Understanding Markets: Point/Counter-Point

July 22, 2010

by Thomas McQuade and Chidem Kurdas

Though it should be obvious to all that markets are of immense benefit to humanity, any appreciation of these institutions is almost always hedged with a perceived need to constrain and regulate—in short, to subject them to conscious outside control.  The reasoning is understandable: the unconstrained pursuit of self-interest can only lead to chaos.

But the preference for constraint through centralized direction betrays a profound misunderstanding of the way markets work.   

Can we explain that claim any better than the volumes already written on the topic?  We find that, when we discuss the issue, we agree on the basics, but differ in emphasis and details—and details matter.  Here is part of our discussion, in point/counter-point format.

Thomas: I’d tackle the issue by emphasizing the futility of conscious control of complex systems.  This is a Hayekian knowledge argument, but I’ll augment it by pointing to an interesting similarity between markets and scientific communities.

The basic interaction in markets is the exchange of goods and services.  This relies on supporting arrangements—institutions such as property, contract enforcement, goods production and money. Those supporting institutions are a separate topic or, rather, many topics.

The market system itself is a large and highly interconnected system of exchanges, capable of generating beneficial effects unintended by the participants.  The latter’s behavior is nonetheless shaped by feedback from these effects.  And it is an open system, subject to and adaptive to changes, including attempts at outside control.

The recognition of some social systems as complex arrangements whose beneficial effects are the unintended consequence of people pursuing personal goals is an extremely important but seriously underappreciated scientific discovery.  The idea has a long history among scholars with a classical liberal bent, from Bernard Mandeville, David Hume and Adam Smith in the 1700s to Hayek in the 20th century.  But it is counterintuitive—surely, to generate societal benefits, people have to act with the intent to produce these benefits or to be constrained to act as if they do.

Not necessarily.  There exist stable social arrangements in which repeated self-interested interactions generate as side effects certain societal benefits.  At the same time, the effects feed back to constrain and channel the behavior of the participants.  It is a characteristic of such systems that they adapt, in unforeseeable ways, to attempts from the outside to alter their functioning.

Take science, for example.  Modern science – meaning the system of interactions between scientists that has predominated since its beginnings in the Royal Society of the late 1600s – does not rely on practitioners being dispassionate seekers of truth.  It relies rather on the unhindered operation of the core institutions of publication and citation, on the self-interest of participating scientists in making use of contributions to further their own research, and on reputations built by producing work useful to peers.

There is no central control, yet these interactions lead to the generation of a relatively stable body of scientific knowledge while allowing for a wide range of personal motives.  Attempts at central control lead to perverse results, Lysenkoism being the most notorious example.  Sensibly, there is no serious call for direct outside intervention in the workings of science.

Market interactions are different from those in science, but the systemic operation is similar.  Here, repeated interactions generate market knowledge in the form of prices for a wide range of goods and services.  This benefits the participants by exposing opportunities for consumption, production, and exchange that would otherwise be difficult to discern.  The system allows for a variety of motivations for interacting.  And it generates feedback about the desirability of particular goods and services in the form of reputation, profit, and accumulated wealth.

In science, where concerns for “fairness” or “equality” do not have the same emotional tug, the inappropriateness of outside intervention is easier to appreciate.  But the similarity should suggest that the argument against intervention in science applies with equal force to markets.

Chidem: That’s a fine explanation of how self-directed systems like scientific communities and markets work.  But it ignores the creative destruction caused by innovation—or in the case of science, by paradigm shifts.  Capitalism as a whole has thrived by creating new markets and destroying existing ones.  To fully understand how it functions, one has to consult Joseph Schumpeter as well as Hayek.

Simple markets have been around for millennia, but economies were largely stagnant over the long haul.  Political powers-that-be controlled economic activity and did not leave scope for much innovation—indeed often punished innovators.  Economies grew and shrank without enjoying the permanent gains in affluence that we take for granted.  Societies were not only materially impoverished but politically downtrodden as well.

What makes advanced capitalist systems distinctive is the profusion of new products, technologies and organizations.  The decentralized structure of market economies was essential for the kind of innovation that accounts for the progress of the past several centuries.  We owe our well-being to not having a single power controlling the economy.

People complain that nobody is in charge—but economies overseen by a central authority remain bereft of productive innovation.  They may build pyramids or, in the modern example of the Soviet Union, send men to space, but don’t create mass affluence.  Neither do they allow the exercise of liberty.

Classical liberals have long highlighted the link between central economic control and general lack of freedom—Milton Friedman perhaps expressed it best.  We should not forget that a dearth of economically meaningful innovation goes with central control.  Markets are complex systems with feedbacks and adaption mechanisms, as you show.  That picture should be complemented with the understanding of how markets relate to freedom and the right to try out new ideas.  … to be continued …

17 Responses to “Understanding Markets: Point/Counter-Point”

  1. Troy Camplin Says:

    I don’t see the disagreement here. The complex adaptive systems/spontaneous orders Thomas discusses are maintained by creative destruction. They are known as “dissipative structures”, and they exhibit such features as global patterns and emergence. Such systems have maximum degrees of freedom and are maximally creative (see the work of Stuart Kauffman on the latter point).

  2. Richard Ebeling Says:

    I’m afraid that I, too, do not see the point-counterpoint, in terms of rationales for limiting markets through regulations.

    The two positions presented are more complements than criticisms of each other as explanations of positive features of the market process.

    The basis of the “Austrian” view for free markets vs. the Neo-Classical “doubts” about the market has more to do with the latter’s equilibrium view of necessary conditions for complete markets with full information that, then, asserts that since existing markets fail to pass this “test” there is a need for government to correct and fill in these “market failure” gaps with regulations.

    (There is an article by Frank Hahn in the 1980s, which appeared in “Lloyd’s Bank Review,” that presented one of the clearest statements of this in the context of claiming that Adam Smith’s case for the invisible hand fails any test of justification due to incomplete markets.)

    Another essential argument by mainstream economists on the “limits” of the market relate to “equity.” That is, that the outcomes of the market are “socially unjust” due to the arbitrariness of initial endowments; or asymmetric information that places some at a disadvantage relative to others; or arbitrary degrees of “market power” that place some, again, at an inherent disadvantage in terms of the relative income share they can earn.

    These are among the leading reasons why many economists who, otherwise, say they appreciate the “efficiency” of markets, consider unregulated markets or unredistributed income shares are examples of “market failure” that need correction so “real markets” can come closer to some version of the Neo-Classical model of an “efficient” and “perfect” market.

    Richard Ebeling

  3. Seth Says:

    I too do not understand the disagreement.

    I did like this: “Political powers-that-be controlled economic activity and did not leave scope for much innovation—indeed often punished innovators.”

    I think that’s at work within businesses. I’ve seen that very thing damage good business. The business is run for the benefit of the politicians in charge and innovation is squelched, which means the new revenue pipeline eventually dries up and the politicians are left holding the bag on an obsolete value proposition.

  4. chidemkurdas Says:

    re Troy’s “The complex adaptive systems/spontaneous orders Thomas discusses are maintained by creative destruction.” Yes, well put.

    But there are different mechanisms involved. There are reinforcing effects — positive feedback effects — to given technology, organizational structures, institutions etc. Then there are changes in this data that do not initially get reinforcement, only after they become established.

  5. chidemkurdas Says:

    re Richard “The two positions presented are more complements than criticisms of each other as explanations of positive features of the market process.” True, they are complementary. But the tendency is to look at one, usually markets working with given technology, organizational structures etc

  6. Thomas McQuade Says:

    I agree that the points made are complementary rather than contradictory. Innovation and “creative destruction” are not only compatible with a view of markets (and science) as adaptive systems, but are (as Troy notes) an integral part of the process of adaptation. The difference is in emphasis: Chidem wants to emphasize the value of individual liberty by noting the correlation between it and societal affluence, while I would rather stress the understanding of the (very particular) types of context in which such liberty is both constrained and co-opted into producing generally beneficial emergent results from arrangements capable of adaptive growth.

    Richard is of course correct in his setting out of the essentials of Austrian-neoclassical disagreement. But I can’t counter those mainstream arguments any better than has already been done, over and over. The adaptive systems approach, besides having some current cachet and much potential for interesting development, seems to me to provide a very good context in which Hayekian ideas could be “discovered” anew. There’s the (small) possibility that changing to a more friendly arena might work better than slugging it out in the same hostile ring.


  7. [...] Organizations Fail Thomas McQuade and Chidem Kurdas have a nice post at ThinkMarkets called Understanding Markets: Point/Counter-Point.  I’m sucker for points and counter-points.  I love that stuff. I don’t think we have [...]

  8. Troy Camplin Says:

    Chidem,

    “But there are different mechanisms involved. There are reinforcing effects — positive feedback effects — to given technology, organizational structures, institutions etc. Then there are changes in this data that do not initially get reinforcement, only after they become established.”

    Yes, but that is also a description of complex adaptive systems — especially creative systems. There is a theory called Bios Theory that is a mathematical theory of creativity related to chaos theory. In it bipolar feedback (simultaneous positive and negative feedback) results in creativity. This is a recent further development of complex adaptive systems that further reinforces Hayek’s spontaneous order economics. So, again, it’s really just complementary.

  9. Roger Koppl Says:

    Thomas,

    Your comments are rather general, as they must be in this sort of context. But that leaves me wondering how extreme an argument you are making. Does your criticism of “central control” extend to social insurance and “regulation”? If so, do we have any examples of large complicated economies that are rich by today’s standard and yet entirely free of such “interventions”? If not, would not the same arguments about our epistemic limits suggest that we do not really know what a totally laissez-faire economy would look like or whether it would really trump the messy reality of mixed economies? (I am not defending the status quo in all its particulars, but I am a defender of Humean status quo bias. Like Hume, I feel violent changes in the political system. You know: piecemeal social engineering is the best we can do.)

  10. Roger Koppl Says:

    I *fear* violent changes . . .

  11. chidemkurdas Says:

    Troy, yes, “that is also a description of complex adaptive systems — especially creative systems.”

    My point, though, is that there is a tendency — which Thomas’ original paper exemplified — to focus on the positive feedback effects. Markets have such mechanisms but just as important, indeed Schumpeter persuasively argued more important, are developments that are not typically subject to positive feedback because the product, technology or organization is novel.

  12. chidemkurdas Says:

    Re complex systems: concepts from physics/biological sciences/mathematics are suggestive metaphors for understanding social orders but on their own insufficient to catch the latter’s distinctive characteristics. The ideas being borrowed now are of course different from the old-fashioned physics that inspired economists in the mid-20th century and still dominates mainstream economics today. Yet the danger is the same–mechanistic applications that leave out key social features.

  13. Thomas McQuade Says:

    Roger, my argument (general and sketchy as it is) is indeed extreme, in your sense. But it is not a proof of anything; it is an attempt to interest readers in a relatively new way of looking at social arrangements, one in which ideas of process, knowledge generation, the value of individual liberty, and the deleterious effects of mechanistic interference would seem to me to find a natural and fertile home. That’s all; it may or may not pan out – science is not predictable.

    One thing that it says nothing about is policy, if by that one means advice to the people in power under the current arrangements as to how they could intervene (or not) better. As for that, I am as Humean as thou in fearing violent changes – but, I suspect, for somewhat different reasons. I see the current political arrangements, with their strong concentration of power, as fundamentally flawed, but I would prefer them to die slowly of irrelevance rather than with a bang of rebellion, chaos, or worse. In the meantime, I am all for piecemeal defensive measures to limit the damage.

  14. Roger Koppl Says:

    Thanks for that helpful reply, Thomas. It makes me wonder if our views aren’t closer than I had previously imagined.

  15. Thomas McQuade Says:

    Chidem, I wonder if you and I have completely different understandings of particular words, like “mechanistic” and “feedback”.

    A mechanism is a deterministic arrangement – where output is predictable given the input and the current state. That’s precisely what adaptive systems aren’t, as inputs in such systems can generate a complex train of self-modifying reactions. So, if I am missing “key social features”, it is not because a mechanistic model is being invoked.

    Novel inputs, in markets as in science, may or may not generate follow-on interactions from those who notice them; and these may or may not be sufficient to make a more lasting alteration to the structure of the systemic interactions. This in itself is not feedback. Feedback is when downstream effects of such interactions loop back to affect those interactions. In general, innovations that make an impression will be subject to both negative and positive feedback, and I don’t see why the mere fact of novelty would preclude positive feedback in this case. Both types of feedback are important, and I am surprised that I appear to have emphasized one over the other.

    If you think the underlying problem stems from the impossibility of predicting the responses of specific individuals, then that’s where the economist’s insight comes in – acknowledging this impossibility, we concentrate on incentives, and talk about changes at the margin. This seems to me to be quite compatible with a (non-mechanistic) “systems” viewpoint.

  16. chidemkurdas Says:

    Thomas,
    What I mean by mechanistic is the tendency to describe how a social order works as if its similarity to a complex adaptive system is enough to make it always work like a complex adaptive system. Useful as this may be as a first approximation, it runs the risk of becoming yet another version of the physics-inspiration that led mainstream economics astray.

    No doubt incentives are important but at the same time human behavior has peculiarities such as being prone to certain biases and subject to path-dependent rules.

    Of course you’re perfectly justified to ignore behavioral specifics when analyzing an issue that does not depend on them.

  17. Troy Camplin Says:

    Complex adaptive systems are agent-based. Thus, the interactions of the agents as they actually interact are taken into account. Thus, behavioral specifics are taken into consideration. The end result is that it acts like a complex adaptive system because it is a complex adaptive system. But each one acts in ways that are unique to its particular structure, while at the same time acting in CAS-specific ways. There is thus an order there, but it is a statistical order, and it is one combined with chaos. The very nature of CAS’s undermines any sort of mechanistic understanding of the economy — or of anything described by CAS theory.

    In the end, spontaneous orders are CAS’s. There is no difference between the two ideas. The latter only has some mathematical modeling, but such models cannot be mistaken for reality, as too much of the math we use now is. CAS theory is a theory appropriate to the level of complexity of biology, psychology, and society/economy/culture, with the only difference being that the agents become more independent at each level. That is important, but by definition it is taken into consideration in the models.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 1,716 other followers

%d bloggers like this: