The Macro-Institutional Delimitation of Economic Complexity

A while back, Brian Arthur of the Santa Fe Institute wrote a short piece on complexity economics (h/t Tom Hickey). I find much of the work Arthur and others have done on increasing returns, path dependence and related phenomena very interesting. My doctoral supervisor’s doctoral supervisor (there has to be a less cumbersome way of expressing that) was Paul David who also had an interest in path dependence and the significance of history in economic development. My supervisor’s influence motivated me to devote chapters in my thesis to, among other things, an application of the theory of path dependence to the distribution of earned income and a consideration of the implications of cognitive limitations within the context of path-dependent economic processes. My (orthodox) postgraduate supervisor’s commitment to cross-disciplinary research also motivated me to integrate insights from liberal moral philosophy, policy history and institutionalism (new and old) into the analysis.

Although very interesting and worthwhile, I think the basic outlook of complexity theory can be (though need not be) problematic if its emphasis on micro-level behavior is allowed to obscure the delimiting role of macro-level decisions, institutions and the socioeconomic system itself. The outlook, when this occurs, can become akin to the way economists traditionally (though less so now, at the cutting edge) depicted money as emerging spontaneously out of barter rather than recognizing its likely historical emergence in coercive debt relations and the central role played by the state, particularly through its imposition of tax and other obligations denominated in a state money. Certainly, those debt relations came out of prior micro-level and macro-level interactions, but once they are embodied in a society’s legislative and political framework, they significantly shape the parameters within which further behavior, complex or otherwise, plays out.

In the third paragraph of the linked article, Arthur writes:

To see how complexity economics works, think of the agents in the economy – consumers, firms, banks, investors – as buying and selling, producing, strategizing, and forecasting. From all this behavior markets form, prices form, trading patterns form: aggregate patterns form.

Arthur’s piece is only short, so not everything could have been covered, but I find this brief description of the economy quite telling. There is no state. Or, at least, the state is too insignificant to mention. There are no institutions. No system as such. There are just patterns that emerge out of spontaneous behavior.

Above all, in this description, the macro emerges out of the micro, and the (likely to be significant) impact of the macro on the micro is left out of view.

If we consider actual economies, this picture seems quite removed from reality. How, to take just one example, did we get suburbia? Was it the outcome of a discovery process of individual agents acting as if independently but with interdependent effects? One day they paused to take a look around and realized they’d created suburbia?

More likely, a relative handful of institutionally powerful actors, with influence inside the corridors of state power, took a decision with the intention to create a mass market for new housing and automobiles, with flow-on demand created for a variety of other consumer durables.

Within the overriding framework of suburbia, there was plenty of room for individual discovery processes to take place. But to suppose suburbia emerged out of spontaneous micro-level behavior would seem farfetched. The decisions taken were almost certainly deliberative, involved a rather small number of individuals and required exertion of state power to put in place. Once the tracks (roads) were laid down, the trains (demand for new housing, automobiles and household appliances) followed. And consumers “spontaneously” discovered a longing within themselves for these things.

The very basis of path dependence – a sensitivity to initial conditions – points to the critical role played by those who set the parameters within which spontaneity is given free play. The formation of those parameters is very far from a spontaneous micro-level activity. It is deliberative, planned and applied at the macro level with the intention of delimiting micro-level behavior, sometimes within quite narrow confines.

It is similar with much of economics.

Does unemployment emerge as the result of the effectiveness of, or interdependencies between, individual actions? No. It is created at the macro level. Job search and matching processes occur within that aggregate constraint, which is imposed by macro-level action.

Is inflation due to complexity of spontaneous behavior? Not really. Regardless of such complex behavior, inflation ultimately depends on macro-level decisions and actions.

Did basic infrastructure or the internet spontaneously emerge? Hardly. The state, with a view to long-term planning, was heavily involved.

The interdependent effects of micro-level discovery processes, which certainly play a significant and interesting role, do so within a framework set in place through macro-level decisions and actions. Within that framework, spontaneity is then free to occur.

I think the most important economic questions concern how to create a macro and institutional framework within which maximal scope is provided for the kind of free (including exploratory) individual and voluntary collective behavior we think should be facilitated or encouraged. In the ideal institutional framework, it would largely be unimportant what patterns of activity actually emerged. After all, those patterns of behavior would be emerging as the result of choices freely made within the framework, as expressions of agents’ desires.

When, in reality, patently negative behaviors are found to thrive within the prevailing framework, that points not only to micro level failures but above all to a flaw in the macro framework that needs to be addressed. Such flaws will not be most effectively removed by waiting for spontaneous changes in behavior at the micro level. The agents whose behavior is most responsible for the negative outcomes might have no personal interest in altering their behavior and might be perfectly satisfied with behaviors or outcomes that society as a whole regards as unsatisfactory. Key changes will be called for at the institutional and systemic levels. Coordination is required.

We can of course say in such a case that the need for macro-level change is revealed by the emergence of unsatisfactory behaviors at the micro level, and that appropriate changes at the macro level will ideally be responsive to, and so be affected by, what is happening at the micro level. But because the micro-level behavior is always embedded within a macro framework, the latter will need addressing.

None of this is to suggest that individual liberty should be given a low priority or that freedom for individual and group expression is not to be highly valued. The opposite is the intention. A central purpose of the institutional framework and macro-level decision-making should be to provide as much scope as possible for freedom of self-expression, subject only to the proviso that as far as reasonable such behavior should not be harmful to others. (I say “as far as reasonable” because not all conceivable types of harm should necessarily be avoided, such as, for example, the offended sensibilities of Amartya Sen’s Prude when learning of Lewd’s freedom to read a racy novel.)

The better the macro framework, the more freedom is opened up at the micro level for socially beneficial behavior and exploration.

We are not enslaved by the collective provision of public infrastructure or the internet. Quite the contrary. These assets dramatically broaden the scope for micro-level discovery and expression.

Similarly, the maintenance of full employment through appropriate macro measures would enable all individuals who desired a job to obtain one, while at the same time enabling the expression of individual preferences over consumption and saving. Only macro policy has the capacity to enable individual saving preferences to be expressed without impacting negatively on the employment prospects of others.

And so on.

Macro policy and macro institutions, appropriately designed, enhance space for spontaneity at the micro level.

4 thoughts on “The Macro-Institutional Delimitation of Economic Complexity

  1. hi
    You write at just the right time for me (path dependent reading?). I read Critical Mass a while back by Philip Ball. He relates similar questions say, about suburbia as you ask and looks at the work of urban geography in exploring these.

    The question about the macro vs micro has also been discussed in the critique of social theory e.g. structural functionalism, or Durkheims problem of trends in suicide. So we are not alone historically to be asking these questions.

    I think a problem does come from posing the ‘patterns’ as structure. Doing so displaces the implications of the micro simply to another level in modelling the complexity of things. If the structure as posed has effects then is it an actor too? If yes, why then do not the same path dependent vagaries & uncertainties apply to the structural actors as to the micro actors? Certainly the state, if we want to equate this with big structure (a problem for theory similarly?) would have other states, oligarchies, big-actor and even micro-actor ‘forces’ (a riot as an example) to more than create an environment of complexity, uncertainty and path dependence. Historians seem to play with these ‘forces’ all the time. Maybe we should look in greater detail at how history is studied too.

  2. “Complexity theory” is necessary for our understanding. Yet, the attempt “to derive macro from micro” on its grounds can be suspected, as other similar attempts, to hide the class or group nature of income and property distribution.

    In epistemological terms, the aim of deriving “macro” from “micro” can be understood as extending concepts and methods of physics to economics based on the following two ideas:
    – The behavior of some physical objects at “macro” scales, e.g. a star, can be explained with just the knowledge of “micro” relations, the interactions among nuclear particles. Furthermore, it is clear that “macro” behavior does not influence “micro” behavior: interactions among nuclear particles are not changed by them being inside a star, a planet or an interstellar gas.
    – This successful approximation in physics should give us explanatory success if applied in economics.

    If one stops to ask why it should give us explanatory success, the answer can only be because it gave us explanatory success in physics. This is not a great answer, but, anyway, extending the idea from physics to economics has as a plausible consequence:
    – The prevailing social arrangements and political decision-making on economic matters cannot be changed because they arise as a result of an uncountable number of micro-behaviors of people, micro-behaviors that cannot be changed themselves.

    The extraction of this consequence is of course most favorable for the social groups that benefit from the prevailing distribution of income and property.

    Now, on purely cognitive grounds, the problem with the argument is that not all phenomena can be successfully explained at macro scales by resorting only to micro “foundations”, not even in physics. Let us imagine a physicist that tries to explain the behavior of a computer that calculates the digits of the pi number. Knowledge of only the nuclear interactions of the particles that constitute the machine will allow her to understand some aspects of the machine’s operation, but are of no help to understand its purpose.

    Humans make computers that calculate the digits of pi so maybe the physicist goes along investigating these curious animals. They are supposedly curious for a physicist because he can observe that, differently from what happens in a star, the “micro” interactions among individuals are mostly dependent on “macro” determinations.

    “Macro” determinations can be described as the total population of humans separating in behavioral groups. The groups are not mutually exclusive, but they all have the common property of determining behavior for those individuals that perceive themselves in the group.

    Some groups are cooperating, some are indifferent and some are conflicting. Conflicts among groups are most of the time waged in language, exploring the inabilities of the human brain to process information, but sometimes they enter more energetic ways of argument like batons, bullets and bombs.

    It seems that in the XIX century, a bearded human, of his name Karl Marx, got for himself a lot of celebrity and hate by observing a type of group separation and conflict that he termed “class war”. The hate was granted because some groups learned long ago, as a famous Chinese general, that the best strategy to win a war is to preclude the enemy from knowing that a war is being waged on him.

    Despite having created a lot of agitation, Marx analysis was very coarse and very limited. Humans separate into much more groups than he considered, these groups evolve with time, and they present structured and evolving patterns of interaction intra and inter-groups.

    History of humankind is mainly a history of the successes of cooperation and the pitfalls of conflicts among groups. Because conflicts are mainly waged in language using what could be called information weapons, the study of the formation of groups and their interactions has not progressed what one could expect in light of progress in other knowledge areas. A main motivation for this comes from a central element of the social strategy of the groups getting highest income being to successfully obscure and hide the actual existence of group separation from the population in general.

    Yet, if humankind is to substantively evolve, then group separation of population, group structuring and evolution, interactions among groups must be clearly understood. This appears as a requisite to devise effective ways for limiting the deleterious effects of group conflicts, in particular of class war. In a sense, we already know how to arrive at this. We know how a reasonable minimum of dignity and wealth can be realized for all citizens in a society. We even dispose of what could be termed a quality standard or a quality norm for societies in written form: the UN Declaration of Human Rights.

    Has complexity theory and complexity economics a role in building such understanding? Surely, it has and already had, be it only by promoting an “organic concept of economics” subsuming the primitive idea of equilibrium or by developing an agent based modeling approach. But no substantive progress can be attained without considering, even if only implicitly, the fact that the groups fighting to maintain their higher incomes unavoidably resort to political maneuvers strongly hurting many other people and blocking human progress, be it a result of deliberation or of unconscious coordination.

  3. Hi Pete,

    “If we consider actual economies, this picture seems quite removed from reality. How, to take just one example, did we get suburbia? Was it the outcome of a discovery process of individual agents acting as if independently but with interdependent effects? One day they paused to take a look around and realized they’d created suburbia?

    “More likely, a relative handful of institutionally powerful actors, with influence inside the corridors of state power, took a decision with the intention to create a mass market for new housing and automobiles, with flow-on demand created for a variety of other consumer durables.”

    One of the criticisms I have seen raised against this line of reasoning comes from believers in methodological individualism. I want to make it clear before proceeding: I am NOT endorsing their views, I’m just trying to warn you of potential objections (or playing the Devil’s Advocate, if you like).

    Your passage above seems appropriate (as I had personal knowledge on the matter of urban development). Who are the “powerful actors”? This question admits different answers.

    At their most abstract (and it seems they often have a preference for the most abstract), methodological individualists would object that only individual human beings have agency. To say otherwise, they would add, is to believe in conspiracy theories and attributes human-like characteristics to things that have no cohesion, agency, and intentionality.

    A possible answer to that objection is that one can identify supra-individual entities which are SUPPOSED to exert influence on urban development. This implies they have some form of agency, cohesion and intentionality.

    That is, urban sprawl could be the result of urban development policies enacted by municipal and state/provincial planning offices: zoning, limits to building heights and urban density, environmental protection, transportation policies, street design and such.

    Entities like these are easily identified (although their “intentions” are not necessarily clear). The problem is that not all “powerful actors” are so easily identified, and their cohesion and intentions are often subject to controversies.

    Have you thought of an answer to these objections? (Again: I’m not trying to fault your reasoning, which I largely share, I’m just trying to provide a different perspective)

Leave a Reply

Your email address will not be published. Required fields are marked *