September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« Yes, This is a Police State | Main | Market Theory and the Price System »


Feed You can follow this conversation by subscribing to the comment feed for this post.

What do you think other people view the market as if not as a process? I'm not clear on how else one would think of the market.

Well - I'm not clear on how else an economist would think of the market. Clearly the general public sometimes sees it as an independent thing with prerogatives of its own.


Seriously, comparative statics is not process. A system of simultaneous equations is not process.

Have you read Franklin Fisher's Disequilibrium Foundations of Equilibrium Economics? Are you looking at the work I recommended to you the other day on creativity, novelty, etc.?

These are not surface issues, but deep issues in the nature of economic theorizing and the problem situation one is dealing with in economics theory.

This was not directed at the general public, but our scientific understanding of the market. I would also recommend Mel Reader's Chicago Economics: Permanence and Change, JEL essay.


I suppose evolutionary epistemology is to foundationalist epistemology what process economics is to static economics.

re: "Seriously, comparative statics is not process. A system of simultaneous equations is not process."

A paragraph of prose on a page is not process either, Peter. Deduction is not a process. No scientific output we produce is process.

The way we treat and talk about these processes varies and there are varying strengths and weaknesses to different treatments.

But as far as the understanding of the market I'm not sure what it could be other than a process of interaction, cooperation, and exchange. I'm simply curious what else you think people think of it as. I'm still not exactly clear on that point.


I think it is wrong to focus on the first part of the quoted sentence rather than the remainder. Granted, they could mean the same thing and there *is* a great deal of overlap - but it is not necessarily the same to study "the interplay of actions" and to study "the process." Those are very different in terms of units of analysis as well as in approach: studying actions and the interplay of actions says nothing about the overall process, its velocity, or its aim. Similarly, studying the process does not necessarily say anything about individual actions or the interplay of actions; it may simply be the study of macro-level changes without regard to the microfoundations of this change.

I can think of many implications of adopting the "wrong" perspective or approach depending on what one seeks to study, and many of them are far from beneficial or even productive. Especially from an Austrian point of view, it makes a whole lot of sense focusing on the remainder of Mises's statement: so that we study the "interplay of actions" primarily and the process secondarily (if at all necessary when we've already analyzed its microfoundations). So I think Misesians/Hayekians do well in not emphasizing the process (except perhaps when directly aimed toward general equilibrium, comparative static types) but action.

You talk about Mises's "analytic focus on process".

I know we have differences on the method of the analytic focus itself. But the subject of that focus - the market process - doesn't seem to be nearly as different, and I'm simply curious how you think other economists see the market if not as a process. Surely you don't think they see the market as an optimization problem any more than you see the market as a pragraph of prose!!


A few other good sources, of which Pete may have already recommended, include John A. Mathews' _Strategizing, Disequilibrium, and Profit_ (available here: and Meir Kohn's "Value and Exchange," (available here:

The argument I believe Pete is making (correctly, I may add) is that economists today may "think" about market process, they don't necessarily consider it relevant, or possibly consider it secondary, to the study of economics. This dichotomy is what Kohn refers to as the value and exchange paradigms. The profession today overwhelmingly tends to emphasize the outcome (what Kohn refers to as "value") rather than the process (what Kohn refers to as "exchange").

This is evident in all areas of economic research, including in law and economics with the Coase v. Posner paradigms; with Buchanan's public choice theory regarding the nature and function of government as a facilitator of exchange (see his book _The Limits of Liberty_); and a host of development and institutional economic theories (North, Greif, Bowles, etc.).

While I appreciate the thought, I really don't think the gap here is a reading list. And to be honest I don't have time to address this reading list anyway.

I'm simply curious about exactly what Peter thinks others think of the market as, because I genuinely don't know what the market would be if not a process and an interplay of actions cooperating under a division of labor. I am genuinely stumped as to what else it would be. And I'm still not clear what else you think others think it is.

The great process philosophy of the 20th century was Thomas Kuhn.

Not perfect, but illuminating -- sort of like Hayek and Mises.

I disagree with Rescher.

Let me give an example. Darwinian biology is VERY HARD. And very complex. And a completely different way of looking at things.

There are many elements to the theory, and at first it wasn't clear what was essential, or what would fill in all of the gaps, or what the evidence was, etc.

But there was core that stood like a mighty oak -- Darwin's book and body of work.

Whatever else Darwinian evolution was, you couldn't mistake that some large core of it was captured by Darwin.

Ditto with Wittgenstein social picture and "process theory" of language, justification, knowledge, etc.

Ditto with Kuhn's social picture and "process theory" of the advance of science.

There is SO MUCH to chew on, and what there is to chew on is provided in the first instance by EXAMPLES and NOT by formal constructs. It's an elephant with many parts, and it is understood threw soaking in the examples, threw chewing on the examples -- which help us to understand why the formal constructs are inadequate, deeply false in some ways.

Was Darwin right about everything? No. Was his work incompete? Yes. Are we still understanding his work better and in new ways. Yes.

But is it still called Darwinian biology for a reason -- and as Mayr & Ghiselin show, much insight is to be gained by grappling with and attempting to cast down and articulate the explanatory fundamentals of Darwin's multi-various process, making use of the original oak as a frame for thinking things though.

To reiterate.

It's hard stuff.

There are MANY elements to it.

There are signal thinkers who have coherently tied together many, many of the parts, in a way that allows us to grapple commonly on that difficult task of tying things together coherently and getting all of the parts in the best shape we can.

So central and giant thinkers ARE a pivot.









I don't think your comments are contributing to what is already a confusing conversation -- you appear to be commenting on another discussion altogether. Although I am sure your comments make good sense to you, I am also quite sure that they make little or no sense to the rest of us.

I am struggling to grasp what Pete means by "the market is a process." Perhaps the word "process" is too vague in this context. All economists, at some level, understand that the market is or has a process. Pete thinks this view of the market as a process set Hayek and Mises apart from their peers, but what he exactly means by that I don't know.

This is almost as confusing as the post on semantics and syntax.


Like a lot of things with economic analysis, simply acknowledging the existence of the market as a *process* is not the same thing as actually integrating that into one's analysis of the market. There's a lot of lip service paid to the things that Mises, Hayek, Kirzner, et al. emphasize, but many economists don't take them as seriously as they should.

A good example of this from a process perspective would be the Austrian stance on monopoly versus the more mainstream one. Whereas the latter sees this as something to be legislated against, the Austrian would view this most likely as a transient situation, as long as free entry wasn't restricted.

Brad W. - I would agree - simply acknowledging it is clearly entirely insufficient.

I'm increasingly regretting asking in the first place, and am increasingly convinced that this is just a (to quote Bryan Caplan) "Hayek said the sky is blue" situation. I'm just curious what other sky color was even on the table as a possibility. I'm glad to see Lee comment with some curiosity as to what Pete is getting at too.

I'm guessing Pete is saying something like: you can't model the unpredictable.

A model assumes everything it implies. There is no room for the emergence of novel and unpredictable implications. But a market process harnesses the creativity of individual minds. The market process acts upon itself and can solve its own transitory imperfections. In a sense, one cannot model a process that creates knowledge.

Viewing the market as static model makes it a toy. It has no spark of creativity, no capacity for self-evolution or knowledge discovery. Imperfections must be patched over with tape and glue, because they are just part of the toy.

Meh ... something like that anyway.


Practically everything in Austrian economics is used (or can be used) in "mainstream" economics to some extent. What I think is novel about it--and I think this is what Pete is talking about--is that Austrians are the only ones who focus on and take seriously the Misesian and Hayekian perspective of the market as a process--that is, they put it in the forefront of their analysis. Sure, everyone knows that people's values are subjective, that the market is a "discovery process" which economizes on decentralized knowledge, that entrepreneurial alertness and discovery are important, and that time figures centrally in all of this, but it seems, at least to me, that many economists ignore these things when doing economics.

re: "Practically everything in Austrian economics is used (or can be used) in "mainstream" economics to some extent."

I am a very strong proponent of this view. Usually Austrians don't like me because they don't like my claims about the potential for a meaningful synthesis.

I agree with you on the nature of Pete's claim. I'm unclear on his reasons for making the claim.

Granted - that's not the only reason why Austrians dislike me.


1. What Austrians are you referring to? If I didn't think you asked good questions I wouldn't respond.

2. This isn't the sky is blue. End state analysis isn't the same thing as process analysis. Again I point to Wagner's discussion of the difference between the neo-Walrasian and neo-Mengerian paradigm.

It is about institutional filters (as discussed by Nozick), it is about entrepreneurial alertness (as discussed by Kirzner), it is about discovery and learning (as discussed by Hayek), and it is about novelty, creativity, etc. (as discussed by Ostroy, Weyl, etc.).

If you see these analytical points of emphasis in a significant stream of contemporary economics please point me in that direction as I'd love to learn about developments in market process analysis that are ongoing.


Here is an example: Pete and I co-wrote a paper critical of Akerlof's "Lemons" model. Akerlof insisted that information asymmetries preclude potential Pareto efficient exchanges from taking place. For example, I know everything about my used car that I wish to sell, and you value a car as I described mine at more than my reservation price. We have a potential Pareto efficient exchange. But, because you are not certain of the true quality of my car, you discount the price you're willing to pay to a point less than my reservation price. Consequently, the trade doesn't take place.

In this regard, Akerlof is looking at an end state - there is some equilibrium price at which all used cars are exchanged, yet due to information asymmetries many of those potentially Pareto superior trades go unrealized and we're left with only the lemons selling in used car markets. Think of sales as being to the left of equilibrium. Information asymmetries in this regard are not an inherent trait of markets that provide opportunities for entrepreneurs to rectify and from which to profit by doing so, they are obstacles to achieving some idealized end state and must therefore be corrected through regulation.

What Akerlof disregards is the discovery process that entrepreneurs engage in every day. Akerlof wrote NOT in the context of, "Wow, markets are replete with information problems, let me explain how markets operate to correct one of them." No, instead his underlying assumption is that because of information asymmetries, markets fail and therefore require regulation to achieve some idealized end state equilibrium, as if one can even be known.

In reality, what has emerged to rectify Akerlog's "lemons" model is a host of businesses like Carmax, Carfax, and Toyota (and most all other car manufacturers) Used Certified. Markets work because of the discovery process where Akerlof argued they failed.

Coase argued for process in essentially the same way with his "Problem of Social Cost." He was effectively criticizing the neoclassical model, which assumed zero transactions costs. This assumption was important - essential - to achieving some idealized end state price and quantity equilibrium. But followers then claimed markets failed (there were essentially unstable) due to externalities, monopoly power, and . . . wait for it . . . information problems. As Coase rightfully argued, if there were zero transactions costs as assumed by the neoclassical model, there can then be no resulting market failures. Any obstacles to potentially Pareto efficient exchanges would be bargained away by the affected parties. Coase emphasized that social costs were minimized through the process of exchange of rights in the form of first assigning these rights in ways that best facilitate bargaining. It was essential to legal rule making that the only way we can ever know who values most possessing the rights to something is through voluntary exchange (i.e., Kohn's exchange paradigm), not some unknown idealized distribution of rights (i.e., the value paradigm).

I have one question and one comment.

The question is "What questions does "market as a process" answer that standard textbook models of markets do not?"

The comment is that statements like "a market is not a place" strike me as somewhat nonsensical. Often they are places. If my wife says she wants to stop by the market I'm not going to say "You know, the market isn't a place; its a process."

Mark: I don't think I understand your criticism of Akerlof. He says right off that "private institutions may arise to take advantage of the potential increases in welfare which can accrue to all parties" from getting around asymmetric information. And he has a whole section on "counteracting institutions."

A big issue on "process" theory is "intelligibility." You should be able to tell who does what. And what people do should be "intelligible" You should be able to understand how a real person might really do the things the model says he would do. Milton Friedman neglected intelligibility in his famous "positive economics" essay. Economics has really changed since 1980, but I think it is probably still true that most mainstream economists are weaker than the Austrians on intelligibility.

"The heart of the distinction is that, in a process analysis, events are represented as taking place sequentially in real time. By contrast, in a neoclassical is normally concerned with an equilibrium situation--an equilibrium defined not as the end-result or rest-state of a process, but as the condition of logical consistency among a group of mathematical relations. Thus a process analysis can be static in the sense that the process may have an eventual equilibrium state. And a neoclassical analysis can be dynamic to the extent that a variable labeled 'time' may enter into relations whose consistency constitutes the equilibrium. But the meaning of dynamic and static are very different in the two cases" (Langlois 1989, p.7).

I like the quotation from Richard Langlois. I should also add that The Economics of Time and Ignorance was all about the difference between process analysis and equilibrium (of one sort or another) analysis. People who haven't read the book should do so lest they put themselves in the position of having to re-invent the wheel.

One reason that process analysis hasn't caught on among mainstream economists is that no process story can be deterministic. We will not be able to describe each stage or phase of a true process in such a way that given the data of the first stage we can perfectly predict the second stage. Something unexpected (both to the agents and analysts) will happen.

I would now go further than we did in the book and say that process analysis is probably best told in words because they have the necessary "imprecision."

Roger, you forgot the next sentence.

"By nature, however, these institutions are nonatomistic, and therefore
concentrations of power -with ill consequences of their own -
can develop."

Doesn't this statement in itself refute the idea of market process? Especially any notion of the importance of discovery and incentives and time?

Mario -- as Kenneth Boulding pointed out in his JPE review of Samuelson's Foundations, mathematical economics with its flawless precision may prove incapable of capturing the world that is captured with the imprecise and vagueness of literary economics. That was Kenneth Boulding the John Bates Clark Award winner and in the pages of the JPE. Can you imagine that argument being made today by the equivalent and in an equivalent place?

Roger -- yes Akerlof makes those statements, but you have to admit that the emphasis of his argument is not on those market institutional responses. Sometimes these things turn on emphasis --- e.g., consider the role of public choice analysis in Krugman's JEP paper on Is Free Trade Passe? Would you say that Krugman "gets" public choice any more than Akerlof "gets" the entrepreneurial element of the market process, and how today's inefficiencies result in tomorrow's profit opportunity.

Emphasis matters, I'd argue.

Lee Kelly -- you need to work harder if you don't get it.

There is a big difference between the Carnap/Hempel formal construct model of science and the Kuhn social process picture of the growth of knowledge.

There is also a big difference between Carnap's formal syntax/formal semantics model of significance / language and the social institution / dynamic process picture of language of the later Wittgenstein.

Similarly Austin & Kelsen vs Hayek & Hume on the nature of law and rules.

I probably will never pick up on what is at stake reading blog comments.

Discovery & growth & learning go beyond a prior set of "givens" explicitly defined and related to other "givens" according to a given set of stipulated formal rules.

This insight is found in Kuhn, Hayek, Mises and Wittgenstein, among others.

What's not to get?

Brad --

It is a difference in theory versus a difference in historical depiction. Mises is talking about theory, and thus the old institutional emphasis on specific historical context (the market is a place) or the neoclassical theory of the market as a situation (e.g., competition as a noun), misses the point he is arguing about the market as an arena of exchange; as an active process of discovery and learning (e.g., competitive as a verb).

Do readers of this blog not puzzle over the issue of perfect competition and the transformation of economics in the 20th century (see Frank Machovec's excellent book) or the disappearance of the entrepreneur in economic theory (see the work of Humberto Barreto, The Entrepreneur in Microeconomic Theory)?

Learning via training & imitation as part of a dynamic & growing social institutional process that can't be captured in a formal construct is bedrock stuff in Hayek, Kuhn & Wittgenstein. (property rules & norms in the case of Hayek).

It's utterly different than a bit of math or logic typed up in a book which someone points to with an ----> and labels a "process".

On Akerlof:

Mark says Akerlof's "underlying assumption is that because of information asymmetries, markets fail." As stated, that's false. He has less confidence than Austrians in entrepreneurial innovation and more confidence in state control or regulation. As Pete correctly says, IMHO, it is matter of emphasis and not that he just. does. not. get. it.

Pete -
re: "This isn't the sky is blue. End state analysis isn't the same thing as process analysis. Again I point to Wagner's discussion of the difference between the neo-Walrasian and neo-Mengerian paradigm."

Right - and in terms of what takes up most attention in terms of the subject of study, I think clearly Austrians spend more time doing process studies. I find a lot of that work vague and hard to use or apply, but it is what it is. My contention was that I don't know any economists who don't treat the market as a process. When we study federal programs at my job, I mostly do the "impact analyses" and other people do the "process analyses". The fact that I have a comparative advantage in empirically assessing impacts doesn't mean I somehow pay lipservice to the existence of the process component. I simply don't know of any economists that don't view the market as a process and I'm not even sure how else you could view the market.

Mark -
I'd repeat to you what I said above to Pete. You are confusing treatments of specific problems and parsimony in modeling specific mechanisms with the idea that they don't think of the market as a process. I hear complaints like this from Austrians all the time about realism in models. Of course there are unrealistic elements. Parsimony in exposition is not the same thing as a simple-minded understanding of the market. This is why we have a frictionless plane in physics, and that's classical mechanics. The emergent order of the market is far more complicated than that. If you want to understand a certain mechanism in the market, you have to theorize it parsimoniously. What you're asking me to believe is that these models are perceived to be true and complete renditions of the economy. That is entirely absurd, Mark, and I've never heard anything remotely like that from anyone that actually uses these models.

Mark -
Akerlof also says quite clearly on the first page: "It should be perceived that in these markets social and private returns differ, and therefore, in some cases, government intervention may increase the welfare of all parties. Or private institutions may arise to take advantage of the potential increase in welfare which can accrue to all parties. By nature, however, these institutions are nonatomistic, and therefore concentrations of power - with ill consequence of their own - can develop"


1. Public intervention is one of many solutions
2. Private solutiosn emerge as well
3. All of these solutions introduce very realy risks.

Now, you're concerned because he doesn't talk about what you would have prefered he talk about. But what is really new or groundbreaking about talking about government failure? Nothing. Nothing at all. Everyone knows government failure. When he wrote this, information asymmetry issues were just starting to emerge in a rigorous way. So he produces a parsimonious model to illustrate one mechanism - no suggestion at all that this is somehow all the market is - just a model to illustrate a mechanism. And on top of that he has a big, bold disclaimer.

I really don't understand what the concern is. Must every paper on every mechanism of any interest have to go through all the avenues that might lead from it? This is silly. The man richly deserved his Nobel.

Oh I see you and Roger already went over that.

Roger left out the next sentence, but then you left out the prior sentence.

If you read both the sentence before and the sentence after it's clear he is talking about risks posed by bother public and private solutions. Which is legitimate, I think. If there weren't risks we wouldn't call these "market failures" and "government failures". We wouldn't even be talking about them.

I don't like the term "market failure" anyway. It's too value-laden for people, as the discussion here suggests. I prefer just refering to "externalities" or "asymmetries". There is no moral content to it - it's just a fact of life to be understood.

So to end where I started: if you don't think other economists think the market is a process what do they think it is?

I find this whole subject very interesting and timely. It's strange how posts here sometimes coincide with some weird tangential thoughts I have been working through (considering I am not an academic).

This week I read Husserl's Vienna Lecture as I am discussing it with some friends tomorrow evening (yeah, I know). When I read a piece like that, I also seek out additional articles about it to compare how other people are thinking about it. Well, in doing so, I came across a very interesting piece by Gabriel Zanotti from the Journal of Markets and Morality titled Intersujectivity, Subjectivism, Social Sciences, and the Austrian School of Economics.

The article seemed very germain to this discussion as the author tries, in part, to tie the understanding of meaning of "market" from Mises, Hayek and Kirzner, to the philosophy of Husserl, Schutz and Gadamer.

At least I found it interesting. Article is here:

"if you don't think other economists think the market is a process what do they think it is?"

It's demonstrable that many economists have mistaken a math construct for "the economy".

That's what they think it is -- its a set of interconnected math functions.

Some others go beyond that -- its a part of the world known via a set of math functions.

Others go even further -- its a part of the world known via a set of math functions whose values are specified by statistical data collection and analysis.


You said, "I hear complaints like this from Austrians all the time about realism in models. Of course there are unrealistic elements." As you probably realize I don't care for "Austrian" arguments that say, in effect,
"They. Just. Don't. Get. It."
I'm with you on that one. But most mainstream economists are still a bit weak on this point about intelligibility, which matters. If your abstraction causes the agents in your model to behave in unintelligible ways, then there is likely something wrong with the model. You've probably abstracted the prince from Hamlet. Exceptions to the rule exist, as Tom Stoppard illustrates. But you should generally have intelligible agents in your model. If I criticize your model for lack of intelligibility, it is no rejoinder to say that all models abstract from many elements of reality.

"Doesn't this statement in itself refute the idea of market process?"


Daniel, it is one thing to pay lip service to the fact the market is a process. But if you use tools that portray static situations, and if you become enamoured of them, then it is easy to forget this fact. Now, I quite agree with you that the mere use of static models is not problematic all by itself, and there are people who can use them while recalling their limits. But it is quite easy to forget those limits, and many economists do so.

In my opinion the problem is NOT with models. And it is NOT with the math, per se. It is with what Alfred North Whitehead called "the fallacy of misplaced concreteness", which mainstream economists are very guilty of. Echoing Greg's point "many economists have mistaken a math construct for 'the economy'". Go ahead, make a model as unrealistic and abstract as you want to (the ERE is such a model), but if it's not *relevant*, or what Roger would say, intelligible, then it's going to have a limited usefulness--just how limited is going to depend. This stuff turns on methodological issues and the purpose of economic analysis, which I think causes Austrians and mainstream neoclassicals to talk past each other sometimes. In any case, process analysis is situated in the real world and it seems that when economists are trying to apply their models they are forgetting (or at least downplaying) this essential component. Although it's probably a simplified conclusion, this to me is the crucial difference between the two.

I completely agree on these points of relevance and intelligibility. I'm not quite sure why this is what the discussion is turning on, because I would have assumed that would have been an uncontroversial expectation.

I can't tell whether Brad W. & Daniel are using "intelligible" as a term of art. I meant to give it a relatively precise definition above, following Machlup. Intelligibility in this sense is about "Verstehen." I don't think it is true that mainstream economists are fully on board with Verstehen, so I don't think you can quite say that the requirement of intelligibility is "uncontroversial." I hasten to point out, however, that it is a matter of degree and emphasis. The bad old days of super-formalistic "neoclassical orthodoxy" are gone, and we should probably stop criticizing "neoclassical economics" as if it had relevance today. With everything so much in flux, we should probably be a bit circumspect about what "they" supposedly do not understand. Nevertheless, it does seem to me that "Verstehen" or "intelligibility" is not appreciated and respected as much by mainstream economists as might suit us Austrian economists.


I concede your criticism regarding emphasis rather than failure. Our point was that Akerlof overstated the problem (the purpose of markets is to solve information problems, and see here ), understated the importance of markets in dealing with asymmetric information, especially risk and means of diversifying risk, and lastly, his concept of competition failed to emphasize discovery and incentives.

Given this, I also agree with your contention that much has changed in the profession over the past forty years (a lot just over the past ten) and don't want to hold such thinking so relevant today. I used this only to exemplify the distinction between end state and process thinking.

Koppl writes: "The bad old days of super-formalistic "neoclassical orthodoxy" are gone, and we should probably stop criticizing "neoclassical economics" as if it had relevance today."

Agreed. And I think many non-Austrians are very much interested in understanding the process. Sure, they might not go the route of Verstehen. But they, too, are trying to understand the process. Those that I have in mind build on Lucas (1986) "Adaptive Behavior and Economic Theory", so the aim is whether and to what extent human or, at least, less-than-hyperrational agents get from their starting point to the rational expectations equilibrium (REE). That is--whether their is a process whereby the REE can be obtained. Some approaches include evolutionary game theory, agent-based computational modeling, and experiments with human subjects. Do these approaches fully capture the sort of process Austrians have in mind? No. My reading thus far suggests they focus less on institutional filters (except for V. Smith, of course) and entrepreneurial alertness, and more on discovery and learning. One might criticize these thinkers on many grounds, but I don't think ignoring processes is one of them.

It's hard to characterize all mainstream economists because there are a lot of differences, but many mainstream economists try to include process in their analyses by analyzing the change from one state of equilibrium to another.

Seems to me that the real problems with mainstream econ are the lack of sound theories of money and capital, and as a result the lack of a decent theory of business cycles.

Many resurrected Keynes during the recent crisis and held him in the saddle with a board as the Christian armies did El Cid to frighten the Muslims. But doing so confessed that they had made no progress in their field in 80 years.

"The bad old days of super-formalistic "neoclassical orthodoxy" are gone"

Tell that to the DSGE macroeconomists!

The claim of Stephen Williamson is that DSGE math can't miss _anything_.

"It's in there" is his claim.

And don't bring up genuine uncertainty, or learning in the sense of changes in knowledge / understanding.

Frydman & Goldberg? Who are they?

Greg writes: "Tell that to the DSGE macroeconomists!"

My claim is not that ALL economists are concerned with processes. Rather, that "many non-Austrians are very much interested in understanding the process."

You are correct: some DSGE macroeconomists don't care about processes. So what? Let them build the equilibrium constructs. Others can consider whether there are processes such that the equilibria they point to can actually be achieved. It is just division of labor and exchange within the academy.

Greg writes: "And don't bring up genuine uncertainty, or learning in the sense of changes in knowledge / understanding."

Surely you don't believe that NO ONE in the mainstream (which I'd define as someone publishing in a top 10 journal) is asking questions about genuine uncertainty (as different than risk in that the underlying probability distribution is unknown).

Google scholar "ambiguity aversion" and you'll find research published in outlets like the QJE and Econometrica asking whether individuals are adverse to genuine uncertainty.

And as I mentioned above, there is a lot of work on learning. For example, almost every macro experiment considers whether subjects learn over the course of the experiment. It has not yet been as well received as the ambiguity research. But give it some time.


I thought that our discussion last week about the MEC is a good example of the issues here.

You said that the MEC according to Keynes definition can fall to zero. You said a lot about the entrepreneur in that discussion.

But, if there are entrepreneurs then they are people who expect that they can earn more than the interest rate from combining capital. The MEC of an economy is defined as the highest MEC of any asset. So, if there are entrepreneurs then the MEC is higher than the interest rate. So, how can either reach zero in the long-run?

What Will said. IMHO, BTW, DSGE is on its way out.

Stephen Williamson wants people like Pete removed from economic departments.

He wants history of economic gone (including contemporary history, e.g. Katrina history) and history of economic thought gone and "literary economics" gone.

He wants to achieve Kuhnian normal science by firing squad.

You are correct Will, I don't believe that.

But I do believe there is a built in professional self interest and an inherent human instinct to force everything back into a closed framework ultimately cashed out in non-opened ended, stipulated givens in a model.

There is also a builtin bias against simply getting the basic explanatory strategy right -- as Hayek said in 1937 most of the work is over once you identify learning as the explanatory cause - all there is left are details with diminishing returns.

But the professional reward mechanism is all about techmicalnadditions doing norm science with the diminishing returns work and there is NO professional reward for getting the global explanatory frame in non-pathological order.

Without Ernst Mayr and philosophers of biology, Darwinian biology would be in a somewhat similar situation -- only difference is the "details" normal science work has added tremendously to understanding the explanatory whole.

Not so much in economics.

Will and Roger are overstating, Greg is underestimating, and Daniel is ignoring the fact that differences in emphasis make mask significant issues of substance.

This business of doing economics is complicated.

Greg --

I much rather have Stephen Williamson that that Noah Smith grad student teaching macro!


As a superficial outside observer, I'm actually a fan of recent directions in NSGE modeling.

I have no beef with Williamson as a math construct builder.

I have problems with him as a would-be scientist providing non-pathological causal explanations of the real world, i.e. I have problems with him as an economist, in the genuine sense, not in the "he does math with entities labeled with economic words" sense.


Seriously: where does either of us overstate? (And, um, cough, cough, I"m still waiting to learn if I got you right on semantics vs. syntax.)

"Information asymmetries in this regard are not an inherent trait of markets that provide opportunities for entrepreneurs to rectify and from which to profit by doing so, they are obstacles to achieving some idealized end state and must therefore be corrected through regulation."

If you'll pardon a comment by a passer-by, this observation reminds me of that riddle where the professor tries to identify the smartest of three students by asking them to determine what color hat has been put on their head. The answer to the riddle turns on the passage of time after the question has been asked. From the fact that no student knows immediately what color hat he is wearing, one student deduces what color hat he is wearing.

The riddle never states precisely how long the winning student waited to determine that there was no clear answer. I always wonder how long I would wait. Our politics often boils down to questions of whether enough time has passed for us to act as if there is no obvious non-governmental solution to any given problem.

If there is an entrepreneurial solution to an information asymmetry (or Standard Oil), it will arise as quickly as contemporary entrepreneurial technologies make possible. At some point, the odds of a solution emerging decline, and the strategic wisdom of opting for regulation grows. We can disagree about how long to wait, but that's not really a matter of philosophy.

One needn't be from one school or another to adopt the view that some problems require government coordination, and that the best way of determining whether a private solution exists is to wait to see if one appears in the absence of constraints thereon. But we should wait only until a consensus emerges that a solution is not likely to arise (even if it might eventually). That strategic judgment, does not, it seems to me, depend on whether one views the market as a "process."

I agree with McKinney when he says above: "It's hard to characterize all mainstream economists because there are a lot of differences..." Nevertheless, I do think the central problem in the modeling of economic behavior is that agents are assumed to have knowledge that there is no way they could acquire. This is the problem with RE models. It is also the problem with Becker and Murphy's rational addiction model.

This issue transcends the difference between equilibrium and pseudo-process models.

This is the essence of Hayek knowledge-problem criticism of the original perfect competition model. But it is also endemic to the neoclassical paradigm as evidenced by the continual relevance of the problem in contemporary models as well.

Another thing... If the mainstream generally understand the points here then why are some of them so protective when others challenge their *models*.

Keynes, for example, made a few quite haphazard and sometimes quasi-static models of parts of the economy and linked them together. In the recent past (i.e. before the crisis) Keynesian economists were in the business of building more careful models to perform the same functions. Now though, the more high-profile ones seem to be in the business of defending exactly what Keynes said.

@Roger -- overstate with respect to how much "mainstream" guys have absorbed process analysis. I have a very restrictive definition of mainstream --- basically what goes on at MIT, Harvard, Chicago, Princeton, and Stanford, and then mimicked elsewhere and what one reads in the pages of the AER, JPE, QJE. All this computability stuff is on the edge, not in the core.

Now I know you know this, but I wonder how many others here have taken the time to read for example Robert Leonard's Von Neumann, Morgenstern and the Creation of Game Theory, and in particular the discussion of process and causal-genetic theory as opposed to functionalist and equilibrium bound theory, and what that means for novelty, creativity, and contextual knowledge.

Moreover, the comments got completely derailed from my original post which was the confusing of a discussion of theorists with the theory itself. The term process means something, it is not just a catch phrase and it relates to the way one does economics. And it has deep roots in the history of the discipline -- again see Leonard's brilliant book.

Mario, agreed, but its seems to me that DSGE modeling is a harmless game. It doesn't seem to me that much policy comes out of it. And if you challenge them on the reality of it they see willing to back down easily.

Look at where everyone jumped in the latest crisis. They fell back on Keynesian prose, not DSGE models, to justify their policies.

I guess I don't see the harm in DSGE models. It seems that when anyone wants to discuss reality they abandon the models for prose.

Here is an article I highly recommend to everyone here: "Organizing Far From Equilibrium: Non-LInear Change in Organizational Fields" -

I will note that the authors probably could have saved themselves a lot of time if they had known of Austrian economics, as they would not have had to discover process theory on their own. But their discoveries about the nature of organizations essentially parallels that already understood by Austrians.

I would argue that the difference between process theoretic economics and non-process theoretic economics has little to do with whether or not one "understands" that the economy is in fact a process, but whether your analysis involves the use of equilibrium in the sense that, to quote the above paper, it "constitutes a point of rest from which there is no endogenous tendency for any individual, firm, or market to change." As a result, they argue that equilibrium "is ill-suited to studying fields in flux." They further ask, what equilibrium can mean when there are "mergers, acquisitions, exists, and CEO turnovers"? More, they argue that "equilibrium-seeking" processes are "evidenced by deviation-absorbing negative feedback loops," which is no doubt a part of any complex process. But insofar as one's analysis assumes equilibrium, only negative feedback loops are being considered. However, there are also positive feedback loops, which by themselves can lock a "community into perpetual disequilibrium" because they amplify deviations. Both processes can be modeled, but what happens, as does in fact happen in real complex processes, if you have bipolar feedback, meaning both simultaneously? Good luck modeling that process (some effort has been made by Hector Sabelli, whose work I know Roger is less impressed by than I am).

To do mathematical modeling, the process has to "hold still" long enough for the math to be able to say something. It would be like if all we ever saw of elephants were photographs, and we studied elephants entirely from photographs. Sure, we "know" they move," but we would be limited to models of how they move. How right would we be? And what would we make of that trunk? Or of how the tusks were used? How to figure out what they eat? Or how they drink? In the end, all the models would be of a stationary elephant. There would be no real understanding of the elephant as a process, no matter what people said about their understanding that an elephant, in fact, is a process and does move, etc.

In the end, the authors of the linked paper pointed out that when one of the authors in a past project proposed to study change in hospitals, "A peer reviewer of our grant proposal asked, "what will you do if the environment does not change?"" The fact that as recently as 1987 someone seriously questioned whether or not the economic environment would change over a 4 year period is astonishing -- and telling. No doubt things have changed since 1987 -- but the fact that the authors of this paper felt they had to tell this story as recently as 2005 and to justify their process-oriented methodology says volumes.

Getting back to Pete's proposition about two intellectual positions in Austrian econonomics, such a difference should be expressed either in the kind of questions that are asked about the phenomena or in the kind of answers that are considered to be acceptable. Can Pete or someone who knows more economics that I do spell out the difference in those terms?

Barry Smith identified some Aristotelian metaphysical propositions in the core of Menger's economics and these are as near as dammit to the metaphysical program that Popper formulated in the philosophy of physics. That is why I think Popper is a huge supporter of Austrian economics in its concern with plans, processes and the emergence of novelties though the activity of entrepreneurs.

The comments to this entry are closed.

Our Books