Monday, June 27, 2022

Muth and Lucas: Call your offices!

 On his Marginal Revolution blog, Tyler Cowen describes the recent "purge" in the trucking industry. The pandemic shift in demand towards goods, as opposed to services, produced a big increase in the demand for trucking, which in turn produced a huge response, including a  big increase in the number of small trucking companies. Now that is all being reversed, precipitously, with trucking companies falling like flies--the purge.

This all sounds very, very "cobwebby" to me. It was Muth's article criticizing the cobweb model that inspired Lucas' "rational expectations revolution." I am old enough to have learned the cobweb model in  a principles course, but I don't know of any modern texts where it appears. We moderns have read Muth and know that rational expectations kill cobwebs.

Let's  syllologize:

If ratex is true, then no cobwebs.

Cobwebs.

Ergo, ratex is false.


qed



27 comments:

Anonymous said...

With no explanation and specific references, this is sadly incomprehensible.

rosserjb@jmu.edu said...

Labor markets are one of the places where cobwebs still exist, and also in cattle cycles.

Anonymous, the classic source is a 1938 paper by Ezekial in the QJE that I think has cobwebs in its title, although the idea is much older, going under different names.

The basic idea is if there sufficient lags in production, with the classic examples involving agriculture, such that one makes a production decision based on a price at one point in time, but the production does not complete until some time later, the price may be quite different. So, high prices induce high production that then produces low prices that then produce low production that then leads to high prices, and so on.

The term "cobweb" comes from looking at the supply and demand curves, with the market essentially cycling around the equilibrium without going to it, always overshooting one way or the other. The ratex argument of Muth was that farmers are smart and know what the equilibrium really is, so they do not over or under produce. But at least one market they still seem to do so in is cattle, where the lags are several years, and the cycle is about 11 years long.

Anonymous said...

The explanation is excellent and appreciated and I will now look to references.

Anonymous said...

https://en.wikipedia.org/wiki/Cobweb_model

The cobweb model or cobweb theory is an economic model that explains why prices might be subject to periodic fluctuations in certain types of markets. It describes cyclical supply and demand in a market where the amount produced must be chosen before prices are observed. Producers' expectations about prices are assumed to be based on observations of previous prices. Nicholas Kaldor analyzed the model in 1934, coining the term "cobweb theorem."

[ This is excellent. From here then we get crop or crop-price insurance. Excellent and important. ]

Anonymous said...

The Cobweb Theorem - The Quarterly Journal of Economics, Volume 52, Issue 2, February 1938, Pages 255–280, https://doi.org/10.2307/1881734

kevin quinn said...

Sorry, anon, for my crypticity--and thanks to Barkley for rectification.

Wasn't Muth's example agricultural. Hogs?

Anonymous said...

Wasn't Muth's example agricultural. Hogs?

[ Advanced technology, including Internet of Things (IOT) applications and agricultural insurance were just what the Chinese used to rebuild and secure pork production and farmers after the experience of African Swine Fever. Pork farming was completely restructured and the restructuring was remarkably fast, and pork is a critical food for the Chinese. ]

Anonymous said...

Remember that early in 1933, Roosevelt produced a program to stabilize and guarantee agricultural prices and production. The Supreme Court would rule the program Unconstitutional, but it would be replaced.

Anonymous said...

This is all excellent!

Anonymous said...

https://fred.stlouisfed.org/graph/?g=KqTz

January 15, 2018

Consumer Prices and Food & Nonalcoholic Beverage Prices for China, 2017-2018

(Percent change)

Anonymous said...

Correcting date:

https://fred.stlouisfed.org/graph/?g=KqTz

January 15, 2018

Consumer Prices and Food & Nonalcoholic Beverage Prices for China, 2017-2022

(Percent change)

[ China maintains a 2-year surplus of important foods. Also seeds and tubers and live-stock breeds are domestically developed, though soybean seeds meant to tolerate low-moisture conditions are now going to be imported from Argentina. ]

Anonymous said...

This was an important post for me, helping me understand quite a lot about the Chinese approach to agriculture, and relatedly to the wildly successful Chinese anti-poverty program. I am grateful for the post and comments.

Thank you so much.

rosserjb@jmu.edu said...

Kevin,

I think Muth did not use a specific example, although the Ezekial paper was supposedly about crops.

A classic example is indeed the old corn-hog cycle, which is a bit more complicated than a cobweb, more like a predator-prey cycle. That one used to exist, sort of five to six years, but seems to have disappeared in more recent years. Farmers getting smarter?

Cattle cycle, with longer lags, still there.

Anonymous said...

https://www.parisschoolofeconomics.eu/docs/guesnerie-roger/muth61.pdf

July, 1961

Rational Expectations and the Theory of Price Movements 
By John F. Muth

In order to explain fairly simply how expectations are formed, we advance the hypothesis that they are essentially the same as the predictions of the relevant economic theory. In particular, the hypothesis asserts that the economy generally does not waste information, and that expectations depend specifically on the structure of the entire system. Methods of analysis, which are appropriate under special conditions, are described in the context of an isolated marlcet with a fixed production lag. The interpretative value of the hypothesis is illustrated by introducing commodity speculation into the system.

2slugbaits said...

I think there's a simpler explanation of the cobweb model. If the slope of the demand curve is greater than the slope of the supply curve, and if agents only set next year's output based on current prices, then the time path is explosive and will not converge to an equilibrium. But my understanding of the cobweb model is that it also describes the reverse; i.e., convergence towards equilibrium when the demand curve is flatter than the supply curve. I don't think the cobweb model is inherently wrong in-and-of-itself. The real question is the extent to which agents are good at game theory.

2slugbaits said...

Barkley Rosser,
I'm guessing that the reason hogs might be vulnerable to the cobweb model is because hogs can't be withheld from the market. They have to come to market based on biology.

Anonymous said...

I'm guessing that the reason hogs might be vulnerable to the cobweb model is because hogs can't be withheld from the market....

[ China always has at least a 2-year store of pork. ]

Anonymous said...

I think Muth did not use a specific example

[ Muth did use pork as an example in the 1961 paper referenced above. ]

2slugbaits said...

Anonymous,
The issue is when hogs must come to the slaughterhouse, not when previously slaughtered pork is sold to consumers.

kevin quinn said...

2slugbaits: yes, as you say, the cobweb idea refers to both cases, damped and explosive cycles depending on relative elasticity of supply and demand, or the stable orbit you get when the two have the same elasticity.

I agree with you re: strategic actors. Muth's confidence that markets aren't prone to over-shooting alternating with undershooting depends on agents acting strategically, forming the unique self-fulfilling expectations about the price following a shock.
This is interesting, I think, because it is in some tension with the picture of agents who are simple price-takers, the picture that underlies the invisible hand theorems.

I will ask an expert on economic dynamics: Barkley, am I making any sense on this last point?

Anonymous said...

The issue is when hogs must come to the slaughterhouse, not when previously slaughtered pork is sold to consumers.

[ The Chinese use pork stores to protect against damaging price swings and to allow pork farmers to plan and be confident in pricing ahead. Simply notice how the Chinese are presently containing food price increases, while providing for steadily increasing agricultural production. ]

Anonymous said...

I'm guessing that the reason hogs might be vulnerable to the cobweb model is because hogs can't be withheld from the market. They have to come to market based on biology.

[ Yes, but the Chinese government buys pork expressly to limit farmer-damaging price declines. ]

rosserjb@jmu.edu said...

Certainly the classic formulation a la Ezekial and older textbooks that used to cover this has to do with the alopes of the supply and demand curves, not even the elasticities. But the positions of those curves depend on expectations, which is what drove Muth to do his paper, also poking at his colleague at Carnegie-Mellon, the late Herb Simon, father of bounded rationality and behavioral economics.

I am on the road about to deliver a paper on the underpinnings of a bunch of this tomorrow morning early in Portland, OR, Western Econ meetings, where two of my three grandsons also happen to be located. Glad the plane flew.

Anyway, the person who first empirically studied the corn-hog cycle was the late Sewall Wright, whom I knew personally. He was also one of the three main creators in the early 1930s of the neo-Darwinian synthesis in evolutionary theory and population genetics, a very long and deep topic I have published on and will discuss briefly in my talk tomorrow morning. As it is, he also did economics in the 1920s while working for the USDA, and his father was an economist. They are identified as the people who first discovered the identification problem in econometrics as they were trying to figure out this exact problem, how to distinguish supply changes from demand changes in data. I happen to possess, Sewall Wright's own personal copy of his long study from the 1920s on the empirics of the corn-hog cycle.

Anonymous said...

Thank you so much:

https://www.jstor.org/stable/pdf/1803183.pdf

September, 1925

The Forecasting of the Price of Hogs
By Charles F. Sarle

The Corn-Hog Industry.

The corn and hog industry is the most important phase of American Agriculture. This industry is largely concentrated in the Corn Belt, which is really the heart of agricultural America. It is composed of northern Missouri, eastern Nebraska, southeastern South Dakota, southern Minnesota, southern Wisconsin and western Ohio arranged about the three states of Iowa, Illinois, and Indiana as a nucleus. This is one of the most fertile large agricultural regions in the world....

Anonymous said...

https://en.wikipedia.org/wiki/Modern_synthesis_(20th_century)

See especially the work of Ernst Mayr, along with that of Wright...

Anonymous said...

http://cob.jmu.edu/rosserjb/THE%20EVOLUTION%20OF%20BEHAVIORAL%20INSTITUTIONAL%20COMPLEXITY.pdf

2017

THE EVOLUTION OF BEHAVIORAL INSTITUTIONAL COMPLEXITY
By J. Barkley Rosser, Jr. and Marina V. Rosser

Anonymous said...

For a sense of the related work of Ernst Mayr, please see:

https://www.nytimes.com/2005/02/05/science/05mayr.html

February 5, 2005

Ernst Mayr, Pioneer in Tracing Geography's Role in the Origin of Species
By CAROL KAESUK YOON