The basis of each modeling is observation. Collecting facts, their selection and connecting in dependencies helping decipher the past and predict the future behavior of the observed system. One of the most influential, often described models is the neoclassical equilibrium model used by the US Federal Reserve (hereinafter FED). Its co-founder, an econometricist Alan Greenspan, who had been the Chair of the Board of Governors of the Federal Reserve System for almost 20 years, describes it in his book The Map and the Territory: Risk, Human Nature, and the Future of Forecasting published in 2013.
The basis of neoclassical modeling is the assumed assumption of how man behaves, specifically his approximation called homo economicus. The human model defined in this way is guided by „rational” assessments of its activities – those that in the long run maximize the benefits of making them, as measured in terms of individual profit. The rational evaluation of the activities carried out by homo economicus leads in turn to conclusions about how the society behaves. This is how the notion of the invisible hand of the market arises, which, while realizing individual benefits, results in an effective market for the production of goods and services. While the prefix „neo” in the term „neoclassical modeling” contains information that – contrary to the considerations of classical economists – under appropriate conditions interference in the acts of invisible hand of the market is acceptable and even desirable because market economy does not always correct itself.
Greenspan postulates the correction of econometric models to include instincts – such as euphoria, fear, panic, optimism and others that interact with rational economic behavior and influence the market
Constructed for the needs of the FED, in cooperation with Greenspan, macroeconomic models contain elements of „toned-down” Keynesianism, monetarism created by Milton Friedman and – as Greenspan euphemistically adds – other, newer concepts in economics(11). Interestingly enough, the very process of constructing a model looks as if model engineers are constantly changing the sets of selected variables and assumptions of equations until they get a result that seems to reproduce historical data in a manner reliable from an economic point of view(12). The selection of variables relevant to such model construction necessarily leads to the elimination of most of them as there are too many variables to be able to construct it effectively. Elimination occurs by trial and error and, ultimately, after sifting through the tangle of details there remains what is most important for a model consistent with the past and suitable for predicting the future.
Greenspan, while dwelling on his observations on modeling helpful in the work of the Board of Governors of the FED, concludes that while the modeling of non-financial sectors of market economies worked relatively well, the finances operate in different leveraged conditions where risk is a much stronger factor than in the rest of economy(13).
At this point, according to Greenspan, there appears a back door of „animal instincts,” that is, behavioral adjustment. Referring to the behavioral sciences achievements, in particular to the work of Daniel Kahneman, Greenspan postulates the correction of econometric models to include instincts – such as euphoria, fear, panic, optimism and others that interact with rational economic behavior and influence the market(14), above all – financial.
And another quote from Greenspan’s book: This does not mean that we should throw homo economicus out with his bathwater: despite numerous evidence of irrational market behavior the data indicate that in the long run free market economies are still driven by rational economic assessments. But of course „in the long run” can, as we know, mean a very long time(15).
Econometric models, Greenspan writes, have become an integral part of the creation of governmental and private strategies of action and remain so today. Despite the fact that in the case of many market crises the FED forecasting system failed, as well as the model developed by the IMF and bank forecasts, plus Greenspan himself declared chances to avoid the last global crisis to be over 50% 24 hours before its eventual outbreak(16), these models still constitute a planning basis that governments and financial institutions refer to.
However, does man really behave as homo economicus adopted in econometric equilibrium modeling approved by econometrics experts even as influential and familiar with behavioral attitudes as Alan Greenspan? The answer is simple: no! The fundamental blow, struck by behaviorists, to the expected utility hypothesis and the axioms of rational choice underlying the classical and neoclassical economics was to show that people do not strive to maximize benefits in a linear manner. The mind of a man, the real one, is non-linear. Daniel Kahneman in Thinking, Fast and Slow describes the research in which the reaction to the proposed benefits was observed. The observations shoved that people prefer the certainty of a minor benefit to even the slightest doubt in the prospective major advantage. People in their thinking often overestimate the probability of occurring less likely favorable (a lottery win) or unfavorable (an unexpected accident) events, and do not substantially appreciate the probability of occurrence of these that are almost certain. The human mind is more focused on avoiding losses than on the valuation of benefits marked with some, not necessarily high risk(17).
Richard Nisbett, a social psychologist, in his highly recommendable book Mindware. Tools for Smart Thinking points out in many examples of behavioral research carried out by psychologists and economists that people think and react in a surprising – from a „rational” point of view – way. For example, in a study led by an economist Roland Fryer teachers (and therefore good candidates for „rational” representatives of humanity) were promised a raise should their students achieved better results. To no impressive end. Yet, when the same amount was paid out at the beginning of the experiment, provided that it would have to be paid back in the absence of students’ progress – a significant improvement was noted. It turned out that the teachers perceived the situation differently when the bonus, the reward received for the unrealized activities, would have to be given back as compared with the situation in which they simply did not receive the bonus for the unrealized achievements(18).
And so this quality of the human mind – non-linearity – in some circumstances changes into non-commutativeness. It would seem that while valuating the human action with a bonus, the arrangement that we will give that bonus before the task is performed, and then, in case of failure, we will take it back, or vice versa, should not matter. And yet it does! Daniel Kahneman in Thinking, Fast and Slow pointed to another interesting example of non-commutativeness – in Germany, the percentage of people who can be organ donors is 12% whereas in a culturally close Austria – 100%. In Sweden it is 86%, in Denmark 4%(19). Why the difference? In Germany and Denmark you must agree to be a potential donor, in Austria and Sweden you have to disagree. For those still in doubt I will give one more example from Thinking, Fast and Slow – doctors were approached with a test-questionnaire in which the risk (identical in value) of two treatment variants was determined in the form of either a survival or a mortality rate. Physicians, after all experts, were choosing as a better option the treatment accompanied with a survival rate instead of a mortality rate, despite the fact that the risk was exactly the same in both cases(20).
In Germany, the percentage of people who can be organ donors is 12% whereas in a culturally close Austria – 100%. In Sweden it is 86%, in Denmark 4%. Why the difference? In Germany and Denmark you must agree to be a potential donor, in Austria and Sweden you have to disagree
Dan Ariely, a behavioral economist known for his excellent books, in The Upside of Irrationality gives another interesting example of non-linearity of the human mind operation – a sense of justice. In clever experiments researchers investigated how people react to the behavior of another person, which they could judge as being unfair to them. And so, the observed sense of justice led experiments participants to give up their own benefits in order to punish those whose actions were previously considered unjust(21).
These and other complex, non-linear and seemingly – from the point of view of „expected utility” and „rational choice” – irrational behavior described by many researchers of human nature – behaviorists – should lead us to deeper reflection on the value of econometric equilibrium models. Just complicating them by applying, as Greenspan suggests, subsequent amendments in consideration of „animal instincts” characteristic of humans – like euphoria, fear, panic, optimism and many others – will give us a sense of understanding the world, but detach from the real understanding of how, when and at what scale the world can still surprise us.
James Rickards, a critic of econometric equilibrium modeling constructed by Greenspan, proposes to use the intensively developing in the last 30 years theory of complexity to describe the socio-economic reality. Man, communities, market and civilization are mutually embedded complex systems, the better understanding of which – as Rickards urges – requires the use of behavioral sciences achievements – Since the complexity of human nature overlaps with the complexity of markets and compounds it(22).
Moreover, the behavioral adjustments postulated by Greenspan to improve the quality of these models are seen by Rickards more as an attempt to harness the achievements of behavioral economics to, de facto, primitive attempts to manipulate the behavior and expectations of market players – namely, as the release of „animal instincts” to nudge the development of modeled social behaviors in the direction desired by central banks and governments(23).
In contrast to the econometric modeling whose axis are sought out and mutually correlated „selected variables” in the aligned long-term trends, the basis for modeling complexity is the magnitude of diversity and intensity of interaction between its elements(24). In the era of globalization, the ever-growing accumulation of knowledge and development of technologies intensifying exchange of information, the diversity of elements, or the “agents,” to use Rickards’ term, of the system as well as the intensity of interaction grows. Also expanding is the complexity and room for action for feedbacks appearing in the system, namely for adaptive behavior.
Moreover, every attempt to thoroughly examine the status of the system will change that system (it can be clearly seen in the referenda organized in different countries and commissioned by ruling elites – Brexit being the current example)
The feedback is endogenous when the system agent learns from his own mistakes, but also exogenous – when he observes the behavior of other system participants. Copying behavior in turn leads to the formation of a „crowd”, a trend and collective reactions. Learning from your own mistakes can lead to the formation of an „anti-crowd.” This in turn may again lead to imitative actions or to combined action strategies depending on individual predispositions and intuitions (sum of life experiences) of system agents.
Rickards formulates several insights based on these premises(25):
1. Capital markets are dynamic complex systems.
2. Complex dynamics is characterized by memory or feedback, referred to as path dependence.
3. Risk on capital markets is an exponential function of the scale.
4. Minor changes in initial conditions in the system translate into huge changes in results.
5. Results of the system operation can be ordered or chaotic.
The first and basic conclusion while assuming that the theory of complexity is a more correct description of socio-economic reality than models based on individuals seeking to maximize benefits and tormented by „animal instincts” is that the behavior of a complex system – phase transitions and/or catastrophic breakdowns – cannot be predicted in the long run. Elements of the system differentiate, enter mutual relations, affect each other and adapt in an unpredictable way. Past behavior does not sufficiently determine future adaptations. Moreover, every attempt to thoroughly examine the status of the system will change that system (it can be clearly seen in the referenda organized in different countries and commissioned by ruling elites – Brexit being the current example). Any attempt to influence the operation of the system, whether by changing the conditions of its operation (central bank policy – the quantity and cost of money), or by influencing the emotions of its participants (announcements of governments, politicians, experts, representatives of central banks) can trigger both predictable, as well as unpredictable consequences. This complex world of neoclassical simplifications is questioned by the simplicity of the complexity theory.