What Does This Tell You About The State of Modern Economic Theory?

|Peter Boettke|

Tyler Cowen links to this paper on capital theory paradoxes.  I recommend reading it closely and thinking seriously about what it tells you about the state of modern economic theory.* 

What do you think?

--------------

*You might find Reuven Brenner's "Extracting Sunbeams Out of Cucumbers" in The Force of Finance (2001) worth reading as well.

Comments

[Juan Carlos]

it tells that economic theory is mathematized beyond reality and any possible comprehension

[Juan Carlos]

i'd very much like to hear what you make of it peter

[Bill Woolsey]

What is the status of the "golden rule?"

Is it a market phenomenon? Is it really true that if there were no population growth (very possible) and no technological improvement (certainly conceivable,) the real interest rate would settle at zero?

Or is this normative? At positive interest rates, the sum of consumption over all periods will be less than it would be at zero interest, and more consumption is better?

I am curious about what Austrian (other?) economists think about this.

Other than that, I think it is good that economists are thinking about multiple capital goods. But I didn't get too far in this particular paper before I decided I have other things I need to do before I can come close to figuring it out.

I did think the doodles were nice.

[david]

"I did think the doodles were nice."

I liked fig 17.2.

[Current]

That's got to be a joke.

[fundamentalist]

The article showcases Machlup's genius. Only someone with enormous insight and communications skills could pack so much economics into such a short speech. His book on the stock market is still one of the best around.

However, I hated to see him use the "acceleration principle". I thought Hayek did an excellent job of putting it to rest in "Pure Theory" and "Profits, Interest and Investment". Essentially, the acceleration principle can work only if the money supply rises rapidly, because without a rapid rise in the quanity of money demand for consumer goods and demand for capital goods and labor cannot all increase at the same time.

Bill: "...the real interest rate would settle at zero?"

Not as long as people place more value on money now than than on money received in the future.

[Rafael Guthmann]

The author of that paper has written a quite famous textbook on microeconomics. He has a very good didactic skills, as this paper uses less math than the average treatment of the topic.

So, what this law tells about "The State of Modern Economic Theory"? I would say that it tells that modern equilibrium theory has long extracted everything that is useful out of the GE approach and now mathematical economists with nothing better to do are making theorems about the less important/interesting stuff.

For example, the author of the paper proves an "anything goes theorem" about the relationship between the variations of the rate of interest and the levels of consumption in the steady state, considering that the maximum level must always be reached at zero interest rates.

To Bill Woolsey: Of course the golden rule tells you that the rate of interest should be zero. But you need to understand that the golden rule is not about maximizing utility, but about maximizing consumption in the steady state. What is the steady state? It is the state reached in about an infinite amount of time (asymptotically), where there is no more capital accumulation. To maximize the consumption output of this state, you need to accumulate capital until its net marginal productivity becomes zero, you will only do that if there is no interest. It is very simple.

[fundamentalist]

I apologize! I was responding to an article on another web site about Machlup!

[fundamentalist]

Albert Einstein stated that "as far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality."

[FC]

Whoever can come up with a mathematical description of Greek finances whose graph resembles a sad clown should win a Nobel.

[Barkley Rosser]

It is not clear that this is current economic theory. It appears to be chap. 17 of an unnamed book, and there are no references more recent than 1982. That was more than a quarter of a century ago, so can hardly be called "current economic theory."

[Mario Rizzo]

I agree with Barkley. Capital theory has been abandoned by current economics. The absence of capital theory tells you something about current economics. It is a dead-end as even Hayek came to see. So what happens to ABCT? Beyond rudimentary observations about capital structure, I am not sure we can say much. This is why I am very sympathetic to Tyler Cowen's revision (expasnsion?) of the ABCT.

[Greg Ransom]

Economists need to think hard about the significance of the fact that production --PRODUCTION -- is AWOL in their efforts to understand the world using the marginali logic of valuation.

Think about it.

For what it's worth, Machlup's brief for Hayek's Nobel identified Hayek's extension of marginalist logic to production as his most significant achievement.

Mario writes:

"I agree with Barkley. Capital theory has been abandoned by current economics. The absence of capital theory tells you something about current economics. It is a dead-end as even Hayek came to see. So what happens to ABCT? Beyond rudimentary observations about capital structure, I am not sure we can say much. This is why I am very sympathetic to Tyler Cowen's revision (expasnsion?) of the ABCT."

[Greg Ransom]

This is basically a myth -- Hayek never renounced the significance of his wok on the marginal logic of production. Quite the contrary.

Hayek had a bizarre way of apologizing for and minimizing just about all of his accomplishments -- a case for psychological biographers. But his off-hand self-deprecation is no basis for evaluating scientific contributions. Hayek was self-undercutting and self-deprecating for no good reason in his Sensory Order and several of his other books -- often apologizing in advance. This is psychology, not a basis for doing science.

Mario writes:

"I agree with Barkley. Capital theory has been abandoned by current economics. The absence of capital theory tells you something about current economics. It is a dead-end as even Hayek came to see. So what happens to ABCT? Beyond rudimentary observations about capital structure, I am not sure we can say much. This is why I am very sympathetic to Tyler Cowen's revision (expasnsion?) of the ABCT."

[Greg Ransom]

Barkley -- what is a reliable account of contemporary production / interest / capital theory.

My impression is that economists today have no really serious or competent or conceptually consistent handle on this domain.

Am I wrong?

And isn't this a bit of a problem a science applicable to a world with production, interest and capital?

By not conceptually consistent I mean not consistent in adhering consistently to marginalist logic, but mixing marginalist/subjective logic and categories with non-marginalist/objective categories and logic of the type found in, say, Ricardian economics and "one good" Ricardian aggregates like "land" or "capital".

[Roger Koppl]

May I ask the naive question? What is capital theory (or the theory of capital) supposed to *do*?

[Greg Ransom]

"May I ask the naive question? What is capital theory (or the theory of capital) supposed to *do*?"

Roger, here are three tasks. Who knows if they can be inter-related a single logically tractable formalism.

1. Lay out the valuational logic for an individual chooser choosing between a number of production methods across time, the way Menger lays out the marginal valuational logic of choice for a chooser in a simple case of two items, as in his _Principles_.

2. Help explain Bohm-Bawerk's problem of originary interest.

3. Explain money interest in the market place.

4. Account for the distribution of income -- who gets what, between labor & owners & lenders & entrepreneurs & arbitrageurs.

5. Help us understand the coordination system of the market as this coordination system works through processes of production.

[Greg Ransom]

Note that #1 can be extended to an imaginary "economy" run by a single all-knowing dictator/god -- i.e. a central planner / math economist, which is the same as a GE construct.

"1. Lay out the valuational logic for an individual chooser choosing between a number of production methods across time, the way Menger lays out the marginal valuational logic of choice for a chooser in a simple case of two items, as in his _Principles_."

[Current]

I don't see how capital theory can be avoided. Every macroeconomic theory that I've seen that doesn't use capital theory implicitly relies on some sort of "homogeneous fund" capital theory.

Hayek's theory may be complicated and Bohm-Bawerks too simplistic, but I think that some sort of theory is needed.

[Greg Ransom]

RE capital it looks to me like economics remains mired in a set of mostly un-perceived pathological problem situations.

It has problems given to it by Ricardo, etc. that make sense only in terms of a classical valuational logic which is both non-marginalist and non-"subjective", e.g. "distribution theory" and "growth theory".

It was Lachman, wasn't it, who pointed out how economists were mixing together the logic of marginalist valuational relations and the logic of Ricardo, e.g. "land" and "capital" as objective and geometrically well behaved aggregate classes -- and they were mostly blind to the incommensurable mess of the stew they kept cooking up.

And the Hayek-inspired growth theory of Harrod and Solow pretended to offer a marginalist model of "capital" expansion by a deeply misleading method which gave them a Ricaro-like objective and geometrically well behaved aggregate class of "capital" on the cheap -- with a one-good class of "capital".

A good case can be made this model of "K" does more to mislead the economics profession, than it does to inform it about the role of production goods in the economy.

When it comes to "interest", it looks like economics doesn't know how to coherently conceive the problem of "interest" as a purely logical problem of choice for a single individual and completely abstracted from the example of money interest in the real world (a single individual choosing between multiple production processes and consumption opportunities across time).

And when economics attempts to engage the multi-person case, the muddle between the purely logical multi-person problem situation and the phenomena of money interest the actual world becomes doubly confused.

Well, those are two. There are others.

If these have been solved recently, or even insightfully discussed recently, please direct me to the reference.

[Peter Boettke]

Good point Barkley. It was probably more the state of the art when I was studying with Michael Alexeev and learning general equilibrium theory, than with what has gone on since. But the author is the author of the standard textbook.

The question was really about the state of the art with general equilibrium theory, not about capital theory per se. But my point, which may not be so novel or even correct, is that "theory" balled itself up into a corner. As I said in a few essays I wrote in the mid-90s, economics has become precisely irrelevant. Perhaps I am fighting a battle from a generation ago and I need to update. But I am concerned when economics becomes so comfortable with the "anything goes".

Pete

[Roger Koppl]

Pete,

Back in 1932, Harold Hotelling showed that Edgeworth taxation paradoxes may arise in a general equilibrium setting. He showed that “a tax on sellers of two commodities may result in both prices being lowered even under free competition.” That's a kind of "anything goes" result, isn' it? I'm not sure how scandalous Mas-Colell's essay is.

[Rafe Champion]

I wonder if his office has a window?

Recall the story about the people, including an economist, who were wrecked on a desert island. They had a store of tinned goods and they searched in mounting desperation for some way to open the tins. "Cheer up" said the economist "assume a tin opener!".

[fundamentalist]

Roger: "May I ask the naive question? What is capital theory (or the theory of capital) supposed to *do*?"

You can't have a good business cycle theory without it. Look at the mess that mainstream econ is in over the latest crisis. Everyone can describe what happened fairly well, but no one has a theory of what happened, except the ABCT, which requires sound capital theory.

[Current]

fundamentalist,

I also think capital theory is needed for growth theory.

[Roger Koppl]

@fundamentalist:

I think you are right. Without a more or less "Austrian" capital theory it is hard to get malinvestment, and malinvestment matters for policy. As far as I can tell, however, we would get everything we need if we pitched all the perplexing stuff about "orders of goods," and "average period of production" and so on. We should just note that different investment goods and capital combinations have different "duration." In other words their values have different degrees of interest-rate sensitivity.

There are at least two issues here. First, it is hard to attach objective meanings to concepts such as "order of good." Second, in a general equilibrium context anything goes.

So let me restate my question:

What is capital theory (or the theory of capital) supposed to do *besides* generate malinvestment in the unsustainable boom?

[Greg Ransom]

In _The Pure Theory of Capital_ what Hayek does is bring to light some of the problems & logic of planning and valuation of goods across time involving non-permanent production goods. Hayek's _Individualism & Economic Order_ essays show how these problems and valuation considerations lie at the heart of global economic coordination and wealth production.

Hayek argues -- persuasively I believe -- that the difference between understanding and not understanding market coordination and the problems of market coordination is the difference between perceiving the coordination and valuation problems of non-permanent goods -- and failing to imagine them. Hayek sees this as at the core of the difference between his own understanding of the coordination process of the market -- and the failures of Keynesian and neoclassical economics to understand these.

Roger asks:

"What is capital theory (or the theory of capital) supposed to do *besides* generate malinvestment in the unsustainable boom?"

[Greg Ransom]

The central difference between Hayek and Lange & Lerner on socialism plays out in different understandings of capital theory.

The central difference between Hayek & Keynes & Samuelson on the business cycle & "macro" plays out in different understandings of capital theory.

The central difference between Hayek & Solow on growth and wealth production plays out in different understandings of capital theory.

Etc.

Roger asks:

"What is capital theory (or the theory of capital) supposed to do *besides* generate malinvestment in the unsustainable boom?"

[fundamentalist]

Current: "I also think capital theory is needed for growth theory."

Yes, I shouldn't have left that out.

Roger: "We should just note that different investment goods and capital combinations have different "duration." In other words their values have different degrees of interest-rate sensitivity."

Roger Garrison illustrates the ABCT with the capital goods/consumer goods trade off graph. It could also be done by calling them future goods/present goods, or capital/labor. It depends on the depth that you want to get into. In my intro class on micro, I show how all three are related and explain the business cycle with the capital goods/consumer goods trade off. It can be that simple. The text book uses the capital goods/consumer goods trade off to illustrate growth over the long haul. I enhance it to explain business cycles.

You have to distinguish between capital and consumer goods at the minimum, but if you want to go deeper into understanding economics, you need to break up capital goods even more. Dividing capital goods into commodities, capital goods, durable goods, and consumer goods is a nice level of disaggregation.

[fundamentalist]

Roger, I just found this in Hayek's "Profits, Interest and Investment":

"But the crude dichotomy of industry into consumers' goods industries and capital good industries is certainly wholly insufficient to reproduce the essential features of the complicated interdependency between the various
industries in actual life. There is every reason to believe that there are as great differences between the position of the different kinds of capital good industries as there are between them and the consumers' goods industries. The capital goods industries are not all equally adapted to supply the consumers' goods
industries with any kind of equipment they may need; they are further organised in a sort of vertical hierarchy."

"This fact is essential for our further argument.
Even the concept of the "stages of production"
which was intended to supply in the place of the crude dichotomy a somewhat finer distinction is not quite adequate for the purpose."

In other words, Hayek found the stages of production insufficient to describe reality.

[Current]

Roger,

Consider the question: Is the business cycle *beneficial*?

Certainly misallocation is involved, but so is forced saving. Wages cannot adjust to the interest rate quickly. I doubt the business cycle is beneficial but I think that we need some capital theory to show that.

And what about growth, as I mentioned?

Now, Hayek's approach may be too complex certainly. But, I think that we need some sort of approach.

[Roger Koppl]

Current:

Evidently I was cryptic. Sorry about that. I was not defending business cycles! I was saying, with you, that "we need some capital theory to show" why business cycles are better avoided. In particular, we need capital theory to show that we get *mal*investment with business cycles. I was also saying that we can keep it relatively simple. We can probably get all we need from a few basics plus duration.

I agree about growth theory. But there, too, I don't think I see why we need the complicated stuff Hayek was waded into in his Theory of Capital. Less is more. We need to wade into the complicated stuff enough to show that it is indeed complicated. We need to do that to warn future researchers off simplistic theories that might lead them into epistemically bogus policies. But there does not seem to be much more to do since more or less anything goes a la Mas-Collel and Hotetlling.

[Current]

I pretty much agree with you. Hayek's "Pure Theory of Capital" is unnecessarily complicated. Much of the problem in that book is caused by Hayek's more-headbreaking-than-normal prose style, and by the tangents he goes into (though the tangents are the best bits). A simpler theory is needed for both growth and business cycles. As you say it's possible to keep it very simple for business cycle theory, but growth perhaps needs a bit more.

[Pietro M.]

Doesn't "anything goes" means "there are no significant restrictions on the dynamic behavior of the economic model?"

It is tantamount to say that general equilibrium, once sufficient complexity is poured in, has no empirical content. That's exactly what Mises thought of his own theory.

Isn't it the same as the Boldrin Montrucchio theorem? Any conceivable dynamic behavior can be justified out of a dynamic general equilibrium model. Which I read as a "whatever you observe, you have at least one DSGE that tells you that you are right" (here I see a possible non sequitur in my argument: the fact that at least one DSGE exists given a certain behavior to be explained doesn't mean that a DSGE with CREDIBLE functional forms and parameter values can be found).

Isn't it equivalent to saying that data can't tell us anything about reality because reality is so complex that the contribution of econometrics to our knowledge is close to nil?

Or, in other words, disputes in economics are rarely if ever settled by data alone, even if theories are constrained to live in the rarefied and abstract world of GE.

[Pietro M.]

Regarding the importance of capital theory: let's scrap everything and see the core of the ABCT idea. In my reading of ABCT, which is "something is surely missing, but it's definitely better than RBC and NUKE models", the core concepts are "structural unsustainability" and "monetary non-neutrality causing structural unsustainability".

The goal of capital theory should be to investigate how it is possible that the economy reaches an unsustainable state which requires a crisis to be corrected.

The goal of monetary theory should be to investigate how monetary shocks can produce the previous result.

The two things together makes a full-blown theory of business cycles which is Austrian in its logical form although not necessarily Austrian in all the details (i.e., maybe a general equilibrium model with some richer dynamic and a wider scope for money may fulfill the same goals, or maybe structural unsustainability can be interpreted in financial terms with less references to production, like in Cowen's book).

In the end, a theory like: "monetary policy causes financial leverage, this causes a boom, financial leverage develops into systemic fragility, at a certain point a crisi shall occur to deleverage" is logically Austrian.

It is not neoclassical, or "Cowenian", because there is widespread inefficiency, it is not neokeynesian because economic policies don't help, or even make things worse. It is neither neoclassical nor neokeynesian because the economy is not a shapeless blob subject to random shocks, but a complex object potentially subject to structural problems.

[Greg Ransom]

You'd think that economists should at least understand why their "science" can't tell us much if anything about an economy using, you know, multiple production goods, as far as chief intellectual tool of their "science" is involved, the logic of marginal valuation.

And you'd think that something could by learned by getting clear about what it can and can't tell us about this matter -- this matter at the very heart of a capital using economic system.

But you don't find really any economists thinking about this (in a non-pathological way -- Solow does not count.)

Most are off doing econometrics and the "scientif" fashion of the week.

In logic and the study of language, the key breakthrough is in understanding the defects and limitations of the dominant "modeling" program -- understanding why it doesn't work or what if can and can't do, is the hinge upon the door to actually understanding turns. This is what you get in the breakthrough work of Wittgenstein.

Note that the bureaucratic and incentive structure of the academy and the philosophy profession still rewards endless reiterations of the old failed "modeling" project, and it doesn't much reward those who devote time to understanding the problems and alternative pioneered by Ludwig Wittgestein.

But at least most logicians and philosophers don't pretend this result doesn't and doesn't have to be answered or acknowledged in some way (if only to attempt to re-bury it).