Capitalism, System Theory, and Complex Adaptive Systems
Copyright © 2005, David A. Epstein.
All Rights Reserved.

November 5, 2005
(in progress)

Note: The following is primarily a theoretical discussion about the merits of free enterprise, not an emphasis upon what has actually transpired in reality. Our modern day economy, for better or worse, is a mixture of State Capitalism, a controlled marketplace, Corporatism, and small business resourcefulness. The arguments below represent the contributions of the pure free market to the entire modern-day system. And after the financial meltdown of 2008, there are certainly sound counterarguments to be made against unfettered Laisez Faire Capitalism.

Capitalism is founded upon sound economic and philosophical principles that not only honor individual liberty and the free exchange of goods and services, but also promote a viable social order that is both just and efficient. The basic tenets of Capitalism hold up remarkably well when an unrestricted free market operates within an integrated socio-economic system; yet the fundamental operations of Capitalist enterprise itself helped, and still help, to create this system environment that it comfortably operates within. But how is it possible for something to assist in the construction of the environment in which it currently resides? This appears to contradict the laws of nature. After all, a building must first have been built before people can occupy it, and only after such an occupancy will the interactions of people necessarily follow. The answers to this question, and many more that will be entertained in this article, lie within the confines of System Theory, and Complex Adaptive Systems in particular. My intention is to sufficiently demonstrate that Capitalism is an economic system that conforms to the principle axioms of System Theory, and not only is compatible with the modern paradigm of the Complex Adaptive System, but in fact thrives and prospers from the potency inherent in that model.

System Theory and Capitalism

As I've stated before in another article, every working system must have the following:

1) inputs,
2) transitional or intermediate states,
3) outputs.

While this is certainly not the total sum of any given system, it is the core of it. Let us describe each of these three ingredients of the economic system known as Capitalism.

1) The inputs to the Capitalist system include land, resources, labor, capital, machinery, money, surveys of potential customer preferences, product ideas, and human ingenuity. In a pure Capitalist society, other than natural limitations of available resources, there are no shortages of inputs into the system. Artificial barriers of a controlled marketplace are the primary obstacles to a consistent flow of necessary inputs for production, services and consumption.

2) Transitional or intermediate states include the various phases of product development and development projects, changes in production or project scheduling, economic fluctuations affecting capital formation or the labor process, and new customer demands for additions or changes to the finalized product before it reaches the marketplace. These are "in-progress" artifacts of the Capitalist system; most of these are invisible to consumers, but they are vital "stepping stones" for creating a successful product offering. Factors such as prudent resource allocation, sound economic decision making, the ability to overcome obstacles, and appropriate readjustment of production levels are engendered to navigate the output into a propitious market arena. These and other factors alter the operating functionality of the overall business process. This will be discussed below.

3) The output includes the end product or service. Some of the revenue generated from customer purchases of these products and services can be reinvested in labor (i.e. wage/salary increases), capital (i.e. machinery and tools), the methods of production (i.e. increased productivity through process improvements), business reorganization (i.e. through implementing "lessons learned" and post-mortem reforms). In these respects, a portion of the outputs form a feedback loop to the inputs of the system.

Later in the article, I'll demonstrate how competition between businesses helps to foster more efficient operations in the Capitalist system.

What is important to understand at this point is that these three components of the system congeal together to form a set of operations that define the system's basic functionality. Businesses will actively or implicitly participate in this process. A group of procedures and instructions will be utilized to transform the inputs into an event-driven process of intermittent and transitional states. Based upon the sequence or concurrence of such states, the values they assume, and the results they generate, the system will "activate" its decision making subsystem to prioritize the results and to choose the "most reasonable pathway" to produce the final output. This multiple-input and multiple-state model is often represented by neural networks. There are undoubtedly such complex networks that describe either the entire capitalist society or an individual business entity.

The system's data flow cycles through a series of repetitions that results in the emergence of distinct patterns of "regularity", which is an intricate composition of operational reliability, stability, and performance. This regularity is expressed in terms of "value" and is a measure of the system's process complexity; it is a function of the various inputs, states, outputs, and time. Here is a simplified example: let I = total inputs, S = summed values of all the states (including both augmentation and dampening factors), O = total outputs, T = total time of operations. Then the process complexity P(I, S, O, T) = I*S*O / t. In contrast, the measure of the system's structural complexity, the inherent complexity of internal organization, is expressed in terms of "pathways" and is a weighed summation of the inputs, junctures between states, and pathways leading to the output. For example, if X = sum of the input pathways, Y = number of junctures between the states, and Z = sum of the outputs pathways, then the level of complexity could equal (1/4)*X + (1/2)*Y + (1/4)*Z. Note that these measures of complexity are not calculated by total permutations of all possible values or pathways; but rather are functional expressions of assumed values or pathways.

Both of these complexity types, in their respective domains, describe the degrees of internal "connectedness"; yet another type of complexity is discoverable as well. If we consider the background or backdrop of a system, and quantify the interactions between the system and it's background (e.g. data flows), then a measure of background complexity can be defined as well. The background provides "hidden" inputs to the entire system. These can be environmental factors, external forces exerted upon the system, various phenomena and filtered sensory inputs, and so forth. In turn, the system can marginally affect its background.

The data flow of the Capitalist process can be boiled down to the following:

(data flow diagram goes here: multiple inputs ==> multiple transition states ==> outputs --> feedback to inputs ==> system / environment interactions)

Complex Adaptive Systems and Capitalism

One of the more exciting intellectual developments in recent years has been the emergence of the study of complexity. A good general description of this subject is found in M. Mitchell Waldrop's book, "Complexity": "Complexity is the science of understanding how independent agents are interacting with each other to influence each other and the whole." These agents interact with each other to form network systems. Alternatively, if we consider the agents to be parts or subsystems capable of existing independently from each other, then they certainly are "networked" together to form a larger integrative system.

There are a few fundamental properties of complex adaptive systems:
1) A networked system will generate greater results than its individual constituent components.
2) New properties, attributes, rules, or behavior emerge from the routine operations of the system.
3) Irrespective of the system/background relationship, such systems are capable of self-organization.
4) Such systems are evolutionary in nature and are adaptive to environmental changes.

Properties 1 and 2:

In Capitalist theory, the phenomenon of "emergence" inherent in the 1st and 2nd properties is best illustrated by the concept of Adam Smith's "Invisible Hand". The basic idea is that when individuals are free to actively pursue what they do best, and not intentionally embrace what is deemed to be the public good, they are in effect working in harmony with each other to promote a higher degree of social order than simply the sum of individual contributions. Transactions conducted in the marketplace that subscribe to the principle of mutual voluntary exchange are the catalyst for generating this higher level of order. The network of self-directed labor and free enterprise creates an emergent system functionalism that extends the base level of production. For example, if firms #1, #2, and #3 manufacture products A, B, and C respectively, and each of the three firms desire to obtain the other products, they will sell these products to each other (e.g. #1 purchases B, #2 gets C, and #3 ends up with A). The extended functionalism is the market of exchanged goods that not only directly benefit the firms involved, but increase the level of operational and structural order within the system, including but not limited to the purchasing, distribution, inventory, and sales activities that are necessary to support these transactions. Even if the labor needed to exchange the goods is internalized in their respective production levels, the firms enjoy the extra benefit of obtaining specialized goods that meet their specific requirements without having to expend additional resources searching for these goods in other markets.

While I have in the past questioned the validity of the invisible hand, thinking it's simply a contrived metaphor without much substance, within the context of this described property of emergence, it would not be improper to view "The Hand" as an unseen mixer of individual ingredients to make a culinary delight! Indeed, the study of complex adaptive systems, as it applies to Capitalism, gives new life to the concept of the invisible hand.

Of course there are important questions to ask about the mechanics of how this invisible hand would work. Some of the main ones are the following:

a) Are the benefits of this invisible hand measurable, and if so, how are the benefits quantified?
b) When the system exhibits what is deemed to be undesirable behavior, or experiences significant failure, to what degree does the invisible hand work hander to revive the system? Alternatively, if the system overheats, does it compensate by slowing things down? What it boils down to is whether this is a system auto-stabilizer.

Properties 3 and 4:

Regarding the 3rd property, that of self-organization, this is the proposition that the system will organize itself, without the assistance of external parties, in such a way as to promote efficient operations. It will self-organize or improve organizational efficiency; in conjunction with its operational processes and based upon its performance, it will trigger some level of error correction and learning, which in turn can lead to reorganization. Through a prolonged period of operation, the system will adapt to changing circumstances, resulting in an alteration of the flow of data processing or functionality. For example, a breakdown of certain components will prompt the system to invoke its repair and recovery subsystem, or compensate by shifting its reliance upon other subsystems. From the 4th property, assuming the system has a sufficient degree of "fitness", evolutionary forces will streamline the system to organize itself more efficiently as it adapts to the changing environment.

Some of the main contributors to the understanding of self-organization include Friedrich Hayek, who elucidated the concept of spontaneous order; Ilya Prigogine, who proposed theories of dissipative systems and irreversible processes;  Humberto Maturana and Francisco Varela, who formulated basic tenets of autopoietic systems; Stuart Kauffman, who investigated boolean networks; and Per Bak who provided the principle of self-organized criticality. There are others who played significant roles, such as John Holland, who pioneered genetic algorithms, or scientists like Gregory Bateson and philosophers like Ludwig Wittgenstein, who had related contributions; but it would be too exhaustive to mention everyone and describe their ideas.

As it applies to Capitalist theory, the 3rd property is best expressed through the realization of an unplanned economy. In contrast to a centralized planned economy where organization, production, decision making, allocation and utilization of resources are actualized through goal-directed edicts formulated by government economic planners, the unplanned economy can only leverage the localized goals of individual agents pursuing their own self-interests. Economic organization isn't enforced top-down throughout the society, but rather emerges from grassroots invocations of mutual voluntary exchange. These exchanges take place between employers and employees in businesses, between producers and distributors of goods, and ultimately between buyers and sellers in the free market.

The operational data flow of this system can be illustrated with the concepts of neural networks. As already noted, these are comprised of multiple inputs and multiple states that lead to the established output. They involve some type of backward propagation which results in error-corrective learning that improves the performance of the system as it undergoes further operational experience. Within the context of this neural network model, workers utilize input resources and design products & services in accordance to input specifications and requirements; but there will be numerous intermediate stages where design changes and revisions need to be made, adaptations to unanticipated circumstances, and so forth, before the recognizable output, the finished product, is finally achieved. Self-organization will be geared towards promoting a more effective set of system operations that will maintain a stable level of functional regularity and sufficiently contend with a sundry of environmental changes. In this context, self-organization and adaptation (the core of the 4th property) go hand in hand. In fact, adaptation is often a euphemism for self-reorganization in response to such environmental changes. Firms will tailor their business strategies and operations to adapt to the changing marketplace.

In theory, businesses will certainly organize themselves to run as efficiently as possible. The primary aim is to organize production and sales such that profits will be maximized, but they will usually strive to obtain other results like smooth-running operations, employee well-being and satisfaction, and normalized relations with other businesses. The networked economy doesn't simply rely upon the self-organization of individual businesses, nor upon a diversity of products and services. Self-replication, a fundamental attribute of autopoietic systems, is another factor promoting wholesale growth. Successful businesses will expand operations by creating other business entities such as new divisions, subsidiaries, franchises, spinoff start-ups, mergers and acquisitions. Competitors will often model their operations upon successful businesses and that too will fuel further growth via these modes of self-replication. Underlying the efficient operation of the networked economy are the principles of evolution: adaptation to changing market conditions, fitness for successfully competing in the marketplace, and variation of business entities, products, production models, marketing and distribution of goods. In the networked Darwinian economy, properties 3 & 4 should smoothly work together.

We should note that Per Bak's wonderful concept of self-organized criticality is readily found in free market conditions. This is the concept that a system will organize itself up to a naturally occurring inflection point, but once it reaches that point, the system will begin to collapse. An example of this is sand that is continually dropped on a pile that grows in size and assume the shape of a cone. One can keep pouring the sand on this pile, and it will retain its shape and structural support; but at some point, the slopes will become too steep and the pile will begin to collapse. As this concept applies to organizations, the key to success is to self-organize and drive the business as far as possible to the criticality point. Somewhere just before reaching that point, optional success will be achieved. This is because a period of decline will occur before the critical point is reached, just as the pile of sand doesn't suddenly collapse, but sand starts to slide down once the conical slopes become too steep. Businesses can strive to achieve maximal success, but will need to realize the consequences of their actions. For example, as a business is experiencing a growth pattern, it will be more prudent to expand its operations by creating new business divisions, units, franchises, or even subsidiaries, rather than over-expanding the core business. This is akin to forming new piles of sand once the natural limit is reached creating the first conal pile. Alternatively, there might be various dynamic factors at work affecting the growth of the business. Economic and business cycles occur throughout a company's lifetime, and a period of growth might be followed by a period of contraction. The company will survive if it can realign its resources to cope with the changing market conditions, just as the pile of sand can survive a gust of wind if it has a solid foundation and supported sides.