Thursday, June 16, 2011

The Market Process and Blue Skies

Peter Boettke has an interesting, if unconvincing, discussion of the Austrian distinction in terms of thinking of the market as a process. He provides an excellent Mises passage: "The market is not a place, a thing, or a collective entity. The market is a process, actuated by the interplay of the actions of the various individuals cooperating under the division of labor".

I asked what he thinks other economists think the market is if not a process, and I'm still not quite sure what his answer to that is. I certainly view it as a process. I'm not even sure what a theoretically viable alternative answer would be!

The discussion takes two diversions that I think are unhelpful, but I suppose might shed light on what he means:

1. They discuss the sorts of issues economists look into
I would agree that Austrians talk about the market process itself a lot more than most mainstream economists. This is true. But I don't think you can draw the same conclusions from that many of the commenters try to draw. At work we evaluate a lot of federal programs. There are two major components to an evaluation: an impact analysis, and a process analysis. I mostly do the statistical work on the impact analysis. What actually happens at these federal programs is usually a black box to me (sometimes we empirically model the process itself, but usually not - we just get at it through interviews and focus groups). My job is just to use econometrics to produce a defensible estimate of the impact a program had. Now, the fact that this is what I spend a lot of my time working on that doesn't mean I don't understand there is a very complex process that is going on in these programs. It's just a scientific division of labor. I'm better at running the numbers. I've interviewed for the process studies on a couple of occasions and it was a good experience, but it wasn't really my strong suit. I'm the only junior staff in my center right now that knows how to implement a regression discontinuity design or do propensity score matching. It figures they assign me to do that stuff.

Evolution is fundamentally a genetic process. There are lots of biologists that aren't geneticists. It doesn't mean they don't accept or understand that evolution is a genetic process. Weather is fundamentally just classical mechanics - it's physics. Meteorologists aren't primarily Newtonian physicists, but this doesn't mean they don't understand the nature of that underlying process.

Where does this market process focus get you, though? Not very far in my opinion. If that's an interesting question to them, that's great. I think there are other more fruitful things to look into, and you can get a lot of mileage answering the really interesting questions out of the parsimony of neoclassical models than you can out of the richness of a poetic prose treatment of discovery, surprise, etc. I have a lot of questions I find interesting and want to answer in my career (1.) why is there excess worker reallocation, and how does it vary over the business cycle? (2.) what role do worker flows play in price level adjustment?, (3.) what is the impact of the job creation tax credits that have been implemented?, (4.) how do scientists and engineers select into technical and managerial occupations?, (5.) why are African Americans so much worse of economically in the United States? (6.) what happened to the American economy between 1919 and 1922?, etc. etc. I don't see how process theorists like Mises or Kirzner can really help me with most of these problems. There could be some applications, but it doesn't seem like the best way to answer the questions that I find truly interesting. Not to mention the fact that market process theory doesn't have any insights into the intellectual history that interest me - the intellectual history of Keynesianism.

What I find scientifically interesting seems like an odd metric for whether I think the market is a process or not.

2. They discuss modeling
The other thing that came up in the comment thread was modeling. Akerlof completely oversimplifies things and never talks about alternatives therefore he doesn't see how complex the market is and that the market is actually a process, etc. etc.. You all have heard the claim before. I don't know why this is such a stumbling block for people. Human society is incredible complex and if we want to understand it scientifically we have to explore questions parsimoniously. Parsimony in exposition is not the same thing as an oversimplified understanding. Yes, models simplify things to highlight specific mechanisms operating in the economy. This is a feature, not a bug. People need to deal with it. We've known about government failure and market efficiency for centuries. This is the foundation of the discipline. Everyone knows this stuff. When Akerlof and Stiglitz were starting to write about information asymmetries that was a genuinely new insight and mechanism they were exploring. Constructing simple models to illustrate the process and understanding the role it plays was excellent science on their parts. It certainly had a higher marginal benefit than someone who spent their time telling us more about what we already know about government failure or market efficiency.


I called this post "the market process and blue skies" because this post from Boettke reminded me of the debate that he had with Bryan Caplan a while back (part 1 is here). I don't always agree with Caplan, but I thought he did incredibly well in this debate - particularly in the debate portion (I was less impressed with the performance during the Q&A, but I imagine that's a lot harder to do). Anyway, one of the things Caplan says is that a lot of Austrian claims are of the "Hayek said the sky is blue" variety. There are lots of claims that Austrians make that basically everyone agrees with and has agreed with even before there was an Austrian school. But if Hayek said it at some point it somehow becomes owned by the Austrians. The market is a process, values are subjective, socialism doesn't work, etc. etc. Anway, it's good stuff.


  1. I certainly view it as a process. I'm not even sure what a theoretically viable alternative answer would be!

    Process orientedness is essentially teleological. The economy exists for some "purpose" and it "functions" a certain way.

    An alternative perspective is network oriented. It is without objective teleology and purpose; the economy doesn't "exist" for any "reason."

    Just as language doesn't "exist" for any "reason" or purpose, neither does the market. Both language and the market are emergent from, and can only be understood from, individual people. That's why I object to depictions of the economy as a process or system. They're useful metaphors when taken intelligently but do harm when taken beyond their application.

  2. On an unrelated note, I was wondering if I could have your email address Dan. I couldn't find it anywhere and was hoping to ask you something. Keep up the good work on your blog!

  3. 'Just as language doesn't "exist" for any "reason" or purpose, neither does the market.'

    Wow, I'd say they both exist for quite obvious reasons: language in order to communicate and the market to promote economic flourishing.

  4. why is process orientation necessarily teleological?

  5. I read Boettke's post very carefully I honestly could not find where he said that mainstream economists don't think the market is a process. Could you at least cite where you're getting this from?

    What I *did* read was that Austrians are more focused on *process analysis* - which has *nothing to do* with saying only Austrians think the market is a process. The mainstream, on the other hand, is more focused on end-state analysis - again nothing to do with whether they would define the market as a process. I'm not sure how you get what your claiming out of this, but I'd love to hear an explanation. I guess if you like to stir things up for nothing........

  6. @Gene: those purposes are discovered ex post - nobody sat down and thought "We need a method of communication, so let's create language!" Contrast this with the Wright brothers having a goal (to fly) and devising a method to achieve that goal (create an airplane) - there's clearly an ex ante purpose there that languages and markets don't have.

    So you're sort of right: markets exist to promote economic flourishing, but that's like saying jagged rocks exist so that animals can scratch themselves - that may be what they accomplish, but nobody sat down and designed it that way. (You may want to point to situations where market economies were established by force, but Mattheus is quite clearly talking about a Hayekian spontaneous order-conception of markets.)

  7. edarinw-
    He wrote "It is this analytical focus on process that sets Mises and Hayek apart from their peers in economics."

    This alone is somewhat ambiguous, of course. Is it only that they are more process theorists or - given the context of the Mises quote - is Pete also saying that they distinguish themselves in seeing the market as a process?

    If you read the comment section it becomes quite clear - he's saying both.

    I would agree they theorize the process more (as I say in this post). I would disagree that they are somehow unique in understanding the market as a process.

  8. Who says that language is an example of spontaneous order, or that it has no purpose? As I understand it, most linguists are Chomskyan and all linguists believe that language is innate to a large degree. They no longer believe in the behaviorist approach, which has been thoroughly discredited. Very few people, except a few fringe movements, take behaviorism seriously. Pinker gives the argument if we were blank slates then there wouldn't have been a point at which we could have "learned" the language to begin with as there was no higher force to impose it on us. Furthermore, language is localized in the broca and Wernicke's areas of the brain.

    This is a way for Peter Boettke and Steven Horowitz to sit around on their prodigious asses and claim there is a connection between various fields and Hayek when there isn't. So if somebody says something that sounds like something Hayek said, Hayek invented it. If Boettke thinks that there is a connection between a group of words used in one field and a group of words used by Hayek, Hayek did the work, etc. This is as crazy as claiming Mises invented a time machine and set the foundations for geometry.


  9. Since the audience is a little different here I thought I'd repost what I said on Coordination Problem.

    The issue Pete is drawing attention to is that some economists may understand that the markey is a process, but they don't fully follow through on that line of thinking. I thought the discussion last week about the MEC is a good example of the issues here.

    Daniel said that the MEC according to Keynes definition can fall to zero. He also 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?

  10. Successfulbuild,

    Greg Ransom sometimes makes the idea of evolution of language idea sound more complex than it really is.

    It's not primarily about the understanding of language, it's about the concrete language.

    In England, for example what Americans call "potato chips" are called "crisps". And what American's call "fries" are called "chips". Each language speaker will use the word that is least likely to cause confusion. Suppose, for example, if 40% of people in a place recognise the word "foo" for something, and 60% of people recognise the word "bar" for the same thing. In that case a speaker who knows both words will tend towards using "bar" rather than "foo". By this kind of process languages become more homogenous. Then there are opposing processes, such as words that more clearly describe concepts and things. Those may be more useful to speakers even if if fewer people understand and they sometimes have to be explained.

    I have no idea if Austrian economists were the first to suggest this, I doubt they were. But I think they were the first to see the similarity with indirect exchange and money.

    This idea is compatible with Chomsky's theory of language comprehension AFAICT.

  11. Wow, I'd say they both exist for quite obvious reasons: language in order to communicate and the market to promote economic flourishing.

    They accomplish certain ends - but that's quite different than saying they exist for Reasons and Purposes like how Plato thought everything existed. Like Watoosh said - there is a dramatic difference between processes that were designed and implemented consciously on the one hand, and networks of actions designed by nobody that nevertheless accomplish goals on the other hand.

    why is process orientation necessarily teleological?

    Because all processes (and systems too) exist and come into being for a specific purpose. They are ex-ante considerations. Processes require linear thinking and goal-orientation: To get from A to Z. The market on the other hand isn't linear and it's not designed to accomplish any goal, or maximize a criterion. It's a case of invention or discovery.

    I have no idea if Austrian economists were the first to suggest this, I doubt they were. But I think they were the first to see the similarity with indirect exchange and money.

    This is the general understanding of how language changes over time and place, and it follows a lot of what Hayek said about emergent order and spontaneous networks in an economy.

  12. re: "Because all processes (and systems too) exist and come into being for a specific purpose. They are ex-ante considerations."

    Restating the claim doesn't answer the question, Mattheus. We understand perfectly well what your claim is.

  13. Daniel, ex-ante considerations are teleological.

    tel·e·ol·o·gy noun \ˌte-lē-ˈä-lə-jē, ˌtē-\
    Definition of TELEOLOGY

    : the fact or character attributed to nature or natural processes of being directed toward an end or shaped by a purpose

    How is this confusing?

  14. Mattheus - we asked you why processes are teleological - why you associate that noun with teleology. And you responded by providing us with a synonym of teleology - by restating your claim.

    Clearly you've attached a lot of insinuation to the word "process" that argosyjones and I are not aware of. We were hoping you would explain what that was.

    You've done this before with stressing that the market is a network rather than an institution, as if there is some obvious reason why those are two conflicting nouns.

    Rather than just repeating your claim, can you justify it? Or else we can all just agree this is a matter of you adding a lot of baggage and implications to specific words.

    Implicit in my question and argosyjones's question, I think, is that we agree the market is not teleological (although it's true human action is purposive).

  15. Mattheus,

    When we say "process" in this context we mean a series of interlinked things that happen. We don't mean a series of things that happen because one person plans them to happen.

    We mean it in the sense of "the process of decay" rather than in the sense of "devise a process for homogenising milk".

  16. No. All actions are not purposeful. This has been verified by cognitive science. Our brains are wired to react instinctively towards certain things because of the process of evolution. The areas of the brain dealing with fear and emotion also dominate the areas of the brain dealing with reason, although the reverse is not true: reasoning rarely plays a role in the structures of the brain that deal with fear. This is so we can react immediately towards threats without having to reason them out.

    This leads neurologists such as Dr. Vilayanur S. Ramachandran to say actions are really post-hoc rationalizations done for reasons other than what we think. If we are not in control of our actions and the unconscious mind plays a large effect, then actions cannot be said to be purposeful. Purposeful implies a direct reason and emotions require no reason to act. There are general rules for action and reaction, such as the conditioned reflexes and so on, but these are not axioms and are not universal.

    Thus, austrian economics is false and bad methodology. Models that persist humans are inherently rational (such as neo-classical economics) also are highly suspect. Neo-classicals respond that these are just "models" but models generally don't contain assumptions that are very much false. There's also the fact that Dennett and others take a more mechanistic view of human nature -- that it's all just neurons firing etc. This also doesn't give much credence to the Austrian philosophy.

    But Austrian economics is not only empirically bankrupt, it's triply false by things based on what we know:

  17. 1. Not all actions are purposeful (see above), have ends, etc.

    2. What we observe inside ourselves is not directly transferable to other people. We can't tell whether there is a decision or merely a reaction to stimuli when another person does something. You cannot get into the minds of other people just because you yourself believe that you think. Thus, "I think, act, etc." doesn't transfer over to all people think, act, etc. based on my thoughts alone.

    What humans perceive in themselves is simply not part of science. Psychologists generally make the distinction between what is perceived within yourself from what you can observe in other humans. This is basic logic. For example, speaking of physics and perception, if we were to try and discover what state of body someone is in when they have a toothache, we cannot do it based on what they tell us alone; it cannot be based on description alone. People with cavities may say they have a toothache, but they may not actually have it. Saying you have a toothache and actually having it are two different things; if they weren't, then you could cure a toothache merely by saying you didn't have one. A dentist would agree this is impossible. Psychologists stress that studies must be reproducible so they can be sure humans are reacting to certain stimuli. Even if I ask someone to pass the table salt, and he passes it, I'm not certain he reacted to my request. He may have just passed it at that moment. Hence, the need for experiments to be reproducible. Psychology is closer to science than economics because it is about something that exists in the natural world: people, and how they behave. It is not just about how they behave in a human constructed system, like economics.

    3. There is no such thing in science as applying axioms and starting from there. There is no such thing as pure certainty in science and contradictions are OK. Thus, statements like "All actions are purposeful" can't be taken seriously as facts in science are ranked more according to their probability of being true -- they never become "axioms" or anything else.

    Now we know that Austrian economics is bad methodology and contrary to science.

    But it's also bad philosophy. The action axiom pre-packages fallacious assumptions things that aren't necessarily present, such as consciousness, free-will, rationality, and it discounts any notion of us being driven by casual forces. Here is Harris' argument against free-will:


  18. I don't really agree with Mises epistemology either. That said, it's more defensible than successfulbuild claims. To Mises acting "rationally" is something much more restrictive than to modern economists.


All anonymous comments will be deleted. Consistent pseudonyms are fine.