Content uploaded by Brian O'Flaherty
Author content
All content in this area was uploaded by Brian O'Flaherty on Dec 20, 2013
Content may be subject to copyright.
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - I996
Implementing WorMow Automation Systems:
Implications for Management Control
Pat Bowe
Dept. Of Accounting, Finance & Information Systems
University College Cork, Ireland
Brian O’Flaherty
Dept. Of Accounting, Finance and Information Systems
University College Cork, Ireland,
Abstract
Automation has revolutionised industry and initiated
profound economic change. In the past two decades,
manufacturing productivity has seen substantial
improvements, yet office productivity has remained
practically unaffected. Workflow Automation Systems,
a new class
of
system which focus on automating
document flows, promises to redress this imbalance.
This study looks at the implication
of
adoption of a
WorkJlow Automation System and in particular the
impact on the managerial control system. The financial
service sector is a main adopter
of
this technology and
many such companies are planning future
implementation
of
workflow automation systems. This
study sets out to explore the extent
of
the control features
that will be included in the systems and attempts to
ascertain the attitude
of
three groups i.e. users,
managers and IT personnel.. Their opinions on the
consequences which this would have
for
the operation
of
the control process in their organisations were then
documented and analysed. The control process model
developed by Child (91 proved useful in the
classification
of
process control activities. Following
analysis
of
the research data an amendment to ChildS
framework is proposed.
Keywords
Workflow automation systems; organisational document
flows; management control process.
Introduction
The level of infiltration of automated computer
systems in manufacturing is now such that further
enhancements to manufacturing processes are becoming
less significant, resulting in ,a new focus on the service
sectors to improve their efficiencies [21]. Similarly,
Reardon [32] has stated that while the last two decades
has seen general manufacturing productivity improve by
over 90%, office productivity has virtually stood still at
5% over the same period [3], [2I] despite the arguments
of Panko [31] that these figures were based on
ungrounded assumptions about the nature of office work.
This issue has been addressed by the development of
information systems which focus on the operation of
office operations: workflow automation systems. The
basic objective of these systerns is to manage the flow of
work around a business so tlhat it is performed quickly
and efficiently, ensuring thalt organisational document
management is executed reliably and inexpensively [22].
As a result, much of the appeal of workflow automation
systems for managers lies in the belief that they add this
a self-steering layer of intelligence and independence to
the process of data distribution [33]. These potential
benefits of workflow automation systems have led
Sprague [35] to state thalt harnessing information
technology to manage documents is one of the most
important challenges facing IS managers in this decade.
13
1060-3425/96 $5.00 0 1996 IEEE
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
Background to Worktlow Automation
Systems
Workflow automation system applications are
relatively recent developments in the information systems
field, and thus far literature on these systems has tended
to be of a highly practitioner-oriented nature [2], [16].
The primary strength of workflow technology is its
capability to elicit great improvement in the efficiency of
operation of the processes which it automates [20], [22].
The result is that descriptions of issues relating to the
productivity elements of workflow software have been
well-documented [3], [17], [34]. Research into the
elements of the workflow system which do not contribute
to this core productivity effect and in particular the
impact of control mechanisms integrated into workflow
systems and their consequences for the operation of the
control process, have tended to be overlooked [21]. This
study aims to address this imbalance by investigating the
consequences of implementing workflow automation
systems for the operation of the management control
process.
The Appeal of Workflow Automation
Systems
The appeal of worldlow automation systems lies in
the often immediate and extensive cost savings which
can be made from their introduction [22], [35]. This
makes workflow systems quite easy to cost justify in
comparison to many other types of information systems
[35]. The benefits which can accrue to organisations can
be both immediate and long-term, both providing
immediate ‘hard’ cost savings, and ‘softer’ benefits in
such areas as the quality of services provided and the
quality of decisions made. Some of the potential benefits
of worMlow automation systems which have been
identified in the literature are displayed in Table 1.
As a result of these benefits the popularity of
workflow automation systems is increasing, with growing
numbers of organisations implementing these systems
and re-engineering their workflow processes. This has
led Charles Babcock [2] to refer to workflow automation
as the ‘next killer application on PCs’.
Table 1. Overview
of
the benefits
of
workfIow automation
systems identi$ed by other authors
l
The number of steps involved in most processes
are dramatically reduced and eliminated,
simplifying the entire document processing system
Pa PI.
* Parallel instead of sequential processing is
supported and parallel tasking abilities may also
be included, thus speeding up the workflow
process [22], [23].
* A reduced the number of errors which occur in the
operation of the process [S], [ 161.
o Enhanced customer service due to a higher
throughput capability, a reduced possibility of
error and the decreased turnaround times required
to perform the same functions [20], [24].
0 Internal productivity increases of typically 30 to
50% [3], [16].
a Enable the streamlining of quality systems in
organisations [3 81.
l
Training of new personnel and retraining of old
personnel on the new systems is made far simpler
[31> [341.
l
Increased flexibility in changing the systems
which are currently in place [3].
l
Significant storage and photocopying cost savings
~241.
The Components of a Workflow Automation
System
The most basic task of a workflow automation system
is to automate the business processes which involve the
transfer and routing of documents from one person to the
next, a task known as programmatic control [24]. Thus
worldlow automation systems always start at the level of
the individual business process, which is the set of linked
activities that take an input and transform it into an
output [ 181.
There are five basic components which are
incorporated in the structure of a workflow system
introduced to automate a business process:
14
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
1. The starting point is the business process
which can be automated as it currently exists
or improved after analysis;
2. These processes are defined as procedures
involving series of tasks which are performed
under a variety of conditions. Each execution
of a process is called an individual case. Some
of these tasks take place in parallel so that
rendezvous points may be necessary to collect
output from several tasks. A system of
deadlines and alarms is usually provided to
support the completion of time-critical
operations;
3. Participants in procedures perform roles and
can be allocated various degrees of authority to
coordinate the approval policies of the
organisations. There is a means of addressing
participants in the system including titles,
roles and references to inbuilt organisation
structures;
4. Workflow Systems allow access to a wide
range of data via a forms processing subsystem
which displays documents within folders;
5. There are reporting and management facilities
which allows management to assess the
performance of subordinates through such
measures as status reports, performance
monitoring tools and audit trails [2 11.
Literature on the first four components is readily
available [3], [7], [21], [34]. With regard to the fifth
component, however, despite worktlow automation
systems literature continually including a control
component in descriptions of the basic features of such
systems, remains largely unresearched.
Managerial Strategies and New Technology
One of the most obvious changes in the sociostructure
of organisations since the early industrial era has been in
the introduction of new technology in the workplace and
the consequent changes in the relative balance between
capital and labour [9]. The four major objectives of
management which are prominent in decisions to
introduce new technology are to reduce operating costs,
to increase efficiency, to secure improvements in quality
and finally to facilitate improvements in managerial
control [9], [lo], [12]. The introduction of new
technology has been identified by a number of writers as
an opportunity for management to increase the scope and
effectiveness
of the control
mechanisms available to it
[l], [lo], [37]. The execution of such a strategy has been
described by Child [lo] and is depicted in Figure 1.
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
Adaplon
Of ne*
fundamental objectives managerial s,ra,sgie* technolow implsmo”,a,ion within
as identified in models --)
t
(or,s”,sd to objwtivos - the labour pro~sss
af capitalism such as: t
. reduce costs
* i”cra4se Rsxibility
cd production !rys,sm
- improve qualily
. entlancs control)
(spsoifie sWects on
~ontrot and skill)
I
EO”,OX, ,prod”a mark‘,,.
hbour market.
technological knowlsdgs,
i”,srve”i”g proc*sse* and
actors (workor and union
r*po”**, “.d”.D offu”c,io”*
and middle managars. va,u(~s
ofwodt organisation dssignsrs)
Figure 1: Representation of the role of managerial
strategy @om Child [9])
It has been argued that, historically, new technology
has been implemented wherever possible not with the
intention of improving the work environment of
employees, but that of reducing the dependence on labour
by improving management’s position of control [ 11, [ 111,
371. This can be traced back to a traditional desire on the
part of management to structure subordinates’ jobs to as
large an extent as is possible in an effort to reduce the
uncertainty and hence the stmss level of their jobs [5],
[ 121, [41]. Indeed, new technology has been found to be
capable of providing a dangerous link between an
oppressive managerial strategy and the labour process
[lo]. The result is that computer systems tend to instill
convergence towards the management style supported by
the system, one which attempts to tighten the control
process whenever possible [l], [5], [29].
Whisler [39] has stated that the preference of
management seems to be to exercise as much control
over the actions of their subordinates as is
technologically and economicallly feasible. For as long as
management seek increased control, new technology will
be implemented and used for the purpose of allowing
them to improve their control Icapabilities [39]. Child [9]
has developed a model for analysing changes in the
operation of the control process. This is a six-stage
model incorporating two groups of three stages each as
depicted in Figure 2. The: first groupings is titled
executive instruction and consists of formulating goals,
setting tasks and deadlines and monitoring progress
towards the achievement of thlose deadlines. The second
grouping is called the feedback side of the model and
involves recording, analysing and evaluating the output
of employees.
15
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
Feedback
Figure 2 The process of management control
Executive Instruction
The executive instruction part of the model involves
those stages in the control process whereby the manager’s
main task is the provision of information to and guidance
of subordinates on the manner in which work is to be
completed [9], [27]. The manager is assumed to have
certain goals, in terms of work that he/she wishes others
to accomplish in a manner which will satisfy or surpass
the required levels of performance. These standards or
rules may take the form of output targets, instructions or
guidelines as to how the task in hand is to be
accomplished [9]. Managers oversee their subordinates
perform this work in the next step, monitoring
performance, with the result of their labour being some
form of output.
Feedback
The feedback part of the control process involves an
upward information flow regarding the job performance
of employees, focusing on the difference between a
system’s desired outcome and its actual performance [ 191,
[25]. The feedback process begins once output of some
form has been generated and is directly or indirectly
measured and recorded [25]. Once measured, this
information is then analysed and evaluated against the
desired goals and standards, generally with a strong
emphasis on productivity, with results meriting
subsequent reward or punishment [9]. This enables
managers to assess the performance of subordinates and
to alter, if necessary, either the goals or the methods
employed to achieve those goals, or both. Thus the
feedback stages consist of a decision-making operation
which compares the measured state of the process with
the desired conditions and decides how the variables
should be manipulated [ 141. The final stage of the
Feedback side feeds back into the Executive Instruction
part of the control process, mainly through influencing
the goal formulation stage, but also the standard setting
phase, thus completing the control process loop [9].
The model is structured so that the control process
stages necessitating the greatest amount of conceptual
tasks are placed at the top of the diagram, with focus
gradually shifting towards more operational tasks as one
moves down through the stages. Thus the goal
formulation and performance evaluation require the
greatest amount of management conceptualisation, and
the two stages regarding the monitoring of operating
behaviour and the recording of output are deemed to be
the most operational [9].
Deficiencies in
WorMow Automation
Systems Research
The primary strength of workflow technology is its
capacity to improve the efficiency of processes which
have been automated [20], [22]. Research into the
elements of the workflow system which do not contribute
to this core productivity effect have tended to be ignored
[21]. As a result, literature in the area has tended to fall
into two categories. The first consists of writers who are
concerned with making generalised statements on the
potential benefits which can accrue to organisations who
are willing to invest in the technology, without
grounding these observations in empirical data. This
would include Bailey [3], Brierley [7], Dzujna [ 131 and
Lavery [22]. The second category is comprised of writers
who are concerned with detailing the analysis, design
and implementation of a single workflow system,
generally one in which the authors have been personally
involved. Included in this category are such authors as
Benson and Cohen [4], Hurwitz 1171 and Schultz [34].
Although these articles are generally insightful on the
design, implementation and operation of workflow
systems, there remains a number of issues which have
not been tackled by either category of research.
As a result of the focus on the efficiency benefits of
workflow automation systems, information on the control
features of these systems is difficult to locate in the
literature. The control elements of workflow systems
tend to be mentioned as only a secondary advantage
which accrue along with the productivity gains [2], [3].
Actual descriptions of the control features which are
enabled by workflow systems technology are also notably
scarce. Yet studies in other areas of the computer field
have indicated that these second-level effects almost
always occur with the integration of computer systems
into an organisation, often with significant consequences
for the operation of the control process [29], [36], [41].
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Orlikowski [29] found that the introduction of
information systems to an organisation can alter the
makeup of the management control system in operation.
This occurred through changes in the relative
combinations of personal control mechanisms, which are
implemented through the direct intervention of human
actors, and impersonal control mechanisms, which are
implemented through utilisation of either the formal,
technical or cultural structures of an organisation. This
study sought to discover whether changes of this nature
were occurring as a result of the implementation of
workflow automation systems and, if so, to determine
what their consequences were for the operation of the
management control process.
Design and Implementation of a Research
Strategy
Although control features had been identified as a
basic component of a workflow system, previous research
has failed to adequately describe the manifestation of this
control component [3], [21]. Consequently, it was
necessary to examine the consequences of the integration
of control features in workflow automation systems for
the operation of the management control process in
organisations. Previous research in other areas of the
information systems field has indicated that the
introduction of information systems can act as a agent of
change in the control process, increasing organisational
dependence on impersonal control mechanisms [29],
[39], [41]. As a result, one objective of this study was to
determine whether this would be the case with workflow
automation systems. Therefore it was desirable that the
research be carried out across a number of management
levels in a number of organisations, to enable accurate
results on the consequences of the implementation of
workflow systems to be established. However, because
very little of the research conducted in the workflow
automation systems area had investigated the
consequences of the integration of a control component
into an organisational workflow process, it was also
desired that the research strategy used would also allow
for the incorporation of qualitative factors.
The satisfaction of these two criteria to the selection
of a research strategy was achieved through conducting
multiple case studies in seven Irish organisations.
Interviews were conducted in a semi-structured format,
and tools such as interview questionnaires, screen
captures and a computer prototype illustrating the control
features being integrated in workflow systems were used
to ensure that interviewees in all of the organisations
shared the same understanding of the background to the
study [40]. A number of pre-test pilot studies had
enabled the design, implementation and testing of a
computerised workflow system, prototype as a means of
illustrating the nature of the control features being
implemented in workflow automation systems. Sample
screens are shown in Figure 3.
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
Figure 3: Sample screens from the research prototype
The selection of suitable research subjects was an
important consideration. Both1 the organisations and the
individuals who were selected for the study were required
to have some experience in the implementation and
operation of workflow automation systems and possess
the intention to introduce further systems of this nature
in the mture. In this way the validity of interviewee
responses was guaranteed. As a result of these criteria,
only financial institutions were used for this study due to
the importance of office processes and potentially large
impact of workflow automation systems on their method
of doing business. Seven institutions were involved in
the study, all of which conformed with the criteria that
worMlow automation systems would have a significant
impact on their method of doing business. It was desired
that the knowledge accumulated would be based on the
experience of people working at all levels of the
organisation. Therefore a minimum of four people were
interviewed in each organisation, representing clerical
workers, lower level manage:ment, IT staff, and upper
level management. This enalbled the development of a
profile of the uses to which workflow technology was
being put in each organisatio8n. Further details on the
interviewees used in the study are given in Table 2.
Table 2: Profile of the Interviewees used in the Study
17
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings
of
the 29th Annual Hawaii International Conference on System Sciences - 1996
The Control Component of Workflow
Automation Systems
Setting Subordinate Tasks and Deadlines
The basic role of control features in workflow
automation systems seems to be evolving. The criteria
which are considered before the introduction of a
workflow system have been expanded to incorporate
consideration of management control facilities. The
variety and scope of these mechanisms of control are also
increasing, with the facilities being introduced
influencing the operation of the control process more
extensively. Consequently, the role of workflow systems
is evolving, with growing consideration of its capabilities
beyond mere productivity concerns, as the their potential
for tightening the control process becomes recognised.
Previous research in other areas of the IS field had
discovered that the implementation of information
systems containing a control element can result in
increased use of impersonal control mechanisms at the
expense of personal methods of exercising control [l],
[5], [29], 1411. Overall, the validity of this finding in the
area of workflow automation systems was upheld by this
study. It emerged, however, that the expected
consequences of the introduction of workflow systems for
the functioning of each stage of the control process were
markedly different. In fact, this study found that changes
in the level of formalisation and the extension of
technical control mechanisms differed at each stage of
the control process. The executive instruction side of the
control process was exposed to a greater shift towards
standardisation and technical control, with a subsequent
reduction in the importance of personal control
mechanisms.
This stage essentially represents an effort on the part
of management to impose boundaries and limits upon the
operating behaviour of employees. As a result, the
introduction of systems of formal and technical control to
this control process stage have been ongoing in the
financial sector for a number of years, due to the
potentially damaging consequences of errors in the
handling of financial transactions. As regards the setting
of subordinate assignments and deadlines, therefore,
interview subjects believed that this would continue to
become increasingly standardised, and that the operation
of this stage would increasingly be incorporated into the
workflow system itself These findings were matched by
a reduction in the amount of management time deemed
necessary to devote to this activity. The importance of
personal methods of setting the standards and deadlines
of employee performance were deemed to be, therefore,
on the decrease.
Monitoring the Operating Behaviour of Employees
Goal Formulation Activities of Management
The goal formulation activities of managers were
identified by interviewees as being one of the control
process stages requiring concentrated human input on the
part of the manager, thus confirming the relevance of the
design of Child’s [9] model. While this task was being
exposed to more in-depth analysis and structuring, its
replication or reinforcement within the format of a
workflow system was deemed to be limited by most
managers and clerical workers. IS staff, however,
believed that the automation of goal formulation tasks
represented nothing more than another challenge for
their skills. In the short term, however, they admitted
that although greater standardisation of this stage would
occur, it would remain largely an area for direct
management control, with limited computerised
assistance or substitution.
Previous researchers had found this stage to be one
where the changes in operation may be the most
profound. This research study indicated that while a
high degree of formalisation would occur at this stage,
and a great deal of the responsibility for subordinate
monitoring would be transferred to the worldlow system,
the estimated decrease in the amount of management
involvement in the operation of the monitoring task was
relatively small. Managers in Irish organisations believe
that the monitoring facilities possessed by workflow
systems will be used largely as a means of improving the
operation of personal control as opposed to substituting
it. Clerical workers and IT staff, however, believed that
tasks in most areas of financial institutions had become
so routinised that there was little need for personal
supervision. Furthermore, the emplacement of
sufficiently stringent standards and computerised
monitoring facilities to oversee the operation of this stage
was believed to be an entirely suitable method of
monitoring the workflow. Clerical workers expressed the
view that developments in the functioning of
organizational work flow meant that
managers had few
tasks to execute in overseeing the operation of this stage
of the control process at any rate. IS staff and top
management concurred, but many expressed the opinion
that lower level managers saw the monitoring task as
being central to their role as a manager and therefore did
not welcome its replacement by either formal or technical
mechanisms of control.
18
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
Recording Employee Output
At the recording of output stage interviewees
generally agreed that formalisation and computer control
were more established here, than at any other stage of the
control process, as the previous research had indicated
[3], [21]. As a consequence, changes in the operation of
this stage based on the introduction of newer workflow
systems, had been perceived to have a minimal effect on
the operation of control over this area. Certainly,
research subjects believed changes in the degree of
personal control exercised to be a very marginal
consideration due to the highly automated and formalised
nature of this stage of the control process anyway.
However, the levels of formalisation and technical
control associated with the recording of output are set to
increase. This was mainly due to the tighter
organisational, national and international standards
being implemented by organisations to standardise the
operation of their transaction recording procedures. It
was believed, therefore, that the need for more extensive
recording and backup systems of financial institutions to
comply with more demanding sets of standards would
result in an increase in the amount of formal and
technical controls coming to bear on this stage of the
control process.
allow for excessive change in its operation, despite the
findings of some of the literature to the contrary [6], [29].
Some degree of standardisation of the procedures
involved in the evaluation OS performance were being
manifested in almost all of the organisations studied,
together with a minimal increalse in the level of technical
control over the tinctioning of this activity. This minor
increase in the extent of impersonal control mechanisms
influencing the operation of this stage of the control
process was expected by top management and IS staff to
result in a diminished amount of management
involvement in the functioning of the employee
evaluation area, although again the changes are not
expected to be significant. Management and clerical
workers in all of the organisations, however, indicated
that they believed that due to its judgmental nature this
stage would continue to be dominated by the direct
personal control. Therefore they expected the operation
of this stage of the control process to continue to be
largely based on the personal evaluation of management.
And, while IS staff and top management strive to
increase the non-managem.ent component of the
operation of this task, in the short term they seem
resigned to this requirement for the use of direct personal
control.
Interpreting the Findings
Measurement aud Analysis of Output
Changes in the control process stage involving the
measurement and analysis of data was deemed by the
interviewees to certainly be noteworthy. Managers had
traditionally imposed a large degree of influence over the
operation of output measurement and analysis in the
functioning of financial institutions. As the variety,
speed of completion and complexity of transactions
continues to increase, however, standardisation and the
automated controls are playing a more important role in
measuring and analysing the output of employees. This
study uncovered sizable increases in the level of
formalisation and a shift towards greater use of technical
controls identifiable in the interviews conducted.
Interviewees at all levels of the organisation admitted
that the process of measurement and analysis of output
had been improved considerable with the drive towards
increasingly automated workflow systems and that even
these marginal changes in the operation of this task were
already significantly easing the burden of personal
performance of this task by management.
Increasingly, organisations are seeking to impose
more formal structures on the (operation of their workflow
processes, but there are limits to the efficiencies which
can be achieved through formalisation. Hence there is a
certain optimal level of stan.dardisation which can be
exercised over any stage of the control process beyond
which the feasibility of further formalisation decreases
and the focus shifts towards searching for opportunities
to replace the standardiseid process with systems
incorporating a greater dependence on mechanisms of
technical control. Thus, for example, the monitoring of
operating behaviour stage had been experiencing
growing standardisation, which had begun to level off,
while the contribution of technical control mechanisms to
its method of operation were expanding at a rapid pace.
Evaluation of Employee Performance
The performance evaluation stage was generally
perceived by interviewees to be an overly intuitive task to
Thus it seems that an increasing influence of
impersonal control mechanisms over the operation of the
control process has tended to occur at two stages. In the
first stage, a workflow system is introduced and, although
management may have to structure and formahse their
behaviour to a greater extent to satisfy standards
introduced in conjunction with the system, its direct
influence on the operation of the control process is
limited. At this phase the operation of the control
process stage becomes more formal, by necessity, in order
to comply with the frequently stringent requirements of
19
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
the system, but the amount of technical control which is
brought directly to bear remains relatively small. As the
control process becomes more structured and formalised,
however, its suitability for replacement by machine
control increases. To compound this, as managers
become familiar with the capabilities of the workflow
system and more confident of empowering it with greater
control over the activities involved, and confidence in its
role as a direct agent of control increases.
greater deal of time interacting with clerical staff and
exploring information system facilities for what they
view as more creative and rewarding work.
The rate at which the formal and technical
mechanisms of control which can be incorporated in the
workflow system, extend their influence over the
functioning of each stage of the control process, depends
on the perceived requirement for human involvement and
the level of technological maturity of the control features
which affect the operation of each individual stage. As
the role of computerisation in the operation of each stage
of the control process becomes more developed, the
extent of influence enjoyed by computers in controlling
the functioning of that stage tends to occur concurrently.
The extent of impact of the control features integrated in
the workflow system on each stage of the control process
is expected, therefore, to increase in the future if IT
personnel continue to implement their agenda of
computerising a greater proportion of the workflow
process. Management and clerical workers, however,
continue to foresee substantial difficulties with increasing
the influence of formal and technical controls over some
stages of the control process where extensive human
involvement is deemed essential such as the goal
formulation and employee evaluation stages of the
control process. As management awareness of the
capabilities of the control features integrated in workflow
systems continues to grow, however, and IT personnel
are dedicated to optimising the influence of worldlow
systems in the operation of the control process.
Overall, the nature of the control task of management
was found to be largely shifting towards responding to
information contained in the system and accordingly
ensuring the compliance of subordinates with the
requirements of the system, a shift which is in line with
previous research findings [41]. The development of
management control systems which operate in this way
also results in the deterioration of the
manager/subordinate relationship through a perceived
reduction in the importance of building networks of trust
and communication in the workplace [26].
Amending the Control Process Model
Child’s [9] model has been used throughout this study
to enable investigation of the impact of workllow
automation systems on the operation of the control
process. Assimilation of the results of the research study
with the control process model yield some interesting
details as to the nature of operation of each of the six
stages of the control process.
Formulating Goals - The results relating to this stage
indicate that its operation is still largely dependent on
direct management intervention in order to function
effectively. Even though some standardisation is
expected, it is largely anticipated that any increase in the
influence of impersonal methods of control over the
operation of this stage of the control process will be
marginal.
As a result of the shit? towards the greater utilisation
of impersonal control mechanisms, brought about by the
introduction of workflow automation systems, the
importance of personal mechanisms of control will
diminish. On this point, however, management groups
disagreed with both clerical workers and IT personnel on
the role of personal control mechanisms in the operation
of the control process. From the perspective of both IT
personnel and clerical workers the role of personal
control mechanisms will continue to diminish in
importance in both the short and the long term. In their
place, they envisage workflow systems gradually
assuming responsibility over the control tasks of
management. However, managers as a whole, do not
believe that there will be a substantial diminution of the
role of personal control mechanisms. Instead, they feel
that the responsibilities which are allocated to the
workflow system will in fact allow them to spend a
Setting Tasks - The findings relating to this stage of
the control process indicate that the degree of
standardisation which is being exerted on the operation
of this area is increasing and that the influence of
technological methods of control over the functioning of
this phase is growing. Consequently, it is expected that
the amount of task and deadline setting activity over
which worlcflow systems will assume control will
increase in the near future.
Monitoring of Operating Behaviour - This stage of
the control process has been the subject of a great deal of
speculation in the information systems field, with
research predicting that the influence exerted over this
stage of the control process by computers would
illuminate the operation of subordinates who use the
system [21], [29], [30], [41]. The findings of this study
indicate that although technology-based control and
increased formalisation are coming to bear on this stage
of the control process, the impact thus far has been
limited and the potential of the technology in this role
may be curtailed by a lack of desire on the part of
20
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
management to utilise it.
Recording Output - Described as a basic function of
workfiow software, recording the output of subordinates
had reached a relatively high degree of technological
sophistication in workflow systems [20], [21]. Changes
in this area were predicted to be of a marginal, though
notable nature, with few further shifts in the direction of
increased exercise of impersonal control mechanisms
expected.
Analysing Output - The measurement and analysis
of data has been described as one of the basic strengths of
information systems 1361. This capability has been
incorporated into workflow automation systems in the
form of the management reporting capabilities which are
being developed. The influence of these capabilities in
analysing data generated by workflow processes in Irish
financial organisations had been seen as being
surprisingly low. This situation is set to change,
however, with some degree of standardisation being
brought to bear on the means by which information is
measured and analysed and the role of the workflow
system in performing these functions is also expected to
grow at the expense of the utilisation of personal control
mechanisms.
Evaluation of Output - The final stage of the control
process involves evaluating the work which has been
completed. This task has been largely performed by
management in the past, with little reference to standard
procedures and little or no involvement on the part of the
workflow system. Although IT personnel were hoping
that they could change this situation in the near future,
the chances of this occurring in the next couple of years
do not appear to be very promising.
Modelling these findings on the control process
depicted by Child [9] shows that the influence of
impersonal control mechanisms in the operation of the
control process seems to moving outwards from the
bottom right-hand side of the diagram in an arc, as
depicted in Figure 4. This arc symbolises the relative
influence of impersonal control mechanisms in the
functioning of each stage of the control process. It’s
position was found to vary from organisation to
organisation, but the trend in workflow systems
development seems to be pushing the arc outwards and to
the left. The result is that to an increasing extent the
operation of each stage of the control process is being
conducted through a combination of impersonal control
mechanisms whose evolution and consolidation is
facilitated by the workflow system.
Feedback
tmportance of Impersonal Control
Figure 4: Amending Child s (9) control process model
Conclusions
Overall, the nature of the management control task is
expected to change radically with the introduction of
workflow systems which include control features. Due to
the increased dependence on the role of the workflow
system in the operation of the control process, the need
for compliance with its requirements will be substantially
raised, a process being continually driven by the need to
conform more closely to national and international
standards embedded in their computer systems. The
supervisory task of the man,ager is also increasingly
becoming a reactive activity, im that to a growing extent
it is based on responding ~to information which is
highlighted by the workflow system. The result is that
electronic supervision is gradually eclipsing the role of
human supervision with the ongoing introduction of
workflow systems. The movement towards greater
dependence on impersonal mechanisms of control is also
diminishing the importance of the manager/subordinate
relationship. This is because the utilisation of integrated
control features is reducing the importance of
maintaining strong relationships of trust and
communication between managers and their
subordinates. Clerical workers, in particular, felt that the
importance of a good relationship with the superior was
weakening as the necessary levels of trust and
communication declined. Management are gaining
greater influence over a wider range of subordinates’
activities and are becoming increasingly results-oriented
with the introduction of workflow systems containing
control features. Clerical staff, meanwhile, will continue
to become increasingly distant from their superiors as IT
personnel press their agenda of increased
standardisation and utilisation of technical control.
21
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
Proceedings of the 29th Annual Hawaii International Conference on System Sciences - 1996
Bibliography
[l]Aldrich, H. and Mueller, S.; The Evolution of Organisational
Forms; Organisational Behaviour, Vol. 4; 1982.
[20]Kimborough, S. and Moore, S.; Message Management
Systems: Concepts, Motivations, and Strategic Eflects;
Journal of Management Information Systems; Fall 1992.
[2]Babcock, C.; Go with the flow, Computer-world; April 1993. [21]Knight, B.; Workj7ow Management: An Overview; Datapro;
October 1993.
[3]Bailey, C.; The Paperless Dream; Network, February 1993.
[4]Benson, B. and Cohen, H.; The AT&T Rhapsody Business
Orchestration Solution; AT&T Technology; Vol. 5, No. 3;
1991.
[22]Lavery, M.; A Survey of Workflow Management Soflware;
Proceedings of the 1992 Document Management
Conference; 1992.
[S]Bjorn-Andersen, N. and Pedersen, P.; Computer Facilitated
Changes in the Management Power Structure; Accounting,
Organisations and Society; Vo1.5, No. 5; 1980.
[23]Marshak, R.; Keyfile: Bringing Imaging and Workflow to
the Desktop; Patricia Seybold’s Office Computing Report,
Vol. 14, No. 11; 1991.
[6]Bjom-Andersen, N., Eason, K. and Robey, D.;
Managing
Computer Impact: An International Study
of
Management
and Organisations; Ablex Norwood, 1986.
[24]McCready, S.; There’s more than one kind of Work-flow
Software; Computer-world; November 1992.
[;?S]Merchant, K.; The Control Function of Management; Sloan
Management Review; Summer 1982.
[7]Brierley, E.; Workflow Today and Tomorrow, Proceedings
of the 1993 DocumentManagement Conference; 1993.
[8]Cafasso, R.; Workflow next rising star?; Computerworld;
March 1991.
[26]Miles, R. and Snow, C.; Organisations: New Concepts for
New Forms; California Management Review; Spring 1986.
[27]Mintzberg, H.; The Structuring
of
Organisations; Prentice-
Hall; 1979.
[9]Child, J.; Organisations: A Guide to Problems and Practice;
Harper & Row; 1984. [28]Nieberding, M.; Image Processing of Health Insurance
Claims; A4C Journal; March/April 1990.
[lO]Child, J.; Managerial Strategies, New Technology and the
Labour Process; In Knights, D., Wilmott, H. and
Collinson, D. (eds.); Job Redesign: Critical Perspectives
on the Labour Process; 1985.
[l l]Coombs, R.; Automation, Management Strategies, and
Labour-Process Change; In Knights, D., Wilmott, H. and
Collinson, D. (eds.); Job Redesign: Critical Perspectives
on the Labour Process; 1985
[29]Orlikowski, W.; Integrated Information Environment or
Matrix of Control? The Contradictory Implications of
Information Technology; Accounting, Management and
Information Technology;
Vol. 1, No. 1; 1991.
[3O]Orlikowski, W.; CASE Tools as Organisational Change;
MIS QuarterZy; September 1993
[31]Panko, R.; Is Offrce Productivity Stagnant?; MIS Quarterly;
June 1991.
[12]Dawson, S.; Analysing Organisations; McMillan; 1992
[ 13]Dzujna, C.; Speeding the Workflow With Electronic Forms;
The Oflce; June 199 1.
[32]Reardon, R.; Essentials of successful corporate electronic
mail; Computer Communications; December 1989.
[14]Evans, R., Impact of computerisation on control; In
Forester, T. (Ed.); The Microelectronic Revolution: The
complete guide to the New Technology and its impact on
Sociev; Oxford; 1980.
[33]Reinhardt, A.; Smarter E-Mail is Coming; Byte; March
1993.
[l S]Hale, D., Haseman, W., and Groom, F; Integrating Islands
of Automation; MIS Quart&y; December 1989.
[16]Hale, K.; What is Production Workflow Management
Software?; Proceedings of the Document Management
Conference; 1993,
[34]Schultz, L.; Computer Programs and Cataloguing: One
Innovative Approach; OCLC Micro; Vol. 4, No. 5; 1988.
[35]Sprague, R.; Invest in Electronic Document Management
Systems; I/S Analyzer, December 1992.
[36]Sprouli, L. and Kiesler, S.; Connections; MIT Press; 1992.
[37]Webster, F. and Robins, K.; Information Technology: A
Luddite Analysis; Ablex Norwood, 1986.
[ 17]Hurwitz, J.; AT&T’s Rhapsody: Targeting Groupware and
Workflow Automation; Patricia Seybold’s O&e
Computing Report, Vol. 13, No. 6; 1993.
[38]Wernham, B.; Groupware meets IS0 9001; American
Programmer, September 1993.
[ 18]Johansson, H., McHugh, P., Pendlebmy, A. and Wheeler,
W.; Business Process Reengineen’ng: Breakpoint
Strategies for Market Dominance; Wiley & Sons; 1993.
[39]Whisler, T.; The Impact of Computers on Organisations;
Praeger; 1970.
[19]Kefalas, A.; Global Business Strategy: A Systems
Approach; South-Western Publishing Company; 1990.
[4O]Yin, R.;
Case Study Research;
Sage; 1989.
[41]Zuboff, S.; In the Age
of
the Smart Machine: The Future
of
Work and Power, Heinemamr Professional; 1989.
22
Proceedings of the 1996 Hawaii International Conference on System Sciences (HICSS-29)
1060-3425/96 $10.00 © 1996 IEEE
