There may be cases in which the medium of account is abstract so that prices are quoted in terms of the abstract medium of account, but in such cases there is a fixed relationship between the abstract medium of account and the real medium of account. Prices in Great Britain were once quoted in guineas, which originally was an actual coin, but continued to be quoted in guineas even after guineas stopped circulating. But there was a fixed relationship between pounds and guineas: 1 guinea = 1.05 pounds.

I understand Scott to be saying that the price level is determined in the market for the outside money. The outside money can be a real commodity, as it was under a metallic standard like the gold or silver standard, or it can be a fiat money issued by the government, like the dollar when it is not convertible into gold or silver. This is certainly right. Changes in the price level undoubtedly result from changes in the value of outside money, aka the medium of account. When Nick Rowe and Bill Woolsey argue that changes in the price level and other instances of monetary disequilibrium are the result of excess supplies or excess demands for the medium of exchange, they can have in mind only two possible situations. First, that there is an excess monetary demand for, or excess supply of, outside money. But that situation does not distinguish their position from Scott’s, because outside money is both a medium of exchange and a medium of account. The other possible situation is that there is zero excess demand for outside money, but there is an excess demand for, or an excess supply of, inside money.

Let’s unpack what it means to say that there is an excess demand for, or an excess supply of, inside money. By inside money, I mean money that is created by banks or by bank-like financial institutions (money market funds) that can be used to settle debts associated with the purchase and sale of goods, services, and assets. Inside money is created in the process of lending by banks when they create deposits or credit lines that borrowers can spend or hold as desired. And inside money is almost always convertible unit for unit with some outside money.  In modern economies, most of the money actually used in executing transactions is inside money produced by banks and other financial intermediaries. Nick Rowe and Bill Woolsey and many other really smart economists believe that the source of monetary disequilibrium causing changes in the price level and in real output and employment is an excess demand for, or an excess supply of, inside money. Why? Because when people have less money in their bank accounts than they want (i.e., given their income and wealth and other determinants of their demand to hold money), they reduce their spending in an attempt to increase their cash holdings, thereby causing a reduction in both nominal and real incomes until, at the reduced level of nominal income, the total amount of inside money in existence matches the amount of inside money that people want to hold in the aggregate. The mechanism causing this reduction in nominal income presupposes that the fixed amount of inside money in existence is exogenously determined; once created, it stays in existence. Since the amount of inside money can’t change, it is the rest of the economy that has to adjust to whatever quantity of inside money the banks have, in their wisdom (or their folly), decided to create. This result is often described as the hot potato effect. Somebody has to hold the hot potato, but no one wants to, so it gets passed from one person to the next. (Sorry, but the metaphor works in only one direction.)

But not everyone agrees with this view of how the quantity of inside money is determined. There are those (like Scott and me) who believe that the quantity of inside money created by the banks is not some fixed amount that bears no relationship to the demand of the public to hold it, but that the incentives of the banks to create inside money change as the demand of the public to hold inside money changes. In other words, the quantity of inside money is determined endogenously. (I have discussed this mechanism at greater length here, here, here, and here.) This view of how banks create inside money goes back at least to Adam Smith in the Wealth of Nations. Almost 70 years later, it was restated in greater detail and with greater rigor by John Fullarton in his 1844 book On the Regulation of Currencies, in which he propounded his Law of Reflux. Over 100 years after Fullarton, the Smith-Fullarton view was brilliantly rediscovered, and further refined, by James Tobin, apparently under the misapprehension that he was propounding a “New View,” in his wonderful 1962 essay “Commercial Banks as Creators of Money.”

According to the “New View,” if there is an excess demand for, or excess supply of, money, there is a market mechanism by which the banks are induced to bring the amount of inside money that they have created into closer correspondence with the amount of money that the public wants to hold. If banks change the amount of inside money that they create when the amount of inside money demanded by the public doesn’t match the amount in existence, then nominal income doesn’t have to change at all (or at least not as much as it otherwise would have) to eliminate the excess demand for, or the excess supply of, inside money. So when Scott says that the medium of exchange is not important for changes in prices or for business cycles, what I think he means is that the endogeneity of inside money makes it unnecessary for an economy to undergo a significant change in nominal income to restore monetary equilibrium.

There’s just one problem: Scott offers another, possibly different, explanation than the one that I have just given. Scott says that we rarely observe an excess demand for, or an excess supply of, the medium of exchange. Now the reason that we rarely observe that an excess demand for, or an excess supply of, the medium of exchange could be because of the endogeneity of the supply of inside money, in which case, I have no problem with what Scott is saying. However, to support his position that we rarely observe an excess demand for the medium of exchange, he says that anyone can go to an ATM machine and draw out more cash. But that argument is irrelevant for two reasons. First, because what we are (or should be) talking about is an excess demand for inside money (i.e., bank deposits) not an excess demand for currency (i.e., outside money). And second, the demand for money is funny, because, as a medium of exchange, money is always circulating, so that it is relatively easy for most people to accumulate or decumulate cash, either currency or deposits, over a short period. But when we talk about the demand for money what we usually mean is not the amount of money in our bank account or in our wallet at a particular moment, but the average amount that we want to hold over a non-trivial period of time. Just because we almost never observe a situation in which people are literally unable to find cash does not mean that people are always on their long-run money demand curves.

So whether Nick Rowe and Bill Woolsey are right that inflation and recession are caused by a monetary disequilibrium involving an excess demand for, or an excess supply of, the medium of exchange, or whether Scott Sumner is right that monetary disequilibrium is caused by an excess demand for, or an excess supply of, the medium of account depends on whether the supply of inside money is endogenous or exogenous. There are certain monetary regimes in which various regulations, such as restrictions on the payment of interest on deposits, may gum up the mechanism (the adjustment of interest rates on deposits) by which market forces determine the quantity of inside money thereby making the supply of inside money exogenous over fairly long periods of time. That was what the US monetary system was like after the Great Depression until the 1980s when those regulations lost effectiveness because of financial innovations designed to circumvent the regulations.  As a result the regulations were largely lifted, though the deregulatory process introduced a whole host of perverse incentives that helped get us into deep trouble further down the road. The monetary regime from about 1935 to 1980 was the kind of system in which the correct way to think about money is the way Nick Rowe and Bill Woolsey do, a world of exogenous money.  But, one way or another, for better or for worse, that world is gone.  Endogeneity of the supply of inside money is here to stay.  Better get used to it.

About these ads

29 Responses to “It’s the Endogeneity, [Redacted]”


  1. 1 Frank Restly November 25, 2012 at 9:04 pm

    A medium of account is what it sounds like, an accounting method. The receipt that you receive at the grocery store is a medium of account. The receipt itself has no intrinsic value, it is simply a record that a transaction took place. It has legal value in the sense that if a disagreement between the buyer and seller occurs at date after purchase, both the buyer and seller have a document that describes the transaction agreement – for instance I get a gallon of milk from grocer in exchange for $3.00.

    A medium of account can exist without a medium of exchange. In a barter system, both parties may still want a written record of a transaction in case either party misrepresents what is being offered. And so a medium of account can be created in any transaction. The change in the value of the medium of account is a direct function of the nominal value of the transactions:

    Medium of account, dM / dt = k * PQ

    As time passes, the voracity of many legal claims diminish ( I can hardly blame the grocer for bad milk if I wait two months before drinking it ). And so k represents a time function for the legal value that the medium of account holds.

    Grocery receipts are just one example of a medium of account. Another would be the equity or debt of a company.

    A medium of exchange is what it sounds like, a medium that is accepted in exchange for goods and services. In this case the change in the value of the medium of exchange is a function of both the nominal value of transactions as well as the velocity of the existing monetary stock:

    Medium of exchange, M = PQ / V

    A store of value is what it sounds like. It is a medium that holds it value relative to the change in a price level:

    Store of value, dM / M = dP / P or ln (M) = ln (P)

  2. 2 David Beckworth November 25, 2012 at 10:24 pm

    David, I share Nick and Bill’s view and it doesn’t require that inside money always be exogenously determined. In fact, it allows for the endogenous destruction of money. That, in fact, is typically a big reason for an excess demand for money. Financial firms are cutting back on intermediation–making new money assets–at the very time the demand for such assets is high.

    Consider, for example, a case where a debtor is paying down his loan and does so by paying out of his checking account. Also assume the bank is cutting back on making new loans because of expected economic weakness. The reduction in the deposit account is not being offset by an increase in new lending. Thus for a given demand for money, there is now an excess money demand. This excess money demand occurred with an endogenous reduction in the money supply. And it is much easier to show empirically that excess demand for inside money is tied closely to recessions. See here for example.

    Finally, I don’t see why there can’t be excess demand for outside and inside money at the same time.

  3. 3 Mike Sproul November 26, 2012 at 7:51 am

    Gold does not show up on the liability side of anyone’s balance sheet, but paper/deposit money, whether convertible or inconvertible, does show up on the liability side of its issuer’s balance sheet, whether that issuer is a private or government bank.

    So while we’re all fine with the proposition that digging new gold out of the ground will reduce the value of gold, it is a mistake to extend that argument to say that the value of the US dollar will fall as the Fed issues more of them. After all, as a new dollar appears on the liability side of the Fed’s balance sheet, a dollar’s worth of bonds will simultaneously appear on the asset side of the Fed’s balance sheet. This is plainly true under gold convertibility, but it is also true when the Fed suspends gold convertibility and maintains only bond convertibility, using its bonds rather than its gold to buy back its dollars.

    We need to change (or dump) our definitions of inside/outside money. Gold is outside money. Paper/deposit moneys are inside money, whether convertible or not. (And remember that convertibility is a matter of degree. All banks, public and private, issue money that falls somewhere in the middle of the convertible/inconvertible spectrum.)

    The defining characteristic of inside money is that it appears on the liability side of its issuer’s balance sheet, and as such its value depends on its issuer’s assets/liabilities, not just on the quantity issued.

  4. 4 JP Koning November 26, 2012 at 9:52 am

    “And while I am on a digression, let me also say a word or two about the terminology.”

    Defining definitions isn’t digression, it’s core. We all have to be on the same page for discussion to make sense.

    As with Scott’s post, I’m still confused about the idea of a medium of account. For instance, you say that “All prices for monetary exchange are quoted in terms of the medium of account.” You also say that “…if the dollar is defined in terms of gold, then gold is the medium account while the dollar is a unit of account.” We know that US shopkeepers in the 1920s quoted prices in dollars, but the dollar was defined as 25.8 grains of gold. So what was the medium of account in the 20s? According to your first sentence, it would be dollars, but according to your second, it would be gold. It seems to me that you’re working with two different meanings for the same word.

    Secondly, you seem to be saying that even if the dollar is defined in terms of gold, we can only say that gold is the medium of account if that definition is implemented by some operational mechanism (like universal convertibility). But take an example in which the dollar is defined in terms of a given quantity of gold, dollars are irredeemable, and the price of dollars is kept at gold parity by open market operations. Is not gold the medium of account? Similarly, you seem to say that the medium of account was not gold during Bretton Woods because locals could not convert their dollars into gold, only foreigners could. But much like the open market operations example, a limited version of redeemablity was sufficient to keep the dollar tethered to gold. It seems somewhat arbitrary to cease calling something a medium of account because the operational mechanism changes.

    I think the best way to go about this would be to give examples. What was the medium of account before the Fed? What was the medium of account from 1914 to the suspension of gold convertibility in early 1933? During the six or so months before FDR redefined the dollar from 25.8 grains to 15.7 grains in 1934? What was the medium of account during Bretton Woods? What about from 1968 to 1971 when there was an official gold price for central banks but an unofficial market price? What about when Nixon finally cut the gold window in 1971? What is today’s medium of account? What if the US pegged its currency to some other currency?

  5. 5 JP Koning November 26, 2012 at 10:32 am

    “Prices in Great Britain were once quoted in guineas, which originally was an actual coin, but continued to be quoted in guineas even after guineas stopped circulating. But there was a fixed relationship between pounds and guineas: 1 guinea = 1.05 pounds.”

    Here’s another example. The pound was defined as a certain number of grains of gold. Shilling and pence each had less grains in them. Guineas were a coin that represented 1.05 pounds. We’ll say that prices in Great Britain were quoted in guineas. Is the medium of account here the price in which prices are quoted (guineas), or is the medium of account the definition of the pound (x gold grains)?

  6. 6 Luis H Arroyo November 26, 2012 at 10:38 am

    Good questions, JP Koning. I Don’t understand the complete idea. I think there should be simpler way to see it.
    I can understand the endogeneity, but cannot see relation with inside / outside Money. I can see that because endogeneity, FED doesn”t control al Money supply. But not any more. If there are other consequences, can’t catch.

  7. 7 David Glasner November 26, 2012 at 1:17 pm

    Frank, You are free to define “medium of account” however you want to. But your definition seems to be different from mine. An accounting method is not the same as how prices are expressed. A medium of account determines the units in which prices are expressed or quoted. So I think that we are just talking past each other now.

    David, I think that I agree with you. Inside money and outside money are substitutes, and in recessions and financial crises, there is a strong tendency for people to lose confidence in inside money and convert their holdings of inside money into outside money. But it isn’t clear to me that there is an excess demand for inside money, there is a shift in demand from inside to outside money which drives up the value of outside money (causing deflation) unless there is an elastic supply of outside money (which is true for a small open economy under the gold standard, or, for that matter, a dollar standard), but not for a large open economy or a closed economy. But, yes, there could be an excess demand at the same time for both outside and inside money. In that case, however, it is the excess demand for outside money that is producing the deflation.

    Mike, I understand (I think) what you are saying. But when the Fed exchanges dollar (n + 1) for a bond, hasn’t it diluted the value of all the n dollars that were previously issued? Isn’t the aggregate value of the (n + 1) dollars is exactly equal to aggregate value of the n dollars?

    JP, I got into trouble with Noah Smith recently for dismissing the importance of definitions, so I guess I have not learned my lesson. Definitions are only important for purposes of disambiguation.

    Gold was the medium of account. A dollar was simply the name assigned to a particular amount of gold. Gold is the medium of account, the dollar identifies a particular quantity of gold and assigns the name dollar to it.
    I agree that it is possible to keep the dollar equal to a particular value of gold by way of indirect rather than direct convertibility, but that was not the case under the Bretton Woods system. There was no free market in gold, the market itself was managed by the central banks. Do a search on the London gold pool.

    The medium of account before 1914 (the Fed didn’t change the medium of account, it was the breakdown of the gold standard when WWI started that changed everything). Gold convertibility was restored after WWI, but it wasn’t clear whether gold was determining the value of the dollar or the dollar was determining the value of gold. As more countries joined the gold standard, the US lost control over the value of gold in the late 1920s and gold became the medium of account with disastrous consequences. After the collapse of the gold standard in the 1930s, the dollar was the medium of account, but gold still was a reference point, so the system was difficult to describe unambiguously, but I would still say that the dollar was the medium of account, a state of affairs that did not change under Bretton Woods, because there was no free market for gold. Because pegged exchange rates generally have no legal standing but are the result of a free choice of the monetary authority, I don’t think it means that the medium of account is not the domestic fiat currency, as long as prices are generally quoted in terms of that currency.

    About guineas, under a gold standard, the medium of account is gold, the unit of account can be either pounds or guineas or both depending on which unit is being used for purposes of price quotation.

    Luis, I hope this helps, but I agree it’s complicated, and, as JP’s questions illustrate, there are plenty of grey areas.

  8. 8 Frank Restly November 26, 2012 at 2:35 pm

    David,

    “Frank, You are free to define “medium of account” however you want to. But your definition seems to be different from mine. An accounting method is not the same as how prices are expressed.”

    Prices are expressed in a unit of account (dollars, pounds, pesos, etc.) not a medium of account. We could invent a unit of account (call it the uneasy-coin), and we could in our collective heads decide what the value of every durable and non-durable good is relative to one uneasy-coin. That does not mean that a single uneasy-coin is ever minted.

    If we instead start minting a medium of account (as opposed to a medium of exchange) called an uneasy-coin then we must also invent some way for that medium of account to have any value (otherwise it is just worthless pieces of metal with a pretty pictures). Since our medium of account will not be used directly as a medium of exchange, the other way it can have value is if we by legal edict bestow value onto it. One way that a government bestows value upon a medium of account is by requiring that taxes are paid in that medium.

    The way I look at things, a medium of account receives its value from a legal authority or government. The value is bestowed upon the medium (hence I used a k multiple on PQ to describe the value of all of the medium).

    A medium of exchange receives its value by mutual agreement between two parties that said medium is acceptable for payment. The value of the medium of exchange relative to the goods that it is buying is arrived at through the bargaining and transaction process.

    From the paragraphs above you seem to imply that no transactions were ever settled directly in gold (gold as medium of exchange). I do not believe that was the case. Gold as a medium of exchange existed prior to paper currencies.

  9. 9 Mike Sproul November 26, 2012 at 2:39 pm

    David:

    There is normally no dilution. Start with a central bank that has 100 oz of silver backing 100 bank notes (dollars). Then let that bank print 1 more dollar and use the dollar to buy 1 more oz. Now we have 101 oz. backing $101, and each dollar is still worth 1 oz.

    Next, instead of using that last dollar to buy 1 oz, let the bank use the dollar to buy a bond that is worth 1 oz. You still have assets worth 101 oz. backing $101, and the dollar is still worth 1 oz.

    Next, instead of that last dollar being spent on a bond worth 1 oz, let it be spent on a bond worth $1 (i.e., denominated in $, not oz.). Nothing changes. You still have stuff worth 101 oz backing $101.

    Dilution only happens when the central bank fails to take adequate backing in exchange for the dollars it issues. For example, if that last dollar was spent on candy that was eaten by the bank chairman, then the bank would have 100 oz worth of assets backing $101, and the dollar would lose 1% of its value because of dilution.

    Only the silver should be called outside money. The paper dollars issued by the central bank can be called inside money, just like the checking account dollars issued by private banks. (But I think the silver should be called base money, while the paper and checking account dollars should both be called derivative moneys.)

  10. 10 JP Koning November 26, 2012 at 6:08 pm

    Thanks for the definitions, David. Let me go read the rest now.

    You mention the London Gold Pool, I’ve actually written a few articles on the subject, here and here.

    I find it odd that when gold convertibility was abandoned, the medium of account switched from being gold to the dollar. Isn’t it nonsensical to define the dollar as the dollar? Why not just say that there was no medium of account? Is it necessary to have a medium of account?

  11. 11 Nick Rowe November 26, 2012 at 8:45 pm

    David: Imagine a world in which there is a medium of account, but no medium of exchange. Because people barter anything with everything. Assume that all prices are fixed in terms of the medium of account. And that all prices are fixed too high, so there is an excess demand for the medium of account, but that all other relative prices are correct. Can there be a recession?

    I say no. Unemployed workers would just barter their labour for the output produced by firms.

    Now imagine a world where one good is both MOA and MOE. All prices are fixed, except the price of peanuts, which is perfectly flexible. So the market for peanuts always clears. There can be a recession. Nobody would say that there cannot be an excess demand for money because you could always get more money by selling peanuts.

    Now assume that the price of peanuts is fixed by the central bank, by making money convertible on demand into peanuts. So the supply of money is endogenous. There can be a recession.

    Now replace “peanuts” with “outside money that serves as MOA”.

  12. 12 HJC November 28, 2012 at 9:30 pm

    David: Perhaps the two opposing views (above) can be reconciled (somewhat). The quantity of inside money changes when there is a change in the quantity of bank lending. The demand for deposits cannot directly change the supply of deposits, only indirectly by affecting the demand and supply for loans. So, for Nick and Bill, there can be an excess demand for inside money (deposits), but with no increase in bank lending this excess demand cannot be met – the supply is fixed or even falling. For you and Scott, the supply is not fixed, but responds to the demand of public for the creation of new inside money (as bank lending). The key is to realise that the inside money changes are driven by the asset side of the banking system’s balance sheet, not the liability side.

    Nick: In your example where “one good is both MOA and MOE”, is that good supposed to be peanuts, or just any other good?

  13. 13 Becky Hargrove November 29, 2012 at 11:02 am

    For me the medium of account is extremely important: it is where the real struggle exists for services valuations and definitions, in the context of assets and wealth creation.

  14. 14 David Glasner November 29, 2012 at 2:02 pm

    Frank, Often the medium of account and the unit of account are identical. But it is possible under a gold standard that there are different units of account representing different units all ultimately defined in terms of gold, which is the ultimate medium of account. At least that’s how I am understanding the terms. I agree that gold has in the past served as both a medium of exchange and as a medium of account. But it is possible to envision a system in which gold no longer served as a medium of exchange even though it remained the medium of account. That in fact is the direction in which the gold standard was evolving in the nineteenth and early twentieth centuries.

    Mike, If the last dollar is used to buy an interesting-bearing bond issued by the government, there is one more dollar of currency and one less bond outstanding. The government and the central bank will not sell back the bond that they acquired, so the public can’t exchange the extra dollar of currency for the bond. So how will the extra dollar be willingly held if there is no dilution of its value given that the real demand for cash balances has not changed?

    JP, But prices are being quoted in dollars and dollars aren’t convertible into gold. By default the dollar must be the medium (and unit) of account.

  15. 15 Frank Restly November 29, 2012 at 4:20 pm

    David,

    “Frank, often the medium of account and the unit of account are identical.”

    No.

    http://www.thefreedictionary.com/medium

    Medium: An intervening substance through which something else is transmitted or carried on.

    http://www.thefreedictionary.com/unit

    Unit: A precisely specified quantity in terms of which the magnitudes of other quantities of the same kind can be stated.

    In the case of gold – medium is gold, unit is pound or ounce. In the case of paper currencies, medium is paper, unit is dollars, pounds, or pesos.

  16. 16 Mike Sproul November 29, 2012 at 6:49 pm

    David:

    Suppose that last dollar was spent by the central bank on a foreign government bond, just to minimize complications. Granting your assumption that demand for real cash balances is unchanged, the simplest answer is that as the central bank issues that last paper dollar, some other dollar will reflux to its issuer. That refluxing dollar might be a checking account dollar issued by a bank, a credit card dollar issued by mastercard, a walmart gift card dollar, etc.

    But what if these channels of reflux don’t work for some reason? The central bank has issued $101 to people who wanted only $100, and it refuses to allow any of its dollars to reflux in exchange for silver, bonds, furniture, or anything else. Then what happens?

    The answer is that the central bank’s refusal to accept a reflux of dollars is an effective default, a withdrawal of backing. If people are suddenly told that the central bank will never again redeem any dollar for any of its assets, then all 101 outstanding dollars will lose all value. But the central bank will probably not do something so extreme. It will probably just devalue, and start allowing its dollars to reflux in exchange for assets worth .99 oz. That way the $101 outstanding will have a real value equal to 100 of the old dollars. Of course, the bank has an extra 1 oz of assets because of its default.

  17. 17 JP Koning November 29, 2012 at 7:57 pm

    David, I’ve read through the rest of your piece and get where you’re coming from.

    I can’t help but feel that the phrase “medium of account” got recruited unnecessarily into the debate. There have been some definitional hassles – at the outset I think Scott mixed up the unit-of-account and medium-of-account. It’s been confusing to follow everything, but I have a feeling that if you all locked yourselves together in a room for a week and discussed the issues, you’d walk out on the same page.

  18. 18 J.V. Dubois November 30, 2012 at 6:54 am

    First, thanks for this article. I see this topic as very important and interesting enough to get more attention.

    Anyways, I still think that somewhere at the core of this issue lies misunderstanding about the definition of what MOA really is. For instance everybody now agrees that in USA “dollar'” is both: MOA (a quantifiable thing that exists either in paper or electronic form) a Unit of Account (a pure concept, a unit in which we measure the MOA). It also serves as MOE.

    However reading what you say do you think it makes equal sense to say that not Dollar, but “CPI consumption basket” is the “ultimate” MOA? It just gets needlessly complicated. I say that in its core, what MOA really is about is relative prices. The selection of actual medium (gold, silver, CPI basket, central bank notes) and of unit (ounces, kilograms, dozens) than serve as an anchor for relative prices.

    And this is the biggest disadvantage of thinking about pure “MOA”. If for instance we say that “medium” of account is something not related to market exchange, like for instance “hydrogen on Jupiter” (or some other commodity that one cannot obtain on market, or that can be obtained only on very dysfunctional market), in such a case MOA gets disconnected from the system of relative prices and it is then becoming less and less useful in its main function – to measure value of one thing available on market compared to something else.

    So saying all that and having this explanation of MOA in our heads, saying that shocks in the supply and demand for MOA cause recession is the same as saying that shocks in relative prices cause recessions.

    For instance let’s say that If our medium of account is steel and unit of account is 1kg of steel. Let’s say that in equilibrium, 1 hour of labor should buy 30 kg of coal or 100 kg of iron ore … and in some roundabout way it should buy us 10 kg of steel. My first question is – everyone who thinks it is MOA that causes recession – what is the mechanism for it? I can imagine a shock that now 1 hour of labor buys (or produces) only 5 kg of steel and labor prices are sticky. Is this the story you guys have in your mind of how recessions come about? That it is real shocks causing issues with relative prices and real economy having trouble adjusting for it?

    For now I will not touch the whole subject of how adjusting for this shock is (it think) impossible and let’s pretend that this MOA is valid and it can be solved by some mechanism. Even then it is still very different from how other people think how demand recession start. The mechanism where MOE plays prominent role because it enables people to hoard MOE. No such thing is possible with pure MOA.

  19. 19 Frank Restly November 30, 2012 at 8:45 am

    J.V.

    “I say that in its core, what MOA really is about is relative prices.”

    But which relative prices? I think that is what is really the key issue. If we are talking about the relative price of an orange versus an apple then the medium of account could be something as simple as a receipt of exchange between two people who barter – I trade you my 5 apples for your 4 oranges. We both keep a receipt to record what the relative price of apples to oranges are. The medium of account is the receipt.

    The problem you get into is trying to decide how a medium of account has value when it is not used as a medium of exchange. For that you either have a legal system or a government that bestows value onto the medium of account.

  20. 20 J.V. Dubois November 30, 2012 at 9:39 am

    Frank: All of them, general equilibrium stuff. So a change in the price of one good will affect the price of every other good out there on the market. Or to put it in other way, you cannot determine the price of an apple unless you account for all other goods on the market.

    So if we say that we measure everything as price of an apple (MOA) with 1kg of apples as a baseline (unit of account) what we did is that we conceptualized the relative price system. You can create a “snapshot” of an economy and express all prices in apples.

    I just want to say that this is purely abstract thing. You may for instance “calculate” the value of minerals on Mars or calculate how much “value” pirates extracted given the current market rates of minerals / copies of songs. Of course if these minerals or songs were ever actually sold on the market, relative prices would change and your prior calculation would not be valid.

    So in the end, it is not “government” that bestows “value” onto medium of account. It is the market. If it is bonds that you use as medium of account, then the MOA value is derived from market value of bonds. How much apples will you get if you sell such a bond on the market? How much oranges can you get for apples? How much labor can you buy for that many oranges?

    And as for your last example – trading apples for oranges. I hope you are not going to argue that if you just happen to remember the transaction then medium of account is your brain, aren’t you? If you take that “medium” definition too far, you are right. But this does not help much.

    For practical reasons whatever happens to be widely used thing that people use when they refer to value is the MOA. When looking on receipt was it written in a way that stated that the price of an apple was 0,8 oranges or that the price of an orange was 1,25 apples. Whatever medium was used to conceptualize the value was MOA. If each side of the trade thought about the exchange n a different way then we have two MOA. On the global market there is a lot of MOA – all the national currencies. Then one has to track the relative prices of these currencies (exchange rates) to be able to compare them if one for instance wants to compare GDP of different countries in a different time.

  21. 21 Frank Restly November 30, 2012 at 10:38 am

    J. V.

    “And as for your last example – trading apples for oranges. I hope you are not going to argue that if you just happen to remember the transaction then medium of account is your brain, aren’t you? If you take that medium definition too far, you are right. But this does not help much.”

    Medium: An intervening substance through which something else is transmitted or carried on.

    No, any medium is a physical substance. And so your brain is a medium. But can two mental recollections of a transaction be different? Sure, one party could have a lapse of memory or just outright lie. That is why a written record is kept and why signatures have a legal basis.

    “So in the end, it is not government that bestows value onto medium of account. If it is bonds that you use as medium of account, then the MOA value is derived from market value of bonds.”

    First, I said a government or a legal system. A government / legal system (normally) does not bestow value on the medium itself, it bestows value on the quantification of that medium. Every standardized measure (kilograms, meters, ounces, miles, etc.) has a legal backing that defines it. See:

    http://en.wikipedia.org/wiki/Plan_for_Establishing_Uniformity_in_the_Coinage,_Weights,_and_Measures_of_the_United_States

    Also, marketable U. S. government bonds are a relatively new invention. Prior to 1950, U. S. government bonds were primarily nonmarketable (could not be resold).

    Finally, if the federal government decides to tell bond holders to stick it, who has changed the value of the bonds – the market or the government? The only reason the federal government cannot do this is that bond holders are legally protected by the 14th Amendment to the Constitution.

  22. 22 David Glasner December 2, 2012 at 6:11 am

    Nick, Thanks for your response, and forgive me for the tardy response; it took me a while until I could take the time to think carefully about your comment.

    First, when you say that all prices are fixed, I understand this to mean that all prices are quoted in terms of the medium of account (aka numeraire?) and are legally set at too high a level. This means that all prices for the MOA are too low, implying that there is an excess demand for the MOA. You then say that relative prices for all other commodities are right. This is a little hard for me to think through. There is an excess demand for the MOA which is offset by excess supplies of the (n-1) other commodities that are being offered for the MOA. People are hoarding the MOA. Am I understanding you correctly?

    You ask whether there can be a recession. Your answer is no, and I think that I would agree. You have already assumed that there is no disadvantage to barter without using the MOA, so everybody is hoarding his MOA rather than bartering it away, but that doesn’t affect barter for the other n-1 commodities.

    From this case, you move to an MOA that is also used as money. You say that all prices are fixed except the price of peanuts. I take it that peanuts are not the MOA, or am I not following you? When you say that all prices are fixed, I take it that you mean that all prices are legally fixed in terms of the MOA. So we have n+1 commodities. Commodity n+1 is the MOA, commodity n is peanuts, and commodities 1 to n-1 shall remain nameless. The prices of commodities 1 to n-1 are legally fixed and quoted in terms of units of the MOA , and the price of peanuts is allowed to adjust to clear the market for peanuts. Do I have that right?

    I am guessing that you think it is somehow important that there cannot be a recession in your first scenario, but that there can be one in your second scenario, and you are attributing this to the existence of a medium of exchange. I would put it differently. There can be no recession in the first scenario, because there are zero transactions costs, which allows all welfare-enhancing trades (including barter exchange) to be executed (except those prohibited by the price control on the good that serves as the numeraire). But in such a world, the effect of the price control can be easily circumvented because there are zero transactions costs.

    You also point out that in the second scenario there could be a recession even though the money supply was endogenous. I agree, but the recession that you are positing seems to me to be a supply-side recession not a monetary-disequilibrium recession.

    HJC, Have you read Tobin’s essay “Commercial Banks as Creators of Money?” If not, you need to read it.

    Becky, I don’t think that the medium of account is intrinsically that important, but there can be situations when fluctuations in the value of the medium of account can have serious repercussions as in the Great and Little Depressions. The trick is to insulate the medium of account from undesirable fluctuations in value.

    Frank, “Medium of account” is a term of art in monetary theory. What a dictionary definition says about the meaning of medium is really beside the point.

    Mike, OK I see your point. On substance, I think we basically agree, but I wouldn’t use the term default as broadly as you do.

    JP, I agree that we may very well be using our terminology inconsistently, so that it is not clear whether our disagreements are substantive or semantic or both.

  23. 23 merijnknibbe December 2, 2012 at 1:55 pm

    As a historical fact: the unit of account existed long before metal coins were invented. And forget about barter. Non-montary economies are not based on barter but upon other coordinating mechanisms, like family and culture. Remember that Herodotus did not only state that (long after the unit of account was introduced as a measure of debt) Phrygia (of King Midas fame) was not just the first country to use (state issued) coins – but also the first country were retail shops sprang into existence. These first coins were made of electrum, a natural alloy of gold and silver, but contained less gold and more silver than the natural alloy found overthere – and don’t think that in those days (Archimedes!) they had not yet invented the noble art of counterfeiting! Clearly, sombody came to the brilliant conclusion that adding the unit of account (‘stater’) to natural electrum enhanced its value and therwith enabled debasement. The unit of account in fact enabled changes in the medium of account. What does this have to do with modern money? For one thing, we might start to imagine and assume a little less and base our stories more on fact. Which brings us to the economic statisticians, who by necessity and for reasons of consistency and coherence have to use precise definitions to estimate money. And hey – it seems that they never had to re-invent endogenous money at all, all of the monetary statistics in a fiat money system are and were based upon the ‘loans create deposits’ idea. In depth reading:

    http://www.ecb.int/pub/pdf/other/manualmfibalancesheetstatistics201204en.pdf?91257748bd9d6e88bd69ea623adceee1

    And when you really think about this it turns out that it are not the banks who create money. Banks and borrowers together do this – banks are the tap (and are allowed to create ‘legal tender’ by law) and borrowers open this tap. The borrowers are in fact the active ones (which means that no central bank will ever be able to pump up NGDP without massive ‘helicopter drops’). One can, again, consult the monetary statistics, which show this in detail. This is what endogenous money in the end is – it’s in the created by us, by borrowing from the Monetary financial Institutions, be it by borrowing for house purchase or by using an overdraft facility. That’s what endogeneity is. In the end, it’s a unit of account backed with our debts as collateral.http://www.ecb.int/press/pdf/md/md1210.pdf

    So, whenever you talk about money, don’t forget to talk about debt. And let our imagination and assumptionation (whatever) be constrained by well defined concepts and metrics which we can – and as they are the guys and galls who produce the data we use – and have to borrow from the economic statisticians. Economics can be a science!

  24. 24 HJC December 2, 2012 at 2:43 pm

    David, thanks for the response. I had skimmed the essay, since you referred to it above. The problem is that comments like “Unlike competing financial industries, commercial banks cannot seek funds by raising rates” are no longer valid and reflect the time he was writing. A lot of the analysis – “demand for bank deposits can increase only if the yields of other assets fall” – rests on this making it invalid, it’s hard to pick out the good bits. It refers to a period before active liability management and securitisation when banks had a lot of government paper on their balance sheet and deposit rate ceilings.

    So I suspect you are guiding me towards statements like “If bank deposits are excessive relative to public preferences, they will tend to decline”. If ‘deposits’ here means narrowly-defined on-demand bank liabilities then of course, they can be converted to non-deposit bank liabilities, but so what? This is simply “satisfy[ing] … portfolio preferences.” But if ‘deposits’ means non-equity bank liabilities (as it often does) then – not directly. The more relevant point is “Bank deposits decline with bank assets” – not the other way around. Bank assets are generally non-marketable so an attempt by the public to reduce bank liabilities (meaning bank assets must fall as well) is a bank run, cf. Northern Rock. At a systemic level – a financial crisis.

    “For bank-created money…there is an economic mechanism of extinction as well as creation, contraction as well as expansion.” This is true, but I don’t think it can be achieved from the liability side if bank money is defined as non-equity bank liabilities, which it surely must be.

  25. 25 David Glasner December 4, 2012 at 2:35 pm

    merijnknibbe, Thanks for your interesting comment, which I largely agree with. However, it is not just how much the public wants to borrow that matters; how much money the public wants to or can be induced to hold is the other blade of the scissors.

    HJC, Tobin’s point was precisely that the conventional account of the money multiplier depended on banks being unable to seek funds by raising rates, so I think that you need to read the essay a bit more carefully. You are right that banks are at risk because their assets are less marketable than their liabilities. That is the transformation that makes banks so important and at the same time makes them unstable. But the two sides of the balance sheets interact. One side does not completely dictate the other side.

  26. 26 HJC December 4, 2012 at 5:06 pm

    Thanks, I will take another look. Our positions are probably not too far apart – I would say that an increase in bank funding costs will increase lending costs thus reduce lending and finally allow a reduction in bank liabilities. From that I suppose I am suggesting that the asset side “dictates” the liability side. Perhaps you prefer to leave our the intermediate steps, but I think they are interesting in their own right.

  27. 27 Lorenzo from Oz December 5, 2012 at 3:12 am

    This is why I love reading your blog, your posts are greatly clarifying and you link to great articles :)


  1. 1 Do monetary statistics take account of securitisation (meta, 1 graph)? « Real-World Economics Review Blog Trackback on December 3, 2012 at 1:19 am
  2. 2 Do monetary statistics take account of securitisation (meta, 1 graph)? | Jo WeberJo Weber Trackback on December 6, 2012 at 2:32 pm

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s




About Me

David Glasner
Washington, DC

I am an economist at the Federal Trade Commission. Nothing that you read on this blog necessarily reflects the views of the FTC or the individual commissioners. Although I work at the FTC as an antitrust economist, most of my research and writing has been on monetary economics and policy and the history of monetary theory. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 286 other followers


Follow

Get every new post delivered to your Inbox.

Join 286 other followers

%d bloggers like this: