Slides - The George Washington University

CONCEPTIONS OF COMPLEXITY
AND IMPLICATIONS FOR
ECONOMICS
Stuart A. Umpleby
The George Washington University
Washington, DC
Three conceptions of complexity
• Complexity -- Systems with large numbers of
variables and many relationships among them
• Self-organizing systems -- Systems composed of
numerous autonomous (usually preprogrammed)
actors or agents or algorithms which interact,
often producing unexpected results (emergence)
• Reflexive systems -- Systems involving knowing
participants acting in systems in which they have
an interest in the outcome
Three conditions define reflexivity
• There is circular reasoning – observe, act,
observe, act…
• The observer is included in what is observed
• Descriptions on two levels are required –
observation and participation
• These conditions seem to involve fallacious
reasoning – circular reasoning, the ad
hominem fallacy -- and self-reference
The situation in economics
• For many years economists have been using an
equilibrium model of economic systems
• Economists assume that if some incident affects
an economy, it will quickly return to a stable
equilibrium
• This model is taken from thermodynamics in
physics
• Equilibrium theory is based on some very
unrealistic assumptions about human behavior
Reflexivity theory as an alternative
• George Soros has proposed reflexivity theory
as an alternative to equilibrium theory
• Reflexivity theory assumes that social systems
are different from physical systems because
they are composed of thinking participants
• Each person observes what is happening,
thinks about alternatives, chooses a course of
action and then participates to implement it
Why has reflexivity theory not been
accepted?
• When I explain reflexivity theory to
economists, most say they have not heard of it
• Some say that reflexivity theory would
encounter logical difficulties
• By “logical difficulties” they usually mean
Russell’s Paradox and self-reference
Are logical difficulties an obstacle to a
more realistic theory of economics?
• This is a question I have been pondering for
several months
• The answer is yes and no, but both yes and no
are complex
• I shall describe the situation as I see it now
Yes, logical difficulties are an obstacle
• Clearly self-reference is a difficult problem
• Paradox is a form of inconsistency and leads to
undecideability
• The informal fallacies, whether known
explicitly or only implicitly, also lead social
scientists to want to avoid a reflexive point of
view
No, logical difficulties are not the
obstacle
• The problem with believing that self-reference is
the primary obstacle to adopting reflexivity
theory is that most social scientists have not
heard of self-reference
• Most social scientists have not heard of the
Russell set – the set of all sets that do not contain
themselves
• Also, some philosophers and mathematicians
familiar with Russell’s paradox interpret it
narrowly and say it is not related to social
systems
Yes, nevertheless
• One possible reply is that many people believe
that there are logical problems, even if they do
not know a formal representation
• The problems have been known for a very long
time, at least since Epimenides’s paradox (circa
600 B.C.) of the Cretan liar
• We have laws governing self-incrimination,
conflicts of interest, corruption
• Hence, we have devised ways of dealing with selfreference through legal arrangements and
judgments about human behavior
The goals of social scientists
• Currently social scientists, particularly
economists, are trying to create a social
science similar to physics
• Paradox, inconsistency, and undecideability do
not fit their image of what science should look
like
• Economists want a quantitative, mathematical
theory, similar to physics, even if the
foundational assumptions are unrealistic
The siren song of physics
• It is not so much the logical problems of selfreference that stand in the way of reflexivity
theory
• Rather the problem is that the image of
physics constitutes a greatly desired “fantastic
object”
• Like ancient mariners, economists do not
worry about being ship-wrecked on rocks,
instead they listen to the music of the lorelei
Underlying assumptions
• Because of the strong desire to create economics
in the image of physics, economists are willing to
adopt many highly unrealistic assumptions
• Some of these assumptions are being corrected
with behavioral economics
• The ergodic assumption may be addressed next
• There are also two philosophical assumptions
that have not yet been addressed by economists
A practical solution to the problem
• As noted, we already have devised solutions
to the problem of self-reference in social
systems through laws against conflicts of
interest, corruption and self-incrimination
• Another corrective action is business ethics
Experts vs. decision-makers
• An expert who is advising a decision-maker can
think of himself as giving objective advice
• However, a decision-maker sees himself as a
participant in the social system and must take
responsibility for his decisions
• Of course the expert is also a participant in that
he seeks to maintain his reputation as an expert
and seeks to maintain his position as an adviser
• But experts usually ignore this kind of
participation in an effort to claim objectivity
Self-reference is an issue for
decision-makers
• When an expert sees himself as uninvolved and
objective, he is saying he is outside the system
• But a decision-maker is deliberately trying to
influence a social system
• Since the decision-maker is inside the system,
there is ambiguity and undecideability
• Is the decision-maker acting to benefit himself,
his firm, society or some combination?
A formal solution to the problem
•
•
•
•
Rather than A or not A
Describe a process
A leads to B leads to C leads to D leads to A
Describing a process implicitly adds time as an
additional dimension
• System dynamics and recursive function
theory are possible representations
Conclusion
• Methods rather than theories may be the best
way to describe management. The observer is
then included
• Economics needs to expand its conception of
science to include the observer and the effects
of economic theories on the operation of
economic systems
Contact information
Stuart A. Umpleby
Department of Management
The George Washington University
Washington, DC
www.gwu.edu/~umpleby
[email protected]
Presented at Capital Science 2012
A conference of the Washington Academy of
Sciences
Washington, DC 2012
March 31 to April 1, 2012