post-autistic economics review
Issue no. 26, 2 August 2004
article 2



      issue 26 contents                            PAE Review index                               home page




Complexity Economics and Alan Greenspan

Lewis L. Smith    (USA)

© Copyright 2004 Lewis L. Smith




Dynamic systems are ubiquitous throughout the universe. They range from nanomachines to the universe itself. And in their celestial form, they have been a subject of human inquiry for at least six thousand years. As a result, many of the themes of concern to complexity researchers have already been studied in astronomy, biology, cardiology, chemistry, computer science, demography, economics, electricity, game theory, mathematics, meteorology, physics, et cetera, albeit in each case from the perspective of a particular scientific discipline.


But it is only recently that we have come to recognize that many dynamic systems long considered “independent” actually constitute a single family, one which we now call complex systems. Examples of the latter are biological species, cardiovascular systems, economies, human societies, neural systems and securities markets. What ties these seemly diverse systems together and how their common features came to be recognized are the subjects of this section.


For example, “the invisible hand”, first noted by Adam Smith in 17761  is a classic example of the “emergent properties” so characteristic of complex systems. Whenever the potential buyers and sellers for a particular good or service reach a critical mass in terms of number, they may spontaneously organize into a decentralized, competitive market which exhibits a coherent set of prices. Moreover, this set may remain in equilibrium for a considerable period of time. As this market coevolves with other markets and with its cultural, ecological and institutional environments, it may not only exhibit types of dynamics which are either unique or those which are common to markets as a class, but also some or all of those which are common to complex systems.


But it was not until the 1970’s that researchers in these diverse fields began to talk to one another sufficiently to overcome the interdisciplinary barriers of concepts, jargon and pride and achieve a painfully won breakthrough in mutual understanding. At some point, some of them came to realize that the apparently distinct objects of their affections had something in common, enough for these features to be fruitfully studied within the scope of a new discipline, which came to be known as “complexity”, among other names. Like the famous blind men, they realized that they were all feeling the same animal (or at least related versions of the same animal) only this “animal” was a good deal more complicated and complex than an elephant!


Today, with a proliferation of business advisory groups, conferences, consultants, fellowships, journals, research institutes, seminars, workshops, et cetera, complexity is here to stay, despite occasional expressions of doubt from within and without the discipline.2  Moreover, the students of complexity have already been able to give useful advice to economists, managers, politicians and others.3  In this, they resemble the “mature” discipline of biology, which continuously contributes to the invention of medicines which cure people and save lives, even though it still cannot tell us what constitutes “life” or how life came about.


Nevertheless, the goals of many complexity researchers seem to have become more modest. Whereas once some of them dreamed of uncovering the equivalent of Newton’s mechanics for all complex systems, many would be content to find for each type, a set of laws governing its dynamics and then some general principles underlying all of these sets. These principles would then become the foundations of a mature discipline of complexity.4


Moreover, as a young discipline, complexity still has many issues to resolve. These include a lack of consensus on basic definitions, on metrics, taxonomy and terminology. There is also an urgent need to improve theoretical constructs and develop new ones, and to perform many more empirical validations.


Some of the open questions yet to be answered include — What exactly do we mean by “complexity” and “complex systems”?  What is an “emergent property” and how does it “emerge”?  How do we model and explain the dynamics of systems which are capable of manifesting such diverse behavior as “lockins”, “multimodal” behavior, “path dependence” and “branch jumps on the possibility tree”, all within the planning horizon of the observer/participant?5

Some outstanding characteristics of complex systems are the following.


  • Once a critical mass of potential participants has been reached, they spontaneously self-organize into a dynamic system of successive hierarchies. This is done by a process of mutual accommodation, without central direction, planning or programming.6

With only modest intelligence, local information and simple rules for interaction, the participants in these systems can generate very complex system behavior. This is due in part to the positive feedbacks which may occur from events in the life of the system. In turn, the latter are due (in part) to the existence of increasing returns to scale, such as those which may be provided by network effects. As a result, there is no need for strong assumptions about the capacity, knowledge or rationality of the participants, although none of these properties are prohibited. (So much for rational expectations!)


  • Once formed, complex systems exhibit surprising properties, called “emergent properties” which cannot be deduced in advance from the properties of the participants, from the rules for their interaction or from any combination thereof. Adam Smith’s “invisible hand” was one of the first to be recognized.7


  • A complex system is likely to spend more time in disequilibrium than equilibrium. And there is no guarantee that departures from equilibrium will be short, in either distance or time. (Keynes lives!)  Moreover, being in equilibrium may even be suboptimal, if it means that you are “asleep at the switch” and are going to be “zapped” by a competitor, as happened to the US auto manufacturers on the eve of the first Japanese assault on their market share.8  Finally multiple equilibriums are also possible.


  • The dynamics of a complex system are best described by non-linear as opposed to linear relationships9, but as yet it is not possible to accurately model the former. The closest one can come are simulations based on cellular automata.10 Despite their limitations, these models produce behavior similar to the observed behavior of real systems.11

An important subfamily of complex systems are both adaptive and evolutionary (CAE systems).12 Some important characteristics of CAE systems are the following. 


  • CAE systems co-evolve with their environment(s) and/or other system(s). Examples are the interactions of deer, soil, vegetation, weather and wolves; biologically healthy lakes being managed for multiple use; and of course, a national economy in a world of other economies, national and international institutions et cetera. This co-evolution is often more “bouncy”, complicated and faster than Darwin imagined for biological species, and may involve symbiosis as well as competition.13


  • In the medium and long runs, the evolution of a CAE system is liable to be unpredictable, in both space and time. To be sure, the inherent characteristics of the participants, their initial endowments and institutions, the environments within which the system operates and phenomena such as lockins and path dependence14, may set a certain “tone” to the system’s evolution and for a while at least, keep it within a fairly compact region of its possibility tree. However, other factors, some like the bumpers in a game of pinball, will eventually set the system off in unexpected directions. These include “visits” to chaotic and random modes, the importance of initial conditions in the case of the former, the unpredictability of outcomes in both cases, increasing returns to scale, political crises, technological innovations, epidemics, wars and branch jumps on the possibility tree.


  • Given the foregoing, the “best estimate” forecasting long favored by American automobile manufacturers and Marxist dictatorships (among others) is “out”, and “scenario planning” is “in”.


  • Long-run optimums cannot be defined and may be multiple. So every investment plan must be “re-optimized” from time to time. And in comparing investment options, strategic merits and robustness against surprise may be more important than an incremental advantage in terms of the internal rate of return.15


  • In a world of CAE systems, a new kind of manager and a new kind of planner are required. Also and for the first time in history, “antenna people” become important. These are people who can detect whether the current scenario is unfolding as planned, shifting under ones feet or turning into something unforeseen, and do so in time for the organization to avoid being ambushed!16



Alan Greenspan


In the last few years, a number of agencies of the US federal government have hired consultants who specialize in applying the fruits of complexity research to strategic planning and/or to management.  In August 2003, there occurred what may turn out to be one of the biggest breakthroughs of all for complexity theory and in a most unlikely place, Jackson Hole, Wyoming.  Moreover, it happened at a symposium sponsored by the Federal Reserve Bank of Kansas City, located in what some consider “the heartland of America”.  Speaking on “Monetary Policy under Uncertainty” and clothing his message in the traditional language of risk management, Alan Greenspan, Chairman of the FRB, expressed numerous ideas which could have come straight out of the mouth of a complexity economist. If my suspicions are correct, complexity economics has partially penetrated one of the greatest bastions of the US economy.


Chairman Greenspan’s talk is only five pages long. Following are a few quotes from this extraordinary document.17 [Italics are mine.]


Uncertainty is … the defining feature of [the monetary] landscape …As a consequence, the conduct of monetary policy … requires an understanding of the many sources of risk and uncertainty that policy makers face …


“…a critical result [of the attempt to achieve this understanding] has been the identification of a relatively small set of key relationships that, taken together, provide a useful approximation of our economy’s dynamics … [However] our knowledge about many … important linkages is far from complete and in all likelihood will always remain so. Every model … is a vastly simplified representation of the world that we experience …


“… a prominent shortcoming of our structural models is that … not only are economic responses presumed fixed through time, but they are generally assumed to be linear


“… also the relationships underlying the economy’s structure change over time in ways that are difficult to anticipate … what constitutes money has been obscured by the introduction of technologies that have facilitated the proliferation of financial products …


“A well-known proposition is that, under a very restrictive set of assumptions, uncertainty has no bearing on the actions that policy makers might choose …These assumptions are never met in the real world.


“… policy makers need to consider not only the most likely future path …but also the distribution of possible outcomes about that path …


“A policy action that is calculated to be optimal … may not in fact be optimal, once the full extent of uncertainty …is taken into account …


“… only a limited number of risks can be quantified with any confidence. And even these risks are generally quantifiable only if we accept the assumption that the future will replicate the past … 18


“… Our problem is not the complexity of our models but the far greater complexity of a world economy whose underlying linkages appear to be in a constant state of flux.


“Rules by their nature are simple and, when [both] significant and shifting uncertainties exist in the [economy, these rules] cannnot substitute for risk-management paradigms …


“… monetary policy based on risk management appears to be the most useful regime by which to conduct policy …”


I wrote the Chairman about this speech and received a Delphic reply from one of his assistants, assuring me that the Chairman will continue to consider such factors in the future!  So if he has embraced much of the complexity message, he is not yet “out of the closet”. His language is veiled, and his “conversion” is incomplete.


As regards the language, one conjectures that he has clothed his ideas in the mantel of risk management, so as not to scare his ex-coworkers on Wall Street.


As regards his philosophy, other positions adopted in recent years show that in some ways, he is still far to the right, in terms of the traditional US political spectrum.19 For example, in 2000, he denounced “irrational exuberance” in the stock markets, then refrained from action, when he could have sent a strong psychological message by raising “margin requirements”, the minimum down payment required for purchases of stocks on credit. Subsequently he opposed tax cuts. But once they were enacted (with 60%  going to 10% of the taxpayers) he called for expenditure cuts in order to balance the federal budget. This of course leaves some 43 million Americans who lack health insurance “out in the cold”.


Nevertheless, “a cat” did “get out of the bag” at Jackson Hole. Let’s see how we can turn this felicitous event to our advantage, in the struggle to replace neoclassical economics with something humane and realistic.





1. Smith, A., An Enquiry into the Nature and Causes of the Wealth of Nations (Glasgow edition, two volumes, Oxford, 1976).


2. Durlauf, S., “Complexity and Empirical Economics”, Feb 2003, < >.


3. In addition to the examples mentioned previously, see also: Allison, M. A. and Kelly, S., The Complexity Advantage : How the Science of Complexity Can Help Your Business Achieve Peak Performance (McGraw-Hill, 1999), Axelrod, R. and Cohen, M. D., Harnessing Complexity (The Free Press, 1999), and Kupers, R., “What Organizational Leaders Should Know about the New Science of Complexity”, Complexity  Sep/Oct 2000, among others.


4. Klüver, J., “The Evolution of Social Geometry”, Complexity 09/01. For a detailed and erudite discussion of complex systems in different disciplines, see Bar-Yam, Y., Dynamics of Complex Systems (Addison-Wesley, 1997).

5. A “possibility tree” is a diagram which charts the possible evolutions of a dynamic system from its present condition in the form of a branching tree, on the assumption that each possibility can be discretely described and is related only to one antecedent and a few successors. An example of a “branch jump” would be if Oman, currently dependent on crude oil, suddenly found it economic to apply the Shell Middle Distillates process to producing diesel, kerosene and naphtha from stranded gas fields at isolated locations in that country.


6. Participants are called “agents” in the literature, which begs the question, Agents of whom? In fact, participants are often independent, as is the case with small business owners and stock-market players.


7. The formation of water from hydrogen and oxygen is sometimes cited as an example of an emergent property in chemistry. In fact, a Martian who knew about the valence electrons of these two gases and how earthly chemical reactions take place, could predict the possibility of water without ever having seen it.


8. The phrase “asleep at the switch” refers to the early days of railroading, when switches in the tracks were manually operated by a “switchman” who spent most of his shift in a small shack along side a telegraph instrument, waiting for word that a train was coming. Sometimes switchmen fell asleep, from alcohol or boredom, occasionally with disastrous consequences.


9. Mateos, R., Olemdo E., Sancho, M., and Valderas, J. M., “From Linearity to Complexity : towards a New Economics”, 2004, < >.


10. For example: Conway’s “Game of Life” in Gardner, M., Life and Other Mathematical Amusements (Freeman, 1983) ; Epstein, J. M., “Agent-based Computational Models and Generative Social Science, Complexity May/June 1999 ; Gross, D. and Strand, R., “Can Agent-based Models Assist Decisions on Large-scale Practical Problems ?”, Complexity  Jul/Aug 2000, and Page, S., “Computational Models from A to Z”, Complexity  Sept/Oct 1999.                                      See also < >. 


11. For two additional candidate characteristics, see Chu, D., Strand, R. and Fjelland, R., “Theories of Complexity”, Complexity  08/03 (2003).


12. The literature frequently refers to complex adaptive systems (CAS), a somewhat looser term which appears to encompass CAE  systems. See Markose, S., “Markets as Complex Adaptive Systems” 09/03 < >.


13. Darwin, C., On the Origin of the Species (Harvard U. Press, 1964).


14. The reality of lockins is controversial and has generated a large literature for which I have not found a good summary. However, path dependence is common in developing countries. For an excellent example, see Reinhart, C., Rogoff, K. and Savastano, M. A., “Debt Intolerance”, NBER working paper #9208, August 2003.


15. Some of the best examples of this kind of tradeoff are unfortunately the least accessible. For example, the debates between “bean counters” and “innovators” within pharmaceutical companies. And the internal debates between bean counters and engineers over energy conservation and renewable energy measures for existing factories.


16. See Smith, L. “Who Matters in a Complex Society ?”, June 2004, Economics Web Institute.  Go to Google Advanced Search, enter < > for English language only. Once “inside” this URL, look for above title and click on underlined word essay below the abstract.


17. 29/default.htm


18. A risk which is not quantifiable is no longer a risk. It is an uncertainty.


19. By our emphasis on disequilibrium, complexity investigators are not only “off the spectrum” but in rebellion against 250 years of economics. After all, in the final analysis, even Austrians, Marxists and Schumpaterians come down in favor of equilibrium as normal, if not also a good thing.


Lewis L. Smith, “Complexity Economics and Alan Greenspan ”, post-autistic economics review, issue no. 26, 2 August 2004, article 2,