Three goods means two markets, if one good is money.

Paul Krugman says: "I guess my question is, why can’t we walk and chew gum at the same time? I like to introduce the subject with IS-LM as a representation of three-market economics, then say look, in practice short-run policy involves setting an interest rate target and adjusting M to get there, so that’s how we’ll do it for the next two classes." (emphasis added).

Nope.

Paul says the same thing in the paper(pdf) he links: "Now equilibrium in a three-good model can be represented by drawing curves that indicate combinations of prices for which each of the three markets is in equilibrium." (emphasis added).

Nope.

Only later in that paper does he say: "Sixty years on, the intellectual problems of doing macro this way are well known. First of all, the idea of treating money as an ordinary good begs many questions: surely money plays a special sort of role in the economy. (For some reason, however, this objection has not played a big role in changing the face of macro)." (emphasis added).

Yep.

In a barter economy with three goods, M, B, and Y, there are three markets. In the first market M is exchanged for B; in the second market M is exchanged for Y; in the third market B is exchanged for Y.

In a monetary exchange economy with three goods, one of which, M, is the medium of exchange. The third market is missing. You can exchange Money for Bonds; you can exchange Money for Output; but you cannot exchange Bonds for Output.

Each good except money has one market, in which it is exchanged for money. That's what "money" means. There is no "money market". Every market is a money market.

Actually, in a monetary exchange economy, it's even more complicated than that. Because everybody specialises in producing a different type of output. You sell your own output for money, then use that money to buy someone else's output.

If we could all just consume our own output, there wouldn't be recessions. The unemployed would just produce, and consume the same goods they produced.

If we could barter our own output, there wouldn't be recessions. Unemployed bakers and brewers would bake bread and brew beer and swap their bread and beer.

Three goods, two markets. But you can't consume your own output. You have to sell all your own output for money, then buy someone else's output for money. That's the macro world of ISLM.

Actually, I'm really pretty sure Paul Krugman knows all this. He just forgot. Stuff happens. But other people don't know this.

Comments

[White Rabbit]

If we attempted to build a real barter economy, wouldn't "super-output", such as oil or steel, which is central to most output creation processes, still be able to trigger general recessions?

I.e. the most generic definition of recessions would be: "when a sudden price increase in an interconnected enough market cascades across other markets and causes a reduction in output".

Money, naturally, is the most central good - but other goods are central too.

This does not reduce the validity of your claim, that modern recessions in diversified, open economies are predominantly caused by money - i.e. caused by the coordinated break-down in promised value of future work performed.

[Donald Pretari]

http://www.guardian.co.uk/commentisfree/2011/sep/25/charlie-brooker-capitalism-and-shreddies
"If the entire global economy goes down the wazoo (or up the wazoo – I guess it depends on where the wazoo's orifice is located), all currencies may be rendered meaningless. But if we adopt some kind of medieval bartering system instead, how will I pay for my Shreddies? I for one refuse to perform sexual favours at the checkout. Will we still have checkouts? Or Shreddies themselves? Even if we do, I bet we won't have the "Frosted" and "Coco" varieties any more. Just plain standard Shreddies, eaten from a bowl fashioned from a dented hubcap, purchased in exchange for a hand job during a massive global war."

[Kevin Donoghue]

"In a monetary exchange economy with three goods, one of which, M, is the medium of exchange. The third market is missing. You can exchange Money for Bonds; you can exchange Money for Output; but you cannot exchange Bonds for Output."

We've been over all this many times, but still that pronouncement seems vacuous to me. If I can sell B for M and M for Y, then nothing in the model is going to change if I can sell B for Y directly. The only market where the no-barter restriction actually makes any difference is the labour market; if agricultural labourers will take payment in corn then the Keynesian notion of involuntary unemployment is hard to justify. But I see no reason why Krugman would quibble about whether bonds can be traded for output.

So yes, IS/LM is three-market economics in the sense that there is rationing in the fourth market, the labour market, so that N is determined by Y via the production function.

"Actually, I'm really pretty sure Paul Krugman knows all this. He just forgot."

I'm sure he knows you feel strongly about it and he must have read Clower's famous paper at some stage, but he has at least hinted that he really doesn't think it matters all that much.

[Nick Rowe]

White Rabbit: I suppose so. If oil production were suddenly cut off, that might cause a big disruption across a wide number of markets, and a fall in output and employment.

But recessions are normally associated with: a fall in prices (or a fall in inflation); it gets harder than normal to sell stuff, and easier than normal to buy stuff. I don't think those two things would happen if the only problem were a drop in oil supply. Probably prices would rise, and it would get harder than normal to buy stuff.

Don: well, with a big enough economic disruption, that's what it might come down to. "What have you got to trade?"

Kevin: "If I can sell B for M and M for Y, then nothing in the model is going to change if I can sell B for Y directly."

That's what I keep thinking about, almost daily. Sure, we are agreed it matters that you can't sell Y for Y, and Y for L. But does it matter if you can't sell B for Y, if the M/B market clears? (But then the M/B market does not clear, for some people, who are borrowing-constrained. Does this affect the answer?) I wish my head were clearer.

[Deus-DJ]

hi, i've visited your blog for the past week...questions...why do you insist on equilibrium models, on using ISLM when conducting analysis, and continue to analyze things in terms of a barter economy especially after publication of David Graeber's book on 5000 years of debt? Was Mirowski not penetrating enough to stop you from conducting silly equilibrium exercises? Is post keynesian criticism of ISLM not enough for you to stop using it? Was a work by an anthropologist not enough to stop you from assuming that before money there was barter?

Why are you even a monetarist when money is introduced to the economy endogenously? There is of course an exogenous component, but the way you model doesn't assume the endogeneity of money at all.

[Nick Rowe]

Deus-DJ: You read like some sort of self-parody of a Post-Keynesian. If you have actually read *any* of this blog over the last week, you must have very poor reading comprehension. You ask why I continue to analyze things in terms of a barter economy???

[Kevin Donoghue]

"But then the M/B market does not clear, for some people, who are borrowing-constrained. Does this affect the answer?"

Once you take account of the fact that some borrowers are constrained you have to put IS/LM aside, I think. At some point Hicks has to pass the ball to Stiglitz.

[Nick Rowe]

Kevin: wouldn't Hicks pass half the ball to (say) Paul Krugman's babysitting coop model? If half the people are borrowing-constrained, they act as if they live in a model with just Money and Output?

[Kevin Donoghue]

Nick, when it comes to models in which different agents face different constraints (some being up against their credit limit for example) it seems likely to get horribly complicated very quickly. I quit studying economics in the late 1980s and came back to it just a few years ago. As I try to get a bit more up-to-date I'm struck by how difficult macro can be even without Keynesian complications. Keynes was obviously wrong about long-run analysis being "too easy, too useless a task" -- maybe it's useless but it sure ain't easy! He thought Ramsey's model was "terribly difficult" but it doesn't seem to have dawned on him that that's where economics was going.

I’m not sure where I’m going with this but I suppose what I mean is that if you want to model agents facing multiple constraints – credit rationing as well as labour rationing – you may be faced with a tangle of different possible regimes, like Malinvaud only more so. As Queen Elizabeth the 1st said to the Earl of Essex when she sent him to sort out the Irish: “I fear you’ll not return from that bog.”

[Dan Kervick]

A theoretical question: Start with the spot markets for the three products. Does the addition of a credit market represent a different kind of market altogether, in which a different kind of product is exchanged? Or is it just a way in which the markets for the three products may be re-organized?

For example, it would seem as though for any two goods A and B, the rules in the market for exchanges of A and B might permit credit-based exchange. That is, people might sometimes exchange an immediate delivery of A at time t for a promise of delivery of B at some later time t'. Between t and t', the seller of B is indebted to the seller of A, and after making delivery the debt is discharged.

Once credit is involved, you can even have meaningful exchanges of A for A. For example X might deliver a certain amount of seed corn to Y in the spring, and in exchange Y might deliver a somewhat higher volume of seed corn to X following the corn harvest. In the modern world, we have a lot of exchanges of money for money that occur in this way.

If we think of credit as just one way of making an exchange, and if we allow that with the addition of credit, A-for-A exchanges are economically meaningful, then if we have two non-monetary products A and B, along with some money M, there are actually six markets to consider: A-for-B, A-for-M, B-for-M, A-for-A, B-for-B and M-for-M.

On the other hand, we might think of all exchanges as spot exchanges, and conceptualize credit as involving the exchange of an entirely different kind of product: For example, the seller of A exchanges A on the spot for an IOU-for-B form the seller of B. Later, the new holder of the IOU-for-B can exchange that IOU for some other product. But typically it is exchanged with the seller of B for some B, at which time the seller of B has a sort of self-cancelling IOU which is extinguished.

If we think of things this way, then whenever we have three fundamental products - including M -involved in spot exchanges, we might also have IOUs for each of those products. This means that there are actually six products being exchanged: A, B, M, IOUs-for-A, IOUs-for-B and IOUs-for-M.

[Nick Rowe]

Dan: "If we think of things this way, then whenever we have three fundamental products - including M -involved in spot exchanges, we might also have IOUs for each of those products. This means that there are actually six products being exchanged: A, B, M, IOUs-for-A, IOUs-for-B and IOUs-for-M."

That's the way I think of it. And if it's a monetary economy with 6 goods, there are (6-1)=5 markets. Or if it's a barter economy, there are (up to) (6-1)x(6/2)=15 markets.

[Ralph Musgrave]

Nick says “If we could barter our own output, there wouldn't be recessions.” Just to be pedantic, an individual or group of individuals who make product X in a barter economy can become wholly or partially unemployed if no one else is able to make product X, and if the above individuals charge an excessive price for product X. Another condition would be that the above individuals cannot turn to producing something other than X.

[J.V. Dubois]

Deus-DJ - Did you read Nick's response to the discussion in this thread: http://blog.mises.org/18371/murphy-replies-to-david-graeber-on-menger-and-money/comment-page-1/#comments

So as I understand it, there is nothing fancy on Graeber's findings. He claims that if you study primitive societies, there is never a spot barter trade found. So the only thing he proves is that most likely the credit existed prior to money exchange, and that money developed for multitude of reasons in various societies. However the key finding remains the same - there is no larger societal structure that can exist without money. Large temple complexes invented money for accounting purposes until it is found as a convenient way to conduct the exchange. In another example he states that relative prices of one good in terms of the other were constructed by law just to impartially settle disputes for property damage and such.

So what. This Graeber's argument has nothing to do with barter economy as usefull tool in thought experiments. And actually, he even admits that barter economies actually exist in societies with previous experience with money (such as POW camps). This fits to 99.999% of the world population - so why not using it as an alternative to money exchange economy?

So just to sumarize, what I've got from Graeber is that if the division of labor in society is flat, and that most interaction are personal and repeating (such as between neighbors or in family), money is not needed. However concept of debt and credit remains. But this is not news to me. I owe my friend a favor without putting a price tag on it. I understand this concept. But if the society reaches certain stage where anonymous exchange becomes a norm, you cannot rely on this purely moral norms. Then you must invent money - for accounting and for exchange.

[GeorgeYuri]

(In response to White Rabbit, who wrote

"..wouldn't "super-output", such as oil or steel, which is central to most output creation processes, still be able to trigger general recessions?..")

http://en.wikipedia.org/wiki/Patent_encumbrance_of_large_automotive_NiMH_batteries

and other related comments, here -

http://www.macrobusiness.com.au/2011/10/molten-salt-solar/

Skim the home page for excellent reporting on economic matters from down under -

http://www.macrobusiness.com.au

[Donald A. Coffin]

I've recently begun re-reading Axel Leijonhufvud's "On Keynesian Economics and the Economics of Keynes. I've just finished the section in which he writes about aggregation issues, using this framework:

Markets/Goods
Consumption goods
Capital goods/equity
Bonds
Money
Labor

How do we aggregate this? He argues that the "Keynesian" aggregation is
Goods (Consumption goods and Capital goods/equity)
Bonds
Money
Labor

While Keynes' aggregation is
"Long" assets (Capital goods/equity, bonds)
"Short" goods (Consumption goods, or what Keynes called "wages-goods")
Money
Labor

So in what he calls the Keynesian model, you have three prices:
Goods prices (proxy: GDP Deflator)
Interest rate
Wage
And the numeraire--money

And in Keynes' model, you also have three prices:
Interest rate (for capital/equity and bonds)
Consumption goods price
Wage
And the numeraire--money

But in the 5-good model, you'd need four prices:
Return to equity
Return to bonds (interest rate)
Consumption goods price
Wage rate
And the numeraire--money.

I've struggled, off and on, since I first read his book, with trying to clarify (in my own mind) which aggregation scheme makes the most sense. What is clear, I think, is that Leijonhufvud is right, that the aggregation scheme does matter. In effect, the question is (as he argues), which is more stable, the price of capital goods relative to the price of consumption goods, or the price of capital goods relative to the price of bonds?

Finally, is it possible to disaggregate that and do a model with more markets/goods successfully, or do the modeling complications overwhelm the gains?

[Too Much Fed]

Nick's post said: "Kevin: "If I can sell B for M and M for Y, then nothing in the model is going to change if I can sell B for Y directly."

That's what I keep thinking about, almost daily. Sure, we are agreed it matters that you can't sell Y for Y, and Y for L. But does it matter if you can't sell B for Y, if the M/B market clears? (But then the M/B market does not clear, for some people, who are borrowing-constrained. Does this affect the answer?) I wish my head were clearer."

IMO, people need to be more specific here. Most people are going to think that sell B for M is going to mean sell an EXISTING bond for medium of exchange (not currency denominated debt). It seems to me that most of the time when people say sell B for M they mean sell a NEW bond for medium of exchange (is currency denominated debt).

[Too Much Fed]

A little off topic but and from:

http://www.calculatedriskblog.com/2011/10/misc-market-and-more-on-household.html

"The red line, here, is median real household income, as gleaned from the CPS, indexed to January 2000=100. It’s now at 89.4, which means that real incomes are more than 10% lower today than they were over a decade ago."

There is why some people are borrowing constrained and most likely should be. The solution is not more currency denominated debt. The solution is to tighten up the labor market to generate some real earnings growth based on the budget of the lower and middle class.

[Jacques René Giguère]

TMF: That was the whole point of the Great Moderation. Squeeze middle and working class income . Let them replace that with debt. Then promise them relief in the form of measly tax cuts to be paid by reduced governemnt services.When the debts get unpaid, hire half the unemployed to beat up the other half. Hose with tear gas, lather with billy clubs, repeat.
Ford rebate for those in TO?

[Nick Rowe]

Ralph: sure, we could get excess demand for some goods, and excess supply for other goods, even in a barter economy. But recessions don't normally look much like that. But yes, you weren't saying this does typically happen.

J.V. Good response. But probably wasted on Deus-DJ. I Googled his name, and found a lot of flyby irrelevant comments on other blogs too. There are thoughtful Post-Keynesians, but he's not one of them.

Donald: good questions. I don't know.

To my mind, there is one important distinction between capital goods and long bonds that often gets overlooked. A capital good is like a long bond that promises a real (inflation-adjusted) dividend, and where that real dividend depends on the future level of demand. So you would expect capital goods prices and bond prices to respond in very different ways to expected future monetary policy. And expected future monetary policy is what is communicated in current monetary policy.

TMF and Jacques: think Permanent Income Hypothesis. We want to borrow when our current income is temporarily low.

Jacques: my memory is bad. Was it you who said that Pierre Fortin said something like: "One equation one market"? What did he say exactly, to the best of your memory? I was thinking about it as I wrote this post.

[Kevin Donoghue]

TMF, re new versus existing bonds, I don't think it matters in this context. In the IS/LM model (or my version anyway) agents are free to finish the current period with B>0 (creditors), or B<0 (debtors), or B=0 in the case of Polonius. In aggregate, private-sector bonds outstanding sum to zero. Government debt may or may not be regarded as net wealth of the private sector, i.e. you can assume Ricardian equivalence if you want to (at least I think so; haven't thought about it much though). New and existing bonds are perfect substitutes. The equilibrium condition is that the closing stock of bonds held by all agents must be the stock they wish to hold, given Y, r, P, etc.

[Jacques René Giguère]

Nick: It was " behind each equation, there must be one behavior" Whether it was leading to one market, one equation, we weren't there at the time. And then I veered into micro for the next 30 years.

[Nick Rowe]

Jacques Rene: thanks.

[Deus-DJ]

J.V. Dubois,

Yes I did(now), and its simply a poor excuse to say "I can believe whatever I want". The implications are huge of course: will economics focus on exchange or production? By focusing on barter our gaze of economics is tilted to believe that money arose through exchange and that exchange is the means by which the economy runs. It should be of no surprise that the people who scoffed at Graeber's work are those like the Austrians(and no doubt many neoclassicals). They don't want their precious worldview crumbling down around them.

Of course my questions to you, Dr. Rowe, are coming from a novice trying to understand why you don't frame your models in terms of endogenous money or why you insist on equilibrium models. I don't know how you think and these posts of yours don't shed light on these questions I have. But really, Mirowski tore down equilibrium as any sort of useful context in theorizing or abstracting about/onto the economy. I really don't understand why you and other neoclassical economists continue using it. Help me understand this as a novice. BTW, the only blog I have posted at under this pseudonym is at naked capitalism or more recently at Scott Sumner's site...which I agree, the posts at Sumner's site were particularly irrelevent(and fly-by).

[Nick Rowe]

Deus-DJ: "Of course my questions to you, Dr. Rowe, are coming from a novice trying to understand why you don't frame your models in terms of endogenous money or why you insist on equilibrium models."

Here's a post I did 6 days ago with endogenous money and disequilibrium analysis:

http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/10/wicksell-and-the-hot-potato.html

It all depends on what the question is I'm looking at.

[Nick Rowe]

Deus-DJ: "The implications are huge of course: will economics focus on exchange or production?"

That's a false dichotomy. If producers can't sell the goods they produce in exchange for goods they want to consume, they are likely to stop producing. We get a recession -- a fall in both production and consumption caused by some failure in exchange. And my belief is that we cannot understand that failure in exchange and fall in production without understanding the difference between a barter economy and a monetary exchange economy.

It seems to me that your beliefs about my beliefs are very inaccurate, and based on some very strange stereotypical views about economists that you have picked up somewhere.

[JP Koning]

"But recessions are normally associated with: a fall in prices (or a fall in inflation); it gets harder than normal to sell stuff, and easier than normal to buy stuff."

Say a recession is associated with widening bid-ask spreads. It's harder to buy stuff (or sell stuff) when spreads are wide. But during a boom, bid-ask spreads are narrow and its easier to buy and sell. Because you can have bid-ask spreads in barter, you can have recessions.

[Too Much Fed]

Jacques René Giguère said: "TMF: That was the whole point of the Great Moderation. Squeeze middle and working class income . Let them replace that with debt. Then promise them relief in the form of measly tax cuts to be paid by reduced governemnt services.When the debts get unpaid, hire half the unemployed to beat up the other half."

Right, the "Great Moderation" was about putting negative real earnings growth on the lower and middle class and then "tricking" them into currency denominated debt to attempt to grow and stabilize an economy. Almost all economists can't seem to understand that too much currency denominated debt (a medium of exchange problem) is actually destabilizing. The economists and central bankers don't expect the currency denominated debt to go unpaid because they believe, imo, that the number of hours of work to be unlimited. That means the workers will work more in the present and in the future. They are shocked to "learn" the number of hours of work is not unlimited and try to come up with some other explanation. They refuse to change their assumptions.

[Too Much Fed]

Nick said: "TMF and Jacques: think Permanent Income Hypothesis. We want to borrow when our current income is temporarily low."

The problem with the Permanent Income Hypothesis is that it is an assumption about future wage income that may not come true. What happens when someone thinks his or her current wage income is temporarily low and borrows, and then that person's future wage income expectations are not met (meaning temporary turns into permanent or semi-permanent, see just above at 9:56 p.m.)?

I would rather consider this process more in terms of personal finance budgeting.

[Too Much Fed]

Kevin Donoghue said: "TMF, re new versus existing bonds, I don't think it matters in this context."

I think it matters because it depends on whether someone stops saving to spend (existing bond) or goes into debt to spend (new bond). I think it also matters because of excess savers and excess debtors.

[Too Much Fed]

A decent post about what is wrong (mostly and in charts):

http://www.businessinsider.com/what-wall-street-protesters-are-so-angry-about-2011-10?op=1

There is some medium of exchange mentioned (debt).

It doesn't include the retirement market.

It doesn't include the idea an economy can go from mostly aggregate supply constrained to mostly aggregate demand constrained without it being considered "unnatural" and needs to be corrected.

[Kevin Donoghue]

TMF, I don't know what you mean by "excess savers and excess debtors." AFAICT these terms don't make sense in the Y, B and M model which Nick is discussing.

[Sergei]

"In a monetary exchange economy with three goods, one of which, M, is the medium of exchange. The third market is missing. You can exchange Money for Bonds; you can exchange Money for Output; but you cannot exchange Bonds for Output."

There is a lot of barter in the real world. That is any collateralized lending. Looking at the H.8 table from the FED I can guess that about 2/3 if not more of ALL banking lending is collateralized (in particular real estate). So if banking system is happy to exchange goods for houses, then I see very little reason why bonds can not be exchanged for output. This means to me that in a monetary exchange economy with three goods, one of which is M, there are at least 2.5 markets if not 2.95. Which on the other hand means that bonds have 0.95 "value" of money for purposes of exchange of output. Is it enough to call them money if even houses are? Well ... at least houses *were* before the crisis.

[Too Much Fed]

Kevin Donoghue, excess savers are entities/people like Warren Buffett and Apple. They don't spend all they make, they don't invest (national account term) all they make, and don't retire to spend down their savings. Excess debtors are entities/people who have enough currency denominated debt that it is affecting their monthly budget.

[Kevin Donoghue]

TMF, why does it matter whether the bond Warren Buffett buys is newly issued?

[Too Much Fed]

KD, I own a bond. Warren Buffett saves and then buys the existing bond from me. I decide to keep saving and not spend. I put the medium of exchange into a bank (added to the model).

Warren Buffett saves and then buys a new bond from me (I went into currency denominated debt). In most cases, I am going to spend that medium of exchange into the real economy.

Warren Buffett already has a stock of bonds. If lower interest rates don't get him to stop saving (sell the bond) and spend or invest (national income/accounts meaning) and don't get him to go into currency denominated debt AND people who have no bonds don't go into currency denominated debt, what happens?

[Kevin Donoghue]

TMF, you have 3 cases:
1. You have more M, less B; Buffett has less M & more B. Y unchanged.
2. You reduce B (say from B=0 to B<0), your spending raises Y; Buffett has more B.
3. Buffett has a capital gain on his B (due lower i) and sits on it; nothing happens.

I don't think we're getting anywhere with this, so I'll bow out.

[Too Much Fed]

KD, I'll add one last comment along with the idea I have probably expanded the model.

Let's say there is positive productivity growth (PPG). It seems to me that in case 1. and 3. real aggregate demand (AD) is the same, so the PPG will be used to reduce employment. In case 2., real AD goes up, so the PPG can be used to expand output.

[Too Much Fed]

EDIT: In case 2., real AD goes up in the present but may fall in the future, so the PPG can be used to expand output in the present but may fall in the future.