Download Complexity, Self-organization, and Political Economy

Complexity,
Self-organization,
and Political
Economy
by Duncan K. Foley
Department of Economics
Graduate Faculty, New
School University
65 Fifth Avenue, New York,
NY 10003
[email protected]
What are “complex systems”?
• Highly organized, decentralized systems
composed of very large numbers of individual components
Examples
• The cell, which organizes the function of
thousands of proteins
• The human brain, organized out of millions of neurons
• Ecological systems made up of thousands
of species
• Biological evolution
• The capitalist economy, composed of millions of human individuals, each a complex entity
1
General characteristics of complex
systems
• Potential to configure their component
parts in an astronomically large number
of ways (they are complex )
• Constantly changing in response to environmental stimulus and their own development (they are adaptive)
• Tendency to stable patterns (they are selforganizing)
• Avoidance of stable, self-reproducing states
(they are non-equilibrium systems)
2
Complexity analysis
• Empirical and inductive
• Uses simplified and abstract models of
complex systems with large numbers of
highly stylized and simplified components
• Generalizes from computer simulations of
particular examples
• Problem of generalization to general properties of systems
3
The Classical political economic vision
• Individual economic actions have unintended social consequences
• Economic life is organized and coherent
in ways that no economic actor envisions
or controls.
• Regularities emerge from turbulent, nonstationary evolutionary processes
4
Competition of capitals (Smith)
• Each capitalist maximizes potential rate
of growth of capital (profit rate) by investing in the most promising line of production
• Capital moves from ceaselessly from low
to high profit rate sectors
• The effect is the emergence of an average profit rate around which actual profit
rates fluctuate
• Competition tends to maximize the profit
rate of the national capital (the wealth of
the nation)
5
Long-period method
• Competition does not lead to an equalization of profit rates at any moment (turbulence)
• Ceaseless fluctuation of prices around longperiod natural prices which would yield
equal rates of profit in all sectors
• Market prices gravitate around natural prices
• The abstract concept of long-period natural prices explains the concrete fluctuations of market prices
• Contrast with neoclassical conception of
attained equilibrium
6
An example of a complex social system
• Brian Arthur’s “El Farol” is a local pub
which is fun to visit when it has no more
than 60 people. Several hundred “regular” customers have to decide whether to
go each night. Each regular has a whole
group of models that predict how many
people will be in the bar. Each regular
customer adopts the model that has best
fit the data of past attendance. In simulations the attendance at the bar hovers around 60 customers each night, reflecting a robust self-organizing tendency.
But there is no equilibrium in the microstate which describes the model and behavior of each particular customer
7
Complexity in social systems
• Similar forces lie behind many social phenomena, such as the distribution of taxis
in large cities, the size and growth rates
of urban centers, the outbreak of wars,
and the like
• The analogy with the Classical theory of
competition is clear. The individual profitseeking capitalists of the Classical story
do not settle on one equilibrium plan or
strategy. But despite their failure to reach
any equilibrium in their own behavior, they
tend to equalize profit rates.
8
Classical vision and complex systems
• The Classical vision is consistent with the
complex systems view
• The disequilibrium behavior of households
and firms enforces gravitation
• Classical competition is robust, since it
works even if competition is not perfect
• Capitalist economy is self-organized by
competition
9
Capital accumulation and the division of
labor: Smith
• Economic development arises from the
division of labor that arises as a result of
the widening extent of the market. The
accumulation of capital extends the market, both by increasing the wealth and
income of the population, and increasing
population itself
• The accumulation of capital is part of a
“virtuous cycle”: accumulation increases
the extent of the market, which in turn
fosters a wider and deeper division of labor, increasing labor productivity, profit
rates, and accumulation
• This self-reinforcing cycle is the basic metabolism of capitalist economic development,
responsible both for its creative triumphs
and its destructive paroxysms
10
Classical and neoclassical visions of the
market economy
• Neoclassical analysis identifies the Invisible Hand with the tendency for competition to achieve an equilibrium with an
efficient use of existing resources
• Neoclassical competitive equilibrium is incompatible with increasing returns to the
scale of production implied by the division
of labor
• Smithian growth and development are irreversible processes characteristic of complex systems
• Smith can explain the metabolic processes,
accumulation and competition, that support the evolution of the capitalist economy, but not its specific technological or
sociological history
11
Malthus and Ricardo
• Malthus described a stable demographic
equilibrium in which high mortality balances high fertility. If mortality falls population increases. The larger population
encounters diminishing returns in the face
of limited land so that the standard of
living falls, raising mortality and lowering
fertility. Wrong, but another example of
self-organization
• In Ricardo’s stationary state the pressure
of capital and population on limited land
raises rent and forces the profit rate to
zero
• Ricardo’s stationary state is not a microeconomic equilibrium, but a self-organizing state of a complex system that
continues to adapt and change, even as
it reproduces the stationary state as a
macroeconomic average
12
Marx and dialectics
• Marx can reach powerful, general, analytical conclusions about capitalist economic development without proposing implausible and limited “models”, and without claiming to predict the behavior of
individuals.
• Marx brought to political economy the
language of “dialectics”, which is best
understood as a language to discuss system complexity and self-organization.
• Complex systems are “determined” by the
propensities and tendencies of their parts.
But complex systems tend to exhibit aggregate features that are the opposite of
the behavior of their components. The
pursuit of profit by individual capitalists,
for example, may lead to a falling average rate of profit in the system. Dialectical language promotes this observation
to the (contested) status of a “law”.
13
The dialectics of complex and chaotic
systems
• Self-organization is robust, or “over-determined”: it can survive the destruction
of one or even several pathways through
which it reproduces itself
• Complex, adaptive systems, like chaotic
systems are “determinate” but not “predetermined”. They have “open” futures
• Self-organizing structures in chaotic systems break down very rapidly, while complex systems can sustain self-organizing
structures over long periods
• Chaotic systems are statistically predictable,
while complex systems create irregular statistical patterns that are impossible to extrapolate
14
Marx and political economy
• Marx is more Smithian than Ricardian.
Marx shared Smith’s view that the essence
of capitalism is its ability to overcome
diminishing returns through the widening social division of labor and technical
progress
• Capitalism institutionalizes technical change
through the efforts of firms to gain cost
advantages
• Marx’s theory is an account of the tendency of capitalist systems to organize
themselves as engines of technical change.
Marx, like Smith, sees capital accumulation as an open-ended, evolutionary process.
15
Marx and capitalism
• Marx believed that capitalism rests on a
contradictory and morally unsustainable
system of exploitation of labor. Smith
believed that workers will share in the
gains of productivity (as they have)
• Marx thought capitalism would eventually have to resolve its class contradictions through revolutionary or evolutionary change
• Complex systems theory suggests that it
is impossible to answer this question. There
is no way to compress the analysis of a
complex system into a model that is any
less complex than the system itself.
16
Method and complex systems: 1
• Complex systems challenge “common-sense”
notions of determinacy, predictability, and
stability. It might seem at first that complex systems are inherently resistant to
systematic analysis
• We cannot hope to model the future path
of a complex system in detail, because of
the intractable multiplicity of its degrees
of freedom
• The phenomenon of self-organization, however, opens up a sphere of possible analysis. It is possible to understand the forces
that make for the self-organization of a
complex system and to model these limited aspects of the system. Classical political economists’ theories of competition, demographic equilibrium, and technical change are examples
17
Method and complex systems: 2
• Knowledge of the self-organizing aspects
of a complex system is valuable, but incomplete. Insofar as the economy continues to function on the basis of commodity exchange, it will organize itself
into markets with prices, and competitive
forces will induce technical change. But
we do not know what sectors will prosper,
what bottlenecks of technical change will
emerge, or even how markets will be organized
• The self-organization of complex systems
presents the promise of analytical knowledge about open-ended, evolutionary processes which are inherently unpredictable.
The triumph of Classical political economy was its uncanny power to discover
these results
18
Method and complex systems: 3
• An analytically based social science which
denies the complex systems vision is forced
to ignore the open-ended, indeterminate
character of human social life
• An open-ended, evolutionary account of
human social life which denies self-organization is condemned to epistemological
nihilism
• The recognition of self-organization as
a pervasive tendency of complex, adaptive offers the possibility of discovering
and analyzing substantive regularities of
complex systems like the economy without hypostatizing them as realized equilibrium states.
19
Self-organization and equilibrium
• A very fruitful notion in science is the
concept of a dynamical system. The relevant aspects of a dynamical system at
any moment in time constitute its state.
The collection of all possible states the
system might be in constitute the state
space. The motion of the system through
time is determined by its current state
• Mathematicians call the rest point of a
dynamical system an “equilibrium”.
• Mathematicians, physicists, and economists
use the term “equilibrium” in significantly
different ways
20
Thermodynamic and economic
equilibrium
• Physicists use the term “thermodynamic
equilibrium” to denote a macroscopic state
of a system that tends to reproduce itself. The orderliness of a thermodynamic
equilibrium at the macro-level reflects its
complete lack of order at the micro-level
• The traditional economic notion of equilibrium requires each household and firm
in the economy to be in equilibrium at a
microscopic level in order for the economy itself to be in equilibrium. The orderliness of an economic equilibrium system at the macro-economic level is a reflection of its complete orderliness at the
microeconomic level
21
Equilibrium and self-organization
• Self-organized, complex, adaptive systems
are not in equilibrium either in the thermodynamic or economic sense. Self-organizing structures are characteristic of
systems that are mathematically neither
locally stable nor locally unstable, which
can sustain and reproduce recognizable
structures over long periods of time. Selforganization cannot occur in a stable dynamical system, which tends to collapse
all structures into the stable equilibrium
state. Self-organization is also unsustainable in a completely locally unstable and
therefore chaotic system
• Subsystems of a complex, self-organized
system can be in thermodynamic equilibrium, even though the system as a whole
is organized far from equilibrium. Our
blood, for example, reaches thermodynamic equilibrium at a measurable temperature
22
Methodological advantages of
self-organization
• The equilibrium point of view associates
micro-level structure or the maximization
of micro-level disorder with observed aggregate regularities. The equilibrium point
of view can explain aggregate regularity
only by positing a corresponding microlevel equilibrium or chaos
• These micro-level predictions are often
incorrect The equilibrium theorist must
either abandon the theory, or insist against
the evidence on micro-level behavior that
is not present in reality. Recognizing selforganization can avoid these dilemmas,
allowing the political economist to investigate the dynamics of self-reproducing
structures without projecting them inappropriately onto the complex and evolving micro-level behavior of households and
firms
23
Methodological problems of
self-organization: 1
• The complex systems approach assumes
that self-organization operates for a wide
range of micro-behaviors. Conventional
economic modeling demonstrates that a
specific micro-level equilibrium will give
rise to a specific aggregate regularity
• The complex systems approach rests on
generalization. If we can demonstrate
equalization of the profit rate in a specific model, do we know that it works in
a large, ill-determined set of possible environments?
• The Classical political economic intellectual milieu was more open to speculation
and extrapolation
24
Methodological problems of
self-organization: 2
• Self-organization occurs only in complex,
nonequilibrium systems that are difficult
to represent as closed, tractable, mathematical models. Self-organization can
be investigated only by simulating highly
simplified and abstract models on a computer. The self-organization demonstrated
may be due either to the general structure of the system or to specific peculiarities of the cases simulated
• Scientific skeptics accept results only in
the domain in which they have been demonstrated. Scientific innovators project or
generalize results demonstrated in a narrow domain to a wider domain on the
basis of intuition or instinct.
25
From Malthus to Darwin to Kauffman
• The intellectual path from the Classical
political economists to contemporary complex systems theory runs through physics
and biology
• Statistical equilibrium concepts in physics
originated in the empirical investigation
of social phenomena in the first half of
the nineteenth century
• Charles Darwin’s speculations on natural
selection began from Malthus’ image of
the struggle for survival. Darwin’s vision
of the evolutionary process is a paradigm
of a complex system. The principles of
evolution, mutation, and fitness are simple, but their consequences on the macroscopic level are varied, path-dependent,
and open-ended
26
The Pyrrhic victory of neoclassicism
• Marginalist economists formulate models
amenable to closed-form analytical solution in imitation of the physics of the
eighteenth century
• Institutionalist economics, following Thorstein Veblen, attempted to found a scholarly discipline on the evolutionary metaphor.
Alfred Marshall’s attempted to synthesize
Classical political economy, marginalism,
and institutionalism
• Marshall and Allyn Young emphasize increasing returns to scale, which are incompatible with neoclassical competitive
equilibrium. An increasing returns to scale
economy is inherently open-ended and pathdependent, like the evolution of species
27
The economic paradigm of physics and
biology
• Economic metaphors have been central
to conceptual development in the “hard”
sciences
• At least as important as the influence of
physical or biological metaphors in economics
• We have a direct existential experience
of the capitalist economy as a complex,
adaptive system, which informs our imagination in dealing with other complex systems in physics and biology
• Classical political economy is directly relevant to the emergence of the complex
systems vision, and the Classical political
economists are its intellectual ancestors
28
Humanity’s struggle to control its fate
• The development of nuclear weapons, the
emergence of global environmental threats
from production, genetic engineering, and
revolutions in information and communications technology, are the form this
struggle takes for our time
• Humankind is an assembly of individuals
whose actions interact in complex ways
to form an aggregate outcome. New medical or agricultural technologies turn out
to have very different consequences through
these complex interactions from the intentions of their promoters
29
Classical political economy and the new
economy
• Innovative capitalism and the distribution
of income: why do labor productivity and
wages rise steadily and in proportion in
capitalist economies?
• Can political economy save us from global
warming: can we design mechanisms to
achieve environmentally sustainable growth
paths for the world economy?
• The new economy and population of the
earth: what mechanisms will stabilize the
population of the earth? How big will it
get? How poor or rich will we be?
30
Lessons of complex system theory
• It is impossible to control complex, adaptive, self-organizing systems by directing
the behavior of the individual entities that
comprise them. Traditional social policy depends on linking individual behavior
and aggregate outcomes. We may nevertheless be able to design policies that
influence the self-organization of society
in particular dimensions
• The complex systems paradigm avoids the
Scylla of utopian fantasies of an end to
the dialectical historical development of
human societies and also eludes the Charybdis of conservative resignation in the face
of moral and social problems. We need a
better understanding of self-organization
to influence it
31