Complexity, Evolution and Organizational Science
By P.M.Allen1, L. Varga1, M. Strathern1, C. Rose Anderssen2, J. Baldwin2 and K.
Ridgway2
(1) School of Management, Cranfield University, Beds MK43 0AL, p.m.allen@cranfield.ac.uk
(2) AMRC, Engineering Dept. Sheffield University,
Abstract
A highly significant property of complex systems is that they create structure at
several levels. This is important because “selection” occurs from the outside in, while
exploration occurs in the opposite direction. In 1976 Allen proposed an evolutionary
criterion which said that for a new behaviour, variable or population to emerge, first
of all a new type had to be “tried out”, and secondly the surrounding system had to be
unstable to its appearance. In addition (Allen and Ebeling, 1983) showed that a more
accurate, probabilistic model of the growth of a new element from initially tiny
numbers was fairly random and therefore the operation of selection at the level of
individual experiments was in fact very uncertain and non-deterministic. But this
seemingly inefficient “messiness” of the initial stages of any evolutionary step in fact
allows the build-up of heterogeneity in populations, providing a capacity to break out
in a new direction or to solve any problem – given enough time. This means that the
core of any sustainable organization in a changing world has to be the ability to create
and maintain a pool of exploratory heterogeneity from which new ideas and
innovations can spring. Therefore this is what, ultimately, evolution will select for.
The paper will provide examples of these ideas at work in organizational evolution
and supply chains, demonstrating the importance of the multi-level complexity that
unwittingly orchestrates this messy process of evolution.
Keywords: Complexity, evolution, co-evolution, organization, cladistics, open
systems, structural attractors, resilience, supply chains, practices
Complexity, Evolution and Organizational Science
Introduction
In this paper we shall show that complexity science provides us with the mathematics
and mechanisms of evolution. Furthermore, because evolution is about the creation
and on-going transformation of structure and function, the understanding of complex
systems is therefore the basis for organizational science. The many different
approaches and ideas that have been developed in exploring the relationship between
complexity and organizational behaviour have been reviewed recently by Maguire,
McKelvey, Mirabeau and Oztas (2006), in the Handbook of Organizational studies.
We shall not attempt to summarize this wide-ranging review, but instead try to put
forward a coherent view of our own that starts from a 1976 paper by one of us (Allen,
1976) which provided a mathematical basis for the condition required for an
evolutionary step to occur. This showed how physical, ecological or human
organizations are shaped by path dependent, historical events within them in dialogue
with their environment, a dialogue which the Second Law of Thermodynamics tells us
must provide adequate resources to maintain or grow the organization or structure.
The context of an organization is of course its physical, economic and technological
environment, as well as other organizations, individuals and the cultural and social
realities of the moment. As Tsoukas and Chia (2002) say: “Firstly, organization is the
attempt to order the intrinsic flux of human action, to channel it towards certain ends
by generalizing and institutionalizing particular cognitive representations. Secondly,
organization is a pattern that is constituted, shaped, and emerging from change.” The
co-evolution of an organization with its context is therefore about the continual to and
fro of modification as the “inside” and the “outside” of the organization restructure
over time, blurring the separation and indeed, sometimes radically re-defining the
boundary. So, organizations may separate into different specialist arms, or outsource
work that was previously done in-house. A supply chain may become the relevant
competitive unit rather than the firm, and indeed, we may see that evolution is
governed by changes within an ecology of interacting entities, none of which control
the changes that are occurring.
In an recent review of complexity and organizational change Burnes, 2005, points out
that there is in fact a very large consensus that organizations are facing unprecedented
levels of change and that their ability to manage change is therefore of great
importance (Brown and Eisenhardt, 1997; Cooper and Jackson 1997; Dawson. 2003;
Dunphy et al, 2003; Greenwald, 1996; Johnson and Scholes, 2002; Kanter et al, 1997;
Kotter, 1996; Peters, 1997; Romanelli, and Tushman, 1994). However, despite this
organizational change has proved to be very difficult, with up to 80% failure rate
(Beer and Nohria 2000; Brodbeck 2002; Bryant 1998; Burnes 2004b; Clarke 1999;
Harung et al. 1999; Huczynski and Buchanan 2001; Stickland 1998; Styhre 2002;
Whyte and Witcher 1992; Witcher 1993; Zairi et al. 1994). This suggests that the
traditional way of looking at and planning organizational change and evolution is
somewhat flawed, and that perhaps complexity can offer us some help in improving
this performance.
Despite the ubiquitous nature of evolutionary processes, we still tend to understand
and make sense of what is occurring by looking at successive “snapshots” of the
organization at different moments, which should be seen as successive “stable”
(temporarily) regimes of operation. These arise and persist (temporarily) when the
interactions between their heterogeneous elements are such as to lead to a flow of
resources from the environment, necessary to maintain the elements and their
coordination. Clearly, any structure or organization can either persist by finding an
environment that will maintain it, or it must adapt itself to draw sustenance from
where it is. The importance of the openness of an organization to its environment
points to the absolutely fundamental significance of the idea of the “fit” between an
organization and its environment. This immediately tells us that the “fitness” of any
organization or structure is a measure of its ability to elicit, capture or merit resources
from its environment. In order to maintain “fitness” in a changing environment then,
it will be necessary for the organization to be capable of actively transforming itself
over time.
This creates a second level of explanation of organizational behaviour. The first level
is that it must be such as to obtain resources from the current environment. The
second is that it must also be capable of changing itself in such a way as to respond to
the changing environment. This environment will consist among other things, of other
organizations with similar objectives, and so we can directly see that there will be two
directions of change: 1) the ability to out compete similar organizations, or 2) the
ability to discover other “niches” which can still command resources, but which
escape the competition. Here we shall use complex systems’ ideas to show how we
can understand organizational behaviour, and reveal the elements and characteristics
that lead to successful organizations.
Evolutionary Drive
“Evolutionary Drive” was put forward some years ago (Allen and McGlade, 1987) as
the underlying mechanism that describes the change and transformation of complex
systems. In this view evolution is driven by the interplay over time of processes that
create micro-diversity at the elemental level of the system and the selection operated
by the collective dynamic that results from their interaction together with that of the
system with its environment. It is fundamentally an ecological perspective on the
evolution of organization and structure, as it is based on the occurrence of successive
instabilities in the collective patterns of interaction.
We can devise a simple computer program to demonstrate this idea, by considering a
population that initially sits at a low point of a fitness landscape, and then has random
variation of individual fitness. Some of these may of course be non-viable, but others
will be distributed among higher and lower fitness. However, those that are of higher
fitness will grow at the expense of those of lower fitness, and so gradually the
population will “climb” the fitness landscape.
This experiment tells us that a population will increase its fitness – climb the hill –
providing only that the behaviour of diverse individuals “explores” possible increased
pay-offs that may exist. Ignorance and error making are very robust sources of such
exploration. Clearly random changes in the design of any complicated entity will
mean that many experiments are simply non-viable, which tells us that there is an
“opportunity cost” to behavioural exploration. However, it does not change the fact
that a species WILL climb the fitness landscape by simply making random “errors” in
its behaviour, providing that these are detected and copied.
Figure (1). A population initially of low fitness will increase its fitness simply by
making random errors in its behaviour.
In a second experiment we can examine “how much” diversity creation (errormaking) wins in the discovery of better performance. In this experiment we launch
two populations simultaneously at the foot of the fitness hill, and see how successfully
they climb it. Here population 1 is assumed to have a 5% exploration rate in character
space, while population 2 has 10%. However, we also make the assumption that of
these “random” experiments, only 2% are actually viable, which means that there is a
considerable “opportunity” cost in exploring rather than making perfect copies.
Initially, population 2 wins, because, despite its cost in non-viable individuals,
diffusing faster is rewarded by the fitness slope but later when the hill is climbed,
faster diffusion in no longer rewarded and population 1 dominates. This sequence of
events is shown in figure (2).
Figure (2). If we have two populations that have different rates of “exploration” then
we find that the relative success changes from early exploration to late exploitation.
Our model shows that when we are in a new domain, and there is much to learn – then
high rates of exploration pay off. However, when we are in a mature system that has
already been thoroughly explored there is no point wasting effort on further
exploration. Of course, we can only know the there are opportunities or not by
actually engaging in exploration, but clearly, unless there is some structural change,
the value of exploration falls with sector maturity, and this will lead exploration
behaviour to switch to exploitation.
Figure (3) The coupled effects of mechanisms producing individual heterogeneity and
differential performance in the collective dynamics leads to Evolutionary Drive.
In a detailed probabilistic model of the initial phases of the growth or decline of any
new behaviour, Allen and Ebeling (1983) revealed great uncertainty. Even much
better behaviours had high probabilities of not surviving, figure (4), while worse
behaviour could take a long time to be eliminated. Because of this, the knife-edge
selection often assumed separating successful and lethal mutations, is seen to be false.
Figure (4). The probability of survival for an initial mutant individual. (Allen and
Ebeling, 1983)
The real picture is much messier than this and allows much greater pools of individual
heterogeneity to be tolerated by the collective dynamics. Instead of a clear
mechanistic amplification of improvements, and the suppression of any behaviour less
than the best, it means that a population of individuals can form a “cloud” of
heterogeneity in the fitness landscape and can cross valleys and find alternative peaks
of fitness to climb. As shown in figure (3) over time, given the existence of an
outward exploration and diffusion in character space, the differential dynamics that
the fitness landscape represents shape the collective structure. Evolution operates
through this dialogue between the internal exploration of behaviours and the fitness
landscape representing their differential growth and decline over time.
This view of evolution implies that over longer time scales it is the imperfect
transmission of behaviours that allows exploration and it is the imperfect selection
operated by the environment that allows the creative heterogeneity to persist inside
systems. Over the long term evolution will select for systems that maintain these
imperfect processes and against those that are seduced by the clarity of either. In short
the macrostructure of the collective system provides a protective shell within which
behavioural exploration can take place, and genuine novelties allowed to develop
through initially poorly performing stages.
In human systems, the typical development of economic sectors and markets, as
shown in the work of M. Hirooka, 2003, both expresses and results from the fact that
initially search is rewarded, and therefore the bundling of components attracts
investments. However, as the sector becomes well-established the pay-offs to new
types of products falls, and so investment falls. It then switches to some other area
that may seem to offer better opportunities. This is exactly what our simple theoretical
model above predicts.
The presence of firms with different levels of exploration and exploitation (errormaking and accuracy) will automatically lead to evolution selecting whichever is
most appropriate. So, evolution will be driven by the amount of diversity generation
to which it leads.
Exploration is
% Complete
no longer rewarded
Reservoir of Core
Technologies
Product
Creation
Product
Diffusion
~ 25years
~10years
Exploration
strongly rewarded
Time
Figure (5).. The evolution of any industrial sector (M. Hirooka (2003)) may be
explained by a similar dynamics to that of figure (2), where investment gradually
switches from exploration to exploitation.
And this brings Evolutionary Drive very close to the ideas of creative destruction and
evolutionary economics expressed initially by Schumpeter (1942); Foster and Kaplan,
2001; Metcalfe, (1998, 1999), as well as to the views of Tsoukas and Chia (2002) who
see organization as an emergent and passing outcome of an on-going evolution. We
shall therefore establish the basis of the complex evolutionary processes that give rise
to the emergence and development of organisations and the behaviours and identities
of the participating elements and individuals.
Our aim here is to show that successful organizations require underlying mechanisms
that continuously create internal micro-diversity of ideas, practices, schemata and
routines – not that they will all be taken up, but so that they may be discussed,
possibly tried out and the either retained or rejected. It is this that will drive an
evolving, emergent system that is characterised by qualitative, structural change.
These mechanisms also explain exaptation (Gould and Vrba, 1982; Gould, 2002)
since the micro-diversity pre-exists the “uses” and “niches” that they later embody.
It firmly anchors success in the future on the tolerance of seemingly unnecessary
perspectives, views, and ideas, since it is through the future implementation of some
of these that survival will be achieved. If we define the diversity within an
organization as the different functional types that are present – the number of different
boxes in a systems diagram - then this System Diversity is emerges through an
evolutionary process resulting from the existence of micro-diversity within the
functional types - at the level below. In other words the organizational behaviour and
the functional types that comprise it now, have been created from the competitive
and/or cooperative interactions of the micro-diversity that occurred within them in the
past. Evolutionary Drive tells us that evolution is driven by the noise to which it
leads.
Complexity and Evolutionary Drive
We understand situations by making creative, but simplifying assumptions. We define
the domain in question (the boundary) and by establishing some rules of classification
(a dictionary) that allow us to say what things were present when. This means that we
describe things strategically in terms of words that stand for classes of object. The
“evolutionary tree” is an abstraction concerning types of thing rather than things
themselves. In order to go further in our thinking, and get more information about an
actual situation, we then consider only the present, and say, what is this system made
of NOW, and how is it operating NOW. This is “operational” not strategic. It
therefore assumes structural stability and takes us away from open, evolutionary
change, to the effects of running a fixed set of processes. If the events considered are
discreet, then the running is according to a probabilistic dynamics, and we have what
is called stochastic non-linear dynamics, where different regimes of operation are
possible, but the underlying elements never change nor learn, nor tire of their
behaviours. Two paths are then possible. Either we can assume that the probabilistic
dynamics moves rapidly to a stationary distribution then we find the ideas of “selforganized criticality” and sometimes power law distributions. But if we assume that
we can use average rates instead of probabilities for the events, then we arrive at
deterministic, System Dynamics. This is in general, non-linear dynamics, and may be
cycles or chaos or at equilibrium, but what happens is certain, simple and easy to
understand.
This set of simplifying assumptions is shown in figure (5). On the left hand side we have the
“cloud” of reality and practice. Here, we are in the realm of non-science, in which people try
to sum their experiences informally, and come up with heuristic rules and folklore of various
kinds to deal with the problems of the real world. Science begins by deciding on a boundary
within which explanation will be attempted, in the context of the environment outside. This
again shows the difficulty that science will have in dealing with open systems.
Figure (6). Successive assumptions that lead to “scientific” understanding of a
situation.
If the environment “responds” to the changes in the “inside” then other systems may
be affected by the changed capacities of the system within the chosen boundary, and
so the understanding will fail over some longer time scale and system and
environment co-evolve. The series of assumptions necessary to arrive at either System
Dynamics, or stationary probabilities are shown in Table (1).
Number Assumption Made
1
Boundary assumed
2
Classification assumed
3
Average Types
4
First Pathway
Statistical Attractors
4
5
Second Pathway
Average events, dynamics of
average agents
Attractors of Non-Linear
Dynamics
Resulting Model
Some local sense-making possible – no
structure supposed.
Strategic, Open-ended Evolutionary –
structural change occurs. Statistical
Distributions part of evolutionary process,
can be multi-modal
Operational, Probabilistic, Non-Linear
Equations, Master Equations, Kolmogorov
Equations – assumed structurally stable.
Statistical distributions can be multi-modal
or power laws
Self-Organized
Distributions.
Criticality,
Power
law
Deterministic
Mechanical
Equations,
System Dynamics – assumed structurally
stable. No statistical distribution.
Study of attractors, Catastrophe Theory.
Non-Linear dynamics with point, cyclic or
chaotic/strange attractors.
Table 1. The General Complexity Framework
There are two different routes to final simplifications:
 Consider the stationary solution of the probabilistic non-linear dynamics –
the Master or Kolmogorov equation – the stationary probability distribution
corresponding to “self-organized criticality”.

Consider the dynamics of the average, mean or first moment of the
probability distribution, and assume that this can be uncoupled from the
higher moments. This leads to deterministic system dynamics – a mechanical
representation of the system. We can then study the attractors of this
simplified system, and find either point, cyclic or chaotic attractor dynamics
as the long term outcome.
The different models on the right hand side arise from making successive, simplifying
assumptions, and therefore models on the right are increasingly easy to understand
and picture, but increasingly far from reality. The operation of a mechanical system
may be easy to understand, and indeed may correspond closely to what we would
define as the “organization”, but that simplicity has assumed away the more complex
sources of its ability to adapt and change. A mechanical model is more like a
“description” of the system at a particular moment, but does not contain the magic
ingredient of micro-diversity that constitutes Evolutionary Drive. The capacity to
evolve is generated by the behaviours that are averaged by assumptions 3 and 4 –
average types and average events – and therefore organisations or individuals that can
adapt and transform themselves, do so as a result of their non-fulfilment of
assumptions 3 and 4. That is the generation of micro-diversity and the interactions
with micro-contextualities. This tells us the difference between a reality that is
“becoming” and our simplified understanding of this that is merely “being”
(Prigogine 1981). These ideas have been applied to describe the evolution of
economic markets (Allen, Strathern and Baldwin, 2007) and to a practical example of
Canadian fisheries (Allen and McGlade, 1987a; Allen, 1998). Now we turn their
application to organizational science.
4 Evolution of Manufacturing Organisations
The study of organizational change and strategy can be looked at by reflecting on the
organizations in terms of their constituent practices and techniques. The changing
patterns of practices and routines that are observed in the evolution of firms and
organisations can be studied using the ideas of Evolutionary Drive. We would see a
“cladistic diagram” (a diagram showing evolutionary history) showing the history of
successive new practices and innovative ideas in an economic sector. It would
generate an evolutionary history of both artifacts and the organisational forms that
underlie their production (McKelvey, 1982, 1994, McCarthy, 1995, McCarthy,
Leseure, Ridgeway and Fieller, 1997). Here we look at manufacturing organisations
in the automobile sector.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Standardization of parts
Assembly Time Standards
Assembly Line Layout
Reduction in Craft Skills
Automation
Pull Production System
Reduction of Lot Size
Pull Procurement System
Operator based machine maintenance
Quality Circles
Employee innovation Prizes
Job Rotation
Large Volume Production
Mass Sub-Contracting by sub-bidding
exchange of workers with supplers
Training through socialization
Pro-active Training Programme
Product Range reduction
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Autiomation (machine paced shops)
Multiple sub-contracting
Quality Systems
Quality Philosophy
Open Book Policy with Suppliers
Flexible Multi-functional workforce
Set-up time reduction
Kaizen change management
TQM sourcing
100% inspection sampling
U-shaped layout
Preventive Maintenance
Individual Error correction
Sequencial dependency of workers
Line Balancing
Team Policy
Toyota Verification Scheme
Groups vs Teams
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Job enrichment
Manufacturing Cells
Concurrent engineering
ABC Costing
Excess capacity
Flexible Automation of product versions
Agile automation for different products
In-Sourcing
Immigrant workforce
Dedicated automation
Division of Labour
Employees are system tools
Employees are system developers
Product focus
Parallel processing
Dependence on written rules
Further intensification of labour
Table (2) 53 Characteristics of manufacturing Organisations
With these characteristics (Table (2)) as our “dictionary” we can also identify 16
distinct organisational forms:
 Ancient craft system
 Standardised craft system
 Modern craft system
 Neocraft system
 Flexible manufacturing
 Toyota production
 Lean producers
 Agile producers
 Just in time
 Intensive mass producers
 European mass producers
 Modern mass producers
 Pseudo lean producers
 Fordist mass producers
 Large scale producers
 Skilled large scale producers
Cladistic theory calculates backwards the most probable evolutionary sequence of
events Figure (7).
Figure (7). The cladistic diagram for automobile manufacturing organisational forms.
Again, in agreement with the ideas of Evolutionary Drive, we shall look at this as
being the result of micro-explorations, and then a differential amplification of systems
with emergent capabilities. We have studied the evolution of the automobile
production industry by conducting a survey of manufacturers, and obtaining their
estimates of the pair-wise interactions between each pair of practices. In this
approach, the microscopic explorations consist in the attempts to connect in new
practices to an existing system, with the object of improving performance and
creating positive emergent capabilities. As has been reported before, we can
understand and make retrospective sense of the evolution of the automobile industry.
We have then been able to develop an evolutionary simulation model, in which a
manufacturing firm attempts to incorporate successive new practices at some
characteristic rate. The “receptivity” of the existing complex determines which new
practice will in fact be amplified or suppressed if it tries to “invade”.
We can develop a model that expresses the growth of a firm as the result of the net
synergy of its internal practices. The general form is:
dP(i)
 b * P(i) * (1  
dt
 Pair(i, j).P( j)) * (1 
j
P(i )
)  m * P(i )
N
Where:




P(i) is the size of practice i.
b and m are common resource input and output terms
Pair(i,j) is the pair matrix expressing the synergy or conflict between practices I
and j
N is some limiting market size
Figure (8). Successive moments (t=3000, 10000 and 15000) in the evolution of a
particular firm. The evolutionary tree of the organisation emerges over time.
One possible history of a firm is shown in figure (8). The particular choices of
practices introduced and their timing allows us to assess how their performance
evolved over time, and also assess whether they would have been eliminated by other
firms. As a result of the different firms experimenting over time, there is an incredible
range of possible structures that can emerge, depending simply on the order in which
practices are tried. But, each time a new practice is adopted within an organisation it
changes the “invadability” or “receptivity” of the organisation for any new
innovations in the future. This is illustrates the “path dependent evolution” that
characterises organisational change. Successful evolution is about the “discovery” or
“creation” of highly synergetic structures of interacting practices.
The model starts off from a craft structure. New practices are launched with an
“experimental” value of 5. Sometimes the behaviour declines and disappears, and
sometimes it grows and becomes part of the “formal” structure that then changes
which innovative behaviour can invade next. The model shows how the 16 different
organizational forms have increasingly high synergy as they change in the direction of
lean and agile Japanese practices. Overall performance is a function of the synergy of
the practices that are tried successfully. The particular emergent attributes and
capabilities of the organisation are a function of the particular combination of
practices that constitute it. Different simulations lead to different structures, and there
are a very large number of possible “histories”. This demonstrates a key idea in
complex systems thinking. The explorations/innovations that are tried out at a given
time cannot be logically or rationally deduced because their overall effects cannot be
known ahead of time. Therefore, the impossibility of prediction gives the system
“choice”.
The competition between different firms exploratory pathways through time means
that those who for one reason or another fail to find synergetic combinations of
practice, will be eliminated. Once again we find the principle of Evolutionary Drive,
where the micro-explorations involving the testing of new practices leads to
microscopic diversity among firms, and in turn these are either amplified or
suppressed by the economic competition.
5. Aerospace Supply Chains
Large commercial aircraft manufacturers traditionally defined and specified exactly
what their first tier suppliers should produce for them. This practice that expressed a
high level of buyer dominance over suppliers has gradually changed as the result of a
world wide reorganization of the industry. Aerospace supply chains have evolved
qualitatively as different practices and conventions have emerged that profoundly
shift the relationships between the participating companies. (Giunta et al, 2000).
Day and Atkinson (2004), argue that high impact suppliers have to commit
themselves to cost reductions for long term business relationships, and create a culture
of continuous improvement. However, Nollet et al (2005) argue that even greater
competitive advantage can only be achieved by producing innovative solutions in
terms of technology and products.
Recent developments in supply chain strategies are therefore based on a focus upon
the sharing of skills and capabilities (Cousins and Spekman, 2003). In the context of
creating technologically more advanced products, supply chains need to become
learning communities with skills in creating and transforming knowledge (Preiss and
Murray, 2005). In that sense, Easterby-Smith et al (2005) highlight the absorptive
capacity of a firm to appreciate and process external knowledge and the learning from
past experiences. Engeström (1987), however, is more concerned about the learning
and development experienced when moving away from the old. And this is about the
creation of the new, i.e. innovation.
Risk-sharing partnerships combine these views in practice by bringing together the
best expertise available to produce totally new concepts of aircraft design. These
partnerships are the response to the need for sharing of skills, expertise and
capabilities. This is driven by the fact that an airframe manufacturer cannot carry such
huge investments in capital and expertise alone. The tight integration of risk-sharing
partnerships can facilitate a totally new aircraft concept (Rose-Anderssen et al, 2008).
This paper therefore focuses on the levels of learning and knowledge transformation
involved in the development of these practices. In order to establish the evolution of
supply chains and their management, we have examined the practices, techniques and
characteristics that exist in aerospace supply chains, particularly in the upper tier.
From our interviews and surveys we have examined what these practices, techniques
and characteristics are and the underlying beliefs that underpin the changes.
5.1 The Dimensions of Supply Chain Performance
An examination of the literature and a series of interviews with managers involved in
the aerospace supply chain of the major civil aviation manufacturers identified the
importance of 5 basic dimensions of performance:
1) Quality;
2) Cost Efficiency
3) Reliable Delivery
4) Innovation and Technology
5) Vision.
The stage in the life cycle of the product, or the market situation, determines what mix
of these is required as the platform or product moves from design and conception,
through initial prototyping and production to an eventual lean production phase. In
addition 27 key characteristics or practices were identified that could characterize
supply chain relationships. A questionnaire was formulated to enquire into the opinion
of important individuals within these key aerospace supply chains in order to
understand better the underlying beliefs that affect the decisions concerning the
structure of supply chains.
The first three dimensions of performance can be seen as those of “operationality” –
the successful functioning of a production system in producing high quality products
reliably and cost effectively. The dimensions 4 and 5 however, are really about the
supply chain future. They are about the creation of the next product that will become
the focus of operational behaviour later. So, these two dimensions concern the
creative, exploratory behaviour without which there will be no long term future.
The ideas used to examine the structure and evolution of supply chains are based on
complex systems thinking under the ESRC NEXSUS project, which developed
evolutionary models of organizational change (Allen, Strathern and Baldwin, 2007;
Baldwin…., ). These were based on a cladistic analysis of the history of an industry in
terms of its constituent practices, characteristics and techniques (McKelvey, 1983;
McCarthy,…). These demonstrate how firms adapt different practices in a path
depende
5.2 Questionnaire on Supply Chain Practices
The questionnaire has two parts. The first concerned the potential interaction between
each pair of practices. The practices that are considered are:
1, Outsourcing competitive advantage
2. Outsourcing what is easily imitated
3. High level of collaborative relationship
4. Arms length relationship
5. Long-term relationship
6. Formal partnership
7. Subcontracting whole systems and sections
8. Flexibility of operations
9. Risk-sharing
10. Sharing knowledge
11. Offsets as part of sales contract
12. Culture of continuous improvement
13. Ability to handle cultural differences
14. High level of dominance over supplier
15. High level of planning and control
16. Easy dialogue with supplier
17. IT system integration
18. High levels of integration in chain
19. Responsive to market change
20. Transparent organisation
21. TQM procedures
22. Just-in-time delivery
23. Lean practices
24. Explorative learning practices
25. Investment in training
26. Supplier development
27. Monitoring Suppliers
Managers were asked to fill in the cells a number from -5 up to +5 that expresses
whether two practices are in strong conflict, or are strongly synergetic.
Characteristics
Strongly synergetic (+5), indiffernt effects
(0), strongly conflicting (-5)
2. Outsourcing what is easily imitated
3. High level of collaborative relationship
4. Arms length relationship
5. Long-term relationship
6. Formal partnership
7. Subcontracting whole systems and sections
8. Flexibility of operations
9. Risk-sharing
10. Sharing knowledge
11. Offsets as part of sales contract
12. Culture of continuous improvement
13. Ability to handle cultural differences
14. High level of dominance over supplier
15. High level of planning and control
16. Easy dialogue with supplier
17. IT system integration
18. High levels of integration in chain
19. Responsive to market change
20. Transparent organisation
21. TQM procedures
22. Just-in-time delivery
23. Lean practices
24. Explorative learning practices
25. Investment in training
26. Supplier development
27. Monitoring suppliers
1. Outsourcing competitive advantage
2. Outsourcing what is easily imitated
-5
3
-5
4
4
1
4
5
5
-5
3
3
-5
4
4
4
0
2
0
4
3
4
-4
3
-3
-3
3. High level of collaborative relationship
-4
4
-5
-4
2
0
-4
5
5
-3
-2
5
-4
-3
0
-4
0
0
2
3
-4
-3
-3
5
0
4. Arms length relationship
-5
5
0
0
3
4
5
-3
3
4
-5
0
5
4
3
5
2
0
2
4
5
1
3
-4
5. Long-term relationship
-5
-4
-4
0
-5
-3
4
-2
-3
4
0
-3
-5
-4
-3
0
3
-3
0
-4
0
-4
4
6. Formal partnership
4
4
3
4
4
0
4
3
-4
-3
4
4
5
3
2
4
5
5
4
4
0
-3
7. Subcontracting whole systems and sections
5
4
5
4
-4
4
3
-4
-4
4
4
4
4
2
3
5
5
4
3
-3
-4
8. Flexibility of operations
5
5
5
-4
4
4
-5
4
5
5
5
4
0
4
4
4
-3
4
-5
-2
9. Risk-sharing
4
2
3
4
3
-3
-3
4
-3
3
5
0
2
3
4
4
2
3
1
10. Sharing knowledge
5
-3
3
4
-4
3
3
4
4
3
0
4
2
3
4
4
-3
0
11. Offsets as part of sales contract
5
3
3
-3
2
5
4
0
0
4
1
4
4
-3
0
5
4
12. Culture of continuous improvement
1
5
2
3
0
-3
-3
-4
0
3
-2
-2
-3
4
4
4
13. Ability to handle cultural differences
0
-4
-3
4
-4
3
0
0
4
0
5
3
4
2
0
14. High level of dominance over supplier
-5
-4
5
-2
3
0
0
-3
0
-3
2
0
4
-1
15. High level of planning and control
5
-4
-3
-4
0
4
-3
-2
-4
-4
0
-4
4
16. Easy dialogue with supplier
0
4
2
-3
3
-3
-3
-4
-3
0
-4
4
17. IT system intergration
3
4
4
0
4
3
5
4
0
5
1
18. High levels of integration in chain
4
-1
0
0
4
0
0
0
2
4
19. Responsive to market change
4
0
4
3
4
3
0
4
-3
20. Transparent organisation
1
2
1
3
4
0
3
-2
21. TQM procedures
0
0
0
0
0
3
3
22. Just-in-time delivery
4
5
4
4
3
-3
23. Lean practices
5
-1
4
4
-1
24. Explorative learning practices
5
4
4
-1
25. Investment in training
4
4
-1
26. Supplier development
5
0
2
Figure (9). A typical response for the Pair Interaction matrix.
In essence the first part of the questionnaire filled in a 27x27 matrix expressing the
supposed synergy or conflict between each pair of practices. The second part of the
questionnaire concerned the emergent properties that they produce, which affects the
competitive performance of the supply chains. These were the effects on quality of
production, the costs, the ability to deliver reliably, the technological and innovative
level and the ability to bring vision and originality to the project. These performance
factors clearly constitute the “selection criteria” of the product likely to come from a
particular supply chain. The pair interactions constitute the internal synergies and
conflicts that affect the trade-offs involved in using particular practices. Our aim here
is to provide a general method to help choose the set of inter-firm practices that will
lead to the desired mix of performance qualities.
The 27 possible practices for supply chain development were rated in the second
questionnaire according to their expected capacity to perform well in terms of
Quality, Cost efficiency, reliability, innovativeness and vision.
In fact we obtained results from three airframe makers, an engine maker, a subassembly manufacturer. These have been brought together in an average outcome to
reflect the general views of the sector concerning the expected utility of thes different
practices.
This means that, on average, managers looking for quality as the overriding quality of
their output would pick practices: 7, 10, 12, 21 and 26. If Cost Efficiency is the
dominant consideration then they would pick: 7, 12, 21 and 26. For Delivery
Reliability: 7, 8, 12, 15, 18 and 22. For innovation and technology the practices would
be: 1, 3, 7, 9 and 10, while for Vision there is a weaker response, and only practices 3,
9 and 10 are chosen.
Characteristics
Rate of characteristic to successfactor criteria
High (9), None ())
1. Outsourcing competitive advantage
2. Outsourcing what is easily imitated
3. High level of collaborative relationship
4. Arms length relationships
5. Long-term relationship
6. Formal partnership
7. Subcontracting whole systems and sections
8. Flexibility of operations
9. Risk-sharing
10. Sharing knowledge
11. Offsets as part of sales contract
12. Culture of continuous improvement
13. Ability to handle cutural differences
14. High level of dominance over supplier
15. High level of planning and control
16. Easy dialogue with supplier
17. IT system integration
18. High levels of integration of chain
19. Responsive to market change
20. Transparent organisation
21. TQM procedures
22. Just-in-time delivery
23. Lean practice
24. Explorative learning practices
25. Investment in training
26. Supplier development
27. Monitoring supplier
Success criteria factors
Product
Cost
Delivery Techn./
Vision for
quality
efficiency precision innovation the future
6.8
6.6
5
7.6
6.2
4.4
6.6
6
0.8
0.4
5.4
5
4.2
7
6.4
1
1.6
0.8
2
0.2
6
5.4
6.2
6.6
5.4
6.8
3.2
5.4
5.4
4.4
7.6
7.6
7
7.6
4.6
4.2
6.6
7
3.8
3.4
5
4.8
2.6
7.4
6.4
7.2
6.2
5.4
8
5.8
3.2
1.6
2.4
2
2.6
8.2
7.8
7.8
6.4
4.4
3.6
3
3.6
5.6
4.8
4.6
5
5
1
1.2
6
3.6
7.4
1.4
1.4
2.6
2.6
2.4
4.6
5.6
1.4
3.8
5.4
3.8
4.4
6.8
5.4
7
3.2
4.6
5.4
6
6.8
5.6
3.4
2.2
3
3.6
2.6
2.2
8.4
6.8
6.4
3.2
3.4
4.2
4.2
8.2
3
2.8
6.4
6.6
6.8
5
4.2
6
3.6
3.4
5.8
4.8
6.6
5.8
4.4
4.4
5.2
7
6.8
6.8
3.4
4
5.8
4.6
6.2
1
1.2
Figure (10). The average ratings of the different practices in terms of expected
contribution to different dimensions of performance.
This gives us a first estimation of how practices can be chosen in order to achieve a
desired supply chain performance. Clearly, for sustainable behaviour of the firm,
however, we may want to have a cost efficient “cash cow” element running a mature
product production but also have a “design team” part that has the visionary and
innovative practices that will be needed to define a future “cash cow”. Therefore one
should envisage aerospace relationships to encompass both lean production elements
and also visionary, design relationships that are used to define new platforms and
products.
5.3 Modelling the Evolution of Supply Chains
The important point about the evolution of systems is that they concern both the
elements inside a system, that constitute its identity, and also the external environment
in which they are attempting to perform and the requirements that are perceived for
successful performance.
In choosing which practices to use to manage a supply chain therefore we need to
represent both the internal interactions of the supply chain and also the competitive
environment in which it is trying to perform. Therefore it needs to consider the
efficiency of its internal practices and their alignment with the external requirements
imposed by the customer, market place, desired performance. We can represent the
desired, or required dimension of performance by using a weighted sum of the five
types of performance.
We can, quite generally, represent the expected actual performance of a particular
bundle of practices as the result of a matrix multiplication of the internal interactions
between them, with the weighted sum of their expected, individual performances. So,
for example the performance requirement might be for maximum emphasis on
innovation and new technology. In that case a single column of performance can be
obtained by multiplying the five columns of figure 10 by the weights: 0, 0, 0, 1, 0. In
that case the single column will reflect the desire for innovation and the current
unimportance of other dimensions.
However, as we have stated, in order to calculate the “real” performance, or at any
rate whether a new practice would take-off if it were launched, we need to multiply
this single column of expected individual performance by the internal terms of
interaction between practices within the organization or supply chain. This will then
produce a column of figures which will be the real contribution of a practice, taking
into account the synergy or conflict with other practices that are present.
1 on the
diagonal
27x27 matrix
1
1
1
1
1
Characteristics
Strongly synergetic (+5), indiffernt effects
(0), strongly conflicting (-5)
2. Outsourcing what is easily imitated
3. High level of collaborative relationship
4. Arms length relationship
5. Long-term relationship
6. Formal partnership
7. Subcontracting whole systems and sections
8. Flexibility of operations
9. Risk-sharing
10. Sharing knowledge
11. Offsets as part of sales contract
12. Culture of continuous improvement
13. Ability to handle cultural differences
14. High level of dominance over supplier
15. High level of planning and control
16. Easy dialogue with supplier
17. IT system integration
18. High levels of integration in chain
19. Responsive to market change
20. Transparent organisation
21. TQM procedures
22. Just-in-time delivery
23. Lean practices
24. Explorative learning practices
25. Investment in training
26. Supplier development
27. Monitoring suppliers
1
1
1. Outsourcing competitive advantage
2. Outsourcing what is easily imitated
-5
3
-5
4
4
1
4
5
5
-5
3
3
-5
4
4
4
0
2
0
4
3
4
-4
3
-3
-3
1
3. High level of collaborative relationship
-4
4
-5
-4
2
0
-4
5
5
-3
-2
5
-4
-3
0
-4
0
0
2
3
-4
-3
-3
5
0
1
4. Arms length relationship
-5
5
0
0
3
4
5
-3
3
4
-5
0
5
4
3
5
2
0
2
4
5
1
3
-4
5. Long-term relationship
-5
-4
-4
0
-5
-3
4
-2
-3
4
0
-3
-5
-4
-3
0
3
-3
0
-4
0
-4
4
1
6. Formal partnership
4
4
3
4
4
0
4
3
-4
-3
4
4
5
3
2
4
5
5
4
4
0
-3
1
8. Flexibility of operations
5
5
5
-4
4
4
-5
4
5
5
5
4
0
4
4
4
-3
4
-5
-2
9. Risk-sharing
4
2
3
4
3
-3
-3
4
-3
3
5
0
2
3
4
4
2
3
1
1
10. Sharing knowledge
5
-3
3
4
-4
3
3
4
4
3
0
4
2
3
4
4
-3
0
1
11. Offsets as part of sales contract
5
3
3
-3
2
5
4
0
0
4
1
4
4
-3
0
5
4
12. Culture of continuous improvement
1
5
2
3
0
-3
-3
-4
0
3
-2
-2
-3
4
4
4
x
1
7. Subcontracting whole systems and sections
5
4
5
4
-4
4
3
-4
-4
4
4
4
4
2
3
5
5
4
3
-3
-4
1
13. Ability to handle cultural differences
0
-4
-3
4
-4
3
0
0
4
0
5
3
4
2
0
1
14. High level of dominance over supplier
-5
-4
5
-2
3
0
0
-3
0
-3
2
0
4
-1
1
15. High level of planning and control
5
-4
-3
-4
0
4
-3
-2
-4
-4
0
-4
4
16. Easy dialogue with supplier
0
4
2
-3
3
-3
-3
-4
-3
0
-4
4
1
17. IT system intergration
3
4
4
0
4
3
5
4
0
5
1
1
18. High levels of integration in chain
4
-1
0
0
4
0
0
0
2
4
19. Responsive to market change
4
0
4
3
4
3
0
4
-3
1
20. Transparent organisation
1
2
1
3
4
0
3
-2
1
21. TQM procedures
0
0
0
0
0
3
3
1
22. Just-in-time delivery
4
5
4
4
3
-3
23. Lean practices
5
-1
4
4
-1
24. Explorative learning practices
5
4
4
-1
25. Investment in training
4
4
-1
26. Supplier development
5
0
2
1
1
1
1
1
5
0
0
0
0
5
7
0
0
2
0
9
0
0
4
0
0
5
0
0
9
0
4
0
2
4
5
=
Actual
1.057664
0
0
0
0
0.805564
1.208345
0
0
0
0
1.463344
0
0
0.767893
0
0
0.834541
0
0
1.553173
0
0.736019
0
0
0.718632
0.854825
Real
Performance
Per Practice
Figure (11). The real performance of the supply chain is given by multiplying the
expected performance by the 27x27 matrix with unity down the diagonal.
This simple mathematical formula allows us to examine how different bundles of
practices can be expected to perform in the different dimensions that may be desired:
Quality, Cost, Reliability, Innovation and Vision. If different practices were entirely
independent of each other then the off-diagonal elements of the pair matrix would be
zero. In that case the matrix multiplication, with unity, 1, running down the diagonal
would simply give the answer that the expected and real performances would be the
same. But, if there are interactions of synergy or of conflict, then these will modify
the result and the performance generated by each practice will be different from its
value when considered in isolation.
From this simple framework we can therefore examine which practices will “take-off”
if we launch them successively, allowing them to grow if the required performance is
increased, but rejecting them if they do not increase performance. In fact, since to
organize any practice there will be some effort involved, we can say that there will be
a cost per practice that will stop us adding practices that do not enhance performance
by more than the cost of organizing the practice.
We have therefore developed an “evolutionary computing” program that runs
different random sequences of successive launching of practices under different,
conditions of performance selection.
6. Results of the Evolutionary Learning Model
Our evolutionary based computer model captures the idea that initially the agents in
the model have no knowledge of the effectiveness of different practices, nor whether
they are in conflict or synergy with others. The simulations therefore proceed by an
agent trying some, essentially random sequence of innovation by attempting the
introduction of a new practice. Each practice is supposed to have a cost in
implementation and so following the introduction of a practice if the performance
does not increase more than the cost of its introduction then it is considered a failure
and is withdrawn.
In this way, agents attempt to introduce successive practices in random sequences,
and some, by luck, increase the performance of the supply chain much faster than
others. The performance is measured by a weighted sum of the five criteria mentioned
above. If we choose for example .96, .1, .1, .1, .1 then the Quality measure will be the
performance optimized, while if it is .1, .1, .96, .1, .1 then it will be the reliability of
delivery that will be optimized.
The model proceeds by calculating the performance (in terms of the weighted criteria)
for the bundle of practices initially present and adds a random practice. If the gain in
performance is less than the cost of the implementation, then the practice is removed
again. In this way, overall performance increases in steps given by the matrices of
performance, depending on the particular sequence that occurs. Our simulation can be
used to explore the impact of the pair matrix interactions. We can run the average
matrices and see how the average view of managers leads to a choice of practices. We
can compare the results when the off-diagonal elements of the pair matrix are all set
to zero, and when they are small – the values are divided by 50 to approximate an
“order of magnitude” lower than the diagonal terms..
The dialogue between the process “inside” an organization or supply chain and the
demands placed on it by the outside world from which it must wrest a living
constitute a complex, evolutionary system. Firms are forced to proceed by “trial and
error” since the collective interactions are difficult to anticipate, and therefore luck
leads to different firms learning at different rates which bundles of practice lead to
what performance, as is shown in figure (12). The cost per practice is set at 6 which
corresponds to the positive rating of the individual practices’ performance. This sets
the minimum “improvement” that a new practice must bring if it is to be retained. In
reality this would mean that some would be eliminated on these grounds and the faster
learners would succeed. This implies that it is important to try out practices
sufficiently fast, and also that luck will still play a role in who is successful.
Different Sequences
Quality Optimized
Performance
80
70
Series1
60
Series2
50
Series3
40
Series4
30
Series5
20
Series6
10
Series7
Series8
0
1
9
17 25 33 41 49 57 65 73 81 89 97
Series9
Time
Figure (12). Different model runs with different random seeds lead to different rates
of learning about the bundle of practices that respond to the performance required.
Repeated runs with different random number seeds have ensured that the system
results are robust. High performance is obtained by only including practices that
provide the right kind of performance and also are synergetic.
The program can be run with weights corresponding to a performance of high quality
production. We find figure (13).
Practices Maximising Quality
12
10
8
6
Cross 0
4
Cross 1
2
0
1
3
5
7
9
11 13 15 17 19 21 23 25 27
Practices
Figure (13). Practices that lead to high quality production.
The program can now be run with the weights now chosen to maximize successively
the different dimensions of performance. The practices retained for cost efficiency,
reliable delivery, innovation and vision are shown in figure (14).
Cost Maximising individual and interacting practices
Practices for Reliability
12
12
10
10
8
8
Cross 0
6
Independent
6
Interacting
Cross 1
4
4
2
2
0
0
1
1
3
5
7
9
11 13
15 17 19
21 23 25
3
5
7
9
27
11 13 15 17 19 21 23 25 27
Practices
Practices
Vision Practices
Innovation Practices
16
16
14
14
12
12
10
10
Independent
8
Interacting
Independent
8
6
6
4
4
2
2
Interacting
0
0
1
3
5
7
9
11 13 15 17 19 21 23 25 27
Practices
1
3
5
7
9
11 13 15 17 19 21 23 25 27
Practices
Figure (14). The different practices retained for the different dimensions of
performance.
The model allows us to also calculate the performance in all five dimensions that
would be associated with high performance in any particular one of them. This is
important because there will be lower limits of acceptability on any particular mix.
6.1 Patterns of Synergy and Conflict
If we consider the practices selected for maximum quality and look at them on the
pair matrix then we see that the practices that have been retained have a net synergy
of 7, and that no practices have been retained that are in conflict. The patterns for the
practices retained to maximize all five dimensions are shown in figure (15). Again
there are no conflicting practices and the overall synergy is 9.02.
Synergy Conflict for Quality
All positive - Net +ve 7
Synergy Conflict for Cost
All positive - Net +ve 9.02
S27
S27
S25
S25
S23
S23
S21
0.09-0.1
S21
0.09-0.1
S19
0.08-0.09
S19
0.08-0.09
0.07-0.08
S17
0.06-0.07
S15
0.05-0.06
S13
S17
S15
S13
0.04-0.05
S11
0.03-0.04
S9
0.02-0.03
S7
0.01-0.02
27
25
23
21
19
17
15
13
9
11
7
5
3
1
27
25
23
21
19
17
15
9
13
7
11
5
3
1
S1
S23
S21
0.09-0.1
S21
S19
0.08-0.09
S19
0.07-0.08
S15
0.05-0.06
S13
0.04-0.05
0.08-0.1
0.06-0.08
0.04-0.06
0.02-0.04
0-0.02
S11
0.03-0.04
S11
-0.02-0
S9
0.02-0.03
S9
-0.04--0.02
S7
0.01-0.02
S7
-0.06--0.04
0-0.01
S5
27
25
23
21
19
17
15
13
11
9
7
S1
5
S3
S1
3
S3
1
27
25
23
21
19
17
15
S17
0.06-0.07
S5
13
0-0.01
S1
S25
S13
9
0.01-0.02
S27
S15
11
0.02-0.03
S25
S17
7
0.03-0.04
S7
S27
S23
5
0.04-0.05
S9
Synergy Conflict for Innovation
Some conflicting practices - Net synergy +ve 11.7
Synergy Conflict for Reliability
All positive - Net +ve 5.82
3
0.05-0.06
S3
S3
1
0.06-0.07
S11
S5
0-0.01
S5
0.07-0.08
Synergy Conflict for Vision
Some conflicting practices - Net synergy +ve 14.4
S27
S25
S23
S21
S19
0.08-0.1
0.06-0.08
S17
0.04-0.06
S15
0.02-0.04
S13
0-0.02
S11
S9
S7
-0.02-0
-0.04--0.02
-0.06--0.04
-0.08--0.06
S5
27
25
23
21
19
17
15
9
13
11
7
5
3
1
S3
S1
Figure (15). a)Quality, net synergy 7. b) Cost efficiency, net synergy 9.02. c)
Reliability, net synergy 5.82. d) Innovation, has some conflicts but net synergy is
11.73. e) Vision has some conflicts but net synergy is 14.4.
6.2. The Increased performance arising from Synergy
The model manages to select practices that perform much better than when the
interactions between practice are not considered. In other words, the practices retained
are on the whole synergetic. We can look at the patterns of synergy that have been
selected.
Quality
Cost
60.00%
80.00%
50.00%
70.00%
60.00%
30.00%
Quality
Synergy
Synergy
40.00%
20.00%
50.00%
40.00%
Cost
30.00%
20.00%
10.00%
10.00%
0.00%
0.00%
1
3
5
7
1
9 11 13 15 17 19 21 23 25 27
3
5
7
Reliability
9 11 13 15 17 19 21 23 25 27
Innovation
60.00%
80.00%
50.00%
70.00%
Synergy
30.00%
Reliability
20.00%
Synergy
60.00%
40.00%
50.00%
40.00%
Innovation
30.00%
20.00%
10.00%
10.00%
0.00%
0.00%
1
3
5
7
1
9 11 13 15 17 19 21 23 25 27
3
5
7
9 11 13 15 17 19 21 23 25 27
Vision
90.00%
80.00%
70.00%
Synergy
60.00%
50.00%
Vision
40.00%
30.00%
20.00%
10.00%
0.00%
1
3
5
7
9 11 13 15 17 19 21 23 25 27
Figure (16).The increased performance for the five dimensions showing where the
strong synergies are.
This shows us how the synergy term is highly significant, up to 70%, even after
reducing the pair interaction term by dividing by 50 to make it an order of magnitude
smaller than the direct performance. Again very high levels of synergy are found, up
to 75%, and these high levels explain why in some cases the performance that is
obtained in trying to maximise quality, actually produces even higher levels of
performance in innovative characteristics. The effects of the patterns of synergy and
conflict are strong, and therefore the results coming from the retention of particular
patterns are not necessarily obvious.
Knowledge of the effects of interaction can therefore be of considerable advantage in
creating a successful supply chain. Even if the pair interaction terms are considered to
be 50 times smaller than the direct effects of a practice there are still synergy effects
of up to 75%. This shows the importance of considering the systemic, collective
effects of any organization or supply chain.
7. Discussion
The method that we have developed is in fact a completely general approach to the
issue of business and supply chain performance. It is based on the idea that a
successful evolutionary step can only occur if an initial experimental initiative is
amplified by the system in which it occurs (Allen, 1976). The initiative here is in terms
of a “practice” – a particular technique, activity or action that is believed initially will
enhance the collective performance of the system. The model allows us to examine
how different practices will contribute to, or detract from, different possible
dimensions of performance that the overall production system requires in order to
succeed with its customers or in the market place.
Essentially this method comes down to setting up a model that uses the “evolutionary
criterion” of 1976 (Allen, 1976; Allen 1994a and b), and tries to launch a succession
of practices under chosen selection criteria. When the interaction between practices is
ignored then we simply see that if we set the costs at 6, for example, then the model
simply selects and retains those practices that have a performance greater than 6 in
their performance matrix. However, once the interactions are included then as we
have seen the synergy and conflicts between practices can lead to up to 75% higher or
lower than recorded in the non-interacting case. Because of this, the exact identity of
the practices that are retained may change as some have greatly increased
performance, while others suffer much decreased performance. Because the model
selects for practices with combined high performance rather than the opposite, we
find that the supply chain practices suggested by the model give rise to considerably
enhanced performance.
The overall view that emerges is that the collective an organization or supply chain
may differ considerably from that of its component parts taken separately. In other
words, the supply chain or organization forms a “system” which can respond to the
particular demands of its selection environment through its collective interactions.
Over time, it will be the ability of the supply chain to maintain different types of
supply chain corresponding to different stages in the life-cycle of its products.
From the discussions and models presented above we can derive some key points
about organizational behaviour and its evolution. Organizational behaviour must be
such as to allow organizational evolution, or the organization will fail. The rules that
allow organizational evolution are:

The presence of mechanisms that produce internal heterogeneity, which will
involve freedom, ignorance and underlying error-making, exploratory
processes

Differential performance needs to be detected and evaluated with respect to
their alignment with higher level goals. This will then provide the selection
process that will amplify or suppress different elements of individual
behaviour.

The relative performance of the organization within the wider environment
needs to be constantly reviewed and evaluated to see how the selection criteria
are changing and how this may affect the capabilities and competences that the
organization needs.

At the level above the organization, the selection criteria will be changed by
the changes occurring both in the organization considered and the others in the
larger environment. This means that we have a multi-level dialogue going on
between the changes occurring inside an organization and the changes
occurring in the larger environment, the industry or sector, and this within the
changing region, national and global economies.

For successful organizations, as Chia and Tsoukas (2002) point out, aggregate
descriptions will always be short term emergent properties of an evolving
system.
Successful management must behave as evolution does and make sure that
mechanisms of exploration and experiment are present in the organization. Though
they will not be profitable in the short term they are the only guarantee of survival
into the longer term. In reality, the organizations that we observe and describe
formally at any given moment are “structural attractors” (Allen Strathern and
Baldwin, 2005), which, if they persist over time, will change qualitatively as
successive organizational forms emerge.
In hard science a new theory must be capable of being falsified, and therefore must
produce testable predictions, which can lead to genuine cumulative knowledge. In
complex systems however, clean predictions are no longer possible and the
knowledge evolves under less severe selection criteria. In ecological and human
systems emergent structural attractors can occur simply because their particular
emergent capabilities/behaviours succeed in getting resources from the environment.
Fashion, lifestyles, art, artefacts and communities of practice can emerge and survive
providing that there is a clientele for them. They are not about being “true or false”
but simply about whether there is a “market” for them. Living systems create a world
of connected, co-evolved, multi-level structures which may be temporally selfconsistent, but will evolve and change over time. In a market place, we find that firms
need to experiment with their strategy in order to find out how to improve profits in
the moving constellation of other firms. Luck plays a role, but learning will be better
than just hoping. Similarly in our study of automobile manufacturing, we can actually
break down organizational behaviour into its “atomic” components of working
practices, skills and techniques. Complexity tells us that as well as the “organization
you see”, the set of practices, there must also be internal agents with the autonomy to
decide to try out new practices, and to choose which they should be. This is really the
role of management, although hopefully it uses information from throughout the
organization. However, the main point is that although a particular bundle of practices
is what is observable at any given time, a successful evolution will require the
additional presence of agents that suggest new practices, discuss how to bring them in
and implement them, and then who evaluate how they are performing.
Organizational behaviour can be many things, but no organization will survive long if
it is not capable of evolving and adapting over time. Complexity science provides the
scientific framework in which we can understand how and why organizational
behaviour can be evolutionary. It is the natural result of evolutionary drive - a multi-
layered co-evolution of the different levels of description and interaction that emerge
in open, non-linear systems.
Acknowledgement
This work was supported by the ESRC RES-000-23-0845, Aerospace project with
Sheffield University.
References
Allen P. M., 1976, Evolution, Population Dynamics and Stability, Proc Nat Acad Sci,
USA, Col 73, No 3, pp 665-668
Allen P. M., 1994a, Coherence, Chaos and Evolution in the Social Context, Futures,
26, No 6, pp583-597, Butterworth-Heinemann.
Allen P. M., 1994b, "Evolutionary Complex Systems: Models of Technology
Change", In "Chaos and Economic Theory", Eds L.Leydesdorff and P.van
den Besselaar, Pinter, London. ISBN 1 85567 198 0 (hb) 1 85567 202 2 (pb)
Allen P.M., 1998, “Evolving Complexity in Social Science”, in “Systems- New
Paradigms for the Human Sciences”, Eds Altman and Koch, Walter de
Gruyter, Berlin, New York
Allen P. M., 2000, Knowledge, Ignorance and Learning, in Emergence, Volume 2,
Issue 4, p78-103. ISSN 1521-3250
Allen P. M., 2001a, Knowledge, Ignorance and the Evolution of Complex Systems, in
Frontiers of Evolutionary Economics: Competition, Self-Organisation and
Innovation Policy, Eds J. Foster and S. Metcalfe, Edward Elgar,
Cheltenham, UK ISBN: 1840645253
Allen P. M., 2001b, A Complex Systems Approach to Learning, Adaptive Networks,
International Journal of Innovation Management, Vol. 5 No 2 (June 2001)
pp149-180. ISSN 1363-9196
Allen P. M. and McGlade J. M., 1987a, “Modelling Complex Human Systems: A
Fisheries Example”, European Journal of Operational Research 30, pp 147167
Allen P. M. and McGlade J. M., 1987b, Evolutionary Drive: The Effect of
Microscopic Diversity, Error Making & Noise, Foundation of Physics, vol
17, No 7, July, pp 723-728. ISSN 0167-9236
Allen P.M. and Varga L., 2006, “Complexity: The Co-evolution of Epistemology,
Axiology and Ontology”, Nonlinear Dynamics, Psychology, and Life
Sciences, Vol. 11, No. 1, pp. 19-50.
Beer, M. and Nohria, N. (eds) (2000). Breaking the Code of Change. Boston, MA:
Harvard Business School Press.
Brodbeck, P.W. (2002). Implications for organization design: teams as pockets of
excellence. Team Performance Management: an International Journal,
8(1/2), 21–38.
Brown, S.L. and Eisenhardt, K.M. (1997). The art of continuous change: linking
complexity theory and timepaced evolution in relentlessly shifting
organizations. Administrative Science Quarterly, 42, March, 1–34.
Bryant, A. (1998). Beyond BPR: confronting the organizational legacy. Management
Decision, 36(1), 25–30.
Burnes, B. (2004). Managing Change, 4th edition.Harlow: FT/Prentice Hall.
Burnes B. 2005, “Complexity Theories and organizational change”, International
Journal of Management Reviews, Vol 7, Issue 2, pp 73-90
Clarke, M. (1999). Management development: a new role in social change.
Management Decision, 37(10), 767–777.
Cooper, C.L. and Jackson, S.E. (eds) (1997). Creating Tomorrow’s Organizations
Today: A Handbook for Future Research in Organizational Behavior.
Chichester: Wiley.
Dawson, P. (2003). Reshaping Change: a Processual Perspective. London: Routledge.
Dunphy, D., Griffiths, A. and Benn, S. (2003). Organizational Change for Corporate
Sustainability London: Routledge.
Foster R. and Kaplan S, 2001, “Creaive Destruction”, Doubleday, New York
Gould, S.J. and Vrba, E.S., 1982, “Exaptation – a Missing Term in the Science of
Form”, Paleobiology 8(1):4-15. ISSN 0094-8373
Gould, S.J., 2002, The Structure of Evolutionary Theory. Cambridge, MA: Belknap.
ISBN 13 978-0-674-00613-3
Greenwald, J. (1996). Reinventing Sears. Time, 23 December, 53–55.
Harung, H.S., Heaton, D.P. and Alexander, C.N. (1999). Evolution of
organizations in the new millennium. Leadership and Organization
Development Journal, 20(4), 198–207.
Hirooka, M. 2003, “Non-Linear Dynamism of Innovation and Knowledge Transfer”,
Paper presented at the ASEAT & I&IR Conference, Manchester
Huczynski, A. and Buchanan, D. (2001). Organizational Behaviour, 4th edition.
Harlow: FT/Prentice Hall.
Johnson, G. and Scholes, K. (2002). Exploring Corporate Strategy, 6th edition.
Harlow: FT/Prentice Hall.
Kanter, R.M., Kao, J. and Wiersema, F. (eds) (1997). Innovation: Breakthrough
Thinking at 3M. DuPont, GE. Pfizer, and Rubbermaid. New York:
HarperBusiness.
Kotter, J.P. (1996). Leading Change. Boston, MA: Harvard Business School Press.
Maguire, S., McKelvey, B., Mirabeau, L. & Oztas, N., 2006- “Complexity Science
and Organization Studies”, in Clegg, S, Hardy, C., Nord, W. & Lawrence, T.
(Eds.), Handbook of Organization Studies (second edition), Sage, London,
UK, 165-214.
Maynard-Smith J. 1979, “Game theory and the evolution of behaviour” Proc. R.
Soc. London B. 205: 475-488. ISSN: Print: 0962-8452
McCarthy I, 1995, “Manufacturing Classifications: lessons from organisational
systematics and biological taxonomy”, Journal of Manufacturing and
Technology Management- Integrated Manufacturing systems, Vol 6, No 6,
pp37-49 ISSN: 1741-038X
McCarthy I, Leseure M., Ridgeway K and Fieller N, 1997, “Building a
Manufacturing Cladogram”, International Journal of Technology
Management, Vol 13, No 3, pp2269-2296. ISSN (Online): 1741-5276 - ISSN
(Print): 0267-5730
McKelvey B. 1982, “Organizational Systematics”, University of California Press CA
ISBN 0520042255
McKelvey B., 1994, “Evolution and Organizational Science” in Baum J, Singh J,
Evolutionary Dynamics of Organizations, Oxford University press, pp314326. ISBN13: 9780195085846, ISBN10: 0195085841
Metcalfe, J. S., 1998, “Evolutionary Economics and Creative Destruction”,
Routledge, London. ISBN: 0415158680
Metcalfe, J.S 1999, “Restless Capitalism, Returns and growth in Enterprise
Economics”, CRIC , Univ. of Manchester
Peters, T. (1997). The Circle of Innovation: You Can’t Shrink Your Way to
Greatness. New York: Alfred A Knopf.
Romanelli, E. and Tushman, M.L. (1994). Organizational transformation as
punctuated equilibrium: an empirical test. Academy of Management Journal,
37(5), 1141–1166.
Schumpeter J. A. 1942, “Capitalism, Socialism and Democracy”, New York:
Harper, 1975 [orig. pub. 1942], pp. 82-85page 84
Stickland, F. (1998). The Dynamics of Change: Insights into Organisational
Transition from the Natural World. London: Routledge.
Styhre, A. (2002). Non-linear change in organizations: organization change
management informed by complexity theory. Leadership and Organization
Development Journal, 23(6), 343–351.
Tsoukas H. and Chia R, 2002, “On Organizational Becoming: Rethinking
Organizational Change, Organization Science, Vol. 13, No. 5, September–
October, pp. 567-582
Zairi, M., Letza, S. and Oakland, J. (1994). Does TQM impact on bottom line
results? TQM Magazine, 6(1), 38–43.