TAMING THE DECISION TIDE WITHIN ORGANIZATIONS:
some theoretical considerations and experimental results
Jan Faber
Wim Scheper
Utrecht University, P.O. Box 80140 Utrecht, The Netherlands
ABSTRACT
A STELLA model of a simple production organization is used to investigate the effects of different types
of coordination on the adaptation of the organization to its changing environment.
1, Introduction
Today's business organizations are in the midst of the information revolution, Information technology
drastically changes the organization's environment, Not only does it affect the economic and technological
characteristics of existing industries, it also generates opportunities for whole new businesses (Porter and
Millar 1985). In order to survive, business organizations have to find ways of coping effectively with the
uncertainty that accompanies this increased environmental turbulence, and in this respect decision making
is a key process, because changing industry structures and the emergence of completely new industries
demand timely decisions; decisions which may eventually result in effective responses and adaptation of
the organization to its altered environment. But as increased turbulence requires more frequent decision
making, decisions have to be made faster.
Traditional command-and-control organizations seem rather ill-equipped to deal with the requirement of
faster decision making. In this type of organization unforeseen problems, which arise from the increasing
rate of environmental change, are usually referred to higher echelons, thereby forcing top management not
only to make strategic choices, but tactical and operational as well. Moreover, operational problems that
reach the top levels will receive the highest priority as they immediately affect the profitability of the
company. If top management increasingly has to deal with operational and tactical decision making, it
will spend less time on strategic decision making, and it will take more time before strategic choices are
made and implemented. Consequently, a structural adaptation of the organization to its changing
environment will take longer, which means that the organizational viability is at risk.
In sum, information technology poses severe threats to traditional command-and-control organizations.
Fundamental changes are required in order to secure their existence. In order to understand what these
changes look like, organizational decision making has to be discussed in more detail; section 2 will be
devoted to this topic. In section 3 a simulation model of a simple product organization will be described.
In section 4 this model will be used to conduct some simulation experiments in order to test some of the
theoretical considerations presented in section 2. In section 5 the outcomes of the simulation runs will be
discussed, and in section 6 some conclusions will be presented.
2. Decision making in organizations
An individual organization member who has to make a decision requires information concerning the
relevant characteristics of the problem he is facing. Information is data that, after it has been processed by
the decision maker, has acquired meaning. In other words: information is meaningful data (Davis and.
Olson 1985). It follows that decision making depends on the acquisition of data and on the processing of
data into information. Accordingly, faster decision making can be achieved by increasing the speed of the
acquisition and processing of data.
As decision makers are members of organizations, the latter statement applies to two different levels of
analysis, namely the level of the individual and the level of the organization. Faster decision making
therefore can be realized by increasing the speed of data acquisition and -processing both at the individual
and the collective level.
Investment in information systems, or better, data systems, can be regarded as an example of increasing
the speed of data acquisition and processing at the level of the organization (cf. Galbraith 1973; Applegate
et al. 1988). The use of data systems will result in the faster acquisition, processing, and storagé of more
data at the organization level. At the individual level, this means that the decision maker will receive
more data (in the form of reports, memoranda, notes, etc.) in a shorter period of time. Yet, the decision
maker is not interested in data as such, but in data that have meaning with respect to the problem at hand.
He therefore has to process the data that is channelled to him, and has to generate the information
himself. Accordingly, his decision making capacity, i.e., the number of informed decisions a decision
maker can make in a given period of time, depends on his cognitive data processing capacity, which is
limited (see Simon 1985; March 1988).
Thus, the central problem is how to overcome the decision maker's cognitive limitations. One might
argue that only relevant data, i.e. information, should be acquired and presented to the decision maker.
Unfortunately, it is often impossible to determine in advance what is relevant and what is not, especially
in the early stages of a decision making process when the problem is not quite clear yet (Mintzberg et al.
1976, Mason and Mitroff 1981, Nutt 1984). Therefore, it is not possible to determine in advance what
should be labeled as information and what not.
Another way of increasing the speed of data acquisition and processing at the level of the organization is
the hiring of (additional) staff personnel. They might help the decision maker in detecting environmental
threats and opportunities. They can also play an important role in the generation and evaluation of
feasible solutions to the perceived problems. The staff findings have to be reported to the decision maker.
Usually, this is a rather time consuming process, which means that with an increasing environmental
turbulence, managerial decision makers can devote less time to other activities and become overloaded
again. Taking into account the considerable costs involved in hiring staff personnel, it becomes
understandable why this alternative for the reduction of data overload is disfavoured by many business
organizations,
A third strategy that can be pursued to improve the speed of decision making is formed by the reduction
of the number of hierarchical levels (Drucker 1988). By reducing the number of hierarchical levels, it will
take less time to channel information to the decision maker. However, like the former strategies, this
strategy cannot overcome the limited cognitive capacity of the individual decision maker.
In conclusion: the decision maker forms a major bottleneck in organizational decision making processes.
More and faster data acquisition and processing at the organization level can easily exceed his cognitive
capacities, which means that data overload will occur, and the risk of making badly informed decisions, or
of making decisions too late will rise. Accordingly, the adaptation of the organization to its altered
environment will be endangered.
An essentially different approach is asked for. Taking the limited cognitive capacity of individuals as a
starting point, the alternative approach focusses on the number of decision makers, i.e., faster decision
making is realized by increasing the number of decision makers. Two alternatives can be imagined,
namely (1) hiring additional decision makers, and (2) transferring decision making powers to
organizational positions that did not possess these powers before. With regard to the first alternative, two
options exist. First, additional decision making positions can be created, which means that additional
hierarchical levels are created. In the second option the number of positions possessing decision making
powers remains constant, but the number of incumbents of the positions is increased.
One may doubt the effectiveness of both options. Increasing the number of hierarchical levels can
negatively affect the timeliness of data. Increasing the number of incumbents of a formal position,
though in theory a possibility, gives rise to potential competency and responsibility problems. In
addition, both options are costly.
The second alternative, which aims at the transfer of decision making powers to positions that did not
have these powers before, fundamentally affects the organization, because it implies a shift in
coordination principle (coordination being the hallmark of organization). Based on (1) a distinction
between superior and subordinate positions, and (2) a distinction between decisions concerning what
activities must be carried out and decisions concerning how implementation has to proceed, three different
coordination principles can be distinguished in the literature on organizations (see Scheper 1991: 149ff.),
The first principle can be called the command-and-control principle. It is characterized by a rigourous
division between decision making and implementation of decisions: decision making is carried out by the
incumbent of the superior position, while implementation of the decisions is performed by the
- 162 -
incumbents of subordinate positions. Coordination is entirely in the hands of the superior, who gives
orders and controls the implementation by monitoring the subordinates’ activities. This coordination
principle, which is graphically shown in figure 1a!, is the dominant principle in so-called command-and-
control organizations (cf. Drucker 1988). In the literature on organizations, Fayol (1916), Taylor (1947),
and the Aston group (Pugh et al. 1963) provide examples of this coordination principle.
The second principle that can be distinguished is called the cybernetic coordination principle, because it is
based on insights from the field of cybernetics. Both superior and subordinates are responsible for
coordination: what should be done is decided by the superior, how it is done is decided by subordinates.
As the superior does no longer decide how activities should be carried out, he can no longer control by
monitoring the individual activities of the subordinates. Control has to be achieved in a different way.
Cybernetics provides a solution. A fundamental insight from cybernetics concems the control of a
system: it is possible to control a system without knowing the system's inner structure. By monitoring
the system's output and affecting the input if the output departs from a prespecified standard, it is possible
to control the system. Translated into an organizational context this means that the superior does not
have to monitor the activities of the subordinates who constitute the system he has to control;
monitoring their output suffices. Organization models in which use is made of cybemetic insights can be
found in the work of Katz and Kahn (1968), Galbraith (1973), and Emery (1976). Figure 1b is a graphical
representation of the cybernetic principle.
dw dh
aw et
4 3
dw dh
yr an oo
Figure la Figure 1b Figure Ic
Figure 1. Coordination principles
‘The third coordination principle that can be found in the literature is called the competence-principle. It is
named after Max Weber's concept of the competence (Weber 1976: 125). A competence comprises a task
domain and decision making powers. The incumbents of subordinate competences possess a high degree
of autonomy, as the superior does not have the right to interfere in matters of content of subordinate
competences. This means that both superior and subordinate possess decision making powers concerning
the what and the how. For example, a department of a university can be regarded as a competence-based
organization. In a university department, coordination is achieved by the development of programmes,
both in research and in education. The head of the department is responsible for the quality of research and
education of the department. Given the impossibility of direct interference and the relative autonomy of
subordinates, negotiations are required to assure good quality of both programmes. By drafting lists of
priorities, which is a process involving superior and subordinate(s), an overall framework can be
constructed for both programmes. Subsequently, subordinates can specify their contributions in detail
using the general frameworks as guidelines. After the specific contributions have been specified, they are
compared with the general frameworks in order to test whether the general conditions they had to satisfy,
are met indeed,
In organizations based on the competence principle, peer control is essential. It should, however, be
noticed that the superior possesses disciplinary powers, which can be used when subordinates exceed their
1The rectangle represents the superior position, the triangle denotes the subordinate position. ‘Dw’, ‘dh’,
and ‘im’ stand for ‘decision making powers concerning what activities have to be carried out’, ‘decision
making powers concerning how activities have to be performed’, and ' implementation’, respectively.
- 163 -
powers. Representatives of this type of organization models are Beer (1979) and Peters (1989). A
graphical representation of the basic assumptions of this type of organization model is given in figure
le*.
With regard to the possibility of data overload, it can be argued that data overload is more likely to occur
in a command-and-control organization than in organizations based on cybernetic or competence
principles. In the former type of organization all unforeseen problems will automatically reach the
superior's desk, whereas in the latter types of organization subordinates are supposed to solve (many of)
the problems themselves. Moreover, as control takes place at a less frequent pace in both the cybernetic
and competence based organization than in command-and-conwol organizations, another source of the
superiors overload is diminished.
We have to remark that cybemetic and competence based organizations require the superior to fulfil tasks
that are related to the creation of ‘unified vision’, ‘shared vision’, or, ‘common understanding’ (Drucker
1988; Wimmer 1991: 381; see also Kolodny 1986). Since subordinates are relatively autonomous, they
may perform activities that turn out to be counterproductive. By creating unified vision, which serves to
guide the common interpretation of key events, it is attempted to reduce their potentially disadvantageous
behavior. Unfortunately, the literature is not very helpful in assessing the time a superior has to invest in
the process of creating shared meaning. Authors on this subject seemingly assume that applying
cybemetic or competence principles together with processes of creating shared understanding is less time
consuming than using command-and-control principles. A further difficulty concerning shared vision
should be mentioned. That is, the conceptualization of concepts like shared or unified vision is, from a
theoretical point of view, not without severe problems (see Scheper 1991, ch. 3).
For these two reasons we will not include processes of creating shared vision in the following two
sections, in which the theoretical analyses presented in this section will be implemented in a simulation
model. We will, however, readdress this topic in section 5.
Due to the impossibility of conducting field experiments concerning coordination principles within
existing organizations, we will use a simulation model containing the theoretical considerations presented
in this section, and confront the model with empirical data. As cybernetic principles are more often
applied than competence principles, in the next sections only command-and-control coordination and
cybernetic coordination are compared by means of a simulation.
3. Simulation Model of a Simple Production Organization
In the previous sections it has been argued on theoretical grounds that a change from command-and-
control coordination to cybernetic coordination within organizations may reduce decision making strains
and delays at the top management level and, consequently, may enhance the adaptability of organizational
operations to environmental disturbances. The postulated effects of this change in coordination principle
will be empirically investigated by implementing it in the specification of a system dynamics model of
an organization. The system dynamics model represents a theoretical model of a simple production
organization. This does, however, not imply that the organization model specified has no connections
with reality. The model has been checked with a descriptive analysis by Hage (1990) of current problems
faced by the Dutch division of the Heineken company, called Heineken-Netherlands, which is a large
lager-beer producer for especially the Dutch and American consumer markets.
The model has neither been constructed to reproduce. historical time series of observations of specific
organizational processes in reality, nor to make predictions of these processes in the near future. Its
purpose is to explicate and illustrate the theoretical arguments presented in the former sections. Due to its
simplified theoretical specification, the system dynamics model cannot be expected to possess any
explanatory or predictive power; this is not to say, however, that the specified system dynamics model
does not serve some scientific purposes. It explicates the assumptions contained in our theoretical
2The superior's rectangle should, of course, also contain the symbols dw, dh, and im. We did not include
these elements for explanatory purposes, i.e., we wanted to show the development from classical
instrumental theory to competence-theory and its accompanying transfer of discretionary powers.
- 164 -
arguments, which are open to further testing and eventually to falsification. The model also overtly
defines the organizational characteristics under investigation, thereby delimiting the area to which our
arguments apply. Furthermore, the specified causal relationships among variables representing various
organizational characteristics elucidate the feedback cycles contained in the model. Finally, simulation
experiments with the model provide insight into the dynamics of decision making, implementation of the
decisions into operations, and operations themselves within the production organization.
The specification of the system dynamics model of the production organization to be-presented is only
concerned with the production of lager-beer for the Dutch consumer market. Since the early 80's the
yearly Dutch consumption of beer has stabilized at around 12 million hectoliters. Furthermore, the Dutch
beer market can be characterized as a market with open competition among various suppliers. Because of
the open competition on this market, the only feasible short-term policy to be pursued by the production
organization modelled is to improve its marketshare by allocating extra budgets into marketing, i.e.
pricing and promotion (advertising). This policy is based on the expected short-term profits to be gained
in the near future, which in turn are based on the profits earned in the recent past. The development of
such an expectations patter occurs in a temporarily stable environment like the Dutch beer market since
the early 80's. So, in our model it is assumed that the production capacity is constant in the short-term
and is fully utilized. Changes in consumer demand reflect themselves in changing sales and inventories of
finished products, and thereby in changing profits, These latter changes are counteracted by setting other
product prices and changing expenditures on advertising. This situation also characterized Heineken-
Netherlands during the period 1983-89. Its marketshare declined from 57.8% in 1980 to 51.7% in 1989 in
a virtually constant Dutch beer market. Only after nine years the board of directors of Heineken-
Netherlands got alarmed and reconsidered its strategic policy. It was then forecasted that if Heineken-
Netherlands would continue its existing policy, marketshare would be lost and the Dutch division of the
Heineken company would become a lossmaking organization in 1994 (see Hage, 1990). The production
organization specified is schematically represented in figure 2. A more detailed specification of the model
in STELLA™ (Richmond et al. 1987) is presented in the appendix to this paper.
varprodcosts
a a
+/ inventorycosts deliv
oo ¥
depreciation —P> investments —————z——>_ expenditures
+
Figure 2. Causal model of a simple production organization
In figure 2 profits perform a key role in bringing production into line with sales, while it is tried to keep
the inventory of finished goods at a minimum, both by price-setting and advertising. Both policies are
3A more detailed discussion of the scientific value of system dynamics models can be found in Hall &
Menzies (1983).
- 165 -
decided on by the top management depending on the amount of profits realized. From the STELLA model
in the appendix it can be deduced, that if, due to whatever external cause, demand declines, it takes two
time periods via sales, revenues, profits, pricing, and advertising before the level of demand will be
accommodated again to the level of production. In the meantime revenues have become less, while the
expenditures increase due to the costs of a growing inventory of finished goods. Consequently, the delay
in the adaptation of demand to production is a costly aspect of an organization coordinated in a command-
and-control mode. If demand changes frequently, the levels of production and consumer demand will be
continuously out of equilibrium. Deviations from the dynamic production-demand equilibrium trajectory
result in permanently lower revenues and higher costs, and, consequently, in smaller profits. And smaller
profits represent a risk to the long-term viability of the production organization.
The central cause of these problems is the delay in adaptation of prices and advertising budgets to changes
in’consumer demand, which results from the centralized command-and-control coordination. As the top
management has to make relatively many decisions (what- and how-decisions), the decisions concerning
the price of lager and the budget for advertising are taken on a three months basis instead of, e.g., a
monthly basis. Quarterly data on profits, inventory of finished goods, and consumer demand provide the
input to this decision making process. The result is a three monthly adaptation of the price of lager and
the budget for advertising. This delay in adaptation to externally induced changes in consumer demand
disables operations to react timely and adequately to environmental changes. This delay in respons to
externally induced changes in consumer demand and, consequently, in expected profits, and in the
inventory of finished goods is specified in the STELLA model as updating the price of finished goods and
the budget for advertising in accordance with the monthly average change in profits, inventory, and
demand after every subsequent three months period. When profits show a negative trend over the last three
months, the price of lager is lowered and the advertising expenditures are increased. When the ratio of the
levels of the inventory of finished goods and consumer demand shows a positive trend over the last three
months, the price of lager is additionally lowered, whereas the advertising expenditures are additionally
increased. A positive trend in profits leads to a higher price of lager and less expenditures on advertising.
In addition, a decline of the inventory-demand ratio leads also to a higher price of lager and to less
expenditures on advertising. If the level of consumer demand exceeds the level of production, the level of
consumer demand is tried to be reduced by setting higher prices of lager and by cutting down the budgets
for advertising.4 Without tempering consumer demand the discrepancy between consumers’ demand for the
product and the availability of the product on the market will grow continuously. Such a growing
discrepancy may lead to such an unsatisfied demand of consumers, that consumers become frustrated and
definitely change over to another brand of lager. The latter effect endangers future market potentials of the
simulated organization, and thereby its viability
In order to overcome the problems created by the delay in adaptation to environmental changes mentioned
above, it has been argued in the former sections that a change from centralized command-and-control
coordination to decentralized cybernetic coordination may help. Such a change in coordination implies a
segregation of the combined decision making powers conceming what activities should be undertaken, and
how these activities must be carried out. Following the argument presented in section 2, this means that
top management should decide what activities should be undertaken, whereas decisions about how these
activities must be carried out are made by lower level decision makers. In the STELLA model this
segregation is specified as follows. The top management adjusts at the end of every three months period
the price of lager and the budget for advertising for the next three months period on the basis of the
average monthly change in profits realized during the last three months period. In the meantime, the
lower level decision makers can introduce minor monthly changes in the price of lager and the budget for
advertising, which are set by the top management for a three months period, in order to keep the
4It should be noted that a relatively high price elasticity is assumed.
SIf the production capacity is not continuously fully utilized an excess of demand can be counteracted in
the short-term by increasing the degree of utilization via an extra shift of labour. Tempering demand via
higher prices and less advertising is not necessary then. This situation is, however, not considered as we
assumed full utilization of the production capacity. If the excess of demand seems to become a long-term
perspective, a strategic policy is called for in the form of plans for investments, financing, marketing,
etc, (see also section 5). +
=
De -
A 2 - 166 -
inventory of finished goods at a minimum during each three months period. Whether the change in
coordination principle within the modelled production organization has the hypothesized effects on
improving the adaptability of the organization's operations to externally induced changes in consumer
demand, which is reflected by a higher rate of profitmaking, will be analyzed in the next section.
4. Simulation of Centralized and Decentralized Coordination
The STELLA model of the production organization has been used in several simulation experiments for
studying the effects of the proposed change in coordination principle. Each experiment covers a
simulation period of 48 months (4 years). The initial values of stock variables and the estimates of
parameters in this model have been derived from data on Heineken-Netherlands for 1989 (see Hage, 1990).
In all simulation experiments the model is ran during the first three months without changes in consumer
demand, the price of lager, and the budget for advertising, in order to build up a quarterly figure of the
profits realized under the assumption of no change in organizational policies. The baserun is the
simulation run with the STELLA model presented in the appendix containing the data just mentioned. In
the baserun it is assumed that (1) there are no externally induced changes in consumer demand and (2) the
adaptation of the price of lager and of the budget for advertising occurs at the end of every subsequent
three months period. This baserun will be referred to as BR. BR forms a reference for analyzing changes
in the dynamics of the model resulting from a change in the coordination principle described in previous
section and externally induced changes in consumer demand.
In the first simulation experiment the speed of adaptation of the price of lager and of the budget for
advertising to consumer demand is accelerated from once in every three months period to once a month.
This scenario is labeled as SC1, and it represents the cybernetic coordination principle. The rates of
adaptation of once in every three months period in BR and once a month in SC1 are also implemented in
the simulation experiments SC2a and SC2b, respectively, in which the level of consumer demand
declines ceteris paribus as a linear function of time by 2.5% over the last 45 months. In the simulation
experiments SC3a and SC3b the level of consumer demand does not decline as the result of a linear
function of time, but changes ceteris paribus as the result of the function depicted in figure 3 (where the
value 1.000 indicates no change).
derarctrge
nes 4
roa 4
1000 4
0.908
0.976
rs
irs Eyes WE Too
Tene
Figure 3. Change in consumer demand in scenarios SC3a and SC3b
The simulation experiments BR, SC1, SC2a, SC2b, SC3a and SC3b will be compared and analyzed on
the basis of the following indicators: accumulated profits, consumer demand, and inventory of finished
goods (labeled inventory 2). The values of these indicators taken during each simulation run are presented
in figures 4a-4f.
- 167 -
semanro-5 serwanzo-se2
ho ms ae re
pun, Bp Tess dt accoroe
een 1 cooreeran
yee | Hewes gteg deen
Ly -anas00 ares Le rene
avs 2 ne eS SS
OS Ap ss mi sa co 7 TH Wa Te
Figure 4a Figure 4b
mete sou SomPARO-Sem8
4) Bass h SS
acme ee yomE
28 = t 9240.09
be Pu TRS
& 7 =e se a CE Tr aH 3 ie
Figure 4c Figure 4d
semanro-seaa sepanio-t2m
dh Bese ae el
ce
2 comannand
URE f teen
yee Spee
ES
We ir ris ccs Ents We
Figure 4e Figure 4f
From these figures the following conclusions can be drawn. The modelled organization becomes a
lossmaking within 4 years if command-and-control coordination is pursued under relatively stable market
conditions (scenario's BR and SC2a; see figures 6a and 6c). Furthermore, the application of the cybemetic
coordination principle within this organization leads in all cases of (no) externally induced changes in
consumer demand to higher profits than the application of the command-and-control coordination
principle. These higher profits result from a better adaptation of the level of consumer demand to the level
of the inventory of finished products and, consequently, from lower promotion related costs. This is best
seen in figures 6e and 6f. The externally induced change in consumer demand, depicted in figure 3, is
better counteracted in figure 4f than in figure 4e. Figure 4f represents the situation in which lower level
decision makers are authorized to decide on deviations from the price of lager and the budget for
advertising set by the top management, in order to counteract the swing in consumer demand more
quickly than the top management could do. Another implication of this situation is that the top
management does not need to take into account changes in the ratio of the levels of the inventory and
consumer demand in their decisions concerning the price of lager and the budget for advertising. This
means that their decision load is reduced from two issues to one (i.e. the change in profits). So, our
theoretical arguments favouring the cybernetic coordination principle over the command-and-control
- 168 -
coordination principle to be applied within organizations facing a turbulent environment are not falsified
by the results of the simulation experiments. >
5. Discussion of the Model Structure and its Simulated Dynamics
The theoretical problem dealt with in sections 1 and 2 boils down to how and why a change from
command-and-control coordination to.cybernetic coordination does represent an answer to the problems
resulting from increasing environmental turbulence due to the emergence of information technology. In
order to "test" the postulated positive effect, a theoretical model of a simple production organization has
been specified. In the basic specification of the model, decision making and the implementation of
decisions is coordinated in a command-and-control mode by the top management. So, the specification of
the model is based on theoretical arguments and is not the result of datafitting. The estimates of the
unknown parameters and initial values of stock variables in the model have been derived from public data
on Heineken-Netherlands (see Hage, 1990). Consequently, the results of the simulation experiments were
Not a priori forced to disapprove the hypothesis of no effect of a change in coordination principle. The
empirical cycle of scientific research has been followed as close as possible.
From the results presented in figures 6a-6f one might draw the conclusion that with respect to
profitmaking cybernetic coordination performs better than command-and-control coordination within an
organization. There is, however, a threshold, under which command-and-control coordination performs
better than cybernetic coordination. The argument in favour of this statement goes as follows. If the
externally induced changes in consumer demand follow quickly upon one another, and if these changes
outweigh each other in the long-term, immediate counteractions as a result of cybernetic coordination to
offset the effects of these changes put a large strain on the organization itself as well as on its finances.
Enormous sums of money will be spent on pricing, in the form of lower revenues, and on advertising, in
the form of higher expenditures, in order to stimulate demand directly after it has fallen randomly. As the
Positive and negative changes in demand outweigh each other on the long-term, three monthly
counteractions as a result of command-and-control coordination will only partially take into account the
counterproductive changes in demand. With this coordination principle the sums of money spend on
pricing and advertising will be less than with cybernetic coordination. These points are illustrated by the
results of simulation experiments, in which (a) command-and-control coordination is applied (figure 5a),
and (b) cybemetic coordination is applied (figure 5b), with random disturbances between -10% and +10%
of consumer demand after the third month. Yet, it must be noted that these results can only be regarded as
tentative and experimental and that further research into the question of a lower threshold between the
effectiveness of cybernetic coordination and command-and-control coordination within an organization is
needed.
semaxro wupok prronmers « senanre muoct orsromancEs »
Te
a Perna
= savenrory?
oT = sa we MEE Ay Te ge sr he
Figure 5a Figure 5b
The descriptive analysis of current problems faced by Heineken-Netherlands as presented by Hage (1990)
contains no detailed information on its organizational structure and dynamics.
- 169 -
In the model presented in section 3, the improved speed of adaptation of organizational operations to
externally induced changes in consumer demand, as a result of changing from command-and-control
coordination to cybernetic coordination, has been specified. We should, however, make an additional
comment. In section 2 it was argued that adopting cybemetic or competence principles must be
accompanied by processes of creating unified vision, because a common construction of reality is regarded
as an essential condition for concerted action in organizations in which these principles are applied. Since
the literature does not provide any adequate clues regarding how to create unified vision, we did not
explicitly model these processes, nor did we model coordination problems among lower level managers,
which can occur if a shared situation detinition is absent, i.e., a small degree of commonality may lead to
contradictory decision making by lower level decision makers.
In order to illustrate this notion, scenario SC] has been expanded by incorporating a three monthly
adaptation of the level of production’, which is inversely related to the average monthly change in the
inventory of finished goods during the previous three months period (see figure 6a). Next, the three
monthly adaptation of the level of production to the change in the level of inventory of finished goods is
changed into into a monthly adaptation (figure 6b).
semanto raepverian abarTarrO >
home ae h mR
AEE] coproe BEd scceot
ye i en
yon ee
OE Sear ET
Figure 6a Figure 6b
Comparison of figures 8a and 8b leads to the conclusion that the monthly decisions and their
implementation by lower level managers in the production and marketing department are
counterproductive. Figure 6a illustrates the situation that if lower level managers in the marketing
department do not succeed in bringing the level of consumer demand into line with the level of production
within every three months period, lower level managers in the production department will bring the level
of production into line with the level of consumer demand after every three months period. So,
coordination of decision making at lower levels of management is crucial for the modelled organization to
operate succesfully. Following Drucker (1988), we might argue that unified vision, understood as a shared
construction of reality, is a valuable asset in this respect.
6. Conclusions
In traditional command-and-control organizations, an increasing rate of environmental change forces top
management to devote more time to strategic, tactical, and operational problems. Since data overload
easily results from these increased decision making strains, the organizational viability is at risk. In order
to develop and evaluate possible alternatives which might lighten the decision burden of top management,
it proved necessary to distinguish between (1) data and information, and (2) the individual and the
organization, As such, the approach presented in this article differs from the classical study of Galbraith
(1973) in which these distinctions are not explicitly made.
We argued that in decision making processes, the major bottleneck is formed by the limited cognitive
capacities of the individual decision maker (section 2). Accordingly, increasing the number of decision
makers by transferring decision making powers to levels that did not possess these powers before was
7In this scenario the assumption of full utilization of the production capacity has been relaxed (cf.
footnote 4).
- 170 -
seen as the most appropriate alternative. In order to test whether implementation of this alternative results
in faster and better adaptation of the organization to its changing environment, a simulation model was
constructed, and several experiments were run. It was shown that, when confronted with a changing
environment, cybernetically based organizations perform better than command-and-control organizations
Yet, a lower threshold seems to exists. Finally, application of cybernetic coordination seems to require
the creation of a shared construction of reality among the lower level managers. Further empirical research
must be conducted to study this phenomenon in more detail.
REFERENCES
Applegate, L.M., J.I. Cash and D. Quinn Mills, 1988, Information Technology and Tomorrows
Managers, Harvard Business Review 66 (nov-dec): 128-36.
Beer, St., 1979, The Heart of Enterprise, Chichester: John Wiley & Sons.
Drucker, P.F., 1988, The Coming of the New Organization, Harvard Business Review 66:45-53.
Emery, F.E.,1976, Futures We Are In, Leiden: Martinus Nijhoff. .
Fayol, H.F., 1916, Administration Industrielle et Générale, Paris: Dunod.
Galbraith, J.R., 1973, Designing Complex Organizations, Reading, Mass.: Addison-Wesley
Hage, G., 1990, Wankelt Heineken?, Intermediair 24 (34): 30-35.
Hall, R.I., and W.B. Menzies, 1983, A Corporate System Model of a Sports Club: Using Simulation as
an Aid to Policy Making in a Crisis, Management Science 29: 52-64.
Katz, D. and R. Kahn, 1968, Open-System Theory, In: Buckley, W.(ed.), Modem Systems Research for
the Behavioral Scientist, Chicago: Aldine Publishing Company.
Kolodny, HF., 1986, Selbstverpflichtung und Flexibilitat: Abkehr vom Modell der Fremdkontrolle, Die
Betriebswirtschaft 46: 603-609.
March, J.G., 1988, Bounded Rationality, Ambiguity, and the Engineering of Choice, In: Bell, D.E., H.
Raiffa, and A. Tversky (eds.), Decision Making: descriptive, normative, and prescriptive
interactions, Cambridge: Cambridge University Press.
Mason, R.O. and LI. Mitroff, 1981, Challenging Strategic Planning Assumptions: Theory, Cases and
Techniques, New York: John Wiley & Sons.
Minwberg, H., D. Raisinghani and A. Théorét, 1976, The Structure of "Unstructured" Decision
Processes, Administrative Science Quarterly 21: 246-75.
Nutt, P.C., 1984, Types of Organizational Decision Processes, Administrative Science Quarterly 29:
414-50.
Peters, F.G., 1989, Organisation und Information: tiber den Zusammenhang von Interessen, Erwartungen,
Kenntnissen und Handlungen, PhD-thesis Utrecht University.
Porter, M.E., and V.E. Millar, 1985, How Information Gives You Competitive Advantage, Harvard
Business Review 63 (july-august):149-160.
Pugh, D.S., D.H. Hickson, C.R. Hinings, K.M. McDonald, C. Turner, and T. Lupton, 1963, A
Conceptual Scheme for Organizational Analysis, Administrative Science Quarterly 8: 289-315.
Richmond, B., P. Vesunoz, and S. Peterson, 1987, STELLA for Business, Lym, NH.: High Performance
Systems.
Scheper, W.J., 1991, Group Decision Support Systems: an inquiry into theoretical and philosophical
issues, PhD-thesis Tilburg University.
Simon, H.A., 1985, Human Nature in Politics: a dialogue of psychology with political science,
American Political Science Review 79: 293-306.
Taylor, F.W., 1947, Principles of Scientific Management, New York: Harper and Row.
Wimmer, R., 1991, Zwischen Differenzierung und Integration: zur charakteristischen Dynamik von
Organisationen mit steigender Eigenkomplexitit, Gruppendynamik 22: 359-389.
- 171 -
APPENDIX: STELLA™ model of a simple production organization
rou :flow variable
© auitan vate
oO ghost variable determined
‘elsewhere in the model
adsexpenditures
proftchange ‘dtatiochange _profitchange ‘eratiochange
- 172 -
CC BY-NC-SA 4.0