The Trade Cycle; why I=S is a bad place to start doing macro.

In the olden days, economists used to talk about the trade cycle. (They meant a cycle in the amount of all trade, not just international trade). They meant a cycle in the amount of exchange. There are fluctuations over time in the amount of buying and selling that people do. In a boom, trade speeds up; and in a recession, trade slows down.

I think we should resurrect that old term. Thinking about the trade cycle is a better way of thinking about short run macroeconomics than how we currently think about it. Nowadays we don't talk about fluctuations in trade; instead we talk about fluctuations in output. It's not the same thing. We are wrong; the old guys were right. Somewhere, maybe around the 1920's or 1930's (I wish I were better at history of thought) macroeconomics took a wrong turn.

This post is a follow-up to my previous post.

Macroeconomics took a wrong turn when macroeconomists started out their analysis with I=S. Wicksell, the Austrians, Keynes, New Keynesians, and many more besides, all start out with I=S. They all started out in the wrong place. "(Desired) investment" means the (desired) amount of newly-produced capital goods and services. "(Desired) saving" means income from output of newly-produced goods and services minus the (desired) amount of newly-produced consumption goods and services. Is I=S the right way to start thinking about short run macroeconomics? Only if you think that short run macroeconomics is about fluctuations in the output of newly-produced goods and services. If you think that short run macroeconomics is about the trade cycle, then I=S is a bad place to start.

Thinking about fluctuations in output is both too narrow and too broad a way of thinking about fluctuations. And that way of thinking about the question to be answered also distorts the sort of theoretical framework we use to answer it.

Another old term for the subject matter of short run macroeconomics is the "conjuncture". It means a coming together. This word has almost disappeared from economics in English, but it survives in French, Spanish, German, Russian, and maybe other languages too. Short run macroeconomic fluctuations are a conjuncture. A lot of different data series tend to move together cyclically. Lucas drew attention to the conjuncture in his 1975 paper "Understanding Business Cycles". He said that business cycles have different amplitudes and frequencies, and it is only the conjuncture (co-movements) of different things moving together cyclically that allows us to say that all business cycles are alike, and so possibly amenable to a common explanation. If each episode were unique, we couldn't really have a business cycle theory. We would have one theory for the 1982 recession, a different theory for the 1991 recession, and so on. And each theory would have just one data point, so it wouldn't really be a testable theory.

Fluctuations in the output of newly-produced goods and services is a bad proxy for the conjuncture. It is both too narrow and too broad. The trade cycle is a better proxy for the conjuncture.

It is too narrow because it ignores trade in goods that are not newly-produced. Things like old houses and old cars and old furniture and old land.

It is too broad because it includes production of goods and services for one's own use. The unemployed worker is not unemployed. He is employed digging his own garden, fixing his own car, doing his own labour market search. He is not unhappy because he is unemployed. He is unhappy because he is self-employed and doesn't want to be. He wants to sell his labour to someone else, but can't, so has to consume it himself. A recession is not a fall in employment; it is a fall in the amount of employment that is traded. A recession is not a fall in output; it is a fall in the amount of output that is traded.

We could even imagine an economy in which employment would rise in a recession. You might have to work a lot harder to grow your own food to feed yourself than if you could sell your labour and buy food.

A house with nobody living in it is unemployed. A house with somebody living in it, who wants to live somewhere else, but can't sell his house, might as well be unemployed. If something disrupts trade in old houses, so owner-occupiers are living in the wrong houses, and consuming the wrong housing services, because they can't trade, that is a recession in housing. It's just like workers who want to trade their labour but can't, and so are consuming the wrong labour services, their own, when they would rather be trading and consuming someone else's labour services. Comparative advantage, and all that.

A recession is not a drop in output and employment. A recession is a failure to exploit the gains from trade.

Microeconomists start out thinking about trade. And not just trade in newly-produced goods and services. Macroeconomists ought to start out thinking about trade. And not just trade in newly-produced goods and services. Trade is supposed to make people better off. If something makes trade harder, and reduces the amount of trade, that would make people worse off. Maybe that's why people seem to be worse off in a recession. It's because there's less trade. And not just less trade in newly-produced goods and services.

Now, what might that something be? What could cause trade to fall and make people worse off? An increase in transportation costs? Nope. An increase in marginal tax rates? Nope. Those make theoretical sense, but don't work empirically to explain recessions.

The obvious candidate is a shortage of the medium of exchange. If there were a shortage of shopping bags, all trade that required shopping bags would be disrupted, and people would be worse off. If there were a shortage of money, all trade that required money would be disrupted, and people would be worse off. Nearly all trade requires money. The exceptions, barter trades, merely prove the rule. As far as I can tell, barter increases in a recession. People resort to barter, where they can, because there's a shortage of money. But trying to trade without money is like trying to carry stuff home from the supermarket without a shopping bag. It's not as good, so you buy less stuff, and you buy different stuff, and it's very inconvenient.

But somewhere macroeconomics took a wrong turn. We started thinking about the output of newly-produced goods and services, instead of the trade cycle. So macroeconomists started thinking about the demand and supply of newly-produced goods and services, not about trade. So they wrote down "I=S" as their equilibrium condition, and started thinking about what determined desired investment and saving. And they focused on the real rate of interest, defined as the relative price of current newly-produced goods to future newly-produced goods.

If you start out by asking the wrong question, your answer will be wrong, even if it looks right. "What causes fluctuations in the output of newly-produced goods and services?" is the wrong question to ask. "Investment, saving, and the rate of interest" is the wrong answer, because it is an answer to the wrong question.

It's a bit like defining recessions as fluctuations in the output of cars, and explaining the business cycle in terms of the demand and supply of cars, and the relative price of cars. Then trying to explain everything else that happens in business cycles as a side-effect of the fluctuating output of cars.

"What causes fluctuations in the amount of monetary trade?" is the right question to ask. "Money" is the right answer, because it is an answer to the right question.

Comments

[david]

So, money causes fluctuations in the amount of monetary trade. Great.

So, what causes fluctuations in the amount of newly-produced goods? ;)

[Frances Woolley]

Nick: '"What causes fluctuations in the amount of monetary trade?" is the right question to ask. "Money" is the right answer, because it is an answer to the right question.'

Perhaps fluctuations in the amount of monetary trade are caused by gizmos. Joe used to pay Bryan to run a retail store. Bryan used to pay Joe to manage his retirement savings. Then someone came up with a gizmo that allowed Joe to guy goods on-line directly from a warehouse, eliminating the need for retailers. This gizmo also allowed Bryan to manage his own retirement savings, eliminating the need for Joe's financial management services.

As a result, monetary trade collapsed. But the cause isn't money, the cause is technological change that reduced the need for monetary trade, by allowing people to do more things for themselves.

An old example something along these lines is the Japan-Sweden comparison. The argument is that a comparison of Swedish and Japanese nominal GDP is misleading, because lots of things e.g. child care, elder care, etc, that are monetized in Sweden are home produced in Japan. Different technology, different amounts of monetary trade.

[Nick Rowe]

david: money! Because when it's harder to trade stuff, including newly-produced stuff, you produce less.

What causes fluctuations in the amount of newly-produced cars?

[Nick Rowe]

Frances: theoretically correct. Empirically plausible for cross-section data. Empirically implausible for time-series data.

[Frances Woolley]

Nick: "theoretically correct. Empirically plausible for cross-section data. Empirically implausible for time-series data."

Is that the answer to the question "summarize the difference between micro and macro in 15 words or less?"

[Nick Rowe]

Frances: LOL. Yep!

[david]

Clearly a stable M2 isn't "enough", so one still has questions of why exactly demand for money might vary. And you can't answer that without invoking more factors.

Besides that. Clearly the composition of what is traded changes over the trade cycle so one still has to provide additional explanations linking those to money; there isn't one homogenous "goods, service, and labour" aggregate that works here. And these explanations all operate simultaneously with monetary dynamics so arguing that one is more fundamental than the other seems... well, not quite pointless but close. It's like asking: is New Keynesian economics more about nominal rigidity or about money non-neutrality?

You have instruments, which you control. You have real variables, which you care about. "Money" as an explanation doesn't suggest anything obvious about choice of instruments or the impact on real variables - all the explanatory work is being done elsewhere.

[marris]

> The obvious candidate is a shortage of the medium of exchange.

Can you say a bit more about this? Do you think the "shortage" can be decomposed into (1) a measurable decrease in the stock of money substitutes (loan and credit collapse), and (2) a desire to hold more of the real thing (since the substitutes becomes less... uhm substitute-able when everyone tries to sell)?

If what you say is true, there are very interesting things that happen in a cycle turn. For example, what "sparks" that set off downward legs of this cycle? In 2008, I think a lot of people were shocked that the Bernanke/Greenspan put did not materialize, as promised.

I think this "put" idea can be easily translated to the credit market. As long as we're on the up leg, it pays to embrace money substitutes. The guys who leave the party too early will leave "high prices" (and the chance to translate them back into base money) on the table.

[Gregor Bush]

Frances,

"As a result, monetary trade collapsed. But the cause isn't money, the cause is technological change that reduced the need for monetary trade, by allowing people to do more things for themselves."

I would say that in this case the "cause" would still be money, if you defined optimal monetary policy as that which prevents monetary trade from collapsing. But I think you highlight an important source of the disagreement between those who share the Rowe/Sumner/Fisher/Cassel/Friedman view of business cycles and those who do not (eg. Stiglitz). Those who take the Rowe view draw a clear distinction between the “initial shock” and the “cause” of the recession. If technological change leads to decreased monetary trade and leaves people temporarily unable to sell their labour at current wage rates, the appropriate response is monetary expansion which increases monetary trade. According to this view, even if the “initial shock” was a bursting of a housing bubble or an acceleration in agricultural productivity, these “real shocks” can never be the “cause” of a collapse of monetary trade (NGDP). A collapse of monetary trade is always and everywhere a monetary (policy) phenomenon.

But most people don’t see it that way. And most economists don’t see it that way. Even most Fed officials don’t seem see it that way. If you ask 100 economists why unemployment is expected to remain disastrously high in 2012 and 2013 and why inflation is expected to remain below the Fed’s implicit target, most of them will tell you the “cause” is the 1997-2006 housing bubble and the resulting 2008 banking crisis. To Rowe, Freidman, Fisher, Cassel and Sumner, that explanation makes no sense.

[Nick Rowe]

david: Yep. My saying "Money" isn't really an answer. It's just the beginning of an answer. Just like "I=S" is only the beginning of an answer.

Yep, and the big picture only gets you so far. Eventually you have to disaggregate and start looking at different bits, that are related to the whole in different ways. That's true for I=S too.

And sure, we care about output of new stuff. But we also care about whether the right people are living in the right houses, or whether they are stuck immobile because the market for old houses has dried up. We care about the gains from trade, and what happens to people when they can't make those mutually beneficial trades. And that's both a deeper and more generally way of looking at the problem than output and employment.

marris: those are good questions, and good tentative answers. I wish I had better answers. I don't really.

Gregor: there's a difference though. If grandparents all suddenly moved in next door, that would solve e.g. Paul Krugman's baby sitting model problem without using money. And the "central bank" should presumably reduce the stock of money, because less would be needed now that people have found a different way to trade.

I agree with the rest.

[Min]

Gregor Bush: " Those who take the Rowe view draw a clear distinction between the “initial shock” and the “cause” of the recession. . . . According to this view, even if the “initial shock” was a bursting of a housing bubble or an acceleration in agricultural productivity, these “real shocks” can never be the “cause” of a collapse of monetary trade (NGDP). A collapse of monetary trade is always and everywhere a monetary (policy) phenomenon.

"But most people don’t see it that way. And most economists don’t see it that way. Even most Fed officials don’t seem see it that way. If you ask 100 economists why unemployment is expected to remain disastrously high in 2012 and 2013 and why inflation is expected to remain below the Fed’s implicit target, most of them will tell you the “cause” is the 1997-2006 housing bubble and the resulting 2008 banking crisis."

I think that perhaps Aristotle covered this with his distinction between material cause (medium of exchange) and efficient cause (shocks to the system). :)

[Luis H Arroyo]

Very good post, indeed.

[david]

Is there even any evidence that the composition of new vs. old in the market varies significantly over the business cycle? I know that inventory stocks vary dramatically with the cycle, but that still entails buyers of the new stuff.

[Jacques René Giguère]

Conjunture survives not only in other languages economics but in political science and politics as well. Union meetings in Québec (and many political reunions) usually begin with an "analyse de conjoncture ( conjuncture analysis). And we definitely don't conflate that with "cycles".
For a proletarian ( one who only has his "force de travail", the origin of the "workforce" concept) the primordial trade, the one without which no other (buying food say) can take place, is the selling of his labor. Proles have no significant assets to sell. If you want to study the problems of por people having income, you must concentrate yourself on newly-produced goods.
In a money economy,selling your labor depends on money being available either in the financial system or in somebody's willingness to part with it.
That's why we conflate trade,employment,income and output.

[Determinant]

As a point of history, I posit that the focus on newly-produced goods and services that developed in the 1920's was a reaction to industrialization and the politics that fell out of the First World War and the Russian Revolution.

Manufacturing was seen as the "Great Thing" and it surpassed agriculture as the great pillar of the economy. Manufacturing is almost entirely focused on production of new goods, not remanufacturing old goods. Factory workers formed a large new class of people who had shown themselves to be susceptible to leftist ideas, from the Paris Commune of 1870 to the Russian Revolution of 1917. It is unfashionable to say that Macro is an attempt to understand the economy and thereby control it, but it is true. It is meant to provide a framework for the governing class to successfully carry out policy and therefore remain in power. That's why macro always has prescriptions for government.

Both the Austrians and the Keynesians show deep traces of their political origins. The Austrian fixation on free markets and stable money is a result of viewing the Austro-Hungarian Empire's customs union and internal free trade area positively and disapproval of the economic collapse that resulted from the Empire's breakup. This dovetails with a focus on "solid money" as the Austro-Hungarian Crown was seen as a well-managed currency with gold backing; the successor currencies did not fare as well.

In contrast Keynesianism's focus on employment comes from British policy after the First World War to return to the prewar unemployment rate of 6%. Demobilized soldiers were provided with a one-year Out-of-Work Donation Policy to tide them over. But the British economy never returned to that 6% rate in peacetime until the late 1940's. Britain had a single, well-established currency and a unified government and still couldn't make it's goals happen. That is the historical motivation for Keynesianism and it is quite different from Austrian economics.

[Robillard]

Nick, thanks for the interesting re-casting of the theory.

I guess when I think of why firms layoff employees, it is because of shortage of money: a shortage of revenue sufficient to compensate the suppliers of labour while still providing an adequate return to capital.

You said that in unemployment, a worker is consuming his own labour because he is unable to trade it to others because of lack of money. I would think that, upon initial consideration, an unemployed worker generally consumes very little of his own labour though. Instead, unemployment is more like being forced to allocate more time to leisure than one would otherwise. Is leisure essentially consumption of one's own labour product?

To put it differently, every individual has preferences in which they can experience utility by consuming the product of their own labour (leisure?) and/or the labour product of others (goods & services). The actual description of the utility function doesn't matter so much, except that at some point on the indifference curve defined by the utility function, there is declining marginal utility to consuming an additional unit of one's own labour product. Each individual rationally seeks to maximise their utility subject to the budget (time) constraint. In unemployment, the individual is unable to exchange their labour product for money, which is used to buy the labour product of others. So instead, the individual consumes more of their own labour product, which because of declining marginal utility, makes the individual less happy than they could be if they were not unemployed.

[primedprimate]

Excellent post as always but I have two different strands of questions.

What causes fluctuations to money? Can the thing that causes money to fluctuate be smoothed out therefore preventing the fluctuations in money altogether or is it better to smooth out money by directly adjusting the quantity of moeny? Assuming that we can and should smooth out money directly rather than smooth out the underlying source of fluctuations, are there welfare implications based on how we choose to smooth out money (perhaps some methods of smoothing money are better than others?) or is that immaterial?

Could waves of pessimism and optimism cause the trade cycles? Perhaps in a recession, there is a coordination problem such that people don't think there are any good opportunities for trade (they would rather consume or invest in the next period) and so they hold on to money instead. The total number of pieces of paper floating has not changed (so money has not fluctuated) but the extent to which they are being exchanged has reduced. Such a situation could be fixed by increasing the number of pieces of paper so that it is no longer worthwhile to hold on to pieces of paper for the purpose of future transactions OR it could be fixed by directly creating enough opportunities for trade that dislodge the economy from the bad equilibrium in the coordination problem.

[Nick Rowe]

Min: "I think that perhaps Aristotle covered this with his distinction between material cause (medium of exchange) and efficient cause (shocks to the system). :)"

I gotta try to remember that.

Luis: thanks!

david: "Is there even any evidence that the composition of new vs. old in the market varies significantly over the business cycle?"

Good question. Not that I know of. I think that both new and old house sales are procyclical, AFAIK. If they were both equally pro-cyclical, so that the ratio of new to old didn't have any obvious cyclical tendency, that would suggest it would be very strange to look at only new house sales. Which is what theory does now.

Jacques: "That's why we conflate trade,employment,income and output."

Maybe, yes. It would perhaps be different in a peasant economy.

Determinant: Maybe. But I wonder how much it also comes from measuring GDP for a war economy? I can't remember when GDP accounting started.

Robillard. I basically agree. I don't think the labour/leisure distinction can really carry the weight we try to put on it. I can use my time myself, or try to sell it to someone else. If I spend time fixing my car or digging my garden, I think of it as leisure. If I were unable to sell my labour, and so were fixing my car or digging my garden because I couldn't earn the money to hire a mechanic or buy food, I might think of the same activities as labour.

[david]

National accounting was pioneered, unsurprisingly, by the communists in the Soviet Union crafting industrial policy. It became widely implemented in the West in response to the Great Depression.

[Nick Rowe]

primedprimate: Thanks!

Those are both good questions. They aren't off-topic, since they are the very sorts of questions to which this post would lead. But I'm still going to duck them, because they are beyond the somewhat narrow scope of this post. (Plus, I don't have any really good answers either!).

david: Hmmm. Which roughly coincides, I think (not sure) to the rise of I=S in macroeconomics.

[Richard H. Serlin]

Ok, so money is big, but I'm still curious about the Stiglitz article:

Do you think you might be taking Stiglitz too literally, like he doesn't literally mean monetary measures can't help at all. He just means that the big root of this is a huge structural shift, and if you don't address it, don't speed the transition, then you'll just keep getting this drip, drip, drip, over decades.

It's like monetary policy helped in the 2000's, but you had to keep doing it for years and years. You didn't really address the cause of the constant drip, drip, drip. You didn't actually re-surface your roof, you just kept bailing with buckets. Sure it had an effect, it kept your house from molding and rotting, but it would be a lot better to really get the roof in good shape.

Certainly I strongly prefer Stiglitz's solutions. These are really high return investments of the kind the pure free market will grossly underprovide due to externalities, asymmetric information, coordination problems,…

Growth economists Charles I. Jones of Stanford and John C. Williams of the San Francisco Fed wrote in a 1998 QJE paper:

Is there too much or too little research and development (R&D? In this paper we bridge the gap between the recent growth literature and the empirical productivity literature. We derive in a growth model the relationship between the social rate of return to R&D and the coefficient estimates of the empirical literature and show that these estimates represent a lower bound. Furthermore, our analytic framework provides a direct mapping from the rate of return to the degree of underinvestment in research. Conservative estimates suggest that optimal R&D investment is at least two to four times actual investment [emphasis added].

– Vol. 113, No. 4 (Nov., 1998), pp. 1119-1135

[Jacques René Giguère]

The first inkling of National Accounting, as well of input-output matrix, dates from the physiocrat François Quesnay's 1759 Tableau économique.
He credits no value to industry or commerce as they are not "producing" anything. Interestingly, the communist National accounting (Gross Social Product) still used in Cuba,based on XIXth century concepts of who produces value, credits industry but not services.
Today, we recognize the value of commercial services but , funnily, ask Chamber of Commerce types and a lot will tell you that indeed farmers,industrialists and merchants (along with their employees) produce value but nurses or teachers are "consumers" (in the sense of "moochers"). How they reconcile that with private schools or hospitals, I have no idea.
The CofC types are always a step behind in their thinking. God knows what they make out of scientific research and ,even worse, artists and philosophers).

[Martin]

" "What causes fluctuations in the amount of monetary trade?" is the right question to ask. "Money" is the right answer, because it is an answer to the right question."

So if the cause is money, why not use another money instead? If the recession in the US is due to a shortage of people not getting enough green pieces of paper with dead white males on the, why don't they use different pieces of paper? Surely, there must be people somewhere who because they cannot get dollars, will substitute these for another medium of exchange?

There is no good reason, aside from legal tender, why people do not accumulate different pieces of paper. Well, to be honest, I can think of one good reason and that is transaction costs of bargaining with people to accept different pieces of paper and have people agree to switch, but is that really equivalent to an excess demand for money or is that something else?

[Max]

This brings to mind something I read in the WSJ yesterday. Some bond traders were complaining about reduced trade volumes in the short term treasury bond market, and attributed it to lower Fed uncertainty, and hence less disagreement between speculators.

So this is something else that can cause trade: speculative disagreement.

[Nick Rowe]

Richard: "Do you think you might be taking Stiglitz too literally,...?"

Possibly. The Vanity Fair piece perhaps wasn't very clear. And sure, some problems are not caused by money, and can't be cured by good monetary policy. (Though bad monetary policy won't help.)

Jacques: my father, a farmer, had physiocratic leanings at times. The key ratio was between the number of people working on the land compared to the total population. That was a measure of surplus value created. Factory workers didn't produce a surplus, they just transformed value. You put food in one end, and cars come out the other.

Martin: "Well, to be honest, I can think of one good reason and that is transaction costs of bargaining with people to accept different pieces of paper and have people agree to switch, but is that really equivalent to an excess demand for money or is that something else?"

That's the reason I would give too. Money is like language. We want to use the same money as everyone else we trade with. It is very hard for a new money, like LETS, to get off the ground. But, AFAIK, new monies like LETS do tend to get more use during a recession, which tends to confirm my perspective.

Max: A lot of things can cause trade to increase or decrease. But when you see a generalised synchronised decline in trade, *plus when you see that it gets harder to sell goods for money and easier to buy goods with money*, one should suspect a monetary cause.

[W. Peden]

Martin,

Under certain conditions, people DO use different currencies. Businesses engaged in international trade have foreign currency holdings. In countries where the currency reaches the level of being unusuable e.g. hyperinflationary countries like Zimbabwe, people use either another fiat currency like the dollar or cigarettes/cognac etc.

(I can't think of any examples of a currency undergoing severe and unstable deflation being abandoned by its users for another currency as opposed to reverting to barter/gift economics. In recessions, it does seem as though "good money" drives out bad.)

[W. Peden]

And, of course, Eastern Bloc type economies have two parallel economies: a hard currency grey/black/official hard currency economy and the official economy proper. For all the supposed link between the rouble and the pound in the Soviet era, a bit of sterling went a long way in the USSR if you could find the right people.

[Jon]

Nick wrote:
"So they wrote down "I=S" as their equilibrium condition, and started thinking about what determined desired investment and saving. And they focused on the real rate of interest, defined as the relative price of current newly-produced goods to future newly-produced goods."

Do you really mean that the Wicksellian natural rate is the discounting of newly-produced goods. I don't remember it that way--and I don't mean that with confidence to contradict you, but I definitely never thought about it as being a rate DEFINED by newly-produced goods. In Austrian literature, it was referred to 'time preference' which is certainly presented a universal rate not something restricted to newly produced goods and services.

So who defined it to be such a thing?

[reason]

"He said that business cycles have different amplitudes and frequencies, and it is only the conjuncture (co-movements) of different things moving together cyclically that allows us to say that all business cycles are alike, and so possibly amenable to a common explanation."

Yes, but I don't think they are all alike at all. This cycle is quite different in its nature from most cycles since the Great Depression (Japan 1990 being the obvious exception). A balance sheet recession is quite different from a recession started by the Fed to reduce inflation. In this case I think your argument is wrong.

[Martin]

W. Peden,

I know they do, that's why I find it difficult to call it an excess demand for money. I implied as much here,

"Surely, there must be people somewhere who because they cannot get dollars, will substitute these for another medium of exchange?"

When we're talking about an excess demand for money, it seems to me that we're talking about transaction costs of switching (or changing prices..). I don't know whether it is correct to say that we're dealing with an excess demand for money rather than that we're dealing with rigidities that can be overcome by treating it as if.

Nick,

Speaking of metaphors, have you heard of the paper by Narayana Kocherlakota 'Money is Memory' http://ideas.repec.org/p/fip/fedmsr/218.html ?

[reason]

John,
I don't quite understand why you think "the real rate of interest" and the "Wicksellian natural rate of interest" are the same thing.

[reason]

Martin,
"There is no good reason, aside from legal tender, why people do not accumulate different pieces of paper. Well, to be honest, I can think of one good reason and that is transaction costs of bargaining with people to accept different pieces of paper and have people agree to switch, but is that really equivalent to an excess demand for money or is that something else?"

Try debt - and money as a unit of account. It doesn't help to offer Cyranian crowns (or whaever) if my contract says I have to pay US$.

[Martin]

Reason,

If your contract specifies $US, you could just as well give them a different currency, buy them a put with the premium you get for selling a call. No damage done. There can only be trouble when there are transaction costs, frictions in the market. Is the problem then an excess demand for money or the frictions? In the absence of frictions this is what you can do, even if there is an excess demand for money.

[Martin]

This is Say's Law by the way: only rigidities can cause gluts. In this view, an excess demand for money only becomes a problem when the transaction costs of switching are too high.

[Ritwik]

Nick

Sorry if I'm being daft here, but isn't 'output' as defined for national income accounts exactly equal to output traded? Re-sale of a house at a higher value is GDP. Cleaning your own house is not. This is true in a recession or otherwise.

[Nick Rowe]

Jon (and reason): in New Keynesian macro (which is now really Neo-Wicksellian macro, after forgotten his name's famous, canonical book [I haven't had coffee yet]) the Euler equation IS curve has saving and investment depending on the real interest rate, which is the nominal rate minus expected inflation. And expected inflation means the expected rate of change on the GDP deflator, which is new goods.

In Wicksell, savings and investment depend on the gap between the market rate of interest and the natural rate of interest.

The rate of time preference proper, which you are talking about, is one of the things that determines the Wicksellian natural rate of interest.

reason: "A balance sheet recession is quite different from a recession started by the Fed to reduce inflation. In this case I think your argument is wrong."

Why? Suppose we bartered old houses, as opposed to buying and selling them for money. Why would a "balance sheet recession" stop people from swapping houses when they wanted to move?

Martin: I've heard of it, but haven't read it. But I vaguely remember Hicks (or someone) making a similar sort of argument decades ago. I find it an interesting argument, but I disagree on, well, sort of philosophical grounds. It's a "functionalist" explanation.

reason: "Try debt - and money as a unit of account. It doesn't help to offer Cyranian crowns (or whaever) if my contract says I have to pay US$."

But, why did you originally agree on a contract to pay US$, as opposed to crowns?

Ritwik: you are not being daft. I thought about talking about this in the post, and decided against.

1. GDP, as defined, includes some output that is not traded. Unsold inventory and imputed rents to owner-occupiers of houses are two examples. There may be more.

2. Resale of an old house at a higher price is not included in GDP. (But if you hire builders to improve the house that is included in GDP.)

3. The theoretical concept of output we use does include output we produce and consume ourselves, without selling it. Economists complain that GDP doesn't measure output the way we want it to be measured (because it's too hard to measure some stuff).

So, it's messy. There are three different concepts:
1. Theoretical output of newly-produced goods and services, whether traded or not. (I=S macro theory).
2. All transactions of both new and old goods. What I'm talking about.
3. GDP as measured. Which is mostly, but not fully, newly-produced goods and services that are traded.

[reason]

"Why? Suppose we bartered old houses, as opposed to buying and selling them for money. Why would a "balance sheet recession" stop people from swapping houses when they wanted to move?"

Is this relevant? And the answer is yes - because they don't own them anymore.

[Kevin Donoghue]

“The obvious candidate is a shortage of the medium of exchange.”

Knew you would say that. Reminds me of a remark about Roger Scruton: there’s no fun in reading his essays about sexual morality, because you just know he’s not going to end up advocating free love. There’s no suspense.

It’s only ‘obvious’ because that’s the way you’ve been thinking for years. Think of something else instead: credit. I do believe that’s what Stiglitz thinks about. As he says, there’s something special about lending. You don’t lend your money to the highest bidder, because the highest bidder almost certainly won’t repay the loan.

A cyclical downturn comes when lenders are suspicious. It’s not money that’s scarce, it’s trust. The market in used bicycles can be quite brisk. But the market in used houses dries up. It’s the big-ticket items – new and old – which trade less freely, because buyers ordinarily require credit to purchase them.

[Andy Harless]

The "trade cycle" concept is useful, but I don't think we can conclude that the medium of exchange, as such, is responsible for the trade cycle. In general, a failure of trade is the result of some price being stuck at a level that does not clear the market. In general, such a distortion will result in certain beneficial trades not being done. More specifically, the trade failures that we call depressions (or recessions) occur when the exchange rate between perishable and non-perishable goods is stuck at a level that does not clear the market, such that a surplus of perishable goods develops.

Labor is infinitely perishable: each person-nanosecond of labor, if it doesn't get used in its particular nanosecond of availability, becomes completely useless in the subsequent nanosecond. Other goods have varying degrees of perishability. But there is one particular good -- which, just coincidentally, happens to be the medium of exchange -- that is both low in perishability and sticky in price. When people decide they want to consume more in the future and less in the present, they shift their demand toward non-perishables. If all goods had perfectly flexible prices, this would simply mean that the price of all non-perishables would rise relative to all perishables until people were satisfied with their consumption path at the available exchange rate. But since one non-perishable good is sticky in price, when people get too patient, they will keep trying to buy this good even when there is no more available. Thus there is a failure of trade.

The solution is to make the sticky-priced good perishable -- in other words, have the central bank target a high inflation rate.

[Adam P]

"reason: "A balance sheet recession is quite different from a recession started by the Fed to reduce inflation. In this case I think your argument is wrong."

Why? Suppose we bartered old houses, as opposed to buying and selling them for money. Why would a "balance sheet recession" stop people from swapping houses when they wanted to move?"

Nick, surely that's clear. Suppose I bought my house by borrowing $100 (the house was worth $100, I made no downpayment). Now the house is worth $50.

My only asset bother than the house is a cash holding of $2.

I do however have enough income to service my current mortgage.

I'd like to sell the house and buy a different one that's also worth $50, or $40, or some other amount greater than $2.

So, can I do this? I'd submit that the obvious answer is that yes I can, which make's it sound like Nick is right.

However, it would appear that I can only do this within the context of my current mortgage. If the bank that holds my current mortgage allows me to switch the house that secures the mortgage then there's no problem, I sell the current house for $50 and use that to finance the purchase of the house I'm moving to. The bank and I are in an unchanged financial position with each other. I still owe $100 against an asset worth $50.

This implies that certain transactions costs have risen on average, what if the house I'm moving to is in another state and the holder of the mortgage doesn't do business there? What if the mortgage was sold in the boom and I can't find out who the current owner is? There are all sorts of reasons why I may want to finance the new house with a different mortgage and this will not be possible without defaulting on the first mortgage and if I do that then this will surely increase the cost of getting a new mortgage.

Thus, a "balance sheet recession" is nothing more than an increase in average transactions costs for transactions requiring credit intermediation. It's just the Bernanke-Gertler effect, an increase in the external finance premium for a set of economic actors, an increase in the cost of credit intermediation.

Most importantly it has nothing to do with a lack of real money balances, there is no excess demand for money, and thus it shows why MV=PY or MV=PT is a bad place to start doing macro.

[reason]

Nick,
"But, why did you originally agree on a contract to pay US$, as opposed to crowns?"

I don't quite understand your point - history matters we use dollars now because we used dollars in the past.

Martin
"If your contract specifies $US, you could just as well give them a different currency, buy them a put with the premium you get for selling a call. No damage done. There can only be trouble when there are transaction costs, frictions in the market. Is the problem then an excess demand for money or the frictions? In the absence of frictions this is what you can do, even if there is an excess demand for money. "

Does this make sense? If I can buy the put I can buy the currency - but that is exactly the problem isn't it. The currency is too expensive so I make a loss on the deal. It is not that dollars are unavailable - it is that there is excess demand for them. You see you made a deal in the past and you expected you could get the dollars. But the other party knowing that crowns are cheap and dollars are expensive is not just going to let you off the hook.

[reason]

Kevin Donahue - good comment.

[reason]

Martin,
I don't think you are really getting it. In order to obtain a different currency - say crowns, I have to sell the services I provide for crowns. But in (say) the US nobody uses crowns so nobody has them. And as nobody has them they are pretty useless for exchange in the US. And everybody at the moment wants more cash and so everybody is trying to save. What you are saying effectively is we should increase exports. Agree entirely. But the international financial system won't play ball and price dollars so that that happens.

[Nick Rowe]

Adam, Kevin, reason: OK, so you are arguing that the transactions costs due to negative equity and difficulty in getting credit is what causes house sales to decline.

My theory is different. House prices are sticky, especially downwards. So when demand falls, the actual price fails to fall quickly to the new equilibrium, and so there's an excess supply of houses, and fewer houses get bought and sold, and the inventory of houses for sale ("unemployed" houses) increases. The fall in demand for houses might be a local or microeconomic problem, or it might be part of a general decline in demand for goods due to a monetary problem.

Now, how could we empirically test between those two theories? (BTW, I would not deny that there is some truth in your theory.) That's a genuine question; I can't think of a good answer right now.

[Nick Rowe]

Kevin: "Knew you would say that."

Yep, I'm sort of predictable like that! I need more sycophantic commenters, who will instead say: "Brilliant insight! I would never have thought of 'money', but now you say it, it's obviously right!" (Just joking.)

"As he says, there’s something special about lending. You don’t lend your money to the highest bidder, because the highest bidder almost certainly won’t repay the loan."

I don't think that's special to credit. You don't (always) buy the cheapest good, because it might be very low quality. General market for lemons problem. It's the medium of exchange that's special.

[Nick Rowe]

Andy: I'm not sure about your distinction between perishables and non-perishables. Don't we see a decline in sales of both perishable and non-perishable goods in a recession?

reason: "I don't quite understand your point - history matters we use dollars now because we used dollars in the past."

Aha! forget my point. I think I now understand your point. Let me restate it, as I understand it. If we were all using dollars in the past, and if we all thought we would all be using dollars in the future, then in the past we would have signed loan contracts payable in dollars. And when the future arrives, that makes it harder than it otherwise would be to switch from dollars to something else. So that creates additional inertia in using any given type of money, over and above any inertia caused by the difficulty of coordinating everybody switching at once.

Good point. Obvious, now you say it. God knows how I missed it in the past, until you made it. You taught me something. Kudos!

[Kevin Donoghue]

"You don't (always) buy the cheapest good, because it might be very low quality."

True, but I don't think the BMW/Skoda premium has much to do with the trade cycle. Rationing of BMWs does happen, but I don't think it's a matter of great consequence from a macro viewpoint. Credit rationing surely is.

[reason]

Nick,
I actually agree with you about sticky prices on houses. I don't think it is either/or, it is both.

[reason]

Nick,
" So that creates additional inertia in using any given type of money, over and above any inertia caused by the difficulty of coordinating everybody switching at once."

If you'd be thinking about the process of a country leaving this Eurozone, this should be obvious.

[Adam P]

""You don't (always) buy the cheapest good, because it might be very low quality."

True, but I don't think the BMW/Skoda premium has much to do with the trade cycle."

Yes, and most importantly the BMW/Skoda premium presumably is to a certain extent a difference in production costs.

[J.V. Dubois]

Kevin: "A cyclical downturn comes when lenders are suspicious. It’s not money that’s scarce, it’s trust". What kind of trust do you mean by this? Surely not the trust that sun will rise again tomorrow, or trust in your wife. The only kind of trust that comes to my mind in this relation is trust that creditor in general will have sufficient income in the future to pay the installment on time. An their ability to so relies on their ability to sell their labor - or as Nick says in this post - even on their ability to trade some of their old assets on the market. And if in recession it is harder to sell anything on the market "because of money" (as Nick claims), then ultimately, it is the money hiding behind the scarcity of trust of yours.

[Kevin Donoghue]

@J.V. Dubois: ...if in recession it is harder to sell anything on the market "because of money" (as Nick claims)....

But Nick's claim is just what I'm disputing. I'm living amidst the wreckage of the Irish economy and I really don't see that the changes I've witnessed have much to do with money, as such. They have a lot to do with the fact that people who were having loans thrust upon them a few years ago by eager bankers now have no chance of obtaining a mortgage. Nick wants the problem to be about money. That's what makes him a monetarist. If he wanted it to be about sin he'd be a Calvinist.

@reason: thanks.

[reason]

J.V. Dubois
Yes - but this may have another cause than money. Maybe the future looks more uncertain than it did. (O.K. We can then argue that this is because the central bank is not agressive enough in managing expectation and isn't responding appropriately to a rise in the precautionary demand for money. This is all getting very chicken and eggish to me.) I think the key in the end is that both fiscal policy and monetary policy are not distributionally neutral in their effects and the we actually care deeply about distributional effects. But I'll just hide in the corner so both sides can keep being distracted and pretend they are only arguing about countercyclical policy.

[reason]

See, I think fiscal policy is just another variant of monetary policy with different distributional implications.

[Andy Harless]

Nick:

"Don't we see a decline in sales of both perishable and non-perishable goods in a recession?"

We see a decline in sales of those non-perishables whose value is largely in the services they provide for current consumption (cars, household appliances, etc.), because people are trying to shift their consumption to the future. A car may not be very perishable, but the services it provides are perishable. So if you wanted to drive more 5 years from now but not today, you wouldn't buy a new car and leave it in your garage for 5 years.

Compared to money, though, even a physical car is perishable, in the sense that it's costly to maintain and loses value over time even if you don't drive it very much. There aren't many goods that retain 94% of their value after 3 years -- especially in risk-adjusted terms. Depending on how you do the risk adjustment, I'm not sure I can think of any non-financial goods that rise to that level of non-perishability. And the things that come close aren't necessarily hard to sell during a recession. (Looking at what the price of gold has done recently, I don't imagine that jewelry has been hard to sell.)

[J.V. Dubois]

Reason: "See, I think fiscal policy is just another variant of monetary policy with different distributional implications" - I am with you on this. And I think Nick and other Market Monetarists would agree too. Macro should be about preventing the fluctuation of Aggregate Demand (and output) in a way which has the lowest negative impact on the long-term growth. Using monetary policy based on the transactions with government bonds fits this role perfectly as it is government who actually decides the redistributive effects, the role of the CB is just to be there to stop things getting so ugly that the stability of the system itself (and thus even the long-term prospects) are in danger. The "funny" thing right now is that by "doing nothing" monetary policy now actually has larger redistributive effect than before. Since we are in a liquidity trap and major Cetral Banks refuse to use "unconventional" monetary policies, the real interest rate is artificially high. This means that they decided to pick a side of creditors over debtors just by not doing enough.

Kevin: Do you find it easier to think that "trust" and not "money" is the culprit wreaking havoc in your country and all over the world? "Trust" is too close to "morality" for me, with potential to steer the discussion back to "we all sinned by taking too much debt and therefore now must suffer the consequences" fatalism that is taking hold especially in Europe.

[Kevin Donoghue]

J.V. Dubois,
Fair point, I'd sooner refer to credit rationing. All I'm saying is that Nick is too eager to make the medium of exchange the focus of attention. To my mind the place to start doing macro is what Keynes calls the state of long-term expectations. That's what changes when the economy has what Krugman calls a Wile E. Coyote moment.

[James]

I have a question. Yes, one could define Y=C+I as present output = currently produced consumption goods + currently produced investment goods. But one could also define Y as total traded (presently produced and previously produced) goods = total traded consumption goods plus total traded investment (capital) goods. Then the identity S=I holds, as does the natural rate concept, Austrian business cycle, IS-LM model, etc.

[Jacques René Giguère]

AdamP:
"Yes, and most importantly the BMW/Skoda premium presumably is to a certain extent a difference in production costs."
It's backwards: high premium on Bimmers enables BMW to make them in Germany. Though "made in Germany" may signals high-quality and so justify the price.

Kevin Donoghue:
"Nick wants the problem to be about money. That's what makes him a monetarist. If he wanted it to be about sin he'd be a Calvinist."
Or the daughter. of an East German Lutheran pastor.

[Scott Sumner]

Nick, You said;

"A recession is not a drop in output and employment. A recession is a failure to exploit the gains from trade."

I'm not quite willing to buy into this. I'm happy saying it's not a fall in output, although in a sense it is, as an inability to exploit the gains from trade will reduce output. I'd prefer to use employment, market employment if you want to avoid the gardening problem. A recession is a reduction in hours worked for pay.

There are lots of things that reduce gains from trade (like occupational licensing laws) that don't necessarily cause recessions. They may just divert labor into less productive areas. In that case GDP is lower but employment is not lower, and there's no recession even though:

1. GDP is lower than otherwise.
2. There are fewer gains from trade.

So given a choice between output, market employment, and gains from trade, I'll take market employment.

[Determinant]

@Jacques and Ken Donoghue:

Original Sin and Man's Sinful Nature (inherent tendency to sin more often then not) is a pan-Christian concept. The Roman Catholic Church is as much into these doctrinal concepts as anyone else. The specifically Calvinist spin is Total Depravity, which is that man is unable to prevent himself from sinning without the saving Grace of God.

This tiptoe through the TULIPs has been brought to you by Determinant, WCI resident Calvinist.

[Nick Rowe]

Scott: suppose the government banned overtime. Or held a compulsory lottery where the winners had to take a forced holiday from paid work. Or banned women working outside the home. We wouldn't call that a recession. And it wouldn't look like a recession, except for market employment falling. We wouldn't see excess supply and a fall in trade across the whole economy.

(Which makes me think: RBC gets the wrong answer, primarily because it asks the wrong question, just like everyone else.)

James: if you redefine I and S that way, then existing theories of what determines I and S would no longer be applicable. So you end up in the same place.

Andy: "We see a decline in sales of those non-perishables whose value is largely in the services they provide for current consumption (cars, household appliances, etc.), because people are trying to shift their consumption to the future."

Let me highlight that "... because people are trying to shift their consumption to the future." bit.

If you start out from the new goods/I=S perspective, then you will see recessions in exactly that way. "They must be caused by people trying to shift their expenditure into the future". I disagree. Take my old house example. When people try to trade old houses, they aren't all trying to shift their expenditure into the future. They are each trying to shift from one house to a different house. But nobody can sell, because nobody will buy. A multilateral barter/house swap could solve the problem, in principle.

[reason]

"When people try to trade old houses, they aren't all trying to shift their expenditure into the future. They are each trying to shift from one house to a different house. But nobody can sell, because nobody will buy. A multilateral barter/house swap could solve the problem, in principle."

Nick, this is going in circles. Why then is everybody trying to increase their money balances? (Because they want to consume more in the future, perhaps?)

[Martin]

Reason,

"What you are saying effectively is we should increase exports."

If that's how you obtain that different currency, yes. I would however say more generally that you need to trade. The original point was however is that it is transaction costs that prevent people from switching to different currencies rather than a literal excess demand for money.

I sincerely believe that any supply of money is optimal as long as people can bargain without transaction costs. The trouble is that they can't and that's why you need a nominal target to mimick what people would agree to otherwise in a world of 'zero transaction costs'.

In other words you need a monetary policy to make neoclassical micro approximately true. I made a similar comment referring to Alfred Marshall's assumption of a stable monetary unit in his principals on Lars' blog. http://marketmonetarist.com/2011/12/23/how-i-would-like-teach-econ-101/#comment-1691

I believe both Nick and Scott made a similar point in the past about a monetary policy to make Say's law (I believe they meant Say's identity, then again I'll defer to their knowledge on this, after all I believe in specialization) to be approximately true.

[Nick Rowe]

Martin: we did mean Say's Law, not Say's Identity. An identity is true by definition. Say's law is a statement about the world that may or may not be true. It won't be true if monetary policy is bad. It will be true (or approximately true) if monetary policy is good. We want a monetary system where supply really does create its own demand.

"I sincerely believe that any supply of money is optimal as long as people can bargain without transaction costs"

"In other words you need a monetary policy to make neoclassical micro approximately true."

That's asking a little bit too much, in both cases. But yes, that's the direction in which we want to go.

[Martin]

Nick,

What I meant is that you want all excess demands to sum to zero by definition through the use of monetary policy. That is Say's Identity (I am of course assuming the absence any other transaction costs, apart from those associated with an excess demand for money).

Say's law on the other hand is always true regardless of monetary policy as it does not say that all excess demands will sum to zero by definition. Say's Law, (as is Stigler's version of the Coase Theorem), is that all excess demands will sum to zero in the absence of transaction costs.

Say's law simply means that you need to reprice all goods, including money, to make all excess demands sum to zero.

That's how I understand that distinction anyway.

[Nick Rowe]

Martin: OK. There are about 101 different definitions of Say's Law/Identity out there!

[Martin]

Haha yes I know, but this one is the correct one :P.

[Min]

Brilliant insight! I would never have thought of 'money', but now that John Stuart Mill said it in 1829, it's obviously right!

;)

And Nick, you are brilliant. :)

[Richard H. Serlin]

I think another aspect of it is that Stiglitz thinks it's better to deal with the structural issues with massive increases in education (how about free bachelor's degree and preschool), infrastructure, alternative energy, basic science and medicine, etc., then to constantly deal with the unemployment, decade after decade, with low real interest rates producing huge consumer spending and McMansion building.

[Max]

Richard, Stiglitz's "solutions" are the standard pious centrist crap, the economic equivalent of eating more vegetables and exercising more. It's what you advocate when you don't have any real ideas.

[Richard H. Serlin]

Putting huge money into extremely high return public investments is a very real solution. I wish we could be so real. Although, yes, it won't solve every problem. There are other things we should do too.

[JP Koning]

Nick: "Now, what might that something be? What could cause trade to fall and make people worse off? An increase in transportation costs? Nope. An increase in marginal tax rates? Nope. Those make theoretical sense, but don't work empirically to explain recessions. The obvious candidate is a shortage of the medium of exchange. ."

This was an enjoyable post and sparked some good comments. Economics is much more intuitive to me if it is about trade, not output. Ad hoc divisions between old and new goods seem odd to me too.

My question is, what is your empirical evidence that recessions are better explained by monetary phenomenon and not non-monetary phenomenon like transportation costs?

[Daniel Jansen]

however imho money can be broken down into an even more fundamental variable that guides trades. that would be information, information itself is not the only thing that moves trades and in fact money is probably the best representation of the variable of information (in a concrete-ish format). when one can then model money as an exchange of information i think one can start building models with a very real and immanent representation of (yeah money). this makes no sense probably but it does to me so f*k it.

[Martin]

JP Koning,

If I may answer this.

If transportation costs were the reason, what would you expect to see? I for example would expect to see high prices for inputs used in transportation as a result of the high cost for transportation itself. However, gas prices are low.

Or as you can see in this graph:

http://research.stlouisfed.org/fredgraph.png?g=41z

As you can see economic activity and demand for energy move together (this is also used to check Chinese growth: one looks at the electricity consumption), it follows that when economic activity collapses the demand for energy will collapse and as a result its price will drop.

[Gizzard]

I like this post Nick. Ive been thinking for days how to respond and there are so many points to address I hope youll excuse the format of my reply.

" Thinking about the trade cycle is a better way of thinking about short run macroeconomics than how we currently think about it. Nowadays we don't talk about fluctuations in trade; instead we talk about fluctuations in output. It's not the same thing. We are wrong; the old guys were right. Somewhere, maybe around the 1920's or 1930's (I wish I were better at history of thought) macroeconomics took a wrong turn."

We dont call it trade, but fluctuations in sales is what we experience, sales are just trades involving money. GDP is sales, employment is dependent on sales

Is I=S the right way to start thinking about short run macroeconomics? Only if you think that short run macroeconomics is about fluctuations in the output of newly-produced goods and services. If you think that short run macroeconomics is about the trade cycle, then I=S is a bad place to start.
Thinking about fluctuations in output is both too narrow and too broad a way of thinking about fluctuations. And that way of thinking about the question to be answered also distorts the sort of theoretical framework we use to answer it."

So you obviously think that in the LONG RUN the output of newly produced goods and services is important. But isnt the long run just a succession of short runs? I think economists go wrong when they talk about short and long runs. If you think its important for an economy to develop NEW goods and services in the long run the place to start that is in the short run. I can imagine an economy where nothing breaks down and so nothing new needs to be produced. There would be no " I " in this economy and this economy would not be a capitalist economy. So are we examining non capitalist economies now?

"Fluctuations in the output of newly-produced goods and services is a bad proxy for the conjuncture. It is both too narrow and too broad. The trade cycle is a better proxy for the conjuncture. It is too narrow because it ignores trade in goods that are not newly-produced. Things like old houses and old cars and old furniture and old land.
It is too broad because it includes production of goods and services for one's own use."

I wasnt aware that my buddy who sells used cars doesnt count towards GDP. He certainly counts towards employment though. So if he were to hire more employees it could occur without a concomitant increase in GDP because his "sales" dont count? I wasnt aware of that. I agree that something needs to account for this activity somehow.

"The unemployed worker is not unemployed. He is employed digging his own garden, fixing his own car, doing his own labour market search. He is not unhappy because he is unemployed. He is unhappy because he is self-employed and doesn't want to be. He wants to sell his labour to someone else, but can't, so has to consume it himself. A recession is not a fall in employment; it is a fall in the amount of employment that is traded. A recession is not a fall in output; it is a fall in the amount of output that is traded."

This is unnecessary and smacks of the "workers just want a vacation" argument that gets trotted out by some Tea Partiers. A full wheelbarrow load of crap. A recession is when sales go down and the falling sales leads employers to lay off workers cuz they dont need as many. LAY OFF workers, not a general strike or early retirement. There is no need to try and "redefine "unemployment. Everyone knows what it is if they have been unemployed.

"Microeconomists start out thinking about trade. And not just trade in newly-produced goods and services. Macroeconomists ought to start out thinking about trade. And not just trade in newly-produced goods and services. Trade is supposed to make people better off. If something makes trade harder, and reduces the amount of trade, that would make people worse off. Maybe that's why people seem to be worse off in a recession. It's because there's less trade. And not just less trade in newly-produced goods and services."

I think its more accurate to say that macro economists look at the overall end result of all that micro trading. What happened in sum is the question macro economists are asking. And in capitalist economies if no one created anything new and people just reshuffled around used stuff we call it "nothing happening". Is this bad? No But its not a capitalist economy if nothing is happening to create new stuff.

"But somewhere macroeconomics took a wrong turn. We started thinking about the output of newly-produced goods and services, instead of the trade cycle. So macroeconomists started thinking about the demand and supply of newly-produced goods and services, not about trade. So they wrote down "I=S" as their equilibrium condition, and started thinking about what determined desired investment and saving. And they focused on the real rate of interest, defined as the relative price of current newly-produced goods to future newly-produced goods."

That happened when we wanted to know when a capitalist economy was doing its job so to speak.

""What causes fluctuations in the amount of monetary trade?" is the right question to ask. "Money" is the right answer, because it is an answer to the right question."

Yes and unless there is INvestment to inject new money into the economy, there will be no new money. The investment must produce something new though or else its just chasing already produced goods and driving up prices

So I think I=S is the right place to start with a capitalist economy that wants new production

[Mike Sax]

"Now, what might that something be? What could cause trade to fall and make people worse off? An increase in transportation costs? Nope. An increase in marginal tax rates? Nope. Those make theoretical sense, but don't work empirically to explain recessions"

At least you're rulling out the ideology that the Great Depression happened when Herbert Hoover raised the top marginal tax rate like Jude Wanniski, et al.

[Mike Sax]

This talk of a shortage in money makes me think of the MMTers who say our fiat money is government monoploy "token" money. I guess you wouldn't see it that way? Money as a government monopoly-if the currency is a fiat currency?

[Nick Rowe]

Thanks Min :)

Richard: "Putting huge money into extremely high return public investments is a very real solution."

You don't need the "extremely". Putting money into high return public investments (assuming they are indeed high return) is a good thing to do anyway, regardless of the state of the trade cycle. We shouldn't have to be adjusting public investments up and down according to the state of the trade cycle. We shouldn't be stopping high return public investments just because the economy is booming. We shouldn't be starting low/negative return public investments just because the economy is slumping. Fiscal policy has got its own job to do; monetary policy's job is to deal with the trade cycle.

(There's a slight exception to this, though it's not really an exception: if real interest rates are low, for whatever reason, then the rate of return on more public investments will exceed the market rate of interest, and those investments should go ahead.)

[Nick Rowe]

Daniel: actually, that does sort of make sense. You can view money as a token that captures information on trades. If an individual holds $100, that means he has sold goods worth $100 more than the goods he has bought. It's a way to keep track of how much each person owes or is owed by the system, without needing some central accounting system.

JP: Thanks!

"My question is, what is your empirical evidence that recessions are better explained by monetary phenomenon and not non-monetary phenomenon like transportation costs?"

That's the right question. I think about it at odd intervals. I don't have any absolutely conclusive empirical evidence. Just that the monetary hypothesis seems to fit the facts, and others, like transportation costs, don't seem to.

1. When the volume of trade declines, it is usually harder than normal to sell goods, and easier than normal to buy goods. (It's easier than normal to sell money, and harder than normal to buy money.) That strongly suggests a monetary cause. It's hard to think of other causes that would have that implication. That is the one big empirical fact that is very salient to me.

2. I *believe* that barter is countercyclical (and that competing monies like LETS etc. are also countercyclical). If I knew for sure that was an empirical fact, I think it would be very salient.

3. None of the other candidates look good. Or, they might work for one recession, but don't work for others, so all those "special explanations" are very ad hoc.

4. Using historical methods, a la Friedman and Schwartz, we can sometimes identify monetary shocks, and they seem to have the predicted effects.

Mike: "At least you're rulling out the ideology that the Great Depression happened when Herbert Hoover raised the top marginal tax rate like Jude Wanniski, et al."

Yep. For two reasons:

1. If you try hard enough, you can always find some such "explanation" for any given recession. There's always something that changed, that might cause a recession. But those "explanations" are all ad hoc. It's different every time. And they ignore the cases where marginal tax rates increase, and there isn't a recession. Stiglitz's "explanation" also looks ad hoc to me.

2. If you increase marginal tax rates, and that reduces the supply of labour, and the supply of goods, you should see excess demand for goods, and rising prices. It doesn't fit the facts.

"This talk of a shortage in money makes me think of the MMTers who say our fiat money is government monoploy "token" money. I guess you wouldn't see it that way? Money as a government monopoly-if the currency is a fiat currency?"

I *do* see it the same way, roughly speaking. Money doesn't necessarily have to be a government monopoly, but in most (all?) countries today, most money is either government money, or (like bank money) is convertible on demand into government money, and so is linked to government money.

[himaginary]

James@01:56PM:"But one could also define Y as total traded (presently produced and previously produced) goods = total traded consumption goods plus total traded investment (capital) goods. Then the identity S=I holds, as does the natural rate concept, Austrian business cycle, IS-LM model, etc."

Nick@02:24AM:"if you redefine I and S that way, then existing theories of what determines I and S would no longer be applicable. So you end up in the same place."

I don't think redefining I and S that way affects that much. When we learn about national accounts in Japan, "Principle of equivalent of three aspects" is always on the front. (BTW, by googling, I found that that principle seems to be popular only in Japan.) That principle can be expressed as:
GDP(Gross Domestic Production) = GDI(Gross Domestic Income) = GDE(Gross Domestic Expenditure)

I think I=S theory mainly concerns GDI=GDE aspect of that principle, and doesn't make much difference whether GDP is actually GDT(Gross Domestic Trade) or GDNP(Gross Domestic New Production), as long as it captures value-addition reasonably well, one way or another. (Perferct capture is impossible in any case, as other commenters have noted.)

[Nick Rowe]

himaginary: but GDP, GDI, and GDE are all measuring the same thing: production of *new* goods and services.

(It's not just in Japan, by the way. There are three ways of measuring the same thing: measure the new goods produced; measure the income from new goods produced; measure the expenditure on new goods produced. And if you count everything correctly, and make the appropriate adjustments, you should get the same number in each of the three ways.)

[Jacques René Giguère]

himaginary:
to add to Nick. Every textbook has it. Usually in chapter 5 or 6...
Retreating into commenting as the nephews and nieces are agreed that uncles are soo boring discussing situation in Europe...

[JP Koning]

Martin,

Thanks. I was wondering what the specific empirical evidence would be proving the monetary nature of recessions rather than what disproves the non-monetary case. Is Nick referring to the empirical claim that, in a recession, it gets harder than normal to sell stuff, and easier than normal to buy stuff?

[himaginary]

Sorry for not being quite clear about what I meant by "popular in Japan" in my previous comment. I was referring to the particular keyword of "Principle of equivalent of three aspects" (or "sanmen-touka-no-gensoku" in Japanese). Of course the notion expressed by that keyword itself is basic economics.

[himaginary]

"Of course the notion expressed by that keyword itself is basic economics."

That said, whether catchy keyword is permeated among people or not may determine the influence of the notion expressed by that keyword, at least to some extent...

[Jacques René Giguère]

himaginary:
After a quick search, it seems that Japanese are deriving some policy conclusions from that boring accounting identity that we don't.

[Greg Ransom]

You have Hayek wrong. Hw doesn't start out with I = S defined as newly produced capital goods.

Hayekmis all about the changing structure of the allocation of both old and new production goods.

[Mariana Abrantes]

Quite right,I=S leaves a lot to be desired. It implies a closed economy, so it might be time for a refresher course in international economics and finance, for understanding the current Eurozone financial crisis, for example.

In open economies, both large and small, any domestic stimulus, in the form of fiscal policy or easier credit, can drain away into higher imports and external debt, as happened with the "cash for clunkers" programs in hapless and helpless countries such as Ireland and Portugal.

For a view on the limits of trade divergence and the adjustment challenges facing the fragile Eurozone peripheral economies, "economic midgets with one policy arm tied behind their back", see the econoblog PPP Lusofonia, http://ppplusofonia.blogspot.com/2011/12/eurozone-crisis-tests-limits-of.html 

[JP Koning]

Nick, I'm a bit late on this one, came via Daniel Kuehn's blog.

(http://factsandotherstubbornthings.blogspot.com/2012/01/cant-let-this-one-slip-by.html#comment-form)

You said: "The trade cycle is a better proxy for the conjuncture."

Does this mean you are a bit closer to Kling's Patterns of Sustainable Specialization and Trade (PSST)?

[Nick Rowe]

JP: I'm not sure. I don't think that follows.