George Soros presents a critique of and an alternative traditional economic theory, which denies the possibility of the sort of housing and credit bubbles that characterize the crash of 2008 in the United States. Soros is charming, disarming, self-effacing (except about his ability to conquer financial markets), never dismissive of other theories, and never aggrandizing his own approach by presenting straw-man versions of other approaches. I came away from this book with a good deal of respect for Soros as a thinker and as a human being.
Soros' central claim is that traditional economic theory holds that competitive markets tend toward equilibrium, and this is false. "The belief that markets tend toward equilibrium," he writes, "...is no better than Marxist dogma. Both ideologies cloak themselves in scientific guise in order to make themselves more acceptable, but the theories they invoke do not stand up to the test of reality." (p. 75) Soros calls this faulty approach "market fundamentalism."
I learned economic theory when I was a graduate student at Harvard. The central model I learned was called "general equilibrium (GE) theory," initiated by Walras in the late nineteenth century, and perfected in the mid-twentieth century by Debreu, Arrow, Hurwicz, Hahn, McKenzie and others. GE theory is the basic, underlying model in all of contemporary economic theory. It is highly abstract, but by carefully specifying the conditions under which market equilibrium obtains, it provides an analytical basis for understanding not only markets, but also market failures (cases where competitive markets cannot exist, or lead to socially inefficient outcomes). If one accepts this model, one then analyzes a real-world economy by assessing where the real economy deviates from the model, and what we might expect to occur in light of of this deviation. There is no assurance that this methodology will be successful (Google the Theory of the Second Best), but generally it is the best we have, and it appears to work well in practice.
At the time the architects of GE theory achieved their successes in the mid-twentieth century, which consisted of proving the existence of equilibrium under very general conditions, they fully expected that the theory would extend to proving stability and perhaps even uniqueness in the course of time. To illustrate just how far GE theory was from a plausible dynamic model, Walras had proposed that equilibrium would be achieved by having an "courtier" (broker) or "crieur" (crier) call out prices and adjust them according to the degree of excess demand or supply in each market, until equilibrium was achieved. The idea of what in English we call the "auctioneer" equilibrating a decentralized market economy is so bizarre and indeed absurd that leaving GE theory at this level would of course be highly embarrassing to economic theory. To add insult to injury, it was shown by Saari in 1985 and Bala and Majumdar in 1992, that even with an auctioneer, and with very generous auxiliary assumptions, general equilibrium prices would be unstable, and indeed chaotic. The fact is that to this day there is no plausible model of general equilibrium exhibiting dynamic stability.
It follows that there is absolutely no reason given by economic theory for anything like the "market fundamentalism" that Soros critiques. In particular, there is nothing in economic theory that suggests the impossibility, or even rarity, of crashes, bubbles, and meltdowns. Nothing, I stress, at all.
Nevertheless, I have noticed that despite the above undeniable truth, most economists are indeed market fundamentalists when it comes to issues of stability of equilibrium (they are not fundamentalists when it comes to market failure and the need to regulate the market economy, however). I still recall the moment I heard Alan Greenspan, former Federal Reserve Chairman tell Congress that "Those of us who have looked to the self-interest of lending institutions to protect shareholder's equity -- myself especially -- are in a state of shocked disbelief." I myself stood in shocked disbelief that a real economist, not some free-market crazy, could harbor such theoretically ill-founded beliefs. But, in fact, some of the most influential and perceptive economic theorists share this same believe. In their book Animal Spirits, for instance, Nobel prize winning economic George Akerlof and distinguished (and iconoclastic) Yale professor Robert Shiller, say that "if we thought that people were totally rational, and that they acted almost entirely out of economic motives, we too would believe that government should play little role in the regulation of financial markets, and perhaps even in determining the level of aggregate demand." (p. 173). This is a shockingly uninformed statement. There is nothing in economic theory that says that rational individuals interacting on markets will produce stable, efficient outcomes! The GE model, which is the general framework for investigating macroeconomic behavior on a theoretical level, says that if there are no market externalities, there are market-clearing equilibria that are Pareto-efficient. However, as has been long understood, this model has absolutely no attractive dynamical properties.
I conclude that Soros is correct, not in his critique of economic theory, but rather in his critique of market fundamentalism, the reigning ideology of mainstream economists. Where this ideology comes from, I do not know. I do not recall being taught it by my professors at Harvard, and I do not believe it is in the leading graduate microeconomic textbooks. This doctrine is indeed central to the "rational expectations" school of macroeconomics, and perhaps this is where the idea comes from. On the other hand, neither Greenspan nor Akerlof and Shiller belong to this school of thought, so the ideology is probably of more general proportions. For the record, Soros' critique of the rational expectations school in this book is quite cogent, and I am in complete agreement with him. Only an academic the Ivory Tower could place credence in so bizarre a theory.
Soros' own analysis of where economics went wrong is incorrect. Soros studied at the London School of Economics at a time when the old Marshallian tradition was prominent, and before the GE theory took hold. The Marshallian school analyzed single markets in terms of supply and demand, and assumed that the determinants of supply and demand were distinct, so the two schedules were independent. Soros attacks this notion by claiming interdependence of supply and demand, and he dead right. However, the GE model explicitly accepts this interdependence, without which it would be easy to supply analytically tractable dynamics and plausible stability conditions. However, the market economy is inextricably interconnected, and there is no possibility of treating demand and supply independently.
Soros thus incorrectly attributes "market fundamentalism" to economic theory, whereas in fact it is an aspect of the ideology of economists, not an implication of the GE theory that they learn and use. Because Soros has not studied modern economic theory, he attributes the ideology to an improper independence of supply and demand, which is a attributed of old-fashioned Marshallian theory, not modern GE theory.
Soros then goes on to propose an alternative that is geared to overcoming the independence of the two sides of the market. He does this by developing a philosophical system in which individuals interact with the world in both a "cognitive" and a "manipulative" manner, the first having the aim of understand, the second of influencing and changing. According to Soros' reasoning, the two functions can operate at cross-purposes. Most important, we can analyze the past using the cognitive function and intervene in the present using the manipulative function, which leads to a situation in which the future cannot be known. This two-way connection between facts and opinions Soros calls "reflectivity." Because of reflexivity, the economy involves fundamental uncertainty of form not recognized in standard economic theory. The impossibility of stability of equilibrium is due this reflexivity.
Soros' argument is too speculative for economists to take seriously. Economists work with models. Someone who does not like the GE model is obliged to find an alternative model that does a better job. Soros does not supply another model, so most economists will simply ignore him (given his business acumen, they will `respectfully' ignore him). However, I have worked in this area of the past six or seven years, and my research lends some serious support to his argument. Let me explain.
The GE model has no attractive dynamical properties, but the institutions it recognizes (markets, prices, consumers, producers, firms, money, capital goods, etc.) really exist and more or less operate the way the theory describes. The real world market economies show significant stochastic behavior (there are lots of more or less random fluctuations) but the fluctuations occur around an equilibrium condition that, while rarely attained, is more or less, on the average, approximated over the medium run, and which changes only in response to changes in underlying technology, resource availability, and consumer tastes. This indicated to me, as it does to Soros, is that the problem with the GE model is that it assumes individuals know too much, or rather, that they share too much knowledge. Rather, as stressed by the great economist Friedrich von Hayek, knowledge is distributed all over the economy, each individual economic actor only knowing a small part of the whole.
My reaction to this situation was to develop a computer model of the economy using what is called agent-based modeling. My model appears as "The Dynamics of General Equilibrium", Economic Journal 117 (2007):1289-1309. This model assumes (a) each individual knows only a small part of the total picture, and in particular, has his own, private estimate of prices, and (b) individuals improve their position as firms, workers, and entrepreneurs, by copying the behavior others who appear to be more successful than themselves, as well as experimenting and learning from variations in their own behavior. There are two main findings to be had from this exercise. The first is that the economy does tend toward equilibrium, and if shocked, tends to return to this medium-run equilibrium. Thus the economists' ideological faith in equilibrium seems vindicated.
However, the second finding is that there are significant excursions away from equilibrium, to the point that disequilibrium is the general conditions, as Soros says. Indeed, these excursions are frequent, and periodically sufficient to produce the sorts of bubbles and crises that we see around us. Moreover, these large excursions away from equilibrium occur without any aggregate macroeconomic shock, and are due to what I call "local resonances" that are characteristic of the sort of complex, dynamical, and nonlinear system that a general equilibrium system seems to be. For an introduction to the economy as a complex system, see Eric Beinhocker, The Origins of Wealth: Evolution, Complexity, and the Radical Remaking of Economics (Harvard Business School Press, 2006). Such local resonances are perhaps the codification of Soros' reflexive tenencies.
In short, I believe Soros is closer to understanding the current crisis than the free-market fundamentalists, the liberal super-regulators, or the behavioral economists who blame human irrationality. My formal model, using agent-based techniques, produces the sorts of outcomes Soros stresses, and it does so for reasons that are analytical refinements of Soros' "reflexivity." His pronouncements should be taken seriously, although considerable analytical refinement will be need to turn them into defensible policy tools.