# Did confusion over S&D cause the crash of 2008?

If I knew Mankiw was going to link to my last post, I wouldn’t have made it so meandering and confusing.  So here is the mop-up, what I really wanted to say.  But first a few clarifications, as I am getting a lot of comments that are challenging whether my question on movies was fair.   Here’s what I was trying to get across:

Question:   Suppose that people buy more of product X when the price is high, and less of product X when the price is low.  Does that violate the laws of supply and demand?

The answer is no.  It’s not maybe, or it depends, it is no, non, nein, nyet, bu shi, etc.  And it doesn’t depend on what the product is, what the two prices are, what the two quantities are, the answer is always no.

Now let’s consider why this is so confusing.  Take two more examples:

1.  In 2007 the price of copper soared to an all time high of \$4.30 a pound.  Consumers increased consumption and bought 26% more copper in 2007 than in 2006, when the price was \$2.85.

2.  In 2007 the price of copper soared to an all time high of \$4.30 a pound.  Producers increased production and sold 26% more copper in 2007 than in 2006, when the price was \$2.85.

Seeing them together, I think almost everyone can see what I am doing here.  But suppose you showed these two statements to students in isolation.  You asked 50 students if statement 1 was consistent with S&D, and you asked 50 students if statement 2 was consistent with S&D.  I’ll bet that many more students would say that statement 1 is not consistent with S&D, despite the fact that they are exactly the same statement, just worded slightly differently.

One problem is that many people assume a term like ‘consumption’ is synonymous with ‘quantity demanded,’ whereas it is equally equivalent to ‘quantity supplied.’

The other mistake is to make inferences from a price change.   Often the inference is that supply shifted, not demand.  When people jump to this conclusion, they make the erroneous inference that the laws of S&D predict that when the price is high, consumption will be low.  Of course that is not a prediction of S&D.

OK, how did this confusion cause the crash of 2008?  The exact same mistake we make fun of students for making is made by journalists and professional economists every time they assert that money was “easy” because nominal interest rates were low.  Actually this mistake is even worse, because it is doubly wrong.

1.  It ignores the identification problem.  Thus it may be the case that low interest rates are caused by easy money (more supply) or they may be caused by less demand, perhaps due to a weak economy.

2.  In the case of interest rates, however, it is even worse, because even if money was easy, it wouldn’t necessarily cause low interest rates.  Money is easy when there is hyperinflation, but interest rates are usually high.  That’s because monetary policy (the supply of money) also affects the broader economy, and hence the demand for money.

Thus when economists like Anna Schwartz and Joan Robinson equate easy money and low interest rates, they are actually making a much worse mistake that our students make when they make inferences from a change in price.  Just so you don’t think I am picking on these two fine economists;  most of us, including me, occasionally slip into this bad habit.  In some cases it is sort of justified, if the two people speaking both understand that it is obvious that either supply or demand is shifting.

Most people did make this mistake last year.  Monetary policy was roughly as tight last year as in 1981, when Volcker raised rates 20% in order to break the back of inflation.  Both monetary policies were tight enough to drive NGDP growth rates down very sharply.  If the business community and the public and journalists and economists had been aware of what was going on, there would have been riots in the streets.  Some of my commenters point to a hard money bias at the Fed.  Who knows if that’s true, but in a way it doesn’t matter because the Fed bends to public pressure, so they never would have been allowed to adopt such a deflationary policy in the midst of a debt crisis if it had been recognized as such.  And the main reason economists didn’t think money was tight was because rates were cut to low levels, just like in 1930.  This is drawing inferences from a price.  It is the same error that our students make in economics 101.

And as one of my commenters pointed out, it’s not just S&D that economists get confused about, it’s also opportunity cost, as this link shows.   So while I appreciate those who said S&D is easy once you understand the difference between demand and quantity demand, I beg to differ.  It is easy to teach supply and demand, just as it is easy to teach opportunity cost.  It is not easy to apply either concept to the real world.  Even for economists.

Tags:

56 Responses to “Did confusion over S&D cause the crash of 2008?”

1. anon
29. July 2009 at 18:53

unless the demand curve hasn’t shifted, in which case the answer is always yes

2. Greg Ransom
29. July 2009 at 21:13

But take a look at, for example, the papers produced by William White at the Bank of International Settlements — there was a whole lot more going into his identification of a Hayekian artificial boom in the mid-2000s than simply an “assertion that money as ‘easy’ because nominal interest rates were low”. Ditto the recent work of John Taylor. Looks to me like you have a straw man going here, Scott, built for rhetorical purposes having to do with debate about the appropriate causal story to explain the boom-bubble & bust.

Here are some facts of economic life that I would suggest are eliminated by stipulation from your own toy blackboard “S&D” macro construct, but which MUST be part of the causal explanatory frame of ANY valid macroeconomic explanation:

1. The fact that capital goods consist in a structure of coordinated production processes across time, built upon the the fact that people choose longer production processes only if these longer processes promise greater output.

2. The fact that the choice of longer or shorter production processes in influenced by the changing cost of credit.

3. The fact that interest rates set below the natural rate can at once push both over-investment in long-period production process (e.g. housing) and over-consumption (e.g. consumer consumption financed by 2nd mortgages).

People who aren’t blind to these facts by the dictate of their blackboard construct — e.g. William White at BIS — put all of this stuff together with all sorts of empirical observations and background understanding when they called an unsustainable boom/bubble and coming inevitable bust in the mid-2000s.

Maybe Anna Schwartz fits your straw man — but economists like William White and John Taylor certainly don’t.

Scott writes:

“the exact same mistake we make fun of students for making is made by journalists and professional economists every time they assert that money was “easy” because nominal interest rates were low.”

3. Catherine
29. July 2009 at 23:21

Maybe a better answer would be “not necessarily.”

4. Current
30. July 2009 at 01:04

Scott: “Monetary policy was roughly as tight last year as in 1981, when Volcker raised rates 20% in order to break the back of inflation. Both monetary policies were tight enough to drive NGDP growth rates down very sharply.”

I understand and agree with the reasoning you give for the loose link between interest rates and monetary policy, especially in unusual circumstances.

However, I still don’t see why you take the fall in NGDP as a demonstration that monetary policy was tight. It seem to me like there are many reasons it could fall.

5. malavel
30. July 2009 at 03:29

I feel like they are both still a bit deceptive. Why don’t you just present two data points instead?

“2007-01-01: price of copper \$2.85; quantity 2 million tons.
2007-12-31: price of copper \$4.30; quantity 3 million tons.
How can we explain this?”

6. ssumner
30. July 2009 at 03:35

anon, But that would be a different question, not the one I asked. I asked whether the information given was inconsistent with S&D.

Greg, I wasn’t talking about the mid-2000s, I was talking about 2008. There is much stronger evidence for easy money in the mid-2000s, indeed I agree with Taylor that money was easy in 2005. But I see very little evidence that money was easy in late 2008. I looked at seven indicators, and they all showed money exceeding, tight.

Catherine, No, the better answer is no. It is completely unambiguous. That information is not inconsistent with S&D.

Current, I favor targeting NGDP, and thus I use it as an indicator of whether money was tight. The terms ‘easy’ and ‘tight’ are always relative to some benchmark, chosen by the observer.

7. Mark
30. July 2009 at 04:01

How could you not know Mankiw was going to link to your blog? You pumped his textbook in your post and his whole blog is a marketing device for his text lol!

8. Jeffrey
30. July 2009 at 04:14

it is hard to assert that monetary policy is easy when nominal gdp is negative and is expected to remain negative in the following quarter and the value of “growth options” is essentially zero. and the velocity of money is so unstable rendering it empirically impaired.

9. JKH
30. July 2009 at 04:34

Clearly a challenging question, given the range of interpretations from reasonable people.

It becomes much easier somehow when:

“Does that violate the laws of supply and demand?”

becomes:

“Does that alone violate the laws of supply and demand?”

Are these the same question? Maybe so.

10. azmyth
30. July 2009 at 04:38

Scott, I think the problem is you are giving a one handed question to a bunch of two handed economists. We just can’t handle “No and leave it at that”.

11. Alex
30. July 2009 at 04:43

On the issue on how intro to economics is taught and what students learn a new way I´m about to try is to use the book by Robert H. Frank, The Economic Naturalist. You can check the first pages on Google books.

12. Bill Woolsey
30. July 2009 at 04:45

Current:

What could cause nominal income to fall other than the demand for money being greater than the quantity of money?

The way to see this involves asking what happens to the money financing the flow of income and expenditure at a given nominal income? People spend less? OK, what do they do with the money?

The key to understanding Sumner’s argument is to see that a failure of the quantity of money to rise to meet an increase in the demand for money counts just as much as a decrease in the quantity of money given the demand.

The situation in question is the quantity of money rising less than the demand for money.

On the other hand, if we imagine a world with price ceilings, I suppose nominal income might fall because people refuse to sell. The effective demand for money rises (because there is nothing to spend the money on and prices cannot be bid up for some reason.

13. Bill Woolsey
30. July 2009 at 05:02

Scott:

Here is the market process that some of us have in mind.

The demand for money rises faster than the quantity of money. People are short of money. People sell bonds to get money. The nominal interest rate rises. “Tight money” shows up as higher nominal interest rates.

For this process not to occur, the demand for money must fall…

But I really think that most people understand that the demand for T-bills was rising as well as the demand for money. The higher demand for T-bills pushes down their interest rates. Persumably, any shortage of money dampened this. The T-bill rates fell less than they otherwise would.

To me this is obvious– the law of demand always says that people buy less at higher prices than they would if prices were lower. But that doesn’t mean if the price is higher today than yesterday, they buy less today than yesterday. Maybe something else is changing that causes them to buy more. It is just that the higher price then dampens the increase in the amount purchased. (I explain this to my principles students.)

So, T-bill yields were falling, but not as far as they would have fallen if the quantity of money had risen as fast as the demand to hold money.

If we add to this the notion that the reason the demand for money was rising was because T-bill rates were falling, then it is a short step to thinking that the “problem” is whatever was causing T-bill rates to fall.

And that was the flow of funds out of the shadow banking system into T-bills. The shadow banking system was Wall Street. So, if we just fix Wall Street, the funds will flow back to the shadow banking system and out of T-bills. T-bill rates will rise again. And the demand for money will fall, and we will have fixed the problem.

You know, I never use the words “tight” or “loose” money. I always talk about the quantity of money and the demand to hold money and imbalances between them.

14. ssumner
30. July 2009 at 05:08

Malavel, The point of the exercise is to show that reality is deceptive. The point is that people need to deal with that. Framing is important, it throws people off. Logically the two statements are the same, when you think about it, but the way they are worded makes them seem different. But the fault isn’t the wording, it is our brains. We think “consumption” is somehow demand, but it is just as much supply.

JKH, Those two questions are exactly the same.

azmyth. People “overthink” these questions.

Alex, Thanks for the tip.

Bill, Yes, and if someone preferred to use the loanable funds market, the low interest rates could be viewed as resulting from a reduction in the demand for loanable funds, and hence would be associated with falling investment.

15. Rain King
30. July 2009 at 05:47

The problem with the first example is one of semantics.

When you state, “In 2007 the price of copper soared to an all time high of \$4.30 a pound. CONSUMERS INCREASED consumption and bought 26% more copper…” the word order indicates that price increased first, followed by an increase in consumption. That situation, if I am not mistaken, DOES violate the law of S&D.

In reality, it is the rapid shift in demand that causes both price and consumption to increase.

Example wording:

1. In 2007 copper consumers increased consumption 26%. The price of copper soared to an all time high of \$4.30 a pound. Consumers bought 26% more copper in 2007 than in 2006, when the price was \$2.85.

16. RebelEconomist
30. July 2009 at 05:52

As I commented on the previous post, it would be better to define the S&D question more unequivocally, and I like JKH’s suggestion here. Asking the question in that way would provide a truer test of your students, but might provide you with less amusement.

17. RebelEconomist
30. July 2009 at 05:56

That said, I must admit that I have argued in the past that to the exam question “expand (a+b)^2”, the answer “(a + b)^2” deserves some marks for lateral thinking!

18. 123
30. July 2009 at 07:36

Great post!
Actually I started reading this blog because it was the only one with the sane approach to monetary S&D.

This post also illustrates a behavioural bias that might influence the correctness of EMH.

Scott, you might like a latest note by Bill Gross, manager of largest bond fund in America. He touches your favourite themes – the importance of 5% NGDP, EMH, “mother of all stock bubbles”. The link is here:

http://www.pimco.com/LeftNav/Featured+Market+Commentary/IO/2009/Investment+Outlook+August+2009+Gross+Investment+Potion.htm

19. Greg Ransom
30. July 2009 at 08:10

Scott, this is Hayekian macro 101. When over investment in long term production finally collides with over consumption (example, too much housing & retail construction & too much 2nd mortgage consumer spending) as we saw in 2007 and 2008 production goods will come to be in shortage driving prices up and driving demand for more money and credit up — the system is unsustainably out of whack.

Folks who read & understood the economics in Hayek like White & Schiff easily saw all this as a natural part of their S&D background understanding of the situation in 2007 and 2008.

There’s no production taking more or less time in you thinking about money and credit so your S&D thinking about 2008 is unavoidably deficient — you’ve got a construct lacking the central causal component of a capital goods using economy, the point made against Keynes by Hayek and against Friedman and Keynes and Lucas, etc. by Roger Garrison.

20. Greg Ransom
30. July 2009 at 08:30

Scott, this is Hayekian macro 101. When over investment in long term production
finally collides with over consumption (e.g. too much construction & too much 2nd mortgage consumer spending) as we saw in 2007-2008 production goods will come to be in shortage driving prices up and driving demand for more money and credit up “” the system is unsustainably out of whack.

If you’d have read Hayek like White & Schiff you would see this as natural and inevitable as part of your S&D background understanding.

There’s no production in your thinking here (processes taking more or less time sensitive to credit rates and money) so your S&D thinking about 2008 is defective.

21. Current
30. July 2009 at 08:52

Bill: “The key to understanding Sumner’s argument is to see that a failure of the quantity of money to rise to meet an increase in the demand for money counts just as much as a decrease in the quantity of money given the demand.”

I agree with you there

Bill: “What could cause nominal income to fall other than the demand for money being greater than the quantity of money?

The way to see this involves asking what happens to the money financing the flow of income and expenditure at a given nominal income? People spend less? OK, what do they do with the money?”

Scott is talking about NGDP here isn’t he. That is:

GDP = compensation of employees + gross operating surplus + gross mixed income + taxes

There are all sorts of reasons why this may fall. Tax revenues may fall because of a decrease in transactions. Profits may fall.

Looked at from the other side:

GDP = private consumption + gross investment + government spending + (exports âˆ’ imports)

Consumption may certainly fall due to more money holding. Also, exports may fall or imports rise.

22. Bill Woolsey
30. July 2009 at 09:54

Current:

Stick to the expenditure side.

When Sumner is talking about “nominal income” he is talking and nominal aggregate demand for final goods and services. I think sales of final goods and services is the better measure, but “nominal income targetting” is how it is usually described in the literature.

Yes, less consumption, more money demand. But firms could reduce gross investment and increase money demand too. Revenues are just accumulated in a checking account. Government could take tax revenues and put them into a checking account rather than spend them. Foreign exporters could take their dollar earnings and put them into a checking account. Foreign importers might just leave money in checking accounts rather than purchase U.S. products.

If, instead, any of these people purchase nonmonetary finanical assets with the money not spent on consumer goods, capital goods, or government goods, then what do those selling the financial assets do with the money? The money doesn’t just disappear into finanial markets. Perhaps asset prices rise, and that makes the demand for money rise. And, so, one way or another, the reason nominal income falls is the demand for money is greater than the quantity of money.

That is the way I see it.

Oh, and if the demand for money is greater than the quantity of money, there is going to be a drop in aggregate expenditure.

The reason incomes–wages, interest and profits, fit in, is that the incomes are generated from revevenus from the sale of the final products.

But the monetary disequilibrium impacts the expenditures end directly.

23. Patrick R. Sullivan
30. July 2009 at 10:49

With regard to monetary policy, S&D errors usually result from making the assumption that ‘interest rates are the price of money’. Which is not true. Interest rates are the price of renting money, not buying money.

The ‘price’ of money is whatever is given up in exchange for ownership of it (usually one’s labor). In inflations such as we suffered in the 70s, most people have no problem seeing that, say, if a loaf of bread that used to sell for \$1 now goes for \$2, the price of bread has doubled. What few understand is that the price of \$1 has fallen by half.

The mistake is common, I’ve seen it in textbooks, and Walter Heller made it (and was quickly corrected) in his famous debate at NYU with Milton Friedman. Not coincidentally, David Friedman has a nice explication of The Price of Money in chapter 22 of his Price Theory:

Contrast that with the difficulty Scott’s students have with the movie ticket price, which is what Tom Sowell calls the Physical Fallacy. A movie shown at 2:00 PM on Wednesday is a different product from the same movie shown at 8:00PM Friday night.

As is a hotel room in Acapulco a different product in August than in January from the perspective of a resident of Int’l Falls, Minn.

24. Current
30. July 2009 at 11:10

Patrick R. Sullivan: “Interest rates are the price of renting money, not buying money.”

Just a little nitpick in an otherwise excellent post.

Interest rates are the price of renting capital not money. That capital is usually issued to the borrower in the form of money, but it need not be.

25. Patrick R. Sullivan
30. July 2009 at 16:10

Well, you can rent money for investment, and for consumption, so I’m not sure of your point. The reason I put quotes around ‘interest rates are the price of money’ is that I’ve seen exactly that statement in textbooks. Specifically, one in a Money and Banking text in use at UCLA in the early 1980s.

26. Steve Horwitz
30. July 2009 at 16:15

FWIW, I’ve been teaching intro for 20 years and have never had the sorts of problems Scott is concerned with in this post and the other. My students don’t seem to have a problem in offering cogent explanations and I plan to test that with a better worded version of the movie example next time I teach Intro.

I do think clarifying demand vs. quantity demanded is important, but it’s not the key. I think the key is carefully explaining the market processes that underlie S&D rather than teaching it as some sort of equilibrium identity (“S&D space”). If we teach the processes of bid and ask and the preference scales that underlie them and if we show how supply and demand curves are just graphical representations of how those preferences interact with prices, students begin to be able to think in process terms and not just “solve the equations” terms.

Thinking in process terms will help with many of these problems, at least that’s my experience.

27. Fletch
30. July 2009 at 23:33

Scott

Ah, yes “”people tend to “over think” these questions.”” I wasn’t sure what was going on from your first post, but after this one I have gained respect for you and your thoughts on economics. Most economists don’t impress me because they don’t seem to be able to speak in a way that normal people can understand. Too much “academic speak.” I got to your post from one who does, N. Greg Mankiw. Now I know of two as I search for more. This subject fascinates me! Thanks for taking the time for this blog to those of us who would not have the opportunity to hear from you otherwise.

Respectfully,
Fletch

28. Current
31. July 2009 at 00:37

Patrick R. Sullivan: “Well, you can rent money for investment, and for consumption, so I’m not sure of your point. The reason I put quotes around ‘interest rates are the price of money’ is that I’ve seen exactly that statement in textbooks. Specifically, one in a Money and Banking text in use at UCLA in the early 1980s.”

I could lend my car to Jeff. I could demand repayment for that debt in money or in a combination of money and automobiles or in carrots.

I could use the money interest rate for all of these transactions by putting the agreements in those terms.

29. Current
31. July 2009 at 05:19

Bill: “Stick to the expenditure side.

When Sumner is talking about “nominal income” he is talking and nominal aggregate demand for final goods and services. I think sales of final goods and services is the better measure, but “nominal income targetting” is how it is usually described in the literature.”

I see what you mean. So, turnover of money for NGDP components is used as a proxy for overall demand for money. I can see how that would work. I’m going to have to think about it some more though.

30. Victor
31. July 2009 at 06:06

Scott,
When you regress log(Q) on log(Price) you get a negative price elasticity of demand , does that means that most of the time the curve that shifts is the supply curve? I´ve got confused,WITH THIS METHOD OF CALCULATING THE PRICE ELASTICITY OF DEMAND, if the demand curve shifts more often or with a greater magnitude than the supply curve i believe that we would get a positive price elasticity of demand. Am I wrong?

31. Greg Ransom
31. July 2009 at 09:40

This is dead on, exposing the difference between the failed explanatory enterprise of economics conceived as nothing but the “givens” found in “blackboard” mathematics, and the successful explanatory enterprise of economics conceived of as a causal explanatory enterprise utilizing our causal background understanding, e.g. Hayek’s account of “the facts of the social sciences” as involving our direct understanding of entrepreneurial learning and choice in the context of changing relative prices. See also the work of Kirzner and Mises.

Steve Horwitz writes:

FWIW, I’ve been teaching intro for 20 years and have never had the sorts of problems Scott is concerned with in this post and the other. My students don’t seem to have a problem in offering cogent explanations and I plan to test that with a better worded version of the movie example next time I teach Intro.

I do think clarifying demand vs. quantity demanded is important, but it’s not the key. I think the key is carefully explaining the market processes that underlie S&D rather than teaching it as some sort of equilibrium identity (“S&D space”). If we teach the processes of bid and ask and the preference scales that underlie them and if we show how supply and demand curves are just graphical representations of how those preferences interact with prices, students begin to be able to think in process terms and not just “solve the equations” terms.

Thinking in process terms will help with many of these problems, at least that’s my experience.

32. ssumner
31. July 2009 at 11:45

Rain king, You said:

“When you state, “In 2007 the price of copper soared to an all time high of \$4.30 a pound. CONSUMERS INCREASED consumption and bought 26% more copper…” the word order indicates that price increased first, followed by an increase in consumption. That situation, if I am not mistaken, DOES violate the law of S&D.”

Actually it doesn’t as commodity prices move faster than consumption. Say the Chinese government announces a stimulus to build 20 million apartments, all needing copper pipes. The price of copper will instantly soar in the copper markets, even before the quantity consumed actually increases.

Rebel, If even most adults who studied economics in school don’t understand S&D (and believe me they don’t) why in the world would I want to make it easy for students to get the answer right, and give them a false sense of security that they understand something that they really don’t understand. I am actually far too easy on students. The movie example did not appear on a test, I asked the class if anyone knew the answer. In many classes not a single person knows.

Here’s the truth, whether people want to hear it or not. Most people who study the laws of supply and demand think that the implication is that when the price is high you’d expect people to buy less. I don’t know how many times I’ve heard people say “don’t you think higher interest rates will slow the economy?”

123, Thanks 123, I notice he likes my 5% more than Bill’s 3% NGDP growth. But to be honest he slightly confuses short and long run effects, so I don’t think Bill need worry.

Greg, I just don’t see any empirical evidence that the time element in production matters. I think NGDP instability is all we need to explain demand-driven business cycles. And if we stabilize NGDP, we are just left with small supply-side cycles, nothing to worry about.

Patrick. Good observations about the price of money. I hate looking at money from an interest rate perspective. One slight quibble. Is gasoline in the summer different from gasoline in the winter? The price is higher because demand is higher. Isn’t that like movies?

Current and Patrick, I see Current’s point, but also note that because cash pays no interest, the nominal interest rate is the opportunity cost of holding cash. So if you use the term “rental cost” loosely, I think Patrick’s definition is also acceptable.

Steve, Thanks for the tip. You might be right. But you might also overestimate how well your students will do in the real world when facing price changes for which the causes are obscure (say in the oil market, where there is so much debate.) Will they fall into the trap of equating high prices with less supply? (And simply assume that it should reduce Q demanded?)

Fletch, Thanks. To be honest my first post wasn’t too clear. I’m not sure the second one was either, but at least it seems to have slightly clarified things.

Current and Bill, I haven’t formed a clear idea of which NGDP measure is best, but I suspect Bill may be right.

Victor, This is important. You can only do that if you know the demand curve hasn’t shifted. Sometimes they do that test after a big excise tax increase, because that’s a marginal cost (hence supply) shift. So you can see consumer reaction. But if demand is shifting that technique won’t work.

Greg, I think you and Steve are right. I also think this shows that we often teach econ too abstractly (including me) Real world examples help clarify the distinctions we are trying to make. In macro I tend to think in terms of examples of business cycles I have studied, or different experiences with different monetary policies, not just mathematical models.

33. Greg Ransom
31. July 2009 at 12:28

Bull-dozed unfinished houses in LA county were direct physical evidence, as are uncompleted resorts in Las Vegas, and empty condo complexes in Florida. I live in a development built across the 2000s — the price of my own house doubled, then fell back to its original price. About half of the neighborhood has turned over in foreclosers, shorts sales, or people losing their jobs in the housing and mortgage and real estate businesses. Note well, Hayek included housing among longer term production goods.

All I have to do is walk out the door to see the empirical evidence .. evidence I’ve been pointing to on the blogs across the 2000s, evidence which allowed be to call the bust in 2007, more than a year before the professional macroeconomists officially called the recession.

Scott writes:

“Greg, I just don’t see any empirical evidence that the time element in production matters. I think NGDP instability is all we need to explain demand-driven business cycles. And if we stabilize NGDP, we are just left with small supply-side cycles, nothing to worry about.”

34. 123
31. July 2009 at 12:30

Scott, you said:
“I notice he [PIMCO’s Bill Gross] likes my 5% more than Bill’s 3% NGDP growth. But to be honest he slightly confuses short and long run effects, so I don’t think Bill need worry.”
I think Gross is just worried that short term effects will last much longer than many believe. My question is when will long run arrive in Japan, and why 19 years are not enough for the long run to arrive?

35. Patrick R. Sullivan
31. July 2009 at 16:07

‘One slight quibble. Is gasoline in the summer different from gasoline in the winter?’

Physically it’s the same (assuming away different blends for summer and winter), but the hotel room in Acapulco is exactly the same physically in winter and summer too. In both cases the surrounding circumstances make the products different.

36. ssumner
31. July 2009 at 18:02

Greg, By evidence I mean the sorts of things you are describing in an economy where NGDP growth is stable. I agree that a lot of bad things have happened, but I think it’s mostly because NGDP growth has been so unstable. If we had had good policy, you would have been partly right, there still would have been a modest sub-prime fiasco. But we’d have 5.5% unemployment, not 9.5%. We’d have missed the second leg down.

123, Didn’t he say permanent? In any case Japan has huge supply-side problems. By 2004 they may have been in a sort of “equilibrium.” I’m not recommending their policy, just saying that one shouldn’t assume the slow RGDP growth is just monetary policy. They have a falling population and an unreformed domestic economy. Sony and Toyota can only take you so far.

The odd thing is that I argued recently with Lee Ohanian that the pain would be very long lasting (years) because of sticky wages and prices. Now I’m on the opposite side of the debate. I don’t have that much of a disagreement with you, and indeed even Bill says that a financial crisis is not the time to suddenly move to a 3% NGDP growth path.

Patrick, I’ve always taught all of those cases as a seasonal shift in demand over time. Of course our dispute here doesn’t affect the answer to the problem at all, I just find it interesting because I’d never really though about how to distinguish a shift in demand and a slightly different product. You may be right.

37. Greg Ransom
31. July 2009 at 19:22

The intertemporal equilibruim costruction was first conceived when Hayek dated and priced fruit as differen goods at different times of year, on analogy with Mises’ pricing
goods differently at different geographical places duentomtransportation costs. The idea is that you could have strawberries or whatever at different times of year they’d simply have very different costs and prices.

Ironically, this seminal development in equilibrium theory was developed as part of Hayek’s effort tomunderstand macro disequilibrium across time made possible by credit, money and banking finance.

38. 123
1. August 2009 at 01:12

“Didn’t he [Bill Gross] say permanent? In any case Japan has huge supply-side problems. By 2004 they may have been in a sort of “equilibrium.” I’m not recommending their policy, just saying that one shouldn’t assume the slow RGDP growth is just monetary policy. They have a falling population and an unreformed domestic economy. Sony and Toyota can only take you so far.

The odd thing is that I argued recently with Lee Ohanian that the pain would be very long lasting (years) because of sticky wages and prices. Now I’m on the opposite side of the debate. I don’t have that much of a disagreement with you, and indeed even Bill says that a financial crisis is not the time to suddenly move to a 3% NGDP growth path.”

I think there is no real disagreement here. Bill Gross said “A 3% nominal GDP [..] means lower profit growth, permanently higher unemployment,..”. Permanently higher unemployment means higher unemployment over the next 5 years – this is an investment note, not a scientific paper.

39. Current
1. August 2009 at 03:31

Scott: “I see Current’s point, but also note that because cash pays no interest, the nominal interest rate is the opportunity cost of holding cash. So if you use the term “rental cost” loosely, I think Patrick’s definition is also acceptable.”

What I was trying to emphasis here is that the use of money, in form of cash or current bank accounts is not so important for debts. For some sorts of borrowing what the borrower wants is cash. But for others the medium in which the capital is transferred to them is not that relevant. Money is merely convienent. And usually once the money is transferred it is immediately used. When I borrowed to buy a flat I spent the money on the flat as soon as I had it.

So, for anything apart from short-term debt money is not that relevant. What is important is the rate of interest.

40. Current
1. August 2009 at 03:36

Scott, Greg,

This brings up the question “were there any countries where nominal GDP grew quickly”?

Does anyone know where on the internet tables of nominal GDP exist, if anywhere?

41. ssumner
1. August 2009 at 05:43

Greg, Thanks for the Hayek citation. I need to think about these issues a bit more, it’s not obvious to me whether this has important macro implications. Unfortunately as I am doing too much already, I can’t study much new right now.

123, I meant Bill Woolsey, which probably confused you, as the other guy is also Bill. But Bill Gross does say “permanent” which is what I remember. So although I agree with his policy views, I think he is a bit too pessimistic about 3%. Forever is quite a long time.

Current, I think we agree. I argue with Keynesians that the very short term interest rate isn’t that important, because the foregone interest on cash is minor compared to broader interest costs in our economy.

Current, Over what time period? Someone told me China’s nominal GDP was up 4% and their real GDP was up 8% over the last 12 months. Was it 123 who mentioned that?

If you are looking for “less bad” results, some Aussie commenters have discussed their relatively good performance in both this recession and the last one. I think they have kept NGDP growth from falling sharply (but obviously there was some fall in the growth rate.)

42. 123
1. August 2009 at 11:01

When Bill Gross says “permanent”, he means “for the next 5 years”. Do you think NGDP will grow more than 3% on average for the next 5 years?

43. Current
1. August 2009 at 11:34

Scott: “Over what time period? Someone told me China’s nominal GDP was up 4% and their real GDP was up 8% over the last 12 months.”

What time period do you think is relevant? Would figures by the quarter show what you mean?

44. Jon
1. August 2009 at 13:01

Current, Over what time period? Someone told me China’s nominal GDP was up 4% and their real GDP was up 8% over the last 12 months. Was it 123 who mentioned that?

Any commentary on the WSJ article on china’s GDP numbers?

45. ssumner
2. August 2009 at 11:34

123, OK, I missed that, but he is sloppy with language.

Because we are in such a deep slump, I think we might get a bit more than 3% over 5 years. But I agree with him that NGDP will be lower than it should be for the next 5 years or so–so I guess you are right, we aren’t far apart.

Current, I forgot the original context. I suppose that I usually argue it is expected NGDP growth that matters, but if you take a fairly short time horizon you usually get a better sense of whether money is too loose, or tight enough for a liquidity trap. Was that the issue—liquidity traps?

Jon, Most of the press just reports Chinese RGDP, and I know that was just reported as being up 7.9% year over year in Q2. NGDP I don’t know about.

46. 123
2. August 2009 at 12:51

Sloppy with language indeed. The only reason I’m so sure that “permanent” means “5 years” is because I am familiar with other writings by Bill Gross and his team.

There are some differences between your and Gross’s views. He thinks that unless the capacity of financial sector returns to previous levels it is hard to see return to 5 percent NGDP growth. He and his colleagues think that “excessive regulation, higher taxation, and government intervention will be among the factors that will constrain the growth of potential (non-inflationary) output” – i.e. they see lower RGDP growth in the future.

47. ssumner
3. August 2009 at 11:40

123, Your last line is the key, I assume he means we can’t get 5% if the Fed is targeting 2% inflation. Because we can always get 5% with enough inflation. I think there is a good chance we will have a productivity slowdown, and perhaps a small population slowdown as well (crackdown on immigration.) So the trend RGDP may fall to 2% or 2.5%. So there are many factors that could enter in to his thinking.

48. 123
3. August 2009 at 11:50

Here is a more sophisticated version of Gross’s argument written by his partner El-Erian:

http://www.pimco.com/LeftNav/PIMCO+Spotlight/2009/Secular+Outlook+May+2009+El-Erian.htm

49. RebelEconomist
3. August 2009 at 12:03

Scott,

We will have to agree to disagree on whether “potentially yes” is a correct answer to your cinema attendance question, but your remark in the comments to the previous post about the incompatibility of a ceteris paribus assumption with a change in price made me think. I suppose that in the idea of a demand or supply schedule there is an implicit assumption that price could be set by an infinitely elastic supply or demand schedule respectively in order to observe the demand or supply response. The ceteris paribus clause is meant to refer to everything other than this shift. People answer your question in the way they do because they are assuming from the way it is phrased that they are dealing with a hypothetical situation in which the price is fixed and the (in this case, since you pose the question in terms of how many people go) demand responds. The point for this post is that good economists such as those at the Fed are unlikely to make such an assumption when they are dealing with the real world, in which price changes could arise from changes in either demand or supply or both.

I hope you took something from the discussion too.

50. ssumner
4. August 2009 at 06:44

123, I think he is probably right that real growth will slow a bit, but I don’t see that as causing inflation, instead I think nominal growth will slow just as much.

He also has no explanation for the sudden collapse after September 2008, other than Lehman. Yet banking difficulties cannot cause a collapse in output if NGDP is growing rapidly (as we saw in 1933). Unless we learn the lesson of what causes recessions, we’ll keep having them.

Rebeleconomist,

“People answer your question in the way they do because they are assuming from the way it is phrased that they are dealing with a hypothetical situation in which the price is fixed and the (in this case, since you pose the question in terms of how many people go) demand responds.”

This might be appropriate if I asked about demand only, and asked how much more or less people wanted to buy, ceteris paribus. But I asked about the theory of supply and demand, in which case obviously the supply and demand curve are equally likely to shift. I wonder if I had quantity go the other way what people would have said. If quantity goes down when price goes up then demand looks fine, but now the supply part of S&D looks off. Someone could say, “wait, why would firms supply more at a lower price?”

Yes, the discussion was very interesting, and I may return to it at some point.

51. Van
4. August 2009 at 07:05

to scott or anyone else: i dont want to belabor, but i am not understanding. in the copper example at the very start of this discussion, what is the point? that is, how SHOULD we understand the two examples? i am not understanding how they are the same…unless we are speaking of different points of time, where we view demand curve shifting right first, meaning increased demand at all prices. tehn supply curve shifting right to meet the new shift in demand. what am i missing? perhaps i learned nothing in all those years of economics classes

52. ssumner
5. August 2009 at 10:44

Van, The only difference in the two statements is the terms ‘consumption’ and ‘production’. But both terms mean ‘quantity’. So if you replace each term with quantity, the sentences are identical. Hence the meaning are the same.

I am confused by your statement about the two curves shifting. A shift in demand does not cause a shift in supply. In both of the two sentences there was a shift in demand, and no change in supply. And in both cases quantity supplied rose.

53. Van
5. August 2009 at 11:38

scott, you are certainly right – i did not mean to imply that a shift in demand begets a shift in supply, but by how the example was worded, that is how i interpreted it. hence the point of my question. but your explanation above makes it clearer.

conversely, how would u write with a “supply shift” example for students? not important.i am forever confused by real world differences between change in s/d versus change in quantity s/d’d.thanks

54. ssumner
6. August 2009 at 06:48

Van, When supply shifts you just present an example where firms supply more despite lower prices. The computer industry is a good example. So have Q go up, and P go down.

55. Dan Gilles
13. August 2009 at 03:54

The question is misleading because of the word “when”. “When” can mean either “at what time” or “under what circumstances”. Replace “when” with “at times when”, and the answer is no. Replace it with “if”, and the answer is yes.

“If” makes the statement causal – a high price causes a high consumption – which is false. Econ students will reasonably assume that the “when” is used in this way because the such questions are normally used to test whether the student understands the causal effects of price shifts on consumer behavior, in isolation, ceteris paribus.

56. ssumner
13. August 2009 at 07:24

Dan, There was no ambiguity when I asked the question in class (it was not a written question.) I described a student going out and doing surveys of movie theater attendance to see how many people attended at each price. Trust me, students do not understand this issue, not do journalists, as anyone who reads the papers can immediately see.