Dr.
Mathias M. Fonkam, Damian Feese, Felicia Ikwu, Ama Oruomabo, Omasirichukwu Udeinya
1. Introduction
Systems dynamics is an offshoot or specialization of systems
thinking that emphasizes the special role of computer simulation for
understanding the dynamic complexity that is an intrinsic part of most
management type problems and finding leverage points within the system from
which fundamental, long lasting interventions might be effected with a hope to
improving the problem situation. Both represent a paradigm shift from the
traditional scientific method. The scientific approach to complexity is
reductionist by nature since it prescribes breaking complex problems down into
their parts with the hope that an understanding of those parts, or indeed a
solution to the sub-problems they represent, can be assembled for the whole.
Like systems thinking, systems dynamics
is a holistic approach to complexity that emphasizes the importance of the
connectedness between the component parts of the problem and the important
relationship between this structure of a problem situation and its perceived
response or behavior over time. Furthermore, Systems Dynamics argues that it is
this structure or connectedness between the system influences which is
responsible for the dynamic character and the emergent properties that are
unique to the problem situation as a whole. Fundamental components of this
system structure that account for much of its dynamic character and complexity
are feedback loops – both positive and negative loops – and delays. Systems Dynamics
recognizes the fundamental limitations of the human capacity for dealing with
complexity, especially dynamic complexity and therefore extends the Systems
Thinking modeling technique of causal
loop diagrams with a more quantifiable technique of stock and flow diagrams which easily translate to a computer
simulation model that can be used to gain better insight on dynamic complexity
and to discover points of leverage
within the system. From the pioneering works of the Systems Dynamics Group at
MIT many common recurring system structures and corresponding patterns of
behavior have emerged that characterize the majority of complex problem
situations. Even more importantly, for each system archetype a point of
leverage is shown, where a change in the system structure will provide a more
fundamental and long lasting improvement. In each case, this point of leverage
appears to be counter-intuitive or contrary to where our intuition will lead us
to want to intervene to effect change in the system. Understanding these system
archetypes and learning to recognize them in problem situations provides a rich
conceptual toolset in the decision maker’s toolbox and to a whole new approach
to tackling complexity and one that embraces the full dimensions of the problem
at hand with correspondingly better results.
We believe that Systems Dynamics holds great promise for
Sub-Sahara Africa where most will agree serious systemic problems exist at all
levels especially within government, business and other social organizations.
More than half a decade since its inception at MIT, adoption of Systems
Dynamics has been rather slow even in the Western world – the US and Europe
from reports by the MIT group. Poor adoption does not negate its relevance and
power as an approach to complexity. We find very scant adoption within
Sub-Sahara Africa, yet it is exactly within this region that the most profound
impact can be had from its adoption. In this short paper we hope to highlight
some of the key elements of this important cross-disciplinary
subject, just enough to arouse and motivate interest in the subject. Systems
dynamics forms the core of our INF 490 – Systems
Thinking & Modeling for a Complex World – a course offering for senior
students on our Information Systems program but one that we hope to extend
across campus as an important cross-disciplinary course open to all majors. World-wide,
the course is especially important for students within schools of business. In
the next section, we outline what systems dynamics is, why it is becoming
increasingly relevant today and some of its key elements & thinking skills.
Section 3 describes the conceptual modeling technique of causal-loop diagrams used to capture and communicate the structure
of problem situations encompassing feedback loops and delays and generic
behavior over time graphs that characterize some of these structures. Section 4
discusses a number of recurring system archetypes that have been elaborated
from the works of the MIT group using various examples. Understanding these
archetypes alone will shed important insights into the systems dynamics approach
and its power to handling complexity. Section 5 outlines some common systems
dynamics principles embodied in everyday language with examples to drive home
the message and the shift in thinking that is required Also discussed in that
section are a number of what a notable author on the subject, Peter Senge
[Ref], has termed the disabilities of the modern organization that most people
in organizations will relate to. Section 6 is our conclusion with pointers to
various initiatives we are putting up here at AUN to promote this important
discipline within the SSA region.
2. The Systems Dynamics Approach & It
Relevance in a Global Economy & SSA Region
Systems dynamics was invented in 1956 by Jay Forrester at MIT as an extension of the general field of
systems thinking that emphasizes the relationship between structure and
behavior and the important role a computer simulation model can play in
shedding insight on system structure and helping identify or discover points of
leverage to effect long lasting change that drives the system in the desired
direction. According to Forrester, systems dynamics combines ideas from
traditional management, control theory and computer simulation. Like the
general field of systems thinking systems dynamics adopts a holistic view of a
problem situation. It extends the graphical modeling technique of causal loop
diagrams with a more quantifiable stock and flow model that can also more
easily be translated into a computer simulation model. A computer simulation model serves as an
important laboratory for learning and experimentation to discover leverage
points within the system that are often counter-intuitive and hence hard to
find relying simply on our mental models. More quantifiable behavior over time (BOT)
graphs of the system under study can also be derived from such simulation
models which will serve as important communication vehicles. Besides these
graphical notations, systems dynamics introduces a methodology with a set of
important thinking skills to enhance adoption of the holistic view point that always
emphasizes the whole but without losing sight of the trees – we learn to see both the forest and the trees. In
outline the methodology has the following phases and corresponding thinking
skills that are briefly explained:
1. Specify
Problem/Issue – which seeks to clarify the what of the
problem and ensure that we are addressing the right problem and using an
appropriate set of tools. As systems dynamics is more relevant for complex,
dynamic time-varying problems, some important relevant thinking skills to adopt
in this phase include; (1). Dynamic
thinking which ensures and enforces a dynamic viewpoint and regards the
visible behavior or performance indicator as part of an ongoing process with a
dynamic character and perhaps even changing patterns of behavior over time, (2)
System as cause thinking which tries
to make explicit influences believed to be part of the process, and about which
some control can be exerted, pulling from the mental databases of the
stake-holders and experts, in the field, and (3) Forest thinking which seeks scope the problem situation both in
time and space to encompass the influences and permit to discover the exact
behavior patterns that characterize their actual interactions.
2. Construct a dynamic
hypothesis or model – focuses on an
explicit model using causal loop diagrams and/or stock and flow diagrams that
encompasses the feedback loops and delays between the system influences and
reference behavior patterns described by behavior over time (BOT) graphs
deriving from the feedback loops. Thinking skills that can facilitate activities
during this stage include; (1) Operational
thinking where some regard is given to understanding the physics of the
problem area, i.e. how do things work? (2) Closed-loop
thinking which emphasizes the key
systems thinking view that dynamics is generated by feedback loops which can be
re-enforcing (positive) or balancing (or negative) and inherent delays in the
transmission of material or information (3) Quantitative thinking – focuses on
including soft influences (human
related such as motivation, loyalty and commitment) that are often important
parts of the system structure but which can easily be ignored because of
difficulty of measuring them. Quantitative thinking not only encourages
inclusion of such variables but teaches how to introduce numeric units of scale
for such quantities that permits their inclusion in a simulation model.
3. Test hypothesis or
model – Through rigorous testing we get to
check the appropriateness of the model, correct any inadequacies and refine the
model to be more in line with the reality it represents. The key thinking skill
to employ during this phase is Scientific
Thinking which is based on the scientific viewpoint that progress in
science is marked by discarding falsehoods rather than ascertaining
“truth”. Adopting this view means we can
easily discover the limitations and usefulness of our models. Rigorous testing
will also help discover leverage points in the system where change can be
introduced for the desired effect.
4. Implement changes &
communicate understanding – the final phase of
the systems dynamics process regards translating the results of the modeling
exercise to one on the real-life situation and begins with the learning and
change in the mental model of the participants and human actors in the system.
The
figure below illustrates two scenarios for managing a complex problem situation
both based on using models. The one on the left is based entirely on employing mental
models whereas the one on the right illustrate where the systems dynamics
methodology comes in to refine and shed better insight on the problem structure
and hence improve the mental model for a better outcome.
Forrester
& his group at MIT have successfully employed the SD methodology to shed
important insights and even make correct predictions on key issues impacting
some of the top business organizations in the world [Refs], government
organizations and even issues of a global scale [Refs] such as pollution, food
supply, population, disease, depletion of non-renewable resources etc. In many
of such problem situations, the approach has been seen to be superior to the
traditional scientific method in terms of the improvements obtained and the
learning engendered. Such global issues affect Africa in equal measure. In
fact, in some cases such as disease the impact on the SSA region is
disproportionately higher. The systems dynamics approach is needed now as never
before due to the increasing inter-connectedness of global business and the
complexity that comes with that and an increasing recognition of a common
heritage for mankind.
3. Systems Dynamics Models – Causal Loop
Diagrams & Behavior-Over-Time Graphs
Many
real systems are comprised of a large number of interconnected elements and
exhibit very complex behavior as they evolve over time. Causal loop diagrams provide
a powerful method of capturing the inter-connectedness of these elements and
gaining some insights on their dynamic complexity. Causal loop diagrams show
how the various elements in the system are connected by cause-and-effect
relationships, the nature of each of which is depicted by the use of the
symbols S (or + sign) and O (or – sign). S (or +) implies that the elements
connected by a specific cause-and-effect relationship move in the same
direction (as customer satisfaction increases, sales revenue increases); an O
(or -) implies that they move in opposite directions (as the workload
increases, my ability to cope decreases). Single, closed feedback loops are of
two and only two types. A reinforcing or positive feedback loop is
characterized by having an even number of Os around the complete loop (with
zero counting as an even number). We often recognize these loops as virtuous or
vicious circles, for they exhibit exponential growth or exponential decline.
The same cause and effect structure can behave in both ways—which behavior
actually takes place in practice depends on how the feedback loop is initiated,
and whether or not an actively spinning loop is subject to a sudden external
shock. A balancing or negative feedback loop is characterized by having an odd
number of Os (negative signs) around the complete loop. These loops exhibit
goal-seeking behavior, often toward an externally determined target or budget.
Sometimes the approach to the goal is smooth, but if there are time delays
associated with the feedback loop, they can exhibit overshoot and undershoot,
causing the system to oscillate, possibly wildly. Most problem situations will
be characterized by several interconnected feedback loops and depict patterns
of behavior over time. However, at any moment in time the behavior of the
system will reflect a dominant loop in the system which is either positive or
negative,.
For
any given problem situation, there is a proliferation of mental models.
Different people have genuinely and sincerely held different views on how the
world works. CLD provide the opportunity to shine a perspective light on these
mental models, share mental models and improve on their content through team
learning, simulation, refining and critiquing the explicit models. Figure below
shows a CLD typical for most business scenarios and comprising 2 feedback
loops….
4.
System Archetypes, their BOTs &
Points of Leverage
A number of system archetypes have been identified
and documented by the MIT Systems dynamics group. These are recurring
structures and behavior patterns for many complex problem situations. Points of
leverage for these archetypes are often counter-intuitive which explains past
failures of intervention efforts. Understanding these archetypes and learning
to recognize them in similar problem situations will therefore shed a
perceptive light on the correct points of leverage within the system and hence
avoid similar traps within the system that appear suggest themselves as
leverage points. We do not cover all the available archetypes to date. This is
an ongoing area of research. In each case we name and briefly document the archetype,
then sketch a causal loop diagram for its structure and behavior over time
graph. Then we discuss leverage points within the system that have been shown
to yield fundamental and long lasting change.
4.1. Fixes that Fail – often phrased as
“Today’s problems come from yesterday’s solutions”
This archetype describes the familiar situation
where interventions to improve a problem situation lead over time to a
worsening of the problem. Initially the fix appears to work, but then after a
delay, which may be weeks or months later, the problem resurfaces, often with
greater force. Its causal loop model is
shown below:
.png)
A fixes that fail structure consists of a balancing
loop which is intended to achieve a particular result, yet the result is foiled
by an insidious reinforcing loop whose effect is delayed as portrayed by the
clock. These two loops interact in such a way that the desired result initially
produced by the balancing loop is, after some delay, negated by the actions of
the reinforcing loop. A BOT graph of this will show slight jerks of improvement
at each intervention point but generally a worsening of the problem over time. An
effective strategy to dealing with this archetype is advanced planning and
thinking through all the unintended consequences of each action, their own
consequences and so on until you can assess the full impact of a decision.
4.2. Shifting
the burden – often described as procrastination
This
archetype captures the structure of a common human behavior pattern which
relates to our tendency to deal with the easy, the obvious and urgent before
one is forced to confront the difficult, the ambiguous and the important. Often
we put off difficult decisions in the hope that they will go away. A good
example is using alcohol or drugs to suppress personal difficulties or
depression. A Shifting the Burden structure is composed of two balancing loops
and a reinforcing loop. It is a very annoying structure because the two
balancing loops act as a single reinforcing loop driving the situation in the
same direction as the reinforcing loop. Both structures support the movement of
the system in a direction generally other than the one desired. The behavior
pattern is very similar to that of Fixes that fail model. Here however, the
fundamental solution is masked both by its inherent time delay, the impact of
the symptom and side-effects of employing the symptomatic solutions.
.png)
A good example of this archetype is Perpetuation by
Self-deception shown below:
The most effective strategy for dealing with a
Shifting the Burden structure is an employment of the symptomatic solution
AND development of the fundamental solution. Thus one resolves the
immediate problem and works to ensure that it doesn't return.
4.3. Limits
to Growth (or Success)
A Limits to Growth structure consists of a
Reinforcing Loop, the growth of which, after some success, is offset by a
action of a Balancing Loop. This model
captures the structure behind the growth and leveling off or decline behavior
of most businesses, cities and other human activities that start off with
growth or success. Its CLD is shown below:
In the model the focus tends to concentrate on the
positive loop since it drives the desired result. The re-inforcing loop
normally dominates for some time too but eventually its dominance is offset and
may even be reversed by the negative loop. However, emphasis may continue to be
exerted on the positive loop for corrective action. The behavior pattern for
this model is an S-shaped growth or even an overshoot and collapse.The best
defense is a good offense. If there is a
Reinforcing Loop operating start looking for what is going to become a limiting factor, and remove it before
it even has a chance to create a substantial impact on results.
4.4. Drifting
Goals – lowering the bar
In this model a gap between a goal and the current
reality is resolved by taking corrective action (the normal path) or lowering
the goal (the drift). Lowering the goal usually closes the gap immediately,
whereas corrective actions usually take time and more effort. A drifting goals structure is composed of two
balancing loops which interact in such a way that the activity of one loop
actually undermines the intended balance the other loop seeks to achieve.
Consider the following example in which I set out to pursue something I want.
The CLD model follows.
There is only one real effective way to deal with
this structure and that is to disconnect the feedback from pressure to
settle for less to what I want so it can no longer subtract from what
I want. Either you want it or you don't, and indecision is your problem
then see Indecision.
4.5. Escalation
An escalation structure is composed of two balancing
loops which interact in such a way as to create a single reinforcing loop. The
action of each loop provides the basis for increased action by the other loop
and the real foundation for this is insecurity resulting in competition.
Examples of problem situations with this structure include the heating up of an
argument between two persons, the US/Soviet arms race or how urgency begets
urgency. In the business environment price wars between competitors provide a
good example. Its CLD model is as follows.
The 2 balancing loops interact to produce an overall
re-inforcing loop that explains the escalation as shown below.
A good leverage point is to engage both parties in
such a way that they begin to see the value in collaboration or to completely
disconnect them into two separate loops that are self-determined.
4.6. Tragedy
of the commons – often described as “The all for one and none for all”
In this model a common resource is over-exploited
competitively by two or more parties generating an overall activity that
negates the gain for everyone. The CLD for this is shown below. Examples of
this structure abound in business and social organizations sharing a common but
limited resource known to both parties and on which their success depends.
Welfare hand-outs creates such systems too.
The most effective strategy for dealing with this
structure is to wire in feedback paths from A's results and B's
results to the resource limit so as A and B use resources their
results promotes the availability of additional resources.
4.7 Success to
the successful
A Success to the Successful structure consists of
two reinforcing structures which interact in such a way as for create a single
reinforcing structure. Consider a situation where there are two project
managers, Jane and Tom, responsible for managing similar projects.
Their manager, Sarah, has a fixed amount of
resources which she allocates to their projects. Initially both projects are
progressing equally well. Then, for some reason, Sarah chooses to allocate more
resources to Jane's project than to Tom's.
The structure can be redrawn as a single re-inforcing structure as
below.
A good example of this that can easily be seen in an
academic context is shown below.
There are actually two strategies for dealing with a
Success to the Successful situation.
1.
Identify the resource(s) being unequally
distributed and balance the distribution.
2.
Disconnect the two reinforcing
structures so they are not dependent on the allocation of shared resource(s).
5. Organizational
Disabilities & Systems Dynamics Principles in Everyday Language
In his best-seller, “The Fifth Discipline – The
Art and Practice of the Learning Organization”, Peter Senge identified a
number of what he describes as learning
disabilities that characterize the majority of organizations which he
classifies as ordinary organizations and typifying the majority of
organizations today. In Senge’s view, and many will agree with him, “the only sustainable competitive advantage
is the rate at which organizations learn”. These disabilities make it very
hard to see the big picture and hence to make interventions that have a lasting
impact on the system as a whole. We only very briefly introduce with a comment
or two but invite the interested reader to refer to Senge’s book for more
insight on each disability.
5.1.
I
Am My Position – we focus so narrowly on our jobs that we confuse them with
our identify and see ourselves incapable of intervening outside the boundaries
of our jobs
5.2
The
Enemy Is Out There – a by-product of the I am my position view that makes
it hard for us to think systemically so we develop a propensity to always look
for someone else to blame when things go wrong rather than see ourselves as
part of the whole and the problem. Out there and in here are often part of the
same system.
5.3
The
Illusion Of Taking Charge –proactiveness is in vogue in management cycles
yet all too often it translates into reactiveness and fighting the enemy out
there rather than looking inwards to see how we contribute to our own problems.
5.4
The
Fixation on Events – this tendency distract us from seeing the longer term patterns
of change that lie behind events and from understanding the causes of those
patterns.
5.5
The
Parable Of The Boiled Frog – coined after several systems studies of
corporate failure revealed the same results that maladaptation to gradually
building threats to their survival is rather pervasive. Many corporations are
better at reacting to events than detecting build up in threats that threaten
their survival, very similar to the nature of frogs.
5.6
The
Delusion Of Learning From Experience – learning from experience remains the
most powerful way of learning but when our actions have consequences beyond our
learning horizon, it becomes impossible to learn from direct experience.
5.7
The
Myth Of The Management Team – all too often, teams in business tend to
spend their time fighting for turf
Recognizing the constraining actions of these
learning disabilities Senge and his group have established a set of laws of the
fifth discipline which are essentially common sayings that capture an important
tenet or viewpoint of systems thinking. We will simply list them here as their
meanings are well known.
a) Today’s problems come from yesterday’s
solutions.
b) The harder you push the harder the
system pushes back
c) Behavior grows better before it grows
worse
d) The easy way out usually leads back in
e) The cure can be worse than the disease
f)
Faster
is slower
g) Cause and effect are not closely related
in time and space
h) Small changes can produce the biggest
results but the areas of highest leverage are often the least obvious
i)
You
can have your cake and eat it too but not at once
j)
Dividing
an elephant in half does not produce two small elephants
k) There is no blame
6. Conclusions
References
- David Bayless & Don Greer, “An Introduction to Systems
Dynamics”, The Venture Dynamics Group, 2004.
- P. Senge, “The Fifth Discipline, The Art & Practice of
the Learning Organization”, Bantam Doubleday Dell Publishing Group,
Inc.,1994.
- John D. Sterman, “Business Dynamics: Systems thinking and
modeling for a complex world” , McGraw-Hill,2000.
- Vensim – A Systems Dynamics Simulation package: http://www.vensim.com/
- Thomas Binder et al, “Developing Systems Dynamics Models
from Causal Loop Diagrams”,
- George P. Richardson, “Insightful Little Models”, College
of Public Affairs & Policy, University of Albany, NY.
- Meadows, D., Meadows, D., Randers,
J., and Behrens, W. 1972. “Limits to Growth”. New York: Universal Books.
- Denis Sherwood, “Seeing the Forest for the Trees, A
Manager’s Guide to Applying Systems Thinking”, Nicholas Brealey Pub., London, 2002
- Barry Richmond, “The Thinking in Systems Thinking – Seven
Essential Skills”, Pegasus Communications, Inc, www.pegasuscom.com.
- John Boardman & Brian Sauser, “Systems Thinking, Coping
with 21st Century Problems, CRC Press, ISBN-13:
978-1-4200-5491-0
- http://www.systems-thinking.org/theWay/theWay.htm
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)
.png)