Skip to content

Complexity in Economics

↓ Jump to responses

Download the WEA commentaries issue ›

By Maria Alejandra Madi

The need for an epistemological change in economics

Traditional epistemological theories have fostered an endless debate on dichotomies characterized by forms of objectivism, on the one hand, and forms of relativism/skepticism on the other. Currently, among the deep global social and cultural challenges, the crisis in epistemology is characterized by a radical questioning of the whole matrix within which such dichotomies have been drawn.

Taking into account the evolution of Economics as a science, the need for a deep epistemological change has already been pointed out by outstanding economists.  Joseph Schumpeter, for example, rejected the kind of economic thought that mainly favours deductive methods of inquiry – based on mathematical reasoning- because this habit generates analytically unrealistic results that are irrelevant for solving real-world economic problems. Also John Maynard Keynes warned that the understanding of economic phenomena demands not only purely deductive reasoning, but also other methods of inquiry along with the study of other fields of knowledge such as History and Philosophy. Today, Schumpeter’s and Keynes’ criticisms could certainly be levelled at those economists whose beliefs ultimately privilege the adoption of a nominalist bias. This is because, in that approach, the dialogue between the economic theories the economic reality turns out to be abandoned not only in academic research but also in the policy making process.

Considering this background, the shift to Complexity in economic thinking can contribute to substantive epistemological insights in order both to face the contemporary theoretical and methodological challenges and to reject the Cartesian theorization of knowledge under an anthropocentric foundational model of rationality, complete order and truth.

The disembedded human subject of Neoclassical economics

As Edward Fullbrook highlights in his recent book Narrative Fixation in Economics, the Cartesian view of human reality has deeply shaped the way Neoclassical Economics theorizes about the economic and social existence (2016, p. 45). Indeed, while emphasizing the relevance of the pure thought of a disembedded human subject, Neoclassical Economics has reinforced the relevance of the Cartesian method of inquiry that moved the so called scientific (true) knowledge out of the general flux of experience.

Descartes reinforced the analytical-synthetic process of reasoning. Following the deductive method of pure inquiry, human knowledge grows through a rigorous chain of ideas. As a consequence, new thoughts arise while the human subject applies deductive reasoning so as to create a chain of ideas that links the most simple to the most complex ones. In this attempt, true knowledge can be obtained. The Cartesian method represents an attempt to extend the mathematical method of inquiry to all of human knowledge in the form of the mathesis universalis.

In the second part of the Discourse of Method, Descartes presented some principles that should be followed in order to acquire knowledge: 1) human beings cannot  admit any ideas that are not absolutely clear; 2) human beings must divide each problem in so many parts as appropriate for its best resolution; 3) human beings should apply deductive reasoning to organize their  thoughts from the simplest to the most complex ones 4) the analytical-synthetic process of reasoning leads to true knowledge.

According to Descartes, the first principle of his method focuses the importance of “never accepting something as true that I clearly don’t know as such” (Discourse of Method, Part II). Indeed, Descartes was inspired by Geometry as a model of Science. As a result, he considered the postulates of Geometry not only as universal and necessary but also as clear and distinctive ideas related to intellectual intuition. Only these clear and distinctive ideas are considered to be the pillars of true knowledge.

Based on the second principle, Descartes builds his research method of analysis that isolates the clear and distinctive ideas from the most complex ones. His emphasis on the order of thoughts strengthens the role of Mathematics in the Cartesian method of pure inquiry. Moreover, the third principle of his method leads to a special kind of organization of thoughts. In his own words, the organization of thought should start “with the simplest and easier to gradually rise, as if by means of steps, to the knowledge of the more composed, and assuming an order between the ones that don’t precede naturally each other” (Discourse of Method, Part II).

Departing from the mathematical method of reasoning, Descartes arrives at the notion of order in scientific thought, that is to say, once the human subject knows the simple elements of a problem, he can assume all the consequences that derive from those first ideas considered as absolutely certain. Those first ideas have the characteristics of clarity and distinction. Besides, they are known intuitively and constitute the pillars on which true knowledge is based.

Finally, Descartes reinforced the analytical-synthetic process of reasoning. Following the deductive method of pure inquiry, human knowledge grows throughout a rigorous chain of ideas. As a consequence, new thoughts arise while the human subject applies deductive reasoning so as to create a chain of ideas that links the most simple to the most complex ones. In this attempt, true knowledge can be obtained.

Moreover, clarity, distinction and order overwhelmed the mathesis universalis that turned out to be considered as the pinnacle of the epistemo-ontological construction of Cartesian thinking.  The mathesis universalis is, according to the Cartesian epistemology, a general method of pure inquiry able to explain everything, regardless of the nature of the objects to be studied.

As E. Gilson (1945) highlighted, the Cartesian method represents an attempt to extend the mathematical method of inquiry to all of human knowledge in the form of the mathesis universalis.  Indeed, this extension is at the center of the a priori foundations of scientific knowledge in Neoclassical Economics. And as a consequence, the challenge is that the dialogue between economic theories and the economic reality turns out to be abandoned not only in academic research but also in the policy making process.

Complexity science and the Santa Fe Institute

Complex systems are self-organized systems in which elements spontaneously organize themselves as the result of a living process of adaptation, reaction, and innovation. Thus, complexity is the science that seeks to understand, in an evolutionary perspective, the characteristics of the process of interaction that occurs between independent agents. According to Waldrop (1992), from the Santa Fe Institute, such complex systems somehow find a way to balance order and chaos. The “edge of order and chaos” sets itself up as the locus of permanent tension between stability and change. Brian Arthur, one of the leaders of chaos science applied to economics research in the Santa Fe Institute, has adopted the hypothesis of increasing returns to technological change and has identified that new technologies promote new interconnections in new dynamic networks that allow us to think of complex economic systems as self–organized systems. Thus, as a result of innovations, economic systems can diversify and increase its complexity.

Thus, Waldrop’s analysis shows that evolution, as a key issue in complex systems, is more than random mutation and natural selection, since it also involves emergence and self-organization. In short, based on non-linear and multidisciplinary thinking, the scientist claims that complexity in life (and in a variety of systems, including artificial life) is due to spontaneous self-organization. In this sense, besides their emergence, it is necessary to observe the power that systems have to create connections. Scientists of complexity point to the existence of interactions that manifest order (repetitive patterns) and random interactions. In between, on the edge of chaos, there is a set of potentially “creative” interactions. Thus, one of the contributions of complexity theory is the concept of complex systems positioned on the “edge of chaos”, in a zone between rigid stability and chaotic turbulence, with high power to create new connections.

In complexity theory, both the natural world and human societies are understood as complex adaptive systems that share certain crucial properties. First, the natural world and human societies are a network of many “agents” acting in parallel. Each agent finds himself in an environment produced by his interaction with other agents in which there is competition and cooperation. In addition, a complex adaptive system has many levels of organization that are constantly learning, adapting and evolving. Indeed, collective behaviour and self-organization define patterns of behaviour. Third, all complex adaptive systems anticipate the future. From the methodological point of view, Waldrop (1992:142) emphasizes the importance of the use of computational systems in the research of complexity theory. The challenge is to build realistic models in order to make complex adaptive systems simulations that do not converge to equilibrium. In effect, different successful generations of elements, organisms and agents tend to modify and rearrange the characteristics of internal models of behaviour in an evolutionary process. In this perspective, learning, evolution and adaptation are part of the same process whose predictive character is based on implicit or explicit hypotheses about the collective behaviour of the elements of the system. In mathematical and computational models, such hypotheses feed simulation practices with the objective of testing the potentialities of complex systems that incorporate random variables, as well as the multiple elements of internal behaviour models and building blocks. This methodology turns out to be crucial to developing not only artificial intelligence (models related to thought processes), but also artificial life (models related to the biological mechanisms of evolution). In both cases, the study of complexity involves the definition of comprehensive and ambitious algorithms and classification systems (Waldrop 1992: 198).

From an epistemological point of view, it should be emphasized that the scientific conception of the Santa Fe Institute emphasizes that the essence of science is its capacity for understanding and explanation (Waldrop 1992: 255). On the other hand, the theory of complexity points out to a conception of science that advances in the direction of convergence between the human, physical, and biological sciences. Research in molecular biology and cognitive science are relevant examples of this attempt. Besides, Waldrop highlights the reformulation of behavioural theories by rejecting the rationality of economic agents as a priori norm and the equilibrium trend as an inherent principle in markets.

A shift to Complexity in economic thinking

An ongoing dialogue between economic theories and the economic reality should be considered in any attempt to build realistic economic theories in a complexity framework. Recalling Brian Arthur’s words, “we are finally beginning to recover from Newton’s ideas.” (Waldrop 1992: 335) Such an epistemological shift from a non-reductionist perspective opens up new perspectives to re-think human behaviour, the boundaries of evolution and the dynamics of institutions embedded in society.

The understanding of the evolution of real-world markets through time – that is irreversible – requires a shift to Complexity in economic thinking that might favour the adoption of

  • A non-anthropocentric approach to economics
  • A social ontology that is rooted in actual experience in the markets
  • A new approach to the rationality of economic agents
  • An evolutionary approach based on the coexistence of laws and change
  • An ontological indeterminism that rejects a necessitarian approach to real-world economies
  • An epistemological fallibilism that rejects absolute truths
  • An endogenous approach to norms and ethics

Considering the relevance of this topic in economics education, students should be aware of the philosophical backgrounds and practical consequences of different epistemological approaches. Complexity in Economics is not just a new label, but represents a way of rethinking economics as a science.  

References

DESCARTES, René.  Discurso do Método. São Paulo: Abril Cultural, 1973.

FULLBROOK, E. Narrative Fixation in Economics, WEA Books, 2016.

WALDROP, M. Complexity: the Emerging Science at the Edge of Order and Chaos, USA: Simon & Schuster, 1992.

WILLIAMS, Bernard. Descartes: The Project Of Pure Enquiry. UK: Penguin, 1978

From: pp.7-9 of WEA Commentaries 8(4), September 2018
http://www.worldeconomicsassociation.org/files/Issue8-4.pdf

Download WEA commentaries Volume 8, Issue No. 4, September 2018 ›

4 responses

  • ishi says:

    I met a professionally trained economist yesterday who works for US gov on agricultural economics. It was a group meeting–everyone thought she was brilliant–but while all her views were ‘outside the box’ (the meeting was about veganism) i thought they were trivial. Some people still believe in a ‘flat earth’. (I do as well but i’m a conventionalist—i have a pciture of the earth on my wall and its a flat map. )

    I see your most recent reference on ‘complexity’ is from 1992. I recently looked at the news and it said it is now 2018. (Actually complexity in economics goes back to 1970’s , 60’s, and before).

  • Barkley Rosser says:

    It would have been useful if this article had noted any of the developments in complexity economics since 1922 aside from Fullbrook’s comments on Cartesianism. Some books edited by me containing a bit more on this readrs might find interesting include _Complexity in Economics_: 3 volumes, Edward Elgar, 2004 and _Handbook of Research on Complexity_, Edward Elgar, 2009. There is much more out there as well.

  • David Harold Chester says:

    The discussion of complexity in economics, especially in macroeconomics, has become a good excuse for a lack of progress in the development of this subject as a true science. There is no means for the examination of every item of the complex field and even were if such a means possible (with a super computer) it would be impossible to understand the results without taking them together in some kind of average sets. However a proper solution is now available for avoiding the problem of complexity.

    The interested or concerned reader is referred to my short paper SSRN 2865571 “Einstein’s Criterion Applied to Logical Macroeconomics Modelling”. Here the aggregate of all the possible kinds of human-society transactions are taken. Careful examination of these kinds of transactions in this paper shows that only 20 are possible and when the last one is a transaction between two sides having the same nature it can be ignored, due to it making no contribution to the flow of money verses property, goods, services, rents, taxes, etc., around the whole system.

    Without making the decision to cover the complexity by dividing its actions into (idealized) KINDS and taking aggregate values there is no way for making progress. But when this method is applied it is possible to model the whole system as shown in my paper above.

  • Rhonda Kovac says:

    “Complex systems are self-organized systems in which elements spontaneously organize themselves as the result of a living process of adaptation, reaction, and innovation.”

    Unlike the phenomenal systems which are the object of the natural sciences, social systems are fundamentally human made, not self-made. Accordingly, we don’t have to accept all of the complexities of an existing economy as givens. We have the potential to eliminate a good deal of them by actively structuring the system to operate by simpler rules.

Respond to this article

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Please note that your email address will not be published.