Table of Contents
FULL TITLE: Dynamics of Competitive Industries: A Micro
Behavioural Framework*
SHORT TITLE: Dynamics of Competitive Industries
Martin H. Kunc
London Business School
Regent’s Park
NWI 4SA London
Tel +44 207 262 5050 x3507
Fax +44 207 7724 7875
email mkunc@london.edu
John D. W. Morecroft
London Business School
Regent’s Park
NWI 4SA London
Tel +44 207 262 5050
Fax +44 207 7724 7875
email jmorecroft@london.edu
*The research reported in this paper has been funded in part by the PhD Program at London Business
School. Additional funding has been provided by Telefonaktiebolaget LM Ericsson and Mars Inc
through the System Dynamics Group at London Business School.
Abstract
Most published work in business dynamics is conducted either at the level of the
individual firm or at the level of an industry comprising an aggregate of similar firms.
However, there are situations where the performance of industries is better understood
by modelling the behaviour of competing individual firms. When firms in the same
industry adopt quite different views of the ‘best set’ of resources and the overall system
of resources in the industry is tightly interconnected, it is important to model the
heterogeneity of rival firms. We propose a micro-behavioural approach that captures the
essential interactions between firms. To illustrate our approach we run a series of
experiments using Fish Banks, Ltd. to show the wide range of firm and industry
performance arising ftom such heterogeneity. We further develop our micro-behavioural
approach into a framework for understanding the dynamics and evolution of industries
based on selected ideas from system dynamics, the resource-based view of the firm and
managerial cognition.
KEYWORDS: INDUSTRY DYNAMICS; MICRO BEHAVIOURAL MODELLING;
COMPETITION AND RIVALRY; INDUSTRY EVOLUTION; MANAGERIAL
COGNITION; FISH BANKS
Introduction
The field of system dynamics’ has adopted two distinct perspectives to analyse the
dynamics of business performance. There are individual firm models and there are
aggregate industry models. But there are few models pitched at an intermediate level.
In firm-level modelling the dynamic behaviour of individual firms is assumed to
be generated endogenously. The competitive environment is represented passively and
exogenously by specifying benchmarks for competitive factors such as delivery delay,
price or quality that reveal the relative attractiveness of the firm’s product or service to
the customer. For example, Forrester’s (1966) widely cited “market growth model”
represents a firm that fails to grow even under the assumption of an unlimited market.
Growth is prematurely stifled by dysfunctional and unintended interactions between
operating policies for expanding salesforce and manufacturing capacity that inadvertently
result in high delivery delay. In such models, the purpose of the exogenous benchmark is
not to mimic the behaviour of competitors but rather to represent, as concisely as
possible, the external standards by which customers judge product attractiveness
(Forrester, 1961). These individual firm-level models (e.g.: Hall, 1976) have been very
important in understanding puzzling, dysfunctional behaviour and underperformance of
firms under circumstances where any individual firm’s actions do not significantly alter
the environment, or where feedback effects to the environment are small within the time
horizon defined by the modeller.
System dynamics researchers have also developed models of aggregate industry
dynamics. For example, Forrester (1961) presented a highly aggregated model of
manufacturing industry to understand the dynamics of supply chains and their
contribution to business cycles. Among others, Sterman (2000) developed a generic
model of the behaviour of commodities industries, which is based on the relationships
between production, inventory, capacity utilization, capacity acquisition, demand and
price. Related to these industry models, Sterman (1987) suggested that the continuous
tule used to describe aggregate behaviour of an industry in system dynamics behavioural
simulation models while it is not an exact statement of how firm-level decisions are
made, may be an acceptable simplification’.
However, not all business dynamics problems can be modelled as individual firms
or as aggregate industries. Industry evolution is one important exception. During the
evolution of industries, the process of mutual adjustment between heterogeneous firms is
particularly relevant because the actions of individual firms sooner or later influence the
responses of other firms in the same industry. In other words, operating policies are
contingent on other firms’ operating policies. Schelling has described similar contingent
behaviour among individuals who comprise social aggregates (1978: p. 14 and 17), of
which rival firms are particular examples:
People are responding to an environment that consists of other people responding to their
environment, which consists of people responding to an environment of people’s responses.
Sometimes the dynamics are sequential... Sometimes the dynamics are reciprocal.
The goals or purposes or objectives [of people] relate directly to other people and their
behaviour, or are constrained by an environment that consists of other people who are pursuing
their goals or their purposes or their objectives. What we typically have is a mode of contingent
behaviour - behaviour that depends on what others are doing." (emphasis added to the
original).
Schelling also offers a hint on the type of analysis necessary to address this view of
the firm-environment relationships (1978: p. 14):
These situations, in which people’s behaviour or people’s choices depend on the behaviour or
choices of other people, are the ones that usually don’t permit any simple summation or
extrapolation to the aggregates. To make that connection we usually have to look at the system of
interaction between individuals and their environment, that is between individuals and other
individuals or between individuals and the collectivity.... Sometimes the analysis is inconclusive.
But even inconclusive analysis can warn against jumping to conclusions about individual
intentions from observations of aggregates, or jumping to conclusions about the behaviour of
aggregates from what one knows or can guess about individual intentions. (emphasis added to
the original)
We propose a modelling framework, suitable for analysing medium-term dynamics of
firms in fast evolving industries (or equivalently, long-term dynamics in slowly evolving
industries) based on contingent behaviour. In this framework an industry is represented
as two or more heterogeneous individual firms, strongly interconnected through their
shared environment. Here the industry environment for any individual firm is
endogenous and includes rival firms as well as shared customers. Firm performance is no
longer judged relative to fixed industry standards but instead evolves from interactions
within a network of heterogeneous decision-makers in rival firms, each configuring a
system of resources or strategic asset stocks to achieve a sustainable competitive
advantage. Moreover, these decision-makers, as boundedly rational actors (March and
Simon, 1958; Morecroft, 1983), do not necessarily agree on which particular
configuration of resources is ‘best’ to serve their shared market (in other words they
perceive the intended system of resources differently). Nor, when deciding which
resources to expand or contract, do they necessarily give equal weight or importance to
known resource imbalances, shortages or surpluses. Therefore, their decisions cannot be
aggregated because they are based on different assumptions.
When there are important enduring ‘cognitive asymmetries’ between decision-makers
in rival firms then individual firm performance cannot be reliably inferred from a single-
firm model. Equally, industry evolution cannot be deduced from an aggregate industry-
level model. For example it is possible that some strategies may be dysfunctional when
they are pursued by most of the firms in an industry, but they can provide a competitive
advantage when only one individual firm as a leader employs them, such as ‘first-mover
advantage’ (Lieberman and Montgomery, 1988),
Our contribution to the system dynamics literature is to present a frog-pond’
theory for the evolution of industries and firm performance (Klein; Dansereau, and Hall,
1994). This type of theory has two distinctive features: the effects of variables are
context dependent, and a comparative process is used to specify heterogeneity among
individuals within the group. Consequently, this paper presents a framework to analyse
the evolution of industries and its effect on the performance of firms from a micro-
behavioural point of view. To introduce the framework, we first present the results of a
series of experiments using Fish Banks Ltd (Meadows; Fiddaman, and Shannon, 1993).
This widely known simulation game of the fishing industry illustrates the meaning of
important constructs like heterogeneous rivals and cognitive asymmetries, and
demonstrates their pervasive effect on firm and industry performance. We then continue
to develop the framework. Finally, we conclude with some implications for system
dynamics researchers and practitioners.
Experimental Setting
The use of the fishing industry as a metaphor to illustrate the framework
A commercial fishery is a self-contained industry comprising natural fish stocks and
multiple rival firms each operating a small fleet of ships. These firms are interconnected
because they share a regional population of fish. Sometimes the interconnection and
mutual dependence is very strong and generates surprisingly complex dynamics in the
real world. The feedback structure of such a fishery is simple but dynamically complex
due to the effect of nonlinearities and interactions between participants. The essence of
the managerial dynamical problem for fishing fleet operators (the firms) is to achieve a
sustainable competitive advantage and growth while maintaining the ‘right’ balance
between the natural renewable resource, fish, and the man-made resource, ships.
However, the ‘Tragedy of the Commons’ (Gordon, 1954; Moxnes, 1998) characterizes
the typical and dismal dynamics of this industry and firms’ dynamic behaviour.
Individual firms, that try to maximise their wealth and their share of the catch, find
themselves engaged in a race to grow until, unexpectedly, the natural resource collapses.
The usual and catastrophic outcome is an ocean without fish and, at the same time, large
idle fishing fleets.
Different firms adopt different resource building policies and strategies. Even in
a deliberately simplified experimental setting, it can be difficult for any individual firm to
interpret rivals’ policies and behaviour, and even more difficult to infer the resulting
diverse industry dynamics. It is a huge challenge for any firm (and management team) to
survive in such dynamically complex conditions where performance is so critically
dependent on an appropriate balance of resources. Here the regeneration rate of fish is a
non-linear function of an imperfectly known fish stock and the catch depends on diverse
motives and actions of rivals as they build and deploy their fleets.
Experiment Description
While Fish Banks is chiefly a role-playing simulation game that illustrates the
management (and mismanagement) of renewable natural resources, we used this role-
playing game to observe competitive behaviour of teams and its effect on the dynamics
of the industry.
The Fish Banks experiments involved 28 teams from Executive and MBA
programs grouped into 5 separate competitive environments. Each team, whose
members were chosen randomly, had to manage a fishing company competing against 5
other teams in the same ocean. All participants received the same information at the
beginning of the game, and we allowed teams 30 minutes to discuss strategy before
making their first decision. Each team’s objective was to maximize asset value by the end
of the game, where asset value is defined as the salvage value of the fishing fleet plus the
accumulated bank balance.
The game is a good illustration of our proposed dynamics of competitive
industries framework due to several distinctive design features:
* each team started with identical internal resources (ships and cash) and received the
same estimates for the size range of the fish population (shared external resource)
and its regeneration rate,
+ all teams had similar productivity per ship and received the same price per fish, so
their income was determined entirely by their fleet size and fleet allocation (resource
configuration) among fishing areas,
all teams had access to the same information: competitors’ actions like fleet
allocation among the fishing areas, competitors’ fleet size and competitors’ total
assets,
the business concept was very simple as well as the set of choices: expand or not
expand the fishing fleet, allocate the fishing fleet between two fishing areas or not
send to fish at all, and trade or buy ships from other teams, and,
there were three uncertainties similar for all teams: fish population size, real catch
rate per ship, and competitors’ intentions/strategies.
the true set of asset stocks or resources was known by the experimenters, and,
consequently, the experimenters could compare this known set with teams’ different
perceptions of the resources and its effects on the decision-making process.
Each management team had two basic decisions: fleet expansion and fleet
allocation. These two decisions parallel basic concepts in the resource-based view
literature: fleet expansion corresponds to the resource accumulation process (Dierickx
and Cool, 1989); and fleet allocation represents the concept of resource configuration
(Teece; Pisano, and Shuen, 1997; Adner and Helfat, 2003). A representation of the
resource system using a stock and flow diagram is shown in figures | and 2.
INSERT FIGURE 1 HERE
Figure | presents the external resources. The external resources consist of two
stocks ‘Fish Population Coastal Area’ and ‘Fish Population Deep Sea Area’ representing
two independent fish populations. Each fish population increases through a regeneration
rate, which depends on the relationship between the actual fish population and a
maximum natural fishery size. Each fish population decreases through a harvest rate
proportional to the allocated fleet size. The fish population in either area (coastal or
deep sea) has a natural limit to growth and can fall to zero if the outflow from harvesting
exceeds the inflow from regeneration over a period of time long enough to deplete the
resource. Thus, the dynamic behaviour of the external resources in this imaginary fishing
industry depends on an adequate balance between regeneration and harvest rate. If teams
build big fleets, the harvest rate will be higher than the regeneration rate and all teams
will lose money. If teams build small fleets, they may not optimise the economic
exploitation of the fish population. The exact size of these two fish populations and the
regeneration rate was not revealed to the teams at any point in the game. Participants
knew only the size range of the initial populations which was 2000-4000 fish for the deep
sea and 1000-2000 fish in the coastal area.
INSERT FIGURE 2 HERE
There are four internal resources. Three of them - named ‘Ships in Coastal Area’,
‘Ships in Deep Sea Area’ and ‘Ships in Harbor’- represent the fleet of each team. Ships
at sea contribute to an outcome (the overall catch) according to their deployment in the
two fishing areas. The catch is equal to the number of ships (the level of the resource)
multiplied the productivity per ship. The intrinsic productivity (“Catch per ship’), which
is equal for all teams, depends non-linearly on the fish population. Each team has two
decisions: fleet size or resource accumulation (shown as a diamond named ‘Team’s Fleet
Size Decisions’), and fleet allocation or resource configuration (shown as a diamond
named ‘Team’s Fleet Allocation Decisions’). Basically, each team decides a goal for its
fleet size, and, either acquires new ships from the shipyard or trades ships with rivals
(the resource accumulation decision is reflected in the flow regulator named ‘Fleet Size
Change’). In addition, each team adjusts its deployment of ships (by controlling the
resource flows named ‘Ship allocated to Coastal area’, ‘Ship transferred between fishing
areas’ and ‘Ships allocated to Deep sea area’). Thus, team’s decisions affect not only the
size of their firms but also the external resource in an uncertain and competitive
environment.
Finally, the fourth resource -‘Bank Account’- reflects the monetary effects of the
resource configuration such as income and operating costs. Increases or decreases in the
bank account are influenced by fleet size decisions (resource accumulation) and the
results of fleet allocation (resource configuration). As mentioned previously, the
resources that define team performance are ‘Bank Account’, ‘Ships in Coastal Area’,
“Ships in Deep Sea Area’ and ‘Ships in Harbor’.
Data Description
We used four data sources to capture the teams’ decision-making processes and
their effect on the environment and firm performance: decision forms, computer
generated results, team notes and subjects’ comments about their team performance. The
duration of the role-playing game is restricted to 10 periods, but normally it is much
shorter because there are no fish left in the ocean. Most of our experiments lasted just
five periods. One experiment lasted only four periods, while another lasted 8 periods.
We analyse the information obtained from the experiments using two variables:
‘Total assets’ and ‘Ships’. ‘Total assets’ measures the overall performance achieved by
each team. Total assets are the sum of the resource ‘Bank account’ and the salvage value
of the team’s fleet (sum of the resources ‘Ships in Deep Sea Area’, ‘Ships in Coastal
Area’, and ‘Ships in Harbor’ multiplied by the salvage value per ship). The final value of
total assets is the outcome of two separate processes: the number (and salvage value) of
ships reflects each team’s emphasis on fleet expansion, while the final level of the ‘bank
account’ captures the ability of the team to manage the resource configuration
effectively. Teams that manage their resources strategically achieve higher net income
and a larger bank balance.
‘Ships’ represents the outcome of the decision-making process controlling
internal resource accumulation. Teams expand their fleets based on their beliefs about the
best structure of the resource system and the information received from the evolution of
external resources.
Results
We found some results surprising. Figure 3 is a scatter plot of internal resources
(ships) and performance (total assets). While a high number of ships often implies low
total assets, due to operating losses caused by the over expansion of the fleet, there is
too much variation in individual team performance to assume that aggressive expansion
is the sole cause of underperformance, or that aggressive expansion can never yield
superior performance (as an aggregate model of fishery dynamics would suggest).
INSERT FIGURE 3 HERE
12
Even with identical initial resources, access to similar information, and subject to
the same natural fish regeneration constraints, teams neither followed a unique strategy
nor achieved identical performance. Some teams with similar final fleet size achieved
polar opposite performance. For example the lowest and highest performing teams (in
terms of total assets) both acquired around 60 ships. Other teams obtained similar
performance with huge differences in resources (for example total assets of
approximately $10,000 for teams which had acquired 10, 20 and 55 ships respectively).
Even two teams in the same experiment with the same number of ships achieved a 15%
difference in performance (teams 3 and 4 in experiment 2). Table | presents year-by-
year results for all five experiments while Table 2 provides summary statistics for each
experiment at the end of year 5.
INSERT TABLE 1 HERE
INSERT TABLE 2 HERE
The results show important variations in resources and performance between and across
experiments as Table 2 depicts. The individual teams in experiments 2, 3 and 5
performed well. They all achieved positive total assets by the end of year 5. Moreover,
the overall performance of teams is similar. The standard deviation of total assets is low
and the mean is high. In contrast the individual teams in experiment 1 and 4 performed
poorly. Many teams were left with negative total assets by the end of year 5 and
collectively they lost money. There were also large differences between the overall
performance of teams and the size of their resources. The standard deviation of both
total assets and ships is high. Surprisingly, we obtained these contrasting strategies and
performances in a relatively simple resource system (only two resources: fish and ships)
with identical information available to players. We therefore need to understand the
factors that have contributed to such dissimilar outcomes. While there are many
potential reasons why firms perform differently, the characteristics of Fish Banks, Ltd.
(see experiment description) reduce the possible sources of dissimilar performance to
only two: 1. each team’s own process of decision making, and 2. the effect of competing
teams’ actions on the effectiveness of a given team’s decisions. In other words, any
individual team’s performance is not only determined by its own decision-making process
but is also contingent on other teams’ behaviour.
Categorising teams’ decision-making process
Given the simplicity of this system of resources we have been able to categorize the
decision-making processes in terms of the observed relationship between fishing fleet size
and fish sales. We identified three distinct styles of decision-making: reactive, proactive
type 1 and proactive type 2 — corresponding to three regular patterns of behaviour.
Figure 4 illustrates these categories, using results from teams 1,2 and 5 in experiment 2.
INSERT FIGURE 4 HERE
These three teams exhibited diverse performance typical of the categories they
represent. The team labelled ‘reactive’ reacted to growth in sales by continuously adding
ships: more sales, more ships. The team labelled ‘proactive type 1’ tried proactively to
achieve and sustain a pre-planned market share by expanding its fleet almost
instantaneously before obtaining any feedback from its actions on the evolution of fish
sales or observing the actions of competing teams. On the other hand, the team labelled
‘proactive type 2’ tried proactively to avoid losses due to an anticipated collapse of the
fish population, and, consequently, not only didn’t expand its fleet but also sold it before
the game finished.
From an analysis of teams’ decisions, notes and comments, we identify the basic
characteristics of reactive or proactive teams. Reactive teams (65% of all teams)
accumulated fleets varying in size between 6 and 20 ships. We could not infer from their
behaviour that they were consciously managing the limited resource (fish population).
Although they seemed to adopt cautious expansion, they myopically kept expanding their
fleet even when industry-wide fish sales were decreasing, as can be seen in figure 4. A
typical reactive team simply followed the local outcome information provided (team fish
sales) without any indication that team members foresaw the industry-wide effect of their
decisions to build internal resource (fleet size) on the external resource (fish population)
or on rivals’ decisions. For example reactive teams increased their fleet when they
observed growth in the volume of fish caught, but they failed to notice related erosion of
the external resource (which could be inferred from productivity per ship). Reactive
teams also configured their resources on the basis of outcome information. For example
they kept their fleets allocated to a particular fishing area until they found there were no
fish left. To summarize, reactive teams seem not to perceive any causal relationship
between their actions and the dynamic behaviour of the external resource. They appear
to follow simple linear cause-effect logic for the dynamic management of asset stocks
based on immediate local outcome information.
15
Most of the proactive teams (35% of all teams) had fleets between 0 and 5 (type
2) or bigger than 20 ships (type 1). Proactive teams seem to have analysed and inferred
causal relationships between internal resources and the external resource. They planned
ahead by judgmentally forecasting the effect of the two principal uncertainties: fish
population size and competitors’ actions. Proactive teams type 2 were concerned with
the effect of the actions of competitors on the external resource (fish population) and
guessed that other teams would build their fleet slowly. Consequently, they built their
own fleet quickly, and allocated it for a short period in one fishing area, then moved it to
a second area, and sold the fleet before the fish disappeared (fleet = 0 ships and high
bank account); or they did not expand the fleet and sometimes sold it, as in the example
presented in figure 4, when they observed aggressive expansion by other teams.
Proactive teams did not wait for outcome information to confirm their forecasts; they
somehow visualized the likely dynamics of the feedback system and acted on their
expectations.
Other proactive teams, which we called type 1, also tried to guess their
competitors’ actions, but they were focused on the effect of competitors’ actions on their
own internal resource accumulation, particularly the bank account. Hence, they built
huge fleets in an attempt to pre-empt other teams without waiting to receive any
outcome information about external resources or competitors’ actions. But only one of
these pre-emptive teams (team | in experiment 2, figure 4) was very successful (55 ships
and total assets of $11,000). This team not only configured its resources by considering
the external resource dynamics, but also benefited from the reactive behaviour of
competing teams in the same experiment. In conclusion, proactive teams appear to
perceive causal relationships between their actions and the dynamic behaviour of the
resources. Using these perceptions, they develop expectations to feed their decision-
making process. Some expectations anticipate the performance of both internal and
external resources; other expectations are simpler and only anticipate the effect of
competition on the performance of internal resources. Table 3 presents some comments
from the teams as anecdotal confirmation of a relationship between styles of
decisionmaking and patterns of behaviour.
INSERT TABLE 3 HERE
While the two main categories (reactive and proactive) help to illustrate that
decisionmaking style and mental models influence performance, their effect on any one
team’s overall performance cannot be directly inferred as a simple recipe for success.
Performance is also contingent on the behaviour of competitors.
The influence of competing teams on performance and the effectiveness of the decision-
making process
Each individual team’s actions are important to its overall performance, but competitor
teams’ actions are also influential because rival firms are strongly interconnected through
their environment. Consequently, we need to consider the relationships between teams’
decision-making processes to interpret industry and team performance. Table 4 shows
the performance of each type of team depending on the proportion of other types of
behaviour in the experiment.
INSERT TABLE 4 HERE
We can observe in Table 4 the effect of the competitive situation (proportion of
teams’ behavioural type) on the performance of each team in the experiments. For
example, in experiment 3 all teams except one adopted reactive decision-making as they
continuously expanded their fleets in response to revenue growth. Teams’ performances
were quite similar (as reflected in the small standard deviation in performance) for two
reasons: (a) they expanded gradually, and (b) they moved their fleets together from one
fishing area to the other area when the first fishing area collapsed. In experiment 5, all
teams except one adopted reactive decision-making, again achieving similar performance.
In this experiment, however, the only proactive team (team 6) exploited other teams’
reactive decision-making by expanding and allocating its fleet aggressively and then
finally selling it to another team. But, in comparison to other proactive teams, team 6
recognised that the natural dynamics of fish regeneration would impose a limit to the
overall number of ships. So the team quickly built a fleet big enough to obtain a
reasonable income and market share without causing a collapse of the fishery. In
experiment 2, three teams adopted reactive decision-making, leaving the exploitation of
the fish population to an aggressive proactive team (type 1) that obtained superior
performance. In this experiment, a second proactive team (type 2) sold its fleet early
expecting a collapse, a move that helped the aggressive proactive team. In experiments 1
and 4, two teams simultaneously adopted very aggressive proactive decision-making
(type 1) trying to pre-empt the other teams. But this duplicate pre-emption caused an
early collapse of fish stocks (the external resource), adversely affecting all teams except
the proactive type 2 teams, which sold or did not expand their fleet expecting the
collapse of the fishery.
To summarise, the results show that the dynamic behaviour of individual firms is
not always a reliable guide to aggregate industry dynamic behaviour in competitive and
tightly coupled resource systems. Heterogeneous decision-makers in rival firms perceive
the industry’s feedback structure differently and adopt different policies and strategies
for resource building and growth.
There are two important implications for system dynamics models of industries
and firms. First, the dynamics of industries cannot always be deduced by modelling an
aggregation of individual firms and by assuming these firms share a common feedback
structure. Second, dysfunctional behaviour of individual firms does not always arise
from flawed internal feedback structure but may also stem from competitive interactions
among rival firms. To address these implications we now develop a general modelling
framework for industries comprising heterogeneous rivals.
Competitive Industry Dynamics: A Micro Behavioural View
In our proposed framework, two main factors determine the performance and dynamic
behaviour of rival firms in an industry. First, the set of resources that define the industry
(both internal to individual firms and external) is important. Rival firms as open systems
not only acquire resources from their environment but also lose resources either to
competitors or through attrition in dynamic interactions with their environment (Warren,
2001; Warren 2002). Thus, organizational survival in competitive industries is based on
the ability to acquire and maintain resources from an environment consisting of rival
organizations, which compete for shared resources or own the resources required for
surviving and prospering.
Organizations’ actions aimed at meeting their own goals can, under conditions of
intense rivalry, affect the resource system of other organizations, thereby generating
reactions that later influence their own resources. External environments are not
completely exogenous but are in part created by the organization and its decisions.
Consequently, organizations have to fit into patterns of resource exchanges with other
organizations forming adaptive systems embedded in feedback processes (March and
Simon, 1958; Levinthal and Myatt, 1994). In these circumstances we need to observe
and model the interactions between firms to understand both industry-level and firm-level
dynamics. Moreover, we conceive of environment-strategy-structure alignment by firms
as a feedback process of mutual adjustment between firms exchanging, sharing and
competing for resources as outlined in figure 5.
INSERT FIGURE 5 HERE
Second, managerial decisionmaking is important. The dynamic complexity of
industries comprising interlocking resources suggests that differences in the way
managers interpret this complexity, set priorities and guide resource building will affect
relative performance and even the survival of firms in competitive industries.
The Role of Management and Managerial Decision-making
In system dynamics, management is viewed as the process of converting information into
action. This conversion process is decision-making. As Forrester (1961, 1994) notes,
“if management is the process of converting information into action, then management
success depends primarily on what information is chosen and how the conversion is
20
executed. The difference between a good manager and a poor manager lies at this point
between information and action”. The difference between a high performing firm and a
less-well performing rival also lies at this point.
In our framework we build on this view of management by separating managerial
decision-making into two distinct information processing components. There is operating
policy to control the acquisition and composition of resources, and there is strategic
resource conceptualisation to define which resources the business really needs.
Operating policy is normally represented as purposive (though myopic)
adjustment of asset stocks or resources through goal-seeking information feedback
(Sterman 2000, Morecroft 2002). It is the essence of the feedback view of the firm.
Decisions stemming from operating policy lead to corrective actions intended to close
observed gaps between desired and actual resources. Defining and monitoring the gaps
(shortages or excesses) in a firm’s portfolio of resources is essentially an information
processing activity. System dynamicists recognise that such information processing is
imperfect, judgmental and behavioural — subject to the practical constraints of bounded
rationality (Morecroft 1985, Sterman, 1985; Sterman, 2000 ch. 13). Every manager has
available a large number of information sources to gauge the firm’s resources. But each
selects and uses only a small fraction of all available information. Through this
behavioural decision-making process, managers collectively build and configure the
resources for competing in the industry. Here desired resource levels are local operating
goals, loosely linked to overall strategy. In a well-designed firm, the achievement of
local resource goals will lead to successful implementation of strategy. But that’s an
ideal world. In reality firms inadvertently adopt operating policies at cross-purposes with
strategy that degrade performance. Underperformance, arising from misperceptions of
21
feedback, has been documented in experimental studies (Sterman, 1989; Paich and
Sterman, 1993), and is the explanation of firm-level performance paradoxes such as
capability traps in process improvement (Repenning and Sterman, 2002) and
implementation failures in product innovation (Repenning, 2002).
Purposive, boundedly rational asset stock adjustment, with misperceptions of
feedback, is a cornerstone of contemporary applied research in business dynamics. We
adopt this approach and add to it a second component of managerial decision-making
that we call strategic resource conceptualisation. We view this conceptualisation activity
as strategic decisionmaking by top managers to define and communicate the resources
they will need to realise their vision of the business.
The process of resource system conceptualisation is difficult to pin-down with
precision, but we believe it is related to top managers’ mental models of the intended
resource system and the expected sources of competitive advantage. In other words,
each manager has a blueprint in his or her mind of the system of asset stocks that drives
performance and dynamic behaviour of the firm over time. Collectively these blueprints
determine the resource building strategy as well as the markets in which the firm
competes. As Senge (1999: 175) suggests “our mental models determine not only how
we make sense of the world, but how we take action.” Mental models affect what we
see, and two people with different mental models can observe the same industry or even
the same firm, and yet define the intended resource system differently. For example, in
the Fish Banks’ experiments we identified three distinct styles of decision-making. We
suggest these styles reflect different mental models for managing a fishing fleet. In other
words, teams conceptualised differently the system of resources that they need to
22
observe and control in order to maximize asset value. Consequently, they developed
diverse strategies’.
Figure 6 is a stock and flow diagram to represent the mental model that we
believe guided ship purchasing in reactive teams. The content of this diagram requires
careful interpretation since we do not necessarily think that a player’s mental model is
literally a stock and flow diagram. Indeed such a one-to-one correspondence is most
unlikely. Rather the figure conveys an impression of the scope and complexity of a
typical reactive mental model for purchasing. The symbols in black capture a myopic, yet
pragmatic view of fleet adjustment informed by the recent history of fish sales growth.
In other words, a reactive team will continue to expand its fleet providing fish sales are
rising. The greyed-out symbols show what’s going on with the fish population, but that
doesn’t directly influence fleet expansion. The assumption is that, for the purpose of
purchasing, reactive teams do not include the fish population in their mental
representation of the system of resources. (However, that’s not to say they are unaware
there are fish in the sea. The problem may be they don’t know for sure how many, and
they are reluctant to guess, so they ignore population and effectively assume it will take
care of itself). A similar myopia applies to competitors. Reactive teams do not appear to
include competitor’s resources in their mental representation for purchasing.
INSERT FIGURE 6 HERE
In comparison, the mental model of proactive type 1 teams was of greater scope
and complexity. It seems likely they considered both their internal bank account and
competitors’ fleets when purchasing. The part of the network shown in black in figure 7
23
portrays this more ambitious mental model. A typical proactive type 1 team viewed
competing teams’ ships as a threat to their own ability to maximize the value of the bank
account. Moreover, they viewed themselves as engaged in a race to expand their own
fleet more quickly than rivals. Consequently they paid attention to competitors’ fleet size
and even tried to anticipate rivals’ fleet expansion. However, despite this extra
sophistication, fish population (which is the main resource affected by a competitive race
and a limit to growth) was probably not included in their mental representation of the
system.
INSERT FIGURE 7 HERE
The system of resources conceptualised by proactive type 2 teams was even more
complex. These teams seemed to identify ships as a necessary resource to drive sales
and, consequently, to accumulate money in the bank account. But, they also worried
about the operating costs of the fleet. In their purchasing strategy, a focus on operating
costs (relative to revenue) caused them to think about the behaviour of the main shared
resource, fish population, and its likely impact on net income per ship. Consequently,
these teams had in mind a more comprehensive model of the system, as figure 8 shows.
Moreover, they still faced an important uncertainty: competitors’ fleet expansion rate. So
their purchasing decisions were contingent on their perception of rivals’ behaviour.
Some proactive type 2 teams played against cautious slow-to-expand competitors. These
teams, operating in a benign competitive environment, were able to achieve high market
share that helped them to maximize their income. Other teams faced aggressive
competitors that built huge fleets quickly. These teams, operating in a hotly contested
24
competitive environment, recognised the need to sell their fleet before the fishery
collapsed.
INSERT FIGURE 8 HERE
In this interpretation of mental models for strategic decision-making we are
implicitly assuming that top managers conceptualise their firm and strategy in terms of
resource building. The strategies of rival firms may be guided by quite different imagined
resource maps that reflect the particular shared vision of their top management teams and
the practical opportunities and threats they perceive. A well-known example in the
system dynamics literature of a competitive industry viewed differently by rivals is
airlines (Sterman, 1988; Morecroft, 1999). While airplanes are a very common and
tangible resource, competitors in the industry deploy them quite differently. For example,
figure 9 depicts how the management of easyJet, one of the biggest low fare airlines in
Europe, conceptualises aircraft usage and cost compared to full fare competitors, the
traditional carriers.
INSERT FIGURE 9 HERE
However, mental models of resource systems are not right or wrong per se.
Rather it is the context in which mental models are applied that determines their
effectiveness. We define the enduring differences between managerial mental models in
an industry as ‘cognitive asymmetries’. Managers can exploit cognitive asymmetries to
find resource system configurations overlooked by competitors, that are nevertheless
25
highly profitable. However, when cognitive asymmetries are small, the key resources -
where managers focus their attention and effort - are similar for all firms, suggesting that
rival firms are likely to follow similar strategies.
To summarize, we view managerial decision-making in competitive industries as
the resultant of two separate components. The first, operating policy, represents how
organisations and functional managers guide the configuration of the resource system
using goal-seeking feedback. The second, resource system conceptualisation, represents
how, through mental models, top managers collectively identify and communicate the
intended resource system.
Conclusion
The field of system dynamics has paid relatively little attention to interactions between
competing firms when analysing the dynamics of business performance. Instead
researchers and practitioners have tended to develop individual firm models or aggregate
industry models. However, competitive interactions can shape the destiny of industries
as well as the performance of individual firms. To illustrate we use the well-known Fish
Banks gaming simulator as a practical example of rivalry among heterogeneous firms in
the same industry. While the ‘tragedy of the commons’ is a typical result of the game, we
observe that some fisheries perform much better than others and that some teams achieve
sustained positive performance over the lifetime of the fishery while others fail
dramatically.
Building on these results we propose a modelling framework for examining the
performance of rival firms in evolving industries. In this framework an industry is
represented as two or more distinctive individual firms, each advocating a different view
26
of strategically important resources and each pursuing somewhat different resource-
building policies, strongly interconnected through their shared environment and shared
customers. Firm performance no longer arises solely from the internal policy interactions
of individual firms but also from interactions among the rival firms and their
heterogeneous decision-makers, as they attempt to configure a unique system of
resources in order to achieve a sustainable competitive advantage.
When there are important enduring ‘cognitive asymmetries’ between decision-
makers in rival firms then individual firm performance cannot be reliably inferred from a
single-firm model. It is then important to explore industry and firm performance under a
behavioural paradigm that explicitly recognises these cognitive asymmetries and their
effect on feedback structure. The system dynamics literature already offers a rich
process for capturing managerial knowledge in feedback models of individual firms and
their internal policy structure (Morecroft and Sterman, 1994; Zagonel, 2002). Our
framework for competitive industries calls for a similar process to capture the different
ways that executives, in rival firms, conceptualise and manage strategically important
resources.
27
REFERENCES
Adner, R. and Helfat, C.E. 2003. Corporate Effects and Dynamic Managerial
Capabilities. Strategic Management Journal 24: 1011-1025.
Barney, J.B. 1991. Firm Resources and Sustained Competitive Advantage. Journal of
Management 17: 99-120.
Barney, J.B. 2002. Gaining and Sustaining Competitive Advantage 2nd Ed. Prentice
Hall: Upper Saddle River, NJ.
Dierickx, I. and Cool, K. 1989. Asset Stock Accumulation and Sustainability of
Competitive Advantage. Management Science 35: 1504-1511.
easyJet. 2003. “Why the big airlines can’t compete with our prices.” obtained on =
January, 2004 from https://www.easyjet.com/EN/about/aboutourfares.html .
Forrester, J.W. 1961. Industrial Dynamics. Productivity Press: Cambridge, MA.
Forrester, J.W. 1966. Market Growth as Influenced by Capital Investment. Industrial
Management Review 9(2): 83-105.
Forrester, J.W. 1994. Policies, Decisions and Information Sources for Modeling . In
Morecroft, J.D.W. and Sterman, J.D. (eds) Modelling for Learning . Productivity
Press: Portland, OR.
Gordon, H.S. 1954. The Economic Theory of a Common-Property Resource: The
Fishery. Journal of Political Economy. 62: 124-142.
Hall, R.I. 1976. A System Pathology of an Organization: The Rise and Fall of the Old
Saturday Evening Post. Administrative Science Quarterly 21:185-211.
Hofer, C.W. and Schendel, D.E. 1978. Strategy Formulation: Analytical Concepts.
West Publishing Company: St. Paul, MIN.
28
Klein, K.J.; Dansereau, F., and Hall, R.J. 1994. Levels Issues in Theory Development,
Data Collection, and Analysis. Academy of Management Review. 19: 195-229.
Levinthal, D.A. and Myatt, J. 1994. Co-Evolution of Capabilities and Industry: The
Evolution of Mutual Fund Processing. Strategic Management Journal. 15: 45-62.
Lieberman, M.B. and Montgomery, D.B. 1988. First-Mover Advantages. Strategic
Management Journal 9: 41-58
March, J.G. and Simon, H.A. 1958. Organizations. John Wiley & Sons, Inc: Chichester,
UK.
Meadows, D.H.; Fiddaman, T., and Shannon, D. 1993. Fish Banks, Ltd. A Micro-
computer Assisted Group Simulation That Teaches Principles of Sustainable
Management of Renewable Natural Resources 3rd ed. Laboratory for Interactive
Learning: University of New Hampshire, Durham, NH 03284;
Morecroft, J.D.W. 1983. System Dynamics: Portraying Bounded Rationality. Omega.
11: 131-142.
Morecroft, J.D.W. 1999. Visualising and Rehearsing Strategy. Business Strategy
Review 10(3): 17-32.
Morecroft, J.D.W. 2002. Resource Management Under Dynamic Complexity . In
Morecroft, John D.; Heene, Aime, and Sanchez, Ron Editors. Systems Perspectives on
Resources, Capabilities, and Management Processes. Oxford: Pergamon;
Morecroft, J.D.W. and Sterman, J.D. (eds). 1994. Modelling for Learning . Productivity
Press: Portland, OR.
Moxnes, E. 1998. Not only the tragedy of the commons: misperceptions of
bioeconomics. Management Science. 44: 1234-1248.
Paich, M. and Sterman, J. D. 1993. Booom, Bust, and Failures to Learn in Experimental
Markets. Management Science. 39: 1439-1458.
Porter, M.E. 1998. Competitive Strategy. Techniques for Analyzing Industries and
Competitors. The Free Press: New York,NY.
29
Prahalad, C.K. and Hamel, G. 1990. The Core Competence of the Corporation.
Harvard Business Review. (May-June): 79-91.
Repenning, N. 2001. Understanding fire fighting in new product management, Journal
of Product Innovation Management. 18 (5): 285-300.
Repenning, N. and Sterman, J.D. 2002. Self-Confirming Attribution Errors in the
Dynamics of Process Improvement, Administrative Science Quarterly. 47: 265-296.
Schelling, T. 1978. Micromotives and Macrobehavior. W.W. Norton & Company, Inc:
New York, NY.
Senge, P.M.1999. The Fifth Discipline The Art & Practice of The Learning
Organization. Random House; London, UK.
Sterman, J. D. 1985. A behavioral model of the economic long wave, Journal of
Economic Behavior and Organization. 6: 17-53.
Sterman, J. D. 1987. Testing Behavioral Simulation Models by Direct Experimentation.
Management Science. 33: 1572-1592.
Sterman, J.D. 1988. People Express management flight simulator: simulation game,
briefing book, and simulator guide. Available from the author at
http://web.mit.edu/jsterman
Sterman, J. D. 1989. Modeling managerial behavior: misperceptions of feedback in a
dynamic decision makindg experiment. Management Science. 35: 321-339.
Sterman, J. D. 2000. Business Dynamics Systems Thinking and Modeling for a
Complex World. Irvine / McGraw-Hill: Boston, MA.
Teece, D.J.; Pisano, G.P., and Shuen, A. 1997. Dynamic Capabilities and Strategic
Management. Strategic Management Journal. 18: 509-533.
Thompson, J.D. 1967. Organizations in Action. McGraw Hill: New York, NY.
30
Warren, K. 2001. Strategy Dynamics and Scenarios of Industry Evolution (DRAFT).
IXX International System Dynamics Conference Proceedings Atlanta,GA.
Warren, K. 2002. Competitive Strategy Dynamics. John Wiley & Sons Ltd: Chichester,
UK.
Wernerfelt, B. 1989. From Critical Resources to Corporate Strategy. Journal of
General Management. 14: 4-12.
Zagonel, A.A. 2002. Model Conceptualization in Group Model Building: A Review of
the Literature Exploring the Tension between Representing Reality and Negotiating a
Social Order. XX International System Dynamics Conference Proceedings Palermo,
Italy .
31
FIGURE 1
Stock and Flow Diagram of the External Resources in Fish Banks, Ltd.
Extemal Resources
Fish Population
Fish Population
Coastal Area Deep Sea Area
Ni nah par ya Harve rate Nf Fish per year neces ate
(Gest does \Coasta} Area Deep Sea jeep Spa
fegeneration 2
Coastal Area Regeneration
= Deep Sea = Deep Sea Area
Effect of fish stock Coastal area Effect of fish population Catch
on Catch per ship ‘Gata ‘on Catch per ship
Maximum Maximum
fishery size fishery size
Coastal Area Deep Sea
oe
32
FIGURE 2
Stock and Flow Diagram of the Internal Resources in Fish Banks, Ltd.
Internal Resources| 6
Effect of fish stock
‘on Catch per ship
Effect of fish population
on Catch per ship
fal Area Catch jeep Sea Area
atch Normal catch per ship Catch
per ship Deep Ss
Deep Sea
5 _—5 5
Ships allocated
to Deep Sea
area
Normal catch per hip
per ship Coastal Sea
Coastal Sea
Shipg allocated is transferred Shi
between fishing in p,
areas
Ships
fea Area
Ships in Harbour
Team's Fleet Size
Decisions
Fleet Size Change
©
. Coastal Area
Catch
Deep Sea Area
‘Catch Ships
in Deep Sea Area
Price per & Price per fish
Bank Account income
Cost per SI
at Deep
Net income
Investment\
Divesture Ships in Harbour a
ols Cost per Ship
at Coastal
Ships
Cost per Ship in Coastal Area
at Harbour
33
FIGURE 3
Performance (total assets) versus Resources (ships) Scatter Plot at the end of year 5.
Total Assets ($)
15000
10000
5000
o¢ * °
oo 26°
*ee 3 $
2
10 20 ¥v 50 60 70
4
¢
o
¢
2
Fleet Size (number of ships)
34
TABLE 1
Evolution of the Value of Assets and Ships Per Experiment from Year | to 5
Experiment?
Year 1 Year 2 Year 3 Year 4 Year 5
Total Assets Ships| Total Assets Ships| Total Assets Ships| Total Assets Ships| Total Assets Ships
Team 1 2530] 4 1520] 10 (7750)] 60 (17730) 60 (29240) 60
Team 2 2530] 4 3770 | 10 2650 | 15 420| 30 (3170)] 30
Team 3 2230] 4 3120] 16 160 | 40 (6380)| 40 (13440)} 40
Team 4 2530] 4 4020 | 10 4630 | 10 4640 | 13 2610| 13
Team 5 2230] 4 3610 | 14 3830 | 24 460 | 24 (4170)| 24
Team 6 2430| 4 3710 12 3600] 20 520 | 28 (5280)|_28
TOTAL 74480 [24 79750| 72 7120] 169 (18070)[ 195 (62690)[ 195
Experiment 2
Year 1 Year 2 Year 5 Year 4 Year 5
Total Assets Ships| Total Assets _Ships| Total Assets Ships| Total Assets Ships| Total Assets Ships
Team 1 040] 5 6670] 55 17,350 | 55 14530] 55 11550 | 55
Team 2 3440] 5 4520] 7 5,780 | 14 6210| 14 5970| 14
Team 3 3440] 5 4720| 7 6,030 | 7 e660] 7 6560] 7
Team 4 3540] 5 4430] 5 5,480| 5 5800] 5 5710] 7
Team 5 3540] 5 4570] 5 5,780 | 5 6390] 2 6790] 0
TOTAL 75000 [ 25 24910] 79 34420] 83 39590 | 83 36580 | 83
Experiment 3
Year 1 Year 2 Year 3 Year 4 Year 5
Total Assets _Ships| Total Assets _Ships| Total Assets _Ships| Total Assets Ships| Total Assets Ships
Team 1 3540] 5 4570] 5 6390] 5 7130] 5 7120] 5
Team 2 3200] 5 4790 | 10 8450 | 12 9560 | 12 11620 | 12
Team 3 3390] 5 4560| 8 7340 | 11 8570| 16 10180 | 21
Team 4 3240] 5 4780 | 14 8470 | 14 9120] 14 6660 | 24
Team 5 3340] 5 4340] 9 8070 | 19 9230 | 19 4440 | 19
TOTAL 16800 | 25 23040] 43 38720] 61 43610 | 66 40020 | 81
Experiment 4
Year 1 Year 2 Year 3 Year 4
Total Assets _Ships| Total Assets Ships| Total Assets Ships| Total Assets Ships
Team 7 630] 4 3140] 24 3130] 24 (i260) 24
Team 2 180] 4 2680 | 55 (605)| 58 (12965) 58
Team 3 2030] 4 3190] 20 (2830) 20 (7010)| 20
Team 4 1830| 4 2340 | 24 (2710)| 43 (13830) 57
Team 5 2330] 4 3290 | 14 2440 | 14 450| 0
Team 6 2730|_ 4 3400|_6 3325] 0 3655| 0
TOTAL 70930 [ 24 78040] 143 2750] 159 (20960)[ 159
Experiment 5
Year 1 Year 2 Year 3 Year 4 Year 5
Total Assets _Ships| Total Assets Ships| Total Assets Ships| Total Assets Ships| Total Assets Ships
Team 1 2730] 4 3530] 6 5770] 9 7110] 12 7230] 16
Team 2 2630] 4 3790| 8 6090] 8 6930 | 10 7740 | 14
Team 3 2680] 4 3700| 7 6480 | 10 7720| 10 7340 | 10
Team 4 2730| 4 3760] 6 6020] 7 6840 | 7 5760] 9
Team 5 2580] 4 3420| 7 5770 | 11 7440 | 15 7620| 19
Team 6 2430 | 4 4110|_12 7670|_12 9400 | 12 10920 |_10
TOTAL 15780 | 24 22310] 46 37800] 57 45440 | 66 46610 | 78
35
TABLE 2
Total Assets and Fleet Size per experiment at the end of year 5.
Experiment 1 Experiment 2 Experiment 3 Experiment 4 Experiment 5
Total Assets Ships [Total Assets Ships | Total Assets Ships [Total Assets Ships otal Asset Ships
Team 1 (29240)] 60 11550 55 7120] 5 (1260)] 24 7230] 16
Team 2 (3170)} 30 5970 14 11620] 12 (12965) 58 7740) 14
Team 3 (13440) 40 6560 10180 | 21 (7010)} 20 7340) 10
Team 4 2610) 13 5710 z 6660} 24 (13830)] 57 s760| 9
Team 5 (4170)) 24 6790 0 4aao] 19 450) 0 7620) 19
Team 6 (6280)| 28 3655|_0 10920] 10
Total (52690)| 195 36580 83 40020] 81} (30960)] 159} 46610 78
Mean (8782)} 33 7316 17 8004] 16 (5160)} 27 7168 13
Std Dev. 11270] 16 2407 2 2877 8 7265] 26 1701 4
36
FIGURE 4
Evolution of Fleet Size for Three Teams in comparison to Overall Fish Sales
(the figure presents results from teams 1, 2 and 5 in experiment 2 and shows three typical
patterns of behaviour also observed in other experiments)
Total Fish Sales in $ Fleet Size (ships)
35,000 :
25,000
15,000
10,000
6,000
1 2 3 4 5
= © ‘Total Fish Sales —=B— Proactive tupe 1 —t=Reactive SH—Proactive type2 |
37
Patterns of behavior: characteristics of the behavior and comments from teams
TABLE 3
illustrating their behavior
Pattern of Characteristics of the Behavior Examples of the Behavior
Behavior
Reactive | Expanded the flect and allocated iramong the | "Still plenty of fish in Deep Sea, so aggressive
two fishing areas based on past events stance, buy more ships" (team 5 in experiment
(changes in the volume of fish caught) 1)
“Target deep sea to start” and “relocate to
coast as deep sea drops off” (team 2 in
experiment 5)
Did not foresee the effect of their decisions to | “[They will] have most boats in deep sea area
build internal resources (fleet) on the external | whilst stocks of fish remain - so order boats
resource (fish population). early" (team 3 in experiment 4)
Proactive | Set up objectives to control the effect of Obtain 25% of the market share of total fish
type 1 | competitors’ actions on their internal resource. | caught so build the fleet to obtain 25% of the
expected total number of ships. (Team | in
experiment 2)
Tried to guess competitors’ actions, and their | "Pay back in 2 years, first years grow
effect on their own internal resource aggressively" (Team | in experiment 1).
accumulation (bank account)
Build huge fleets in an attempt to pre-empt __| "Our strategy is to build a huge fleet
other teams without waiting to receive any immediately, pillage the fishery quickly and
information about external resources or not expand our fleet after the initial build"
competitors’ actions. (Team 1 in experiment 4)
Proactive | Planned ahead by inferring the effect of one of | "Sell ships in round 2 and be a bank
type 2 _| the two main uncertainties: fish population _| afterwards/ship trading." (Team 6 in
size.
experiment 4)
Tried to guess competitors’ actions, and their
effect on their external resource accumulation
(fish population)
“Not to be in the business as people overfish”™
(Team 6 in experiment 4)
Did not increase its fleet, or build a small fleet
and allocated it for a short period in one
fishing area, then moved it to a second area,
and sold the fleet before the fish disappeared
Move aggressively by expanding fleet quickly
because everybody would do it, go to fishing
areas not exploited, and finally sell the fleet
(Team 6 in experiment 6)
38
Effect on team performance of competing teams” decision-making processes.
TABLE 4
Team's type of
Proportion of teams by behavioural type
z 5 in it Perf
behaviour | *Periment -—Resctive | Proactive type 1 | Proactive type 2 ‘erformance
1 50% 33% 17% Lower than proactive type 2 but better than
proactive type 1
9) 9 5 [Lowest performance among their
2 60% 20% 20% sieuers
Reactive 7" - " In average good performance, with two
2 80% O% 20% lamong the best performers
4 20% 40% 40% Lower than proactive type 2 but better than
proactive type 1
5 83% 0% 17% IGood performance
1 50% 33% 17% Lowest performance among their
competitors.
a 2 60% 20% 20% [Best performance among their competitors
= 3 80% 0% 20% NA
tye 9p 7 5 Lowest performance among their
4 20% 40% 40%
competitors.
5 83% 0% 17% NA
1 50% 33% 17% Best performance among their competitors
2 60% 20% 20% Est ie proactive type 1 but better than
oi Better performance Than Towest reaalve
type 2 3 80% 0% 20% esine Ps
4 20% 40% 40% Best performance among their competitors
5 83% 0% 17% Best performance among their competitors
39
FIGURE 5
Feedback view of environment-strategy-structure alignment
Structure Environment
Rivals in Industry
40
FIGURE 6
Reactive teams’ conceptualisation of the resource system
Fish Sales Past Fish Sales
Growth
Own
Fleet Adjustment
‘Own
Fleet Expansion
Rate
41
FIGURE 7
Proactive teams type 1 conceptualisation of the resource system
Competitors
Bank Account
Competitors
Income
Competitors wn Share of
‘Share of Total Total Fishing
Fishing Fleets Fests
Competitors Competitors
Revenues ‘Ships
Competitors
Feet Expansion
Rate
¢
“otal Fah Expected Compattors!
Sales Fleet Adjustment
wn
Competitors peed
‘Ships
Own
Feet Expansion
Rate
an Share of
wn
Total Fishing
wn
Fleet Adjustment
Rroveoues —=
wn
Bank Account
“Target Share of
Own Total Fishing
Income Fats
42
FIGURE 8
Proactive teams type 2 conceptualisation of the resource system
. Competitors ——
Cone Bank Account ie
Fish Population pisinsy Operating Costs
o_6— @ ©
Competitors’ Cost per Ship
Ships
aoe Competitors
ape jon
Effect af Fish pane Fleet Expansi
pat Rate
Population on
Catch per ship
Total Fish
Catch ee
Own Fleet Expansion
Ships Pei
own
Cost per Ship Floet Adjustment
own
Bank Account
Actual
Gatch Target
@ ue) rosie rot See
‘Onn Own,
Income Operating Costs
FIGURE 9
easyJet’s, a low fare airline, conceptualisation of the usage of their main resource,
airplane, compared to a full fare airline (easyJet, 2003)
telesales staff, these seats don't
exist because of
business closs,
advertising 737-200 on
& cabin crew has 109 seats!
pilots in flight
Sielng
extra cabin
crew to serve
business class
& travel agent
commission
ground
handling
insuronce &
airport
londling fees ticketing costs
aircraft reservation
ownership fees & expensive
cost airports
air traffic lower aircraft
control fees utilisation
delays at
congested
airports
maintenance
fuel same costs as
easyJet
ENDNOTES
' Similarly, the strategy field offers both industry-level (Porter 1988) and firm-level
(Thompson, 1967; Hofer and Schendel, 1978) analysis to explain sustained differences in
the performance and profitability of firms. In particular the resource-based view
suggests that firms’ unique internal resources and capabilities are responsible for
differential performance (Wernerfelt, 1989; Prahalad and Hamel, 1990; Barney, 1991;
Barney, 2002).
? But Sterman also added that aggregate decision rules may not always be appropriate in
industry models, and other methods, such as direct experimentation, are needed to close
the gap between micro-knowledge of individual decisions and the macro-behaviour of
aggregate models and systems.
* Here the term frog-pond captures the essential comparative or relative effect.
Depending upon the size of the pond, the same frog may be small (if the pond is large) or
large (if the pond is small).
‘It’s important to be aware that we are saying a mental model is a representation of how
something specific works (or is believed to work) — in this case a good way to build a
fishing fleet that maximizes the asset value of the firm over the duration of the game. If
one were to ask players for their mental model of fish population (i.e. what determines
the fish population or how does population ‘work’?) it would of course be different to
their mental model for fleet expansion. The fact that players have a mental model for fish
45
population, and may even understand population dynamics, does not necessarily mean
they use this knowledge when devising a ship purchasing strategy.
Back to the Top
46
CC BY-NC-SA 4.0