Saturday, May 9, 2015

Jason Smith — On the use of hypotheses: or, what do you get when you assume non-ergodicity?

Jason Smith replies to Lars Syll (and Paul Davidson).
What comes out of assuming ergodicity? All of basic thermodynamics and much of basic economics. If we assume economic (or thermodynamic) systems aren't ergodic -- what does that give us?

Essentially, assuming non-ergodicity is analogous to the assumption that I(A) < I(B) in the information transfer framework (ergoditicy is the assumption that I(A) ≈ I(B) ... the information in the two macro observable is the same, from which you can derive supply and demand). 
What can we get from the assumption I(A) < I(B)? Nothing. 
That is to say that while ergodicity is a useful assumption, non-ergodicity is a completely useless assumption. It doesn't prove that economies are quasi-periodic chaotic systems or that they are some other kind of complex system -- you need evidence for that! Show us a model that that is empirically successful. Or at least more empirically successful than assuming ergodicity.
Yes, that's a point that Keynes made and which Davidson and Syll elaborate:
Many thanks for sending me your article I enjoyed it very much. I am sure these matters need discussing in that sort of way. There is one point, to which in practice I attach a great importance, you do not allude to. In many of these statistical researches, in order to get enough observations they have to be scattered over a lengthy period of time; and for a lengthy period of time it very seldom remains true that the environment is sufficiently stable. That is the dilemma of many of these enquiries, which they do not seem to me to face. Either they are dependent on too few observations, or they cannot rely on the stability of the environment. It is only rarely that this dilemma can be avoided.
Letter from J. M. Keynes to T. Koopmans, May 29, 1941
Of course, that is a bit of a hand wave but Keynes is much more specific about it in other places. But the idea is that the basis for neoclassical assumptions is too non-representation of the subject matter to yield a useful methodology. Neoclassical methods are not useful for telling us what we really need to know, in particular for policy formulation. Keynes proposed a new economic method based on a monetary production economy in which money is non-neutral and uncertainty dominates.

Keynes was not only a theoretician but an economic "engineer" who are active in the world of policy at the time of the Great Depression and his ideas are credited with saving the day — other than by neoclassical economists that have sought to "correct" this, for which the world is now suffering another prolonged contraction.

In the Keynesian view, econometric models are essentially a waste of time. According to old Keynesians and Post Keynesians, Paul Samuelson "bastardized" Keynes by introducing key assumptions that Keynes specifically rejected.

In this view, macroeconometricians should be doing something else, like looking for types of models that actually are useful, like the stock-flow consistent approach developed independently by James Tobin and Wynne Godley, and set forth in Godley & Cripps, Macroeconomics (1983) and Godley and Lavoie (2007, 2nd ed. rev., 2012). See Lavoie (2010).

Information Transfer Economics
On the use of hypotheses: or, what do you get when you assume non-ergodicity?
Jason Smith


MRW said...

Jason Smith doesn't make any sense to me.

Paul Davidson does. You can actually listen to him discuss it here:

Dan Kervick said...

Yes, that's a point that Keynes made and which Davidson and Syll elaborate

The point Keynes makes in the quoted paragraph has nothing to do with the distinction between ergodic and non-ergodic systems, and he never described his position in those terms. Economists who have adapted the"non-ergodic" terminology to make some other vague point about irregularity or unpredictability, Knightian/Keynesian uncertainty are mostly speaking gibberish.

Tom Hickey said...

That is true. It was Paul Samuelson that intiallyi began the debate about ergodicity in his neoclassical synthesis.

Paul Samuelson [1969] has written that if economists hope to move economics from “the realm of history” into “the realm of science” they must impose the “ergodic hypothesis” on their theory.[1] In other words Nobel Prize Winner Paul Samuelson has made the ergodic axiom the sine qua non for the scientific method in economics. Lucas and Sargent [1981] have also claimed the principle behind the ergodic axiom is the only scientific method of doing economics.

Following Samuelson’s lead, most economists (e.g., Lucas, Cochrane, Sargent, Stiglitz, Mankiw, M. Friedman, Scholes, etc.) and economic textbook writers either implicitly or explicitly have assumed that observable economic events are generated by an ergodic stochastic process.

Paul Davidson, A Response to John Kay

The debate is about a controversy that emerged after Keynes and Keynes did not address it specifically. Davidson continues:

In his The General Theory, John Maynard Keynes stated that classical economists

“resemble Euclidean Geometers in a non Euclidean world who, discovering that in experience straight lines apparently parallel often meet, rebuke the lines for not keeping straight – as the only remedy for the unfortunate collisions which are occurring. Yet in truth there is no remedy except to throw over the axiom of parallels and to work out a non Euclidean geometry. Something similar is required today in economics."[3]

In this analogy comparing Euclidean geometry in a non-Euclidean world to classical theory in our world of experience, Keynes was alluding to the fact that in the classical analysis the future is presumed to be known and therefore free markets are efficient since they produce full employment (the equivalent of the “parallel lines”). Yet significant and persistent unemployment (the “unfortunate collisions”) occur in the real world. Accordingly, classical economists rebuking the lines in the real world for not keeping straight is equivalent to blaming the workers for their unemployment problem because workers would not accept lower wages.

In creating a “nonEuclidean” economic theory to explain why these unemployment “collisions” occur in the world of experience, Keynes uses the logical deductive method but he had to deny (“throw over”) the relevance of several classical axioms for understanding the real world. The classical ergodic axiom which assumes that the future is known and can be calculated as the statistical shadow of the past was one of the most important classical assertions that Keynes rejected.[4]

Keynes's general theory is a deductive method of analysis. Keynes’s concept of uncertainty about the economic future requires the economic system to be generated by a nonergodic process. At the time of his writing The General Theory, Keynes did not know of the ergodic stochastic theory that was being developed by the Moscow School of Probability in the 1930s. Nevertheless in his criticism of Tinbergen's [econometric] method , Keynes [1939][5] wrote that Tinbergen's method is not valid for any economic forecasting because economic data “are not homogeneous” over time. Non homogeneity is a sufficient condition for nonergodicity.

Tom Hickey said...

BTW, there is already a heterodox approach to economics based on thermodynamics called thermoeconomics.

Thermoeconomics is based on the proposition that the role of energy in biological evolution should be defined and understood through the second law of thermodynamics but in terms of such economic criteria as productivity, efficiency, and especially the costs and benefits (or profitability) of the various mechanisms for capturing and utilizing available energy to build biomass and do work.[6][7]


Thermoeconomists maintain that human economic systems can be modeled as thermodynamic systems. Then, based on this premise, theoretical economic analogs of the first and second laws of thermodynamics are developed.[8] In addition, the thermodynamic quantity exergy, i.e. measure of the useful work energy of a system, is one measure of value.

Economic systems

Thermoeconomists argue that economic systems always involve matter, energy, entropy, and information.[9] Moreover, the aim of many economic activities is to achieve a certain structure. In this manner, thermoeconomics applies the theories in non-equilibrium thermodynamics, in which structure formations called dissipative structures form, and information theory, in which information entropy is a central construct, to the modeling of economic activities in which the natural flows of energy and materials function to create scarce resources.[1] In thermodynamic terminology, human economic activity may be described as a dissipative system, which flourishes by consuming free energy in transformations and exchange of resources, goods, and services.[10][11]


See Economics and Thermodynamics: New Perspectives on Economic Analysis, edited by Burley and Foster.

I have previously called attention to economic methodology based on Ilya Prigogine's work on dissipative structures as well as evolutionary biology as suggested by David Sloan Wilson

Y. Shiozawa (Osaka City University), Economy as a Dissipative Structure

John Foster, Why is Economics not a Complex Systems Science? Discussion Paper No. 336, December 2004, School of Economics, The University of Queensland

MRW said...


Thanks for the link to Paul Davidson's "A Response to John Kay." Missed that.

I've been listening to and reading a lot of Paul Davidson recently. I am now firmly in the Paul Davidson camp. He makes a lot of sense to me.

The Appendix in his latest book, "THE KEYNES SOLUTION The Path to Global Economic Prosperity" is worth the price of the book.

The Appendix is called, "Why Keynes’s Ideas Were Never Taught in American Universities." He demolishes Samuelson, and rightly so.

Magpie said...


"Keynes proposed a new economic method based on a monetary production economy in which money is non-neutral and uncertainty dominates."

Where did Keynes (as opposed to his followers, whether New, Post, bastard, legitimate, fundamentalist, whatever the label) do that?

I suppose that may sound kind of snarky, but it's a genuine question.

Dan Kervick said...

BTW, there is already a heterodox approach to economics based on thermodynamics called thermoeconomics.

More crackpots. These people, and most of the other people applying physical systems theories to economics, are dumber than the dumbest neoclassicals.

Assuming human beings and their societies are physical systems, then of course thermodynamics, quantum dynamics and all and any other dynamic theories apply to them. However, almost nothing of useful economic interest can be derived from such applications, since human economic systems are one very special subclass of such physical systems, and the behavior we are interested in depends on the peculiarities of human nature, human preferences, human emotional responses, human rationality, etc. which cannot be read off of a bare thermodynamic or quantum or other physical characterization of the system.

Tom Hickey said...

“Keynes proposed a new economic method based on a monetary production economy in which money is non-neutral and uncertainty dominates."

Where did Keynes (as opposed to his followers, whether New, Post, bastard, legitimate, fundamentalist, whatever the label) do that?

I suppose that may sound kind of snarky, but it's a genuine question.

Keynes did not do so specifically, but this is is the way his work is now summarized in terms of his assumptions that conflicted with neoclassical assumptions.

Keynes did not use the term "monetary production economy" in the GT but he did distinguish between a monetary economy and a non-monetary economy;

The perplexity which I find in Marshall's account of the matter is fundamentally due, I think, to the incursion of the concept 'interest', which belongs to a monetary economy, into a treatise which takes no account of money. 'Interest' has really no business to turn up at all in Marshall's Principles of Economics,— it belongs to another branch of the subject.

Professor Pigou, conformably with his other tacit assumptions, leads us (in his Economics of Welfare) to infer that the unit of waiting is the same as the unit of current investment and that the reward of waiting is quasi-rent, and practically never mentions interest, which is as it should be. Nevertheless these writers are not dealing with a non-monetary economy (if there is such a thing). They quite clearly presume that money is used and that there is a banking system. Moreover, the rate of interest scarcely plays a larger part in Professor Pigou's Industrial Fluctuations (which is mainly a study of fluctuations in the marginal efficiency of capital) or in his Theory of Unemployment (which is mainly a study of what determines changes in the volume of employment, assuming that there is no involuntary unemployment) than in his Economics of Welfare.
— GT, p. 155

"But as soon as we pass to the problem of what determines output and employment as a whole, we require the complete theory of a monetary economy." — GT, p. 232 emphasis added.

In a complete theory of a monetary economy, money is not neutral in the long run, as neoclassical economics assumes.

Subsequently, some economists referred to what Keynes called a "monetary economy" as a "monetary production economy." I don't know who first used the phrase wrt to Keynes, but it seems to have arisen from the distinction that Marx drew between a production economy —C-M-C' — and a monetary production economy — M-C-M' — which Marx identified with "capitalism." See Wray, Theories of Value and the Monetary Theory of Production:

Keynes identified a monetary economy as one in which expectations of the future influence decisions taken today; or, one in which money is a subtle device for linking the present and future; or one in which production begins with money on the expectation of ending with more money later: in his lectures in the early 1930s, at the time that he was just beginning work on the General Theory (GT), Keynes referred to Marx's famous M-C-M'.

Tom Hickey said...

There is, however, a second, much more fundamental inference from our argument which has a bearing on the future of inequalities of wealth; namely, our theory of the rate of interest. The justification for a moderately high rate of interest has been found hitherto in the necessity of providing a sufficient inducement to save. But we have shown that the extent of effective saving is necessarily determined by the scale of investment and that the scale of investment is promoted by a low rate of interest, provided that we do not attempt to stimulate it in this way beyond the point which corresponds to full employment. Thus it is to our best advantage to reduce the rate of interest to that point relatively to the schedule of the marginal efficiency of capital at which there is full employment.

There can be no doubt that this criterion will lead to a much lower rate of interest than has ruled hitherto; and, so far as one can guess at the schedules of the marginal efficiency of capital corresponding to increasing amounts of capital, the rate of interest is likely to fall steadily, if it should be practicable to maintain conditions of more or less continuous full employment — unless, indeed, there is an excessive change in the aggregate propensity to consume (including the State).

I feel sure that the demand for capital is strictly limited in the sense that it would not be difficult to increase the stock of capital up to a point where its marginal efficiency had fallen to a very low figure. This would not mean that the use of capital instruments would cost almost nothing, but only that the return from them would have to cover little more than their exhaustion by wastage and obsolescence together with some margin to cover risk and the exercise of skill and judgment. In short, the aggregate return from durable goods in the course of their life would, as in the case of short-lived goods, just cover their labour-costs of production plus an allowance for risk and the costs of skill and supervision.


Tom Hickey said...


Now, though this state of affairs would be quite compatible with some measure of individualism, yet it would mean the euthanasia of the rentier, and, consequently, the euthanasia of the cumulative oppressive power of the capitalist to exploit the scarcity-value of capital. Interest to-day rewards no genuine sacrifice, any more than does the rent of land. The owner of capital can obtain interest because capital is scarce, just as the owner of land can obtain rent because land is scarce. But whilst there may be intrinsic reasons for the scarcity of land, there are no intrinsic reasons for the scarcity of capital. An intrinsic reason for such scarcity, in the sense of a genuine sacrifice which could only be called forth by the offer of a reward in the shape of interest, would not exist, in the long run, except in the event of the individual propensity to consume proving to be of such a character that net saving in conditions of full employment comes to an end before capital has become sufficiently abundant. But even so, it will still be possible for communal saving through the agency of the State to be maintained at a level which will allow the growth of capital up to the point where it ceases to be scarce.

I see, therefore, the rentier aspect of capitalism as a transitional phase which will disappear when it has done its work. And with the disappearance of its rentier aspect much else in it besides will suffer a sea-change. It will be, moreover, a great advantage of the order of events which I am advocating, that the euthanasia of the rentier, of the functionless investor, will be nothing sudden, merely a gradual but prolonged continuance of what we have seen recently in Great Britain, and will need no revolution.

Thus we might aim in practice (there being nothing in this which is unattainable) at an increase in the volume of capital until it ceases to be scarce, so that the functionless investor will no longer receive a bonus; and at a scheme of direct taxation which allows the intelligence and determination and executive skill of the financier, the entrepreneur et hoc genus omne (who are certainly so fond of their craft that their labour could be obtained much cheaper than at present), to be harnessed to the service of the community on reasonable terms of reward.

— GT, p. 289-90

Tom Hickey said...

Keynes did not address money "neutrality" or "Non-neutrality." at least as a major topic, in those terms. But he did address the concept and its assumption by previous economists. It lies at the foundation of his concept of a monetary economy.

The theory which I desiderate would deal … with an economy in which money plays a part of its own and affects motives and decisions, and is, in short, one of the operative factors in the situation, so that the course of events cannot be predicted in either the long period or in the short, without a knowledge of the behaviour of money between the first state and the last. And it is this which we ought to mean when we speak of a monetary economy.

J. M. Keynes A monetary theory of production (1933), quoted by Lars Syll.

This contradicts the assumption of money neutrality, where holds that money only affects nominal variable (prices, wages, inflation) but not real variable of input and output.

See also Bill Mitchell, Money neutrality – another ideological contrivance by the conservatives

Tom Hickey said...

“By ‘uncertain’ knowledge, let me explain, I do not mean merely to distinguish what is known for certain from what is only probable. The game of roulette is not subject, in this sense, to uncertainty; nor is the prospect of a Victory bond being drawn. Or, again, the expectation of life is only slightly uncertain. Even the weather is only moderately uncertain. The sense in which I am using the term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest twenty years hence, or the obsolescence of a new invention, or the position of private wealthowners in the social system in 1970. About these matters there is no scientific basis on which to form any calculable probability whatever.” We simply do not know!

— Keynes 1937: 213–214)
in J. M. Keynes, “The General Theory of Employment,” Quarterly Journal of Economics 51 (1937): 209–223. ( source)

Here are a couple of interesting post about Keynes on uncertainty.

"In the part of The General Theory that was lost to standard economics as it evolved into the neoclassical synthesis, Keynes put forth an investment theory of fluctuations in real demand and a financial theory of fluctuations in real investment. Desired portfolio composition and thus financial relations in general are most clearly the areas of decision where changing views about the future can most quickly affect current behavior. This responsiveness is true not only for ultimate units like business firms and households but also for specifically financial institutions like commercial banks, investment banks, etc. But the future is uncertain. To understand Keynes it is necessary to understand his sophisticated view about uncertainly, and the importance of uncertainly in his vision of the economic process. Keynes without uncertainly is something like Hamlet without the Prince." — Hyman P. Minsky, John Maynard Keynes, 1975 p. 57

Lot of other apt quotes, too.

Hyman Minsky on John Maynard Keynes
Doug Noland
The Credit Bubble Bulletin

BTW, Nolan was a sound money, anti-Keynesian but the crisis apparently sent him to reading Minsky and Keynes.

Here is an interesting post comparing Keynes and the Austrians on uncertainty regarding expectations.

Economic Thought
Keynesian Uncertainty
Jonathan Feingold

Magpie said...

Thanks, Tom.

Magpie said...

I've read with particular interest the post by Jonathan Finegold Catalán (for short, JFC: the name is a bit long). He linked to Daniel Kuehn and to Chidem Kurdas (who has an opposite opinion).

I'm sure JFC and Kuehn are very knowledgeable about both Austrian and Keynesian economics, and surely about a great many other things. Both sound like agreeable young men.

However, to be honest, I'd bet those posts are not among their proudest achievements. Hopefully I'm not being rude, but I think they have very little idea what they are talking about. And it shows: Try to make sense of JFC's post (Kuehn's is much easier to understand: whatever it was JFC said, Kuehn agrees). Mind you, it's not their fault.

Now, I might be mistaken. Maybe it's just my fault: I did not understand what their deepest meaning was.

But I did understand Kurdas' meaning. And, after all is said and done, after all ceremonial claims of originality are made, I've seen many big-shot Austrians repeating exactly Kurdas' opinion. One may agree or disagree with that opinion, of course. But there's a there there.

With utmost respect, there's no there with JFC and Kuehn.

I don't think "their" uncertainty is any different from Keynes' uncertainty. To his credit, JFC writes:

"Some of the [Austerian] economists in this second category see the fundamental difference between them and the 'stimulus' camp as the latter’s inability to consider the role policy-induced uncertainty has on entrepreneurial action."

Magpie said...

Let's take this, from Keynes:

"The sense in which I am using the term is that in which the prospect of a European war is uncertain ... About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know!"

Let's try to figure what Keynes tried to say there.

I can be tall (say, 6 ft tall or more) or short (less than 6 ft). I know which of these is true, but you don't. You know I must be either tall, or short, but not both. If asked which of them is true, you'll have "no scientific basis to form any calculable probability whatever". You simply do not know! That's uncertainty.

Take my word for it: It must always be true that P(T) + P(S) = 1.

If you knew with absolute certainty that I am tall, you would say that my probability of being tall -- P(T) -- is 1: P(T) = 1. Under this hypothesis, because you knew I am tall, you can also tell I was not short: P(S) = 0.

Let's call that Possibility A and write that down:


Viceversa: If you knew with absolutely certainty that I am short, you would say P(S) = 1 and P(T) = 0.


Note something: you are still uncertain about which of these two possibilities is true. It means you have no reason to believe one or the other is more likely to be true. You simply do not know! If I put a gun to your head, and threatened to shoot you if you didn't pick one, I think we both could agree that you would pick one, at random.

If enough people were selected, put alone in separate rooms, and, without knowing what the others were choosing, forced to pick one of these two possibilities having exactly the same information you have now, is there any reason to believe anything different that about half of them would have picked A and the other half B? Not, that I can think.

In other words, 1/2 of the population votes A, the other 1/2 B. Those are probabilities.

What this means is that, even when nobody knows the real probabilities of something, one can assign probabilities that maximize their uncertainty. Statisticians can do that. Mind you, sometimes they make mistakes, other times they get it right. This is how it works. As information becomes available, one revises those estimates: one adds to one, and subtracts to the other.

Now, you know I am not Keynes' fan (in fact, I feel the deepest contempt for the man and not much more respect for the thinker, either). But if even I (a boorish proletarian) am capable of understanding that, then surely you must conclude that the Prophet himself was able to see it.

On this, I think Keynes is not at fault, but some of his followers.