Wicksteed, stocks and flows

Flows are very very small relative to stocks. Each of us demands a flow of air to breathe, but since the flow demand is very very small relative to the stock supply of air, air is a free good. OK, that analogy is not perfect, so let me build a little "model".

Time t is continuous. There exists a stock M(t) of green bits of paper coloured bits of polymer. M(t) varies over time, but is always strictly positive.

The government prints a flow G(t) of new green bits of paper, and adds that flow to the stock.

The government also burns a flow T(t) of green bits of paper. Each individual agent must give the government a specified (by the government) flow of green bits of paper, on pain of imprisonment, which the government then burns. The sum of all agents' donations to the government add up to that flow T(t) when aggregated across all agents.

dM(t)/dt = G(t)-T(t).

Assume M(t) > 0 for all t.

Returning to my free air analogy: G(t) is the flow of oxygen produced by trees; T(t) is the flow of oxygen we consume when we breathe; and M(t) is the stock of oxygen in the atmosphere.

There exists a competitive equilibrium in which the price of green bits of paper is zero and the whole stock M(t) is lying on the sidewalk. Agents only pick up a flow of bits of paper when they need to donate them to the government, and they do so immediately, and so hold zero stocks. Since the price of bits of green paper is zero, and there is always a stock lying on the sidewalk, there is no incentive for an individual agent to pick any up until immediately before he has to pay his taxes. (Strictly, individual agents would be indifferent about when they picked up the bits of paper off the sidewalk, but if there were the tiniest cost of holding bits of paper they would leave them lying on the sidewalk until the very last second.)

Flows of T(t) are very very small relative to the stock of M(t). A demand for a flow does not mean a demand to hold the stock.

What might cause that equilibrium to break down?

1. Monopoly. One agent might try to corner the market in bits of green paper by picking them all up and selling them to other agents at a profit maximising monopoly price. But that monopoly position is highly contestible. Unless the first agent could pick up all the paper instantly, he would know that all other agents would start picking up paper as soon as they saw him trying to corner the market. So the result would be either Cournot or Bertrand oligopoly. With a large number of agents, the equilibrium Cournot or Bertrand oligopoly price would approach the cost of picking the paper up from the sidewalk. If there were any costs (including opportunity costs of foregone interest or storage) [update, of holding bits of paper] there could not exist an equilibrium in which no bits of paper were lying on the sidewalk.

2. Individual agents get some benefits from holding positive stocks of paper. Perhaps because the timing of their donations to the government are uncertain or fluctuate over time, and it is costly to pick up a large number of bits of paper in a short time, or else the cost of picking up bits of paper varies over time. So agents pick up all the bits of paper when their costs of doing so is low, and hold an inventory of bits of paper in advance of anticipated donations to the government.

Hmmm. That's starting to sound like a transactions demand for money function. But if we have a demand to hold money for transactions purposes, we can still get an equilibrium in which money is held and valued even if T(t) is always zero.

(Wicksteed said (HT David Glasner) that fiat money has value because the government requires it for payment of taxes. I'm saying that's not sufficient. And it's not even necessary, in the sense that if my local recycler accepted fiat money, even at a negligible price, that would be enough to replace T(t)>0 in ruling out the equilibrium in which fiat money has a price of zero.)

Update: David Glasner helpfully posts the relevant passage from Wicksteed here.

Comments

[Kailer]

I resent the Americo-centrism embedded in the phrase "green bits of paper". If you must refer to the national currency ironically, in the name of Canadian nationalism, I insist you refer to it as "coloured bits of polymer".
Thank you,

[Nick Rowe]

Kailer: you're right, sorry! Actually, I might even have been thinking back to my Brit childhood!

[Nathan Tankus]

medium of exchange follows, not leads, unit of account.

[david]

Yet governments can, and have, altered the medium of exchange without enforcing alterations in the unit of account. What's more, in very small open economies it is frequent to see non-domestic-unit-of-account media of exchange sloshing their way through the system (US dollars, currencies of major neighbors, etc.). Unit-of-account explanations have trouble dealing with environments where multiple currencies operate simultaneously, since people clearly can alter said units on the fly if they are habitually used to it.

I don't see why money can't have value for an assortment of reasons without any universally-dominant one.

[J.V. Dubois]

As the Bitcoin experience showed, fiat money can clearly have value without any government support or without explicit or implicit convertibility to any goods. So I would say that David's bubble explanation of money is closer to reality.

[Nick Rowe]

david: "I don't see why money can't have value for an assortment of reasons without any universally-dominant one."

OK. But those reasons must include a demand to hold stocks of it, if there exists a positive stock supply. A flow demand, by itself, doesn't work. Otherwise the stock of oxygen in the atmosphere wouldn't be a free good.

[Nathan Tankus]

Um, the experience of Bitcoin doesn't show that at all. First, it really took off when certain businesses started to allow people to pay in Bitcoin. Second, it was a bubble, but not in the currency sense. people bought in because they thought it would retain (untraceable) value in the future and that proved to be self fulfilling for a relatively short period of time. It didn't really get anywhere close to being a currency.

[Nick Rowe]

LETS, and what the Greeks are up to recently, might be better examples. But this post is about the stock/flow consistency of the Wicksteedian explanation.

[J.V. Dubois]

Nathan: I do not get it. First, I consider anything that has value because people think that other think it has valuee - is a bubble. From baseball cards, land on the moon, possibly part of the "value" of real goods (gold, oil, real estate). Thus fiat monies (especially private ones) are by definition bubbles. So you had intrinsically worthless money being used to pruchase goods and services for years. If this is not a currency, I do not know what is.

[Simon van Norden]

I'm confused by statements like "The flow is very, very small relative to the stock." I think I understand the units for the stock (# pieces of coloured polymer). But flows are defined relative to a unit of time. What time units are you using to make that comparison?

I'm sure you're right that the need to pay taxes is not a necessary for money to have value....but I also can't think of anyone who claimed that it was. Do you think Wicksteed did? As for the sufficiency of the condition, think of the world in which fiat money was free; why would the government make coloured pieces of polymer legal tender? I thought the point of the exercise (from the govt.'s perspective) was to acquire real resources. To put it another way, why would the government require payment of taxes in the form of a free good (e.g. sunlight)?

[Nick Rowe]

Simon: When I say "the flow is very very small relative to the stock", I am, of course, speaking very very loosely, because flows and stocks don't even have the same dimension, so can't strictly be compared at all. Start in discrete time, so flows and stocks have the same dimensions (the flow is the discrete change in the stock from one time period to the next). Then make the units of time smaller and smaller. If we are describing the same system, but observing it at a finer and finer measure of time, the flows per period get smaller and smaller as the measurement period gets shorter and shorter.

I'm sure you can do the formal calculus better than I can! You are just overthinking it, by thinking I'm speaking more rigourously than I am.

"I'm sure you're right that the need to pay taxes is not a necessary for money to have value....but I also can't think of anyone who claimed that it was."

Hang around wilder parts of the econoblogosphere long enough (or just watch this comment thread!) and you will discover lots of things that many people claim about the positive value of fiat money resting on (necessary or sufficient?) the need to use it to pay taxes.

"Do you think Wicksteed did?"

Actually, it's ages since I read Wicksteed, and I can't really remember precisely what he said. But I think he is the original source of this idea, in its various current forms.

[Nick Rowe]

Simon: "I thought the point of the exercise (from the govt.'s perspective) was to acquire real resources. To put it another way, why would the government require payment of taxes in the form of a free good (e.g. sunlight)?"

I echo your thoughts. Governments normally want to collect taxes (and buy stuff) in the same good that people are using as medium of exchange. What is endogenous here?

[Nathan Tankus]

@J.V. Dubois: The difference between you and me is I focus on the liability element of money while you are focusing strictly on the asset element of money. I view the origins of money as charges/tithes/taxes etc that led people to seek out an entity or person's liabilities to be desired (so that they could in turn cancel what they owed). From this point of view the fact that a liability is not a physical asset but a social relation does not necessarily say anything about that liability's "value" to someone or something else. It is an abuse of language to call anything that has value because other people think it has value a bubble. a bubble is the unsustainable increase in the price of an asset.

for the ancient near east see here:http://www.amazon.com/Debt-Economic-Renewal-Ancient-Near/dp/1883053714

for the middle ages in China see: http://www.amazon.com/Fountain-Fortune-Monetary-Policy-1000-1700/dp/0520204085/ref=sr_1_2?s=books&ie=UTF8&qid=1335371027&sr=1-2

for medieval Europe see: http://umiss.lib.olemiss.edu:82/search~S0?/taccounting+historians+journal/taccounting+historians+journal;T=tally/1%2C2%2C0%2CB/frameset&FF=taccounting+historians+journal;T=tally&2%2C2%2C

[Nathan Tankus]

@Nick said" Actually, it's ages since I read Wicksteed, and I can't really remember precisely what he said. But I think he is the original source of this idea, in its various current forms."

Actually this idea is ancient and cross societal. see my links above. There's also some interesting discussions in ancient Greek writings. Not to mention it is pretty universal in ancient languages for the word "symbol" to be about breaking an object in two like a tally stick (as in a symbol for a social relation-or a debt). They even called Charlemagne's currency "imaginary money".

[Simon van Norden]

Nick: thanks for the feedback.

If you're keen on using a stock-flow model to build intuition or model formally, just remember that statements about stocks relative to flows have implications for the dynamics of the flows (and vice-versa.) If you tell me that stocks are relatively "big", I understand you to mean that net flows have been pretty persistent or smooth. If you tell me that the relative size of the stock influences its price, I'd expect that changes in the persistence of flows will change prices. So, for example, if variations in the level of govt. demand T are unpredictable but big, that will tend to make stocks small relative to flows.

Does that help your intuition at all?

[Nick Rowe]

Simon: see here for a statement of the position I am disagreeing with. An irate horde will descend on this blog post any minute, just as soon as they have gathered their strength for the counterattack ;-).

Yep. Normally, when we say "flows are small relative to stocks", we mean that the stock doesn't fluctuate much. In this case, what I really mean is that stocks and flows have different dimensions, so can't really be compared. It doesn't matter in my model if M(t) fluctuates a lot, as long as it always stays strictly positive.

[Nathan Tankus]

@Nick: you've really stepped it in this time! just teasing. I think stated the position pretty well above.

[Nick Rowe]

Nathan: thanks! Actually, I wanted to let Simon know that I was arguing against a widely-held view, but at the same time I didn't really want to start an all-out frontal war with anyone. Because sometimes it's better to argue about particular issues without forcing people to take sides to defend their personal home turf, if it's at all possible. So I used Wicksteed as my....stalking horse(? is that the right word? Probably not). I don't think there are many people who identify themselves as "Wicksteedians".

[Nathan Tankus]

@Nick: lol typo. I was touting my own horn! I meant I did above :)

you've been fair though.

[Nathan Tankus]

*tooting

grr me and spelling are not agreeing this week.

[Mike Sproul]

Bits of paper issued by a private bank can be handed over to that bank in exchange for 1 oz. of gold. Bits of paper issued by a government can be handed over to that government in exchange for staying out of jail. The bank's gold is its asset, and the government's ability to put you in jail is its asset. If the government lost its ability to put you in jail, the bits of paper would lose value, just as they would if the bank lost its gold.

So the bits of paper are no different from any other financial security. they are backed by the issuer's assets.

[Nick Rowe]

Mike: Aha! The extreme Neoclassical backing theorist joins the extreme Post Keynesian chartalists!

The more bits of paper you pick up of the ground and donate to the government the more gold you get in return.

The more bits of paper you pick up off the ground and hand over to the government the less jail time you serve?? Nope. Agent i only has to pick up and hand in Ti(t) bits of paper to get zero jail time. You can't get negative jail time by overpaying your taxes.

[MikeB]

So, Professor Rowe, was the marking of exams so stressful that you are starting a jousting match for some R&R;) Seriously, I've been there, marking can be quite a bear, especially if you care as you seem to.

To the defense of those not to be named. It would seem to me that Professor Wray has always said sufficient, but not necessary. I'm an outsider to the profession so I can't weigh in on the technicals (if there are any). It was not evident to me how your approach above leads to not sufficient. From here forward, I expect to be a passenger only, but I wanted to make a couple of notes:

1. Does it matter? From where we are today and how we go forward with the best policy, does it matter if taxes drive money. It seems to me we've been distracted before. Something about 1+1-1=1.

2. I've worked with Engineers who enjoyed the engineering a little too much and never really got anything finished. I am getting the same sense from some of these economics discussions. The analysis can be fun and interesting, as can the history - I get that. But are we moving things forward for the public good.

3. Those not to be mentioned have some interesting theories and proposals. They appear to have a way forward that has the potential to alleviate so many ills in our world today. Maybe the Nick Rowe theory also moves us forward and I just don't know enough about it. Maybe there are others with equally good potential. It just seems where we are now is not a good place. There are things that need doing. Lots of people who need things to do. Whether it is foreign investors in our natural resources or the brain drain we used to hear so much about, it looks like we are giving our real resources away. What if those not to be named have at least some of it right. Shouldn't we give them a fair shake. BTW, Prof Rowe, you do seem to be one of the few who will engage in a discussion and you should be commended for that.

Let the fireworks begin.

[Nick Rowe]

MikeB: "Does it matter? From where we are today and how we go forward with the best policy, does it matter if taxes drive money."

Indirectly, I think it does. This is one of the roots of the question of whether money is different from other government liabilities, like bonds. Which in turn is at the root of the question of the difference between monetary policy and fiscal policy. And I think (hope?) policy would be better if we understood that distinction better. So what looks on the face of it to be a rather arcane question, about why fiat money has value, might matter for policy.

Funnily enough, these sorts of questions seem a lot less arcane nowadays than they did before 2008. Back then, many of us put aside our theoretical worries; it all seemed to work OK in practice, so it didn't really matter if or why it worked in theory. "Just move the interest rate up or down to keep inflation on the 2% target, and forget about what's inside the black box!" Before 2008, I had set aside my theoretical qualms, and was working on little details about exactly how much to move interest rates up or down to keep inflation at the 2% target.

Those not mentioned aren't bad guys. But their theories and policies do remind me of the ones we abandoned in the 1970's, because they didn't work so well either.

[Nick Rowe]

Aha! Here we go. Are you watching, Simon?

[Mike Sproul]

Nick:
I suppose you can get that result under a carefully constructed set of conditions, but those conditions don't resemble the historical situations where paper money was actually issued. For example, in 1690 the Massachusetts Bay colony had unexpected bills to pay (wages payable to its soldiers), and it paid them by printing up its own paper shillings, each ostensibly worth 1 shilling in English coins. The plan was that the paper shillings would be gradually retired over the following years as they were received by the colony in payment for taxes. The paper shillings were the colony's liability and the taxes receivable were the colony's asset. The shillings were thus valued like any other liability.

Any business could do the same thing. A plumber could print up bits of paper that say "This bit of paper is good for $1 worth of my plumbing work". He could use those bits of paper to pay some of his bills, with the promise that he would gradually retire the bits of paper over time as people pay them to the plumber in exchange for his work. The bits of paper are his liability and his plumbing services are his asset.

In both cases, the value of the bits of paper is determined on regular accounting principles. I suppose you could say that the gradual retirement of the bits of paper creates a demand for them, but why bother with that when the accounting explanation is so obvious?

[Alex Godofsky]

Nick, T(t) isn't some fixed quantity determined in advance. T(t) is determined roughly as a proportion of the total private flow of money. If the price of money were 0 then T(t) would be infinite.

This isn't a story of how money got its nonzero value, but it's a story of how it can keep it. As the price of money falls the flow of money burned increases, acting as a restoring force.

[Ritwik]

There's a very interesting section on the 'price level' in an unrelated NYTimes blog today that may have a bearing on this.

http://6thfloor.blogs.nytimes.com/2012/04/24/eight-things-i-learned-from-reading-every-last-word-of-the-economist/

To quote,

"The 1,000 shillings note exchanged for roughly $0.13 when Gen. Muhammad Aideed employed a printing firm to reproduce the note in 1996. As the number of notes in circulation grew, the exchange value fell to just $0.03, which is the cost of producing an additional note. Since the exchange value equals the cost of production, forgers can no longer profit by increasing the supply. Today, the Somali shilling is a commodity money. Its supply is governed by the cost of ink and paper required to produce a note. From Letters."

My take on this (no idea if it really makes sense) - http://ritwikpriya.blogspot.co.uk/2012/04/value-of-money-price-level.html

[Nick Rowe]

Mike: "In both cases, the value of the bits of paper is determined on regular accounting principles. I suppose you could say that the gradual retirement of the bits of paper creates a demand for them, but why bother with that when the accounting explanation is so obvious?"

Because accounting principles do not determine the rate of return at which assets will be willingly held. You are familiar with the Marshallian concept of the "demand price" of a good (the price at which people will be just willing to buy the quantity supplied)? Now think about the analogous "demand rate of return" for an asset. Accounting cannot tell us the demand rate of return for a Picasso, an IBM share, or a dollar bill, and those three are all different, and depend on the relative quantities of each, and on a lot of other things too.

If those paper shillings were equally useful as gold as a medium of exchange, their demand rate of return would be equal to gold's. If they could only be stored, awaiting for eventual redemption, their demand rate of return would be lower than gold's, and they would initially be priced at a discount to gold, with that discount falling over time as the redemption date approached, making their actual rate of return exceed that of gold.

Not all assets look equally good when you hang them on your living room wall. Not all assets are equally good when you try to use them as a medium of exchange.

[Nick Rowe]

Alex: OK. That's a very different theory, but let's run with it. Suppose P(t) is the price of money (the reciprocal of the price level). Let T(t) be determined by T(t)=S(t)/P(t) where S(t) is the *real* value of taxes (which are fully indexed for inflation).

Suppose we keep G(t) fixed in nominal terms. The government prints G(t) bits of paper and sells them at the competitive equilibrium price P(t) for G(t).P(t) real goods and services. That rules out P(t)=0 as an equilibrium, because if P(t)=0 then T(t) would be infinite and M(t) would immediately fall to 0.

I can't do the math, but this is what I think the equilibrium for M(t) and P(t) would look like: It will depend on the (actual and expected) time-paths for G(t) and S(t). In some periods M(t) would be zero, and in other periods it would be positive. If M(t) were positive, then M(t) could act just like a real return (indexed) bond, with the government able to buy a quantity of goods whose present value would equal the present value of future taxes up till the point where M(t) is next 0. When M(t) hits the 0 constraint, we get G(t)=T(t) so the government can buy an amount of goods equal to current taxes.

If G(t) is indexed for inflation too, like G(t)=g(t)/P(t)and if the demand rate of return on M were exogenous, then we get an overdetermined model. The present value of g(t) would have to equal the present value of S(t). It's the long run government budget constraint.

Of course (see my response to Mike above), if M is in fact used as a medium of exchange, then the demand rate of return on M will not be exogenous, and will depend negatively on real money balances M(t).P(t).

[Nick Rowe]

Ritwik: my reading of that episode is that it contradicts Chartalism. This was fiat money issued by a defunct government that was still used as money. The only way that the defunctness of the government mattered is that there was no government that could prevent counterfeiting. So counterfeiting converted a fiat money into a commodity money. But then, in a sense, *all* paper monies are commodity monies, in that the value equals the marginal cost of producing new notes. It's just that the marginal cost is very high for a counterfeiter, if you risk imprisonment.

[Oliver]

Bits of paper issued by a private bank can be handed over to that bank in exchange for 1 oz. of gold. Bits of paper issued by a government can be handed over to that government in exchange for staying out of jail. The bank's gold is its asset, and the government's ability to put you in jail is its asset. If the government lost its ability to put you in jail, the bits of paper would lose value, just as they would if the bank lost its gold.

So the bits of paper are no different from any other financial security. they are backed by the issuer's assets.

Posted by: Mike Sproul | April 25, 2012 at 04:13 PM

Mike: Aha! The extreme Neoclassical backing theorist joins the extreme Post Keynesian chartalists!

The more bits of paper you pick up of the ground and donate to the government the more gold you get in return.

The more bits of paper you pick up off the ground and hand over to the government the less jail time you serve?? Nope. Agent i only has to pick up and hand in Ti(t) bits of paper to get zero jail time. You can't get negative jail time by overpaying your taxes.

Posted by: Nick Rowe | April 25, 2012 at 04:39 PM

I think one must not confuse a causal relation with causal direction. I'm no pro, but think that Chartalists, who btw. readily concede that private credit precedes government fiat money, thus admitting that taxation is neither sufficient nor necessary to drive CREDIT, would largely agree with the backing theory as stated above in substance, but not in causal direction.

For a private, non-government backed (i.e. no lender of last resort, no deposit insurance, etc) bank to be able to issue liabilities (bank notes) into a market, it needs to have some asset before this takes place. This asset might be gold, it might be the fortune and good name of the private banking family or whatever, but that asset precedes the liability. To the extent that the asset is money, we get an endless regression, begging the question of origin. To the extent that it's some other asset, that begs the question of who decides whether that's legal. I see all fingers pointing to some representative of the cpmmunity...

Owing to its nature as a monopolist of force as well as being the official representative of all of its citizens, a recognised government has the legal power to reverse the above logic. The liability can (must) be spent first. It can (must) simultaneously set a price by demanding something specific in return, say a day's worth of labour and can then keep up future turnover, irrespective of any competing currencies or other forms of credit, by imposing tax liabilities AND offering (usually limited) means by which to obtain tax receipts by which to extinguish them (govt. jobs).

Another aspect is, that all assets are not the same. Insuring credibility by promising to convert into a real, finite, and not readily accessible physical asset such as gold is substantly different from saying that the value of the currency is backed by the willingness and ability of the receipt holders to provide work in return for further receipts, or else... It is both a difference in degree - private banks have intangible assets, too - otherwise there'd be constant bank runs - and kind - banks don't usually offer borrowers jobs to work off the debt, nor can they enforce payment other than by legal, i.e. government backed, means. In the end, governments are different from corporations. Equating them is implicitly committing a fallacy of composition. A government liability is also my liability because it's my government and if I question the credibility of my money I am implicitly questioning my own ability/willingness to deliver goods to the comunity and/or questioning my ability to vote people into government who can live up to their responsibilities as managers of the currency. Bank credit, on the other hand, is a legal contract between three, usually anonymous and separable, parties (debtor, creditor and bank owners). Tom has no means of knowing what Dick or Harry are up to and thus demands other types of assets as security.

[Nick Rowe]

Oliver: thoughtful comment. But to my mind you are missing what I think is the crucial question: of all the many different assets in existence, and of all the many forms of IOUs in existence (both private and government), which one will people *choose* to use as money?

The medium of exchange is just one of many assets. Credit is just one proper subset of those many assets. Which asset will people choose to use as a medium of exchange? Will they chose an IOU, or a real asset? If they choose an IOU, will it be a government or a private IOU?

[Alex Godofsky]

If G(t) is indexed for inflation too, like G(t)=g(t)/P(t)and if the demand rate of return on M were exogenous, then we get an overdetermined model. The present value of g(t) would have to equal the present value of S(t). It's the long run government budget constraint.

Yes, though operationally the government basically allocates a fixed nominal amount for spending and a fixed proportion for taxes each period. The CB (effectively) sells bonds to mitigate the swings in the price level based on deficits/surpluses. The interest rate on government debt informs the fiscal authority whether it needs to reduce the deficit or has room to increase it, etc. I don't think I'm really saying anything new here.

(Though, the LRBC may not be precisely satisfied if there is forever some stock of money or bonds that the public wants to hold.)

Of course (see my response to Mike above), if M is in fact used as a medium of exchange, then the demand rate of return on M will not be exogenous, and will depend negatively on real money balances M(t).P(t).

Is M used as a medium of exchange? Leaving aside physical currency, don't banks just do a version of your own "pick it up off the sidewalk just when you need it" when settling transactions each day (by borrowing it at the federal funds rate), and otherwise (in normal times) hold essentially nothing above their required reserves? The medium of exchange is private money indexed and convertible to base money.

[Mike Sankowski]

Hi Nick,

I am pretty sure the "taxes drive money" idea needs further distinctions to make it useful. Taxes nearly exclusively drive the choice of the unit of account. Taxes drive the medium of exchange to a lesser degree than it drives the unit of account.

Taxes drive the value of the "store of value" function of money to a very small degree - and I agree with your take on the usefulness of the "store of value" function of money.

Nearly every economic transaction in the U.S. is given some USD value. However, a far lesser extent of these transactions use 100% cash to settle the transaction. It's very common to exchange bonds, real estate, firm equity, cars, and other financial and real world goods as part of a deal. For example, the recent purchase of instagram by facebook was a combination of cash and facebook equity. The medium of exchange in this case was (cash + firm equity) in exchange for (firm equity).

We had a huge discussion about S = I + (S-I) over at monetary realism. Some part of the discussion was about how the nominal value of (S-I) gives a reference by which to understand the nominal value of S and I. Most of S and I aren't cash, but are denominated in the unit of account of cash, which is driven by taxation.

Then, JKH went into detail about how the stocks and flows interact. It seems to have some bearing on this discussion.

It's important to note raw taxation isn't the final arbiter of the "usefulness" of currency. The agreement of people to accept the legitimacy of the taxation gives the unit of account widespread value, or not.

Few people around the world would accept any North Korean currency as having value, even though their ability to tax North Koreans is not questioned at all. The legitimacy of the North Korean government is not widely accepted by the people of the world or even in North Korea.

Many people accept the legitimacy of the U.S. government and it's one of the more powerful if more diffuse reasons the USD is used in places like Zimbabwe as their currency.

[Nick Rowe]

Alex: "(Though, the LRBC may not be precisely satisfied if there is forever some stock of money or bonds that the public wants to hold.)"

I see what you are saying there, and basically agree, but I want to be picky and rephrase it, longwindedly, like this:

'The demand rate of return on money varies negatively with the real stock of money. The demand rate of return on bonds may (or may not) vary negatively with the real stock of bonds. If so, we cannot specify the LRBC independently of the time-path(s) of money (and maybe bonds too) because the interest rate(s) used in calculating the Present Values will depend on those time path(s).'

(We normally duck this issue by: assuming government bonds have an exogenous demand rate of return; adding in the Present Value of seigniorage (central bank profits) as an additional source of revenue.)

(The above was a bit off-topic.)

"Leaving aside physical currency, don't banks just do a version of your own "pick it up off the sidewalk just when you need it" when settling transactions each day (by borrowing it at the federal funds rate), and otherwise (in normal times) hold essentially nothing above their required reserves?"

Physical currency was what I had in mind here. If we ignore physical currency, and assume that reserves vanish in the limit, and interpret my M(t) as government money, then we have M(t)=0. Governments get no revenue from printing money, because they don't print any (or any they do print is immediately redeemed for taxes or bonds). If the government validates expectations of future prices (a big if) then the government budget constraint is exactly like the budget constraint of any private corporation. The only difference is that the government controls the medium of account and can impose a surprise inflation to depreciate the real value of its bonds. If all media of exchange are produced by private banks, then those banks will earn any seigniorage from printing money (and there won't be any seigniorage if the banking sector is competitive, because it will all be competed away).

That isn't a bad model (aside from ignoring currency), but it is a very far cry from the Chartalist model, just because of the LRGBC. The government's budget constraint is exactly like that of a private corporation, except it has a clause which allows it to pay whatever % of the value of the debts it chooses, but people trust it not to invoke that clause.

[Alex Godofsky]

Currency is just a form of reserves with substantially larger transaction costs for "picking it up off the sidewalk".

[MikeB]

Prof Rowe
I know I promised to just watch - oh well. Do you know of the buckeroo program? Is it a simple model that proves sufficiency? Yes there are lots of holes in the model - but you seem to be OK with simplified models to study specific issues.

[Alex Godofsky]

(sorry to double post but I didn't think of this until after) Also, no one pays their taxes in currency. Currency is almost just like another private money, but one the government has a monopoly in for historical reasons / because it's easier for a government to prosecute counterfeiters/etc. / network effects of an easily recognizable and verifiable currency.

I don't think the macroeconomy would change much if government-issued green paper were replaced with green paper issued by your bank.

[Nick Rowe]

Mike: To my way of thinking:

1. People *choose* what medium of exchange to use, out of a very wide menu of assets they could use.

2. People *choose* what medium of account to use, out of a very wide menu of assets and goods and services they could use.

Those two choices are tied. Sometimes we choose a totally unrelated medium of exchange and medium of account (like in border zones or hyperinflations) but rarely.

We choose as individual pairs of trading partners, but our individual choices depend very much on what other people around us that we might trade with in future are using. The government is just one of those trading partners, albeit a big one, given the sizes of T and G in GDP.

Governments have power over laws and contract enforcement, and might use those powers to influence our choices of media of exchange and media of account. But only to influence. Witness cases like Cuba, where people used US dollars even when it was very illegal.

[Nick Rowe]

Alex: I tend to agree with you. But, if the government liabilities were not used as either medium of exchange or medium of account, I don't think the government could control monetary policy. (Except, maybe, by specific taxes or subsidies on private issuers of money). And this is *very* far from the chartalist world.

MikeB: I don't know the buckeroo program. I'm bad at both math and computers. Most economists are much better than me. So I just muddle along, trying to specialise in doing stuff in which I have a comparative advantage, and leave that stuff to other economists, who seem very happy to do it!

[Nick Rowe]

Hmmm. It just ate my comment. If that happens to you, and your comment doesn't immediately appear, just backarrow, forwardarrow, then repost. (I can always delete one if you accidentally double-post the same comment.)

[MikeB]

Prof Rowe
I see this as the necessary argument. I think it is evident that lots of things (including USD in Cuba) have value without government obligation. Still, it seems that at least some Cuban Pesos(?) are needed to fulfill obligations to Mr Castro. That sounds sufficient to give some value. We can probably argue about how much value and what percent of transactions happen in Pesos, but isn't that a different discussion.

[Oliver]

which one will people *choose* to use as money?

I'd say nowdays they tend to choose bank liabilities (deposits)denominated in the government's unit of account. This combination makes sense as such liabilities can be used to hord, trade, make payments to other private entities and to settle government liabilties without beiong subject to exchange rate risk. Deposit insurance and modern payment systems also help make this a very safe and comfortable solution. Depending on one's needs, say as an ancient polynesian tradesman or whatnot, and on the technical maturity of the banking system, preferences will obviously have varied and evolved across time and locations.

Why the government's unit of account, you now ask? The Chartalist answer, as you know, is the somewhat Marginalist 'taxes drive currency'. My preferred way of looking at it, is that one can separate between two principles that are both at work in parallel.

The first, which economics seems mainly focussed on, is an open source kind of world driven by supply-demand, convenience, trust and other abstract principles between more or less equal entities. This world knows no borders, no hierarchies and no responisbilities.

The other is a world of hierarchies and individual and institutional responsibilities, usually codified in law, but also always at work in other, more implicit, social arrangements.

My guess is, we need both and that the latter should be given enough spacve to trump the prior if whings go awry if we want an economy to work more smoothly than say in the Euro zone. Moreover I think it isn't legitimate for economics to act as though the latter realm doesn't exist, at least not if it wants to be relevant in the overarching way it claims to be.

[MikeB]

http://www.huffingtonpost.com/warren-mosler/the-umkc-buckaroo-a-curre_b_970447.html

Basically, you need buckeroos to graduate. Earn them for community service (1/hr), buy them or trade them.

[Nick Rowe]

The Cuban Peso is an interesting case.

The "regular" peso can be used in government shops to buy things at prices fixed by government, which you might say gives it value. But: the amount you can actually buy with your ration is very limited, so if you have a lot of pesos, but can't buy anything with them, then *at the margin* those pesos are almost worthless. The official exchange rate is roughly 1 peso = 1 USD, but I saw the black market exchange rate go from around 40 to 120 in a few months, then suddenly fall to around 20 when they allowed semi-private markets to open so that marginal pesos could actually be used as a medium of exchange and buy something.

The "convertible" peso can actually be used to buy off-ration goods at much higher (above world) prices. It is (or was, things change so fast) acceptable at par with USD in special "dollar stores".

Long story.

OK. Those buckaroos sound like a convertible currency, in the sense that the price to buy them is fixed. If I replaced my assumption that G(t) is exogenous with an assumption that the government was a buyer of certain good and paid a fixed price Pg, that would pin down the price of money above zero. But M(t) would be zero unless there was some sort of lag between earning G(t) and spending T(t).

Oliver: Your two principles sound sensible to me. Yes, as an economist, I tend to focus on the first. But one has to recognise that media of exchange do tend (roughly) to coincide with national boundaries. Why is there so little currency substitution? Why do nearly all trades in Canada use Canadian rather than US dollars?

[MikeB]

"*at the margin* those pesos are almost worthless"

Almost is not zero, therefore it seems to me that the obligation did suffice to impart some value. Sorry if I am being obtuse. As I said, I am not a content matter expert here.

I've seen in explained that public spending should be a certain size compared to the economy to be able to effectively manage the ups and downs of the business cycle. What if this certain size also affects the value. I think it is safe to say that the official peso is pretty small portion of total spending in the Cuban economy. What if we look at public spending at all levels of government for OECD countries. If memory serves, these numbers are in the 30-60% GDP range. Does that change the story?

Again, how does all this help us. We seem to be in a perpetual near recession. 1/10 want full time work but can't get it. We've made so many productive gains that very few of us need to work to provide all the things we need. But, then those in need can't afford and the ugly cycle begins.

[Mike Sankowski]

Nick -

"1. People *choose* what medium of exchange to use, out of a very wide menu of assets they could use.

2. People *choose* what medium of account to use, out of a very wide menu of assets and goods and services they could use.

Those two choices are tied. Sometimes we choose a totally unrelated medium of exchange and medium of account (like in border zones or hyperinflations) but rarely."

I totally agree with this - it's more to try and help the broader discussion along. When talking about money, it's easy to get confused. I think taxation very strongly drives the unit of account, and less strongly drives the use of a medium of exchange - even if nearly every exchange ends up with a value given in the unit of account!

The instagram/facebook deal was a "$1bn" deal, not a "$200m cash plus a mountain of stock" deal, even though the second description is more accurate.

Why does this matter to the big discussion? Because I think it takes a stab at providing an framework to answer this question:

"But to my mind you are missing what I think is the crucial question: of all the many different assets in existence, and of all the many forms of IOUs in existence (both private and government), which one will people *choose* to use as money?

The medium of exchange is just one of many assets. Credit is just one proper subset of those many assets. Which asset will people choose to use as a medium of exchange? Will they chose an IOU, or a real asset? If they choose an IOU, will it be a government or a private IOU?"

Taxes drive people to choose a unit of account, because if they don't have a good case to justify their level of taxation, the government will provide one.

The facebook/instagram case is a taxable event for instagram and facebook. They will be very careful with the valuations of their equity in the said unit of account. And this careful accounting of the unit of account is driven because you have to pay taxes in said script.

Distinguishing between force taxes place to adopt a unit of account vs. the force taxes place to use a medium of exchange helps to make it clear what and where we can use policy tools to make an impact.

Then, if we're talking about stocks and flows of money, it quickly becomes obvious taxes aren't the only thing driving the actual demand for the unit of account. And then you end up with your last few lines of this post, and other sources of possible demand for money, which is one reason we broke off and formed MMR.

And Oliver: great comment! I think this is an important distinction to make between credit and government money, even if they get mixed up as soon as they are created.

[Nick Rowe]

MikeB: "Almost is not zero,..."

It could also be used as a medium of exchange on the black market.

"I think it is safe to say that the official peso is pretty small portion of total spending in the Cuban economy."

It would be (especially in the 1990's which is when I'm talking about) very safe to say the exact opposite.

"Again, how does all this help us. We seem to be in a perpetual near recession. 1/10 want full time work but can't get it."

In the 1960's we used to think that unemployment was purely a demand-side problem, and could be cured with looser fiscal and/or monetary policy. Then in the 1970's we learned from Friedman, Phelps, and most importantly from experiencing the failure of those policies, that it wasn't. And most of us economists have moved on, and left That 70's Show. Some haven't.

Off-topic. End of this discussion.

[Nick Rowe]

Mike S.: We are almost on the same page.

"Taxes drive people to choose a unit of account, because if they don't have a good case to justify their level of taxation, the government will provide one."

That's a new idea to me. I think I disagree. Not all taxes had to be paid in, or were measured in, the common unit of account. Forced labour (including conscription) is one example. Tithes (a tenth part of the produce) are a second. But it probably does have some force.

My view is that when people haggle over prices, or set prices, or try to decide whether to buy or not, they always want to translate into a common unit of account. When I go back to England, I keep on translating back into Canadian dollars, because nowadays I speak "Loonie", and have forgotten how to speak "Pounds Sterling".

[Oliver]

Why do nearly all trades in Canada use Canadian rather than US dollars?

I would say, not for 'economic' reasons (i.e supply & demand, trust etc.) but for legal/social reasons, the most important of which is captured in the expression 'tax driven currency'.

[Mike Sproul]

Nick:
"Because accounting principles do not determine the rate of return at which assets will be willingly held..."

Empirically, rate of return has almost no effect on people's desire to hold money. Anyway, the Mass. paper shillings generally traded at par with silver shillings (except for some later inflation), so there was no rate of return in the normal sense. The return they provided was just liquidity.

So suppose the colony will collect 5 silver shillings in taxes in perpetuity. (You'd call this a flow.) At a market R=5%, the present value of taxes receivable is 100 silver shillings. (You'd call this a stock.) The backing theory says the government could print 100 paper shillings and give them to people in the street, and no inflation would happen because 100 paper shillings are adequately backed by 100 shillings of taxes receivable. The backing theory would add that the colony could print another 200 paper shillings and spend them on foreign bonds worth 200 shillings. This would still cause no inflation, since the 300 paper shillings are adequately backed by 100s of taxes receivable plus 200s of foreign bonds. Wouldn't your model imply the paper shillings would lose value as their quantity tripled? That would lead to all kinds of absurdities.

[Nick Rowe]

Mike Sproul: "Empirically, rate of return has almost no effect on people's desire to hold money."

Hang on. Are you saying the demand for money is perfectly interest INelastic? Or perfectly interest Elastic?

I would love to believe the former, because then I could be a really hard-line old-time monetarist treasury view type! But I don't.

I thought that you believed the latter, since it would make your views internally consistent (in my eyes).

I think neither is true. Look at Zimbabwe. Real M/P did not fall to 0 when inflation increased and the real return on holding Zim dollars fell and became negative. But it did fall, and eventually fell to zero, when inflation got high enough. The money demand curve slopes down. You need to assume it's horizontal for your views to work.

[philippe]

Nick,

What do you think about the idea that rather than government having to issue bonds, interest could simply be paid on reserves by the central bank? The overall effect would be no different, would it?

The government could then either deficit spend by just issuing new money directly (which would require major institutional changes), or the central bank could purchase all government bonds as/after they were issued.

In both cases, bank reserves would 'pile up' over time, but the (base) interest rate could be maintained at whatever level was deemed appropriate by paying interest on reserves. Monetary policy could be pursued much as it is at present (without having to constantly engage in OMOs). The only real difference would be that there would be no more government debt as such, no more debt crises or 'running out of money'.

The fact that government 'debt' is not necessary strikes me as being the most significant Chartalist point, especially at present. We can disagree about all the other stuff - whether taxes drive money or whether it might be best to have a permanent ZIRP, for example. Recognising that government debt is something of a fabrication makes it possible to think in a more open-minded way about which policies to pursue, and to focus on 'real problems'. What do you think?

[Mike Sproul]

Oliver:

"Owing to its nature as a monopolist of force as well as being the official representative of all of its citizens, a recognised government has the legal power to reverse the above logic. The liability can (must) be spent first."

It's irrelevant which comes first, and a monopoly of force is unnecessary. People deposit 100 oz. of silver into a banker's vault and he issues 100 paper receipts ('dollars'). The silver existed before the dollars, but they both became assets and liabilities to the banker at the same time. The banker then prints and lends another 200 paper dollars in exchange for borrowers' IOU's worth 200 oz (presumably backed by land). The land existed before the 200 dollars, but they became assets and liabilities to the banker simultaneously. A local landlord then issues 300 of his own paper dollars, each of which he promises to accept for a dollar's worth of rent on his land. The land existed before the dollars, but in this case the liability is created after the asset. (The landlord and his rents is meant to be analogous to the government and its taxes, by the way.)

[Nick Rowe]

philippe: The Bank of Canada already pays interest on reserves.

Take the current system, where governments issue bonds. Then:

1. Pass a law that says that only banks may own government bonds.

2. Make the term to maturity on government bonds shorter and shorter and shorter...

And we get to your system.

1. seems to me to be unwise.

2. would certainly be unwise for a regular corporation, municipal government (or Eurozone government). It may or may not be unwise for a government with its own money. It depends on risk preferences regarding the term structure etc.

But if we ignore 1 and 2, there is absolutely no difference.

In fact, what you have just described is rather similar to the New Keynesian/Neo Wicksellian perspective.

There is no magic bullet of free lunches for Chartalism here.

[Oliver]

@ Mike Sproul

I think you're confusing bank assets and bank capital. The latter must be there before the bank can expand its balance sheet by creating assets (loans) and liabilities (deposits/bank notes). Example: bank A has 2'000.- worth of gold as a capital base off of which it can extend say 100'000.- in loans by issuing 100'000.- worth of bank notes to the borrower who then passes them on to the builder in exchange for a house. The house now belongs to the borrower, whose net worth remains unchanged because he is also stuck with a debt obligation to repay principal plus interest. The banks' retained earnings from the loan go into expansion of the capital base, say by buying more gold. Etc. etc.

[Mike Sankowski]

"My view is that when people haggle over prices, or set prices, or try to decide whether to buy or not, they always want to translate into a common unit of account. "

I agree - my earlier point was taxation has different levels of impact on the different functions of money.

And Nick, you're running close to violating Pyke's Corollary of Godwin's Law: "The first person to bring up Zimbabwe in a discussion of economics loses the argument and ends the debate." ;)

https://twitter.com/#!/traderscrucible/status/195473348778672128

[Nick Rowe]

But Zimbabwe's such a great example! It's recent. Plus, if you want to estimate the effect of X on Y, it really helps to have some really big changes in X!

[Mike Sproul]

Oliver:

I didn't mention bank capital, but of course you're right that a banker must have some capital of his own before customers will do business with him.

[Mike Sankowski]

LOL! I had just talked about Zimbabwe Godwin's law this morning with a friend and here you are!

But this is a good discussion. We had a similar discussion over at MMR - we don't think the state theory of money is complete as it currently stands. Stephanie Kelton has a paper which goes into the basics of state theory of money which might be interesting.

http://cas.umkc.edu/economics/people/facultyPages/wray/courses/Econ601%202012/readings/Bell%20The%20Role%20of%20the%20State%20and%20the%20Hierarchy%20of%20Money.pdf

[philippe]

Nick,

"1. Pass a law that says that only banks may own government bonds."

Can you explain why this is relevant?
As far as I can tell, whether the bonds are owned by banks or companies or private individuals makes no real difference. In each case, people own bonds so as to receive a decent rate of return. If that comes from interest paid directly on reserves or bonds makes no difference to them. People abstain from taking their investment money elsewhere because bonds offer 'risk free' returns. Interest paid directly by the CB provides exactly the same service.

"2. Make the term to maturity on government bonds shorter and shorter and shorter"

This strike me as being irrelevant too, as bond holders can always sell their bonds quickly and easily if they want to spend their money or invest it elsewhere.
Owning a bond is no constraint on spending or investing in alternative 'instruments'. People stay in bonds because they offer 'risk free', decent returns.
If the government were to stop issuing bonds and the CB were simply to pay interest on reserves, investors would still have the same opportunity - i.e. risk free investments with a decent rate of return. Where's the difference?

[philippe]

p.s. I'm not sure it's about free lunches.

Government deficit spending can always be inflationary, and large trade deficits are risky and problematic in many different ways. Too much government spending vs private sector spending may have deleterious effects on overall productivity.
These are some 'real problems' I can think of off the top of my head, not being an economist. I would appreciate further info on other 'real' problems. "Running out of money" or not being able to repay debts denominated in the currency you issue are not 'real' problems.

[Nick Rowe]

philippe: If 1 and 2 make no difference, then there is no difference between what governments are doing now (using bond-finance) and what you want them to do (finance all spending through reserves, and paying interest on reserves).

[philippe]

Yes, that is what I am suggesting, although only as regards deficit spending.

I don't necessarily want governments to pay interest on reserves as an alternative to bond-financing, although that is an option. I simply think it might be important to realise that the two are essentially the same. If one accepts this idea, then the question of 'debt sustainability' becomes pretty much redundant. Other, 'real' concerns become more important.

Again, I would appreciate your input on this subject.

[Nick Rowe]

At some time in the future, inflation will be a constraint. The bigger the debt incurred now, the more taxes will have to be increased at that future time. Just because inflation is not a danger (in the US) today, does not mean that today's debt is not a problem.

End of that topic.

[philippe]

If only our political leaders described the (potential future) issues as clearly as you...

[Mike Sproul]

Nick:

I'm saying the demand for money is interest inelastic. I didn't think that was controversial. But I don't normally think in terms of supply and demand for money, just as stock market analysts don't normally think of the supply and demand for GM stock. The focus is on assets and liabilities.

You lost me on the Zimbabwe dollars. You brought that up in relation to interest elasticity right? But the question I was really wondering about was whether you think that the tripling of the money supply I described would make the paper shillings lose value.

[Ritwik]

Mike Sankowski : 'Taxes drive the unit of account'.

What about international trade? What monetary shift happens when Iran announces that it will trade half of its oil in the currency of its trading partner rather than USD?

[Ritwik]

Or, somewhat tangentially, to say that

1) Iran - India oil trade contract is now denominated in Rupees rather than USD? (numeraire)

2) Iran will accept rupees rather than dollars for its oil (exchange medium)

Are these two substantively different wrt monetary theory?

[Nick Rowe]

Mike: "I'm saying the demand for money is interest inelastic."

Seriously, you don't believe that. If you did believe that, you would be a Quantity Theorist, but you aren't. You believe that the demand for money is (close to) perfectly interest-elastic, whether you realise it or not.

Let i be the interest rate paid on other assets minus the interest rate paid on money. (Either both nominal or both real, it doesn't matter).

Write the money demand function as Md(t)=P(t).L(i(t),Y(t)).

As the elasticity of Md wrt i approaches zero, that money demand function approaches Md=kPY (assuming for simplicity a unit income elasticity). Set Ms=Md and we get the crude QT.

At the other extreme, as the elasticity of Md wrt i approaches infinity, we get i=0 (or some constant), so that we get M(t)/P(t)= Present Value of dividend stream plus buybacks of money, just like for bonds or shares, that must pay the same rate of return as the market. Which is your theory.

And the Zimbabwean experience (not to mention a load of estimates of money demand functions) says that both extremes are incorrect.

[Nick Rowe]

If the interest-elasticity of the demand for money were zero, a doubling of the supply of shillings would halve the value of each shilling, regardless of whether backing changed. If the interest elasticity is infinite, then doubling the supply of shillings leaving backing constant would halve the value of each shilling, and doubling the backing at the same time as doubling the supply would leave the value of each shilling unchanged.

[philippe]

Nick, this quote explains what I was trying to say yesterday:
(Apologies for length)

Scott Fullwiler:

“Instead of the complexities of the TT&L system and bond sales, the more direct and more efficient method of interest rate support would be for the Fed to simply pay interest on reserve balances. With interest –bearing reserve balances (IBRBs), absent offsetting Treasury or Fed operations to drain excess balances created by a deficit, the federal funds rate would simply settle at the rate paid on reserve balances. The nature of Treasury bond sales as offsetting, interest-rate support rather than finance would be obvious. While the private sector is offered an interest-bearing liability of the government in the presence of a deficit to support a non-zero interest rate target, this does not necessitate that the Treasury sells bonds….

…with IBRBs replacing bond sales the Fed’s independence and ability to exogenously adjust the federal funds rate target would be uncompromised and would simply require increasing or decreasing the rate paid on IBRBs. Regardless of the quantity of excess balances, the federal funds rate would not fall below this rate and would continue to influence other rates in the economy via arbitrage…. Holders of deposits so desiring could convert to short term, private liabilities – just as they can now – the rates for which would be set primarily via arbitrage with the Fed’s target. Long term rates – just as now – would be largely dependent upon the current and expected future paths of short term rates. Those desiring fixed- instead of flexible-rate investments – perhaps banks holding IBRBs – could do so through swaps that would be priced similarly. Without Treasuries, government agency securities and swaps could emerge as benchmarks for pricing of private assets, as is increasingly the case already and for which they are better suited than Treasuries anyway. In sum, with a deficit the transmission of monetary policy via IBRBs is identical to that with non-interest bearing reserve balances and bond sales to drain excess reserves.

With IBRBs, all Treasury securities could eventually be replaced; the interest rate on the national debt would then be the rate paid on IBRBs. Treasury securities themselves are simply fixed-rate liabilities and from the private sector’s perspective not functionally different from IBRBs aside from the flexible-rate nature of the latter. Note that consideration of IBRBs demonstrates how interest on Treasury debt is determined: with IBRBs and no securities issued, the interest rate is the rate paid on IBRBs; where short-term securities are issued, as above those rates are set largely via arbitrage with the Fed’s target; as longer maturities are issued, again as above these rates are set largely via arbitrage with the expected path of the Fed’s target. The “crowding out” view of the loanable funds market is irrelevant; the rates on various types of Treasury debt are set by the current and expected paths of monetary policy and according to liquidity premia on fixed-rate debt of increasing maturity. Since long-term rates are normally higher than short-term rates, total interest on the national debt would be significantly reduced if IBRBs eventually replaced Treasuries….. Indeed there is no inherent reason for Treasury liabilities to exist across the entire term structure except as support operations for longer-term rates.

Deficits unaccompanied by bond sales are disapprovingly labelled “monetization,” although there is no meaningful difference from when bonds are issued. A government deficit always creates net financial assets for the private sector; that is, when a deficit occurs, by definition the total credits to recipient bank accounts due to government expenditures are greater than the total debits from bank accounts to pay taxes.
…..

Whether bonds are sold, the ability of banks to finance further private spending is unaffected by debiting reserve balances or deposits created through deficit spending; recall that since loans create deposits, if there are willing, creditworthy borrowers then desired spending is financed in any event. For deficits, what matters for the determination of aggregate spending and inflation is not whether bonds are sold but whether the deficit is too large given the private sector’s desire to net save.”

http://www.cfeps.org/pubs/wp-pdf/WP38-Fullwiler.pdf

[Mike Sproul]

Nick:

You are assuming the correctness of the quantity theory, which is the very point in dispute. You are also not giving the logic of the backing theory a chance to work.

Start with dollars that are backed and convertible into gold at $1=1 oz. I hold $100 for liquidity purposes. If the interest rate is 5%, that means I sacrifice $5/year in exchange for the convenience of always having $100 with me. If the interest rate rose to 10%, my $100 of pocket money is now costing me $10/year. This is still negligible compared to the convenience of always having $100, so at most I might cut my cash holdings to $95. My demand for money, with R on the vertical axis, is nearly vertical, so for simplicity let's say it's perfectly vertical. It does not follow that doubling the quantity of dollars will halve their value. Remember that as the issuing bank issues another $100, it gets another 100 oz worth of assets. The bank is thus just as able as before to maintain convertibility at $1=1 oz. The trouble is that you are drawing the demand for money as a function of the interest rate, when interest rates have hardly anything to do with the demand for money. You could just as well write the demand for strawberries as a function of the interest rate.

But now how do you reconcile this statement:

"a doubling of the supply of shillings would halve the value of each shilling, regardless of whether backing changed."

with the writings of the intelligent, urbane, sophisticated scholar quoted below?

"1. Start with a monetary system where the central bank's paper money (or an electronic equivalent) is 100% backed by gold reserves, and each note can be redeemed at a fixed price for gold...(insert 9 steps, gradually relaxing assumptions)...We now have a modern inflation-targeting central bank."

[rsj]

Nick,

But for most people, including the U.S. central bank, "money" is defined as money of zero maturity which includes checking accounts and money market accounts. The M that you describe is really bank reserves, and is not "money" at all to the non-financial sector, in the sense that whatever reserves are held by your bank is not money held by you, who are in the non-financial sector.

If you consolidate the financial sector with the government sector, then changes in the quantity of bank reserves disappear. The central bank sets interest rates, and based on those interest rates (as well as transaction demand, risk aversion, etc.) the non-financial sector makes a portfolio choice of how to allocate their claims on the banking sector -- whether they want to hold bank deposits, bank bonds, or bank stock.

I think it would be great to re-work all of the previous analogies in the following format: Assume that all banks are owned by the central bank, that every household has a deposit account at the one government bank. All private sector monetary transactions consist of writing checks on the CB accounts. At the beginning of each period, the CB announces an interest rate, and promises to

1) lend as much as is demanded at that rate (during the period)

2) sell as many bonds as are demanded paying that rate (during the period)

Then the growth of money (e.g. CB deposits) would be the difference between 1) and 2), which is very similar to your original model, except that the CB has no control over quantity, and only has control over the rate.

Now if a lot of people borrow from the CB making certain assumptions about what their return on investment will be, when those return expectations fail, you may get into a situation in which no one wants to borrow even at a rate of zero, just because the risk of not being able to pay the loan back is too high. That means -- unless we have fiscal policy or net exports in which the government or someone else starts to borrow from the CB -- that the first term is negative. Now for the total supply of money to grow, the second term needs to be positive, so that more people prefer to hold central bank deposits rather than central bank bonds. But as central bank bonds are risk free, the investment uncertainty that causes private sector borrowing to decrease does not necessarily result in a decreased demand for central bank bonds. So it may well be the case that the money supply is shrinking regardless of what the central bank does, even at zero rates.

If you dis-aggregate the CB and private banks, then you would see the above as a large increase in bank reserves that is not matched by an increase in non-financial deposit claims on banks, which is pretty much what we have now.

[rsj]

the above should be "2) sell as many bonds" not "2) sell as money bonds" damn auto-correct.

[I edited it. NR]

[Saturos]

Sorry, but isn't all demand for money ultimately a flow demand? As you say elsewhere, "everyone holds inventories of money. We buy money only in order to sell it again... I might buy a fridge even if I could never sell it again. I would never buy money if I could never sell it again." And isn't the opportunity cost of holding money the nominal interest rate, not the inverse price level? So the "stock demand for money" would be a function of the interest rate, wouldn't it?