MANAGEMENT CONTROL

Theories, Issues and Practices

MANAGEMENT CONTROL
Theories, Issues and Practices

Edited by
Anthony J. Berry, Jane Broadbent and
David Otley

--
MACMILLAN
Selection, editorial matter and Chapters 1-6
© Anthony J. Berry, Jane Broadbent and
David Otley 1995; other chapters © individual
authors 1995
All rights reserved. No reproduction, copy or transmission of
this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or
transmitted save with written permission or in accordance with
the provisions of the Copyright, Designs and Patents Act 1988,
or under the terms of any licence permitting limited copying
issued by the Copyright Licensing Agency, 90 Tottenham Court
Road, London W1P 9HE.
Any person who does any unauthorised act in relation to this
publication may be liable to criminal prosecution and civil
claims for damages.

First published 1995 by

MACMILLAN PRESS LTD
Houndmills, Basingstoke, Hampshire RD21 2XS
and London
Companies and representatives
throughout the world

ISBN 978-0-333-57243-6 ISBN 978- 1-349-239 12-2 (eBoo k)

DOI 10.1007/978-1-349-23912-2 .1007/978-1-349-23912-2

A catalogue record for this book is available

from the British Library.

10 9 8 7 6 5 4 3 2 1
04 03 32 31 00 99 98 97 96 95
To those who have supported us throughout the project and
continue to give us their support
Contents

Preface ix
Notes on the contributors xii
Acknowledgements xvi
PART I THEORIES OF CONTROL
1 The domain of organisational control 3
Anthony J. Berry, Jane Broadbent and David Otley

2 Approaches to control in the organisational literature 17

Anthony J. Berry, Jane Broadbent and David Otley

3 Structures of control 27
Anthony J. Berry, Jane Broadbent and David Otley

4 Procedures for control 41

Anthony J. Berry, Jane Broadbent and David Otley

5 The context of control 61

Anthony J. Berry, Jane Broadbent and David Otley

PART II ISSUES OF CONTROL

6 Accounting systems and control 77
Anthony J. Berry, Jane Broadbent and David Otley

7 Divisional control 95
M. Broadbent and J. Cullen

8 Strategic control 119

Alan Coad

vii
viii Contents

9 Control of embedded operations: spanning traditional boundaries 139

Anthony J. Berry

10 Economics and control 153

Willie Seal

11 Performance indicators and control in the public sector 163

Peter Smith

12 Organisational culture and control 179

Kim Langfield-Smith

PART III PRACTICES OF CONTROL

13 Management control systems of Japanese companies operating
in the United Kingdom 203
Istemi S. Demirag

14 Management accounting: the Western problematic against the

Japanese application 221
Karel Williams, Colin Haslam , John Williams, Makoto Abe,
Toshio Aida andItsutomoMitsui

15 Management control in an airline 238

C. Wilkinson

16 Control and the National Health Service: some psychology of

managing health care with cash budgets and cash limits 255
Derek Purdy

17 Management control in schools 271

Jane Broadbent

18 Management control in the financial services sector 283

Kim Soin

19 Manufacturing accountability 299

T. Colwyn Jones and David Dugdale

20 Endpiece 324
Anthony J. Berry, Jane Broadbent and David Otley

Bibliography 330

Index 350
Preface

Following the publication of two monographs, New Perspectives in Management

Control (Lowe and Machin, 1983) and Critical Perspectives in Management
Control, (Chua, Lowe and Puxty, 1989), members of the Management Control
Association conceived the idea of collaborating to produce a textbook. The
book was intended to provide a broader and more flexible text on manage-
ment control that those which were available at the time. The book which we
have produced reflects this original aim, along with the interests and idiosyn-
cracies of the various contributors. As such it should not be seen as providing
the last or the definitive word on management control, but instead as a
starting-point. The book does accept that managers are central in the process
of managing, but it does not accept the concept of 'managerialism' uncriti-
cally. It seeks to provide ideas which will stimulate a wider search for knowl-
edge, rather than close down possibilities. .
The book is aimed at those attending both MBA and specialist masters
courses as well as advanced undergraduate courses. It should also provide
food for thought for practising managers. We do not aim to provide a pre-
scriptive and structured course outline, but to provide a resource which can be
'dipped into' on a flexible basis. The volume falls into three parts, each with a
rather different focus and each of the chapters is self-standing and can be
amalgamated into courses to suit individual preferences.
In Part I, the aim is to extend the boundaries of management control and it
therefore explores the various approaches to control which have been adopted
by those writing on the subject. In Chapter 1 we examine the domain of organ-
isational control, focusing on systems approaches. Chapter 2 reviews some of
the approaches to control which have their roots in organisational literature.
In Chapter 3 the way in which organisations and tasks are structured in order
to achieve control is discussed, while Chapter 4 looks at the more detailed
procedures which can be used. Finally in this Part, chapter 5 explores the
context in which control exists. It discusses the extent to which organisations
are constrained by external forces yet are able to reject or buffer the effect of
unwanted changes.
ix
x Preface

It should be stressed that Part 1 does not seek to prescribe the 'one best
way'. We do not believe that this is possible. Thus the material merely seeks to
show the various dimensions of control which need to be considered by the
practising manager as well as the different ways in which control can be con-
ceived. Our hope is that a greater set of possible solutions can be generated as
wider understandings are achieved.
In Part II we narrow the focus rather more and examine particular issues of
control which may be encountered in various different types of organisation.
Again we do not pretend that the list is exhaustive of the issues which could
be raised, but we do feel a range of topical and relevant areas have been
addressed. Chapter 6 examines the general use of one of the most pervasive
tools of control, accounting. Chapter 7 reviews the issue of how to control in
divisionalised companies and Chapter 8 examines issues of strategy. In
Chapter 9 the issues of controlling companies which exist in close relation-
ships with others, those which are embedded in production chains, is
addressed. The contribution that economics can make to control is examined
in Chapter 10. In Chapter 11 the use of performance indicators as a means of
control in the public sector is the area for discussion. Chapter 12 reviews the
literature which has seen culture as either a means to control or an element
which impinges on the possibilities of achieving .control. Following on from
this are two chapters which consider the impact of another culture, the
Japanese, on controls in particular situations. Chapter 13 looks at the way
Japanese subsidiaries operating in Europe are controlled and Chapter 14 takes
a broader look at the extent to which it is possible to speak of a 'Japanese
approach' to control.
Part III is more specific again, this time focusing on control in particular
organisational settings. The settings chosen represent a variety of contexts and
the different authors have taken a variety of approaches. Some chapters are
stand-alone case studies whilst others combine case material w ith more theor-
etical reflection. The five chapters cover the airline industry, the NHS, schools,
financial services and manufacturing. Finally, Chapter 20 reflects on the book
as a whole.
There are other issues which the book has not been able to cover. Issues of
gender and power are ones which we have not addressed directly and no
doubt many other aspects could be listed. The two monographs mentioned
earlier cover important areas and readers are recommended to explore these
books to fill in some of the omissions or to delve more deeply into other topics
of interest. We have also sought to provide a comprehensive bibliography to
enable deeper study of particular areas. Let us stress once more that this book
is merely a starting point, offering some ideas about organisational control,
but never pretending to have all the answers. The sheer complexity of human
society and the creativity of individuals suggests we never will have all the
answers to achieving total control. Our belief is that this is probably a good
thing.
 Preface xi

In writing this book we have had the support of a great number of people.
Direct contributors are listed separately and our debt to them is obvious, but
there are others who have given background support and assistance without
which we could not have completed this project. Mandy Lowndes has pro-
vided secretarial support cheerfully and efficiently and David Simon has
patiently read and commented on early drafts of some chapters. Members of
the Management Control Association have commented on other chapters
which have been presented at some of our regular workshop meetings. Our
thanks go to everyone concerned.

ANTHONY J. BERRY
JANE BROADBENT
DAvmOrLEY
Notes on the contributors

Makoto Abe is Associate Professor in the Economics Department of the

University of Oita in the South of Japan. By training and background, he is a
specialist on labour and employment relations in Japanese manufacturing.
His interest in comparative studies was stimulated by a year's sabbatical
in Britain.

Toshio Aida is Professor in the Sociology Department at Hosei University in

Tokyo . His activities and publications span a broad range in the social sci-
ences . At present, he is particularly interested in Japanese overseas manufac-
turing and the changing balance between Japanese direct investment in
Western market access and in Asian low wage production.

Anthony J. Berry is Senior Lecturer in Management Development in the

Manchester Business School, which he joined after some years in the British
and the American aircraft industries. His research interests are management
development, consultancy and management control.

Jane Broadbent is a lecturer at the Sheffield University Management School,

where she teachers Management Accounting and Management Control. Jane
is a member of the Chartered Association of Certified Accountants, a
qualification gained whilst working as an accountant in the National Health
Service. Her research interests focus on the reforms in the public sector and
the role of accounting in promoting change.

M. Broadbent holds the post of Head of Department of Accounting and

Finance at Manchester Metropolitan University. He is a Fellow of the
Chartered Association of Certified Accountants. He has written texts in man-
aging financial resources and cost and management accounting. His research
interests focus on the interface between accounting mechanisms and manage-
ment control systems in large organisations.

xii
 Notes on the contributors xiii

Alan Coad is a Principal Lecturer at Sheffield Hallam University. Formerly a

management accountant in both private and public sector organisations, his
main research interests lie in exploring the interrelationship of strategy and
management accounting.

J. Cullen is Principal Lecturer in Management Accounting at Sheffield Hallam

University. He joined the university in 1983 after working in several line man-
agement and senior management posts in industry. He is the co-author of a
book called Managing Financial Resources and has written numerous case
studies, articles and chapters with other colleagues. His current research inter-
est include the impact of environmental uncertainty on management control
systems, the links between management control systems and strategy, man-
agement control systems in multinationals, management control systems in
small businesses and issues surrounding business failure situations.

Istemi S. Demirag is Lecturer and Grant Thornton Fellow in Accounting and

Financial Management at the Sheffield University Management School. His
current research interests include financial aspects of innovative management,
strategy and performance evaluation and management in multinational com-
panies. He lectures in financial accounting and management of international
business.

David Dugdale holds the post of Associate Dean in the Bristol Business
School, a faculty of the University of the West of England, Bristol. He spent
sixteen years in industry before joining Bristol Polytechnic in 1987. CIMA-
qualified, Dr Dugdale has worked as a management accountant in industry
and has developed several research interests in the field of management
accounting. Working with Colwyn Jones he has published a number of papers
in the professional and academic press. Together with Colwyn Jones he Ras a
current research project on 'Throughput Accounting'.

T. Colwyn Jones is Principal Lecturer in the School of Sociology, and Research

Fellow in the Centre for Social and Economic Research (CESER) at the
University of the West of England, Bristol. Since 1975 he has been teaching
sociology to accounting undergraduates. He is the author of an undergraduate
text, Accounting and the 'Enterprise: A Social Analysis (1995) and a number of
papers (many with David Dugdale) on accounting rationality, accounting and
technology, investment appraisal, and change in costing systems. His current
research focuses on the role of management accounting in changing manufac-
turing organisations, especially the development of 'Throughput Accounting'.

Colin Haslam is Reader in the Business School at East London University.

Before studying under John Williams at Aberystwyth, he worked for British
xiv Notes on thecontributors

Nuclear Fuels and has published on nuclear waste. He leads the statistical
work for team research with John and Karel Williams which benefits hugely
from his unequalled grasp of official sources.

Kim Langfield-Smith is a Senior Lecturer in the Department of Accounting

and Finance at Monash University, Clayton, Australia, with main teaching
interests in management accounting and control systems. Prior appointments
were at the Universities of Melbourne and Tasmania. Before entering acade-
mic life she worked as an accountant in several commercial organisations. Her
research interests are in the area of management control systems. Current
topics include studying the interface between business strategy and manage-
ment control systems, and investigating new developments in performance
measures. She has publications in journals in both the accounting and man-
agement fields .

Itsutomo Mitsui is Professor in the Economics Department at Komazawa

University in Tokyo. He has mainly researched the problems of small and
medium enterprises and the organisation of supplier networks. As well as
writing for an academic audience, he has acted as consultant and adviser to
Western organisations such as the European Community. This is the second
project on which he has collaborated with Colin Haslam and Karel Williams.

David Otley is KPMG Peat Marwick Professor of Accounting at the Lancaster

University Management School, and General Editor of the British Journal of
Management. He has taught and researched management control for many
years and is joint author (with Clive Emmanuel and Ken Merchant) of the text
Accounting for Management Control (1990). His current interests are in the use of
performance-related pay, and in the impact of corporate strategy on manage-
ment control systems design and use.

Derek Purdy is Reader in Accounting at the University of Reading. His work

has been published in a wide variety of journals which relate accounting and
management issues. He has published extensively about accounting for
convertible debt. His main research interest lies in the effects of financial
management accounting data upon people, their organisations and control
issues, generally construed from a psychological perspective.

Willie Seal graduated in economics from the University of Reading and

worked for Touche Ross as an articled clerk. He has lectured at Bath,
Nottingham Trent and Sheffield Hallam Universities. He is now Research
Leader in Management Accounting and Corporate Finance in the School of
Financial Studies and Law at Sheffield Hallam University. One of his main
research interests has been the application of institutional economics to
accounting and management control (which was the subject of his doctoral
 Notes on thecontributors xv

thesis). More recently he has led a research project on accounting in the Czech
Republic and the relationship between management control and professionali-
sation in UK banking.

Peter Smith is Senior Lecturer in Economics, Finance and Accountancy at the

University of York. He has worked in local government and at the University
of Cambridge medical school. His research interests are public sector
efficiency and public finance, topics on which he has published extensively.

Kim Soin is a researcher in management accounting at Sheffield Hallam

University. Her primary research interest is activity based costing in the UK
financial services sector.

C. Wilkinson is the Deputy Director (Academic Affairs) of Southampton

University Management School. Prior to this appointment in 1993, he was a
Lecturer in Accounting and Finance at Lancaster University for seven years .
His role at Southampton University Management School is the organisation
and management of the full-time and part-time MBA programmes and teach-
ing courses in the areas of accounting and finance, control and strategic plan-
ning systems. His research interests are focused upon the design of control
systems, particularly the development of organisational and managerial per-
formance measures and reward systems.

Karel Williams is a Reader in the Department of Accounting and Finance at

the University of Manchester. He studied nineteenth-century history and
published Pauperism to Poverty (1981) before being radicalised by British
deindustrialisation and taking up contemporary issues through team work
with Colin Haslam and John Williams.

John Williams has recently retired from a chair at the University of Wales,
Aberystwyth, after some thirty years as historian of industrial South Wales.
He has published extensively on South Wales coal-mining and edited the stan-
dard collection of Welsh historical statistics. Since Why are the British Bad at
Manufacturing? (1983) he has worked mainly on more current issues with
Karel Williams and Colin Haslam.
Acknowledgements

The editors and publishers wish to thank the following for permission to
reproduce copyright material:

Harvard Business School Case Services, for Figure 7.1, from B.R Scott, Four
Stages of Corporate Development - Part I (1971); Basil Blackwell, for Figure 7.3,
from M. Goold and A Campbell, Strategies and Style: The Role of the Centre in
Managing Diversified Corporations (1987); Harvard Business Review, for Figure 7.2,
from RS. Kaplan and D.P. Norton, 'The balanced scorecard - measures that
drive performance' (1992), © the President and Fellows of Harvard College, all
rights reserved; Pitman, for Figure 8.1. from RD. Stacey, Strategic Management
and Organisational Dynamics (1993); Accounting, Organizations, and Society, for
Figure 12.1, from E.G. Flamholtz, T.K. Das and AS. Tsui, 'Toward an integra-
tive framework of organisational control' (1985), for Figure 12.2, from J.G.
Birnberg and C. Snodgrass, 'Culture and control: a field study' (1988), and for
Figure 12.3, from R. Simon, 'The role of management control systems in creat-
ing competitive advantage: new perspectives' (1990); Sage, for Box 12.1, from
G. Morgan, Image of Organisation (1986). Simon & Schuster, for Box 12.4, from
R.T. Pascale and A.G. Athas, The Art of Japanese Management (1981); Melbourne
Case Study Services, for Box 12.6, from A. Sinclair and J. Baird, SPC: New Deal
(n.d.): Prentice-Hall, for Box 19.2, from AG. Hopwood, Accounting and Human
Behaviour (1974); Penguin Books Ltd for an extract from The Art Japanese
Managment by Richard Tanner Pascale and Anthony G. Athos (Allen Lane,
1982); Elsevier Science Ltd for an article from Long Range Planning, vol. 20,
no. 5, pp. 42-52: 'Managing diversity: strategy and control in diversified British
companies's by M Goold and A Campbell (1987) and extracts from Accounting
Organizations and Society, vol. 10, no . 1; vol. 13, no. 5; and vol. 5, no. 2; An
earlier version of Chapter 9 was published in the Leadership and
Organisational Development Journal, 15 July, 1994, with permission.

xvi
 Acknowledgements xvii

Every effort has been made to trace all copyright-holders but, if any have been
inadvertently overlooked the publishers will be pleased to make the necessary
arrangement at the first opportunity.
 PART I

Theories of Control
 CHAPTER 1

The domain of organisational

control
Anthony J. Berry, Jane Broadbent and David Otley

Introduction

Organisational control concerns everyone. Whether you are a manager

attempting to run a department, a politician trying to frame legislation to
control multinational corporations, or just an individual affected by the activ -
ities of the many organisations that have an impact on you, organisational
control is a fundamental issue of modern life. In what ways are organisations
controlled? By whom and how? And how can we influence what they do?
These are some of the questions that this book tries to answer.
We shall approach the topic in a questioning and critical fashion, although
primarily from the point of view of a manager who has a role in an organisa-
tional hierarchy. What control mechanisms are available to managers and how
may they best operate the levers of power? But in trying to answer these ques-
tions we shall also need to adopt the points of view of other organisational
participants and ask how controls affect what they do. Indeed control systems
often take on a different complexion when viewed from the perspective of
those being controlled rather than that of those doing the controlling. In addi-
tion, we will sometimes step outside the organisation and address the issue of
corporate governance, or how organisations themselves are controlled by
external interest groups. But our discussion will centre on the use of controls
by managers within organisations, a topic which has become known as
management control.
Many people have attempted to define the term 'management control' and
we shall examine some of the alternative definitions that have been proposed
in this and subsequent chapters. But for the moment let us begin with a simple

3
4 The domain of organisational control

but widely applicable definition: 'Management control is the process of

guiding organisations into viable patterns of activity in a changing environ-
ment: Thus managers are concerned to influence the behaviour of other
organizational participants so that some overall organisational goals are
achieved. Of course, this does not preclude managers from taking other
actions that advance only their own personal self-interest, and which may
even detract from overall goal achievement. Nor does it imply that organisa-
tional goals are fixed or even well understood by most participants. But,
without some control mechanisms, organisational behaviour would degener-
ate into a composite of unco-ordinated activities that are unlikely to possess
the cohesion necessary to allow continued organisational survival.
As can already be seen, discussion of organisational control raises funda-
mental issues concerning the nature of human organisations and the activities
that occur within them. These issues cannot be avoided, although it is all too
easy to let such debate prevent discussion of the more practical issues of
control systems design and implementation. We shall attempt to steer a middle
course by pointing out issues and problems as they arise, yet continually
keeping in mind the central aim of this book, namely, to set down some of the
principles governing the design and use of managerial control techniques
within managed organisations in both the public and private sectors.

The domain of control

In analysing a controlled system, it is first necessary to define the boundaries of

that system. These boundaries are not laid down by some external agency, but
are open to definition by the analyst. Thus we may choose to consider the control
of an individual, either by themselves or by some external agent, the control of a
group of people, the control of an organisation or the control of a whole society.
Secondly, it is necessary to consider who is exercising control. Again this may be
an individual or an organisation; or it may even be an organisation apparently
exercising self-control through a set of designed control mechanisms that relate
to no easily defined individual or group. The boundaries of the controlled
system and those between controller and controlled are thus essentially arbi-
trary, but they are not unimportant. Some ways of looking at a system may be
more helpful than others, so it is up to the analyst to make an appropriate choice.
In general we will draw our systems boundary around an organisation as a
legal entity, but there will be occasions when we wish to include within our
boundaries groups such as customers or creditors who would usually be con-
sidered as external to the organisation. We will also generally consider the
controller to be a manager or a group of managers, such as a board of direc-
tors. But we would wish to emphasise that these are arbitrary choices which
can be varied if a more fruitful analysis can be obtained thereby.
 Anthony J. Berry, Jane Broadbent and David Otley 5

There is also the question of what the system's goals are. In what sense do
organisations have goals and how can we establish what they are? This has
been an issue of considerable interest to organisation theorists, but no conclu-
sive answer has been reached. In a well-known article, Cyert and March (1963)
assert: 'Individuals have goals; organisations do not', although the remainder
of that article is devoted to discussing how the concept of an organisational
objective can be made meaningful. Certainly there is great difficulty in coming
to any conclusion on the issue. Whose goals are we considering and how
stable are they over time? What agreement is necessary between participants
before we can accept a goal as belonging to the organisation rather than to a
group of individuals within it? Indeed who should be regarded as participants
and who are external parties?
One solution was provided by Barnard, in his construct of the purposive
organisation. Another answer was provided by Cyert and March themselves,
who spoke of the goals of the 'dominant coalition' as being essentially the
goals of the organisation. While it would be too limiting to believe that any
organisation had only one purpose or, indeed, that the only purposes of
members of organisations were those of the dominant coalition, it is a helpful
notion to regard organisations as purposive. In that sense, control includes
both problems of regulating the process of the formulation of purpose, and of
regulating the processes of purpose achievement. The schools of managerial
thought which would claim some of this territory would include corporate
strategy and policy formulation.
J.D. Thompson (1967) suggested these problems could be further under-
stood through three major themes; firstly, the establishment of purpose; sec-
ondly, the pursuit of effectiveness; and thirdly, the struggle for efficiency. By
the establishment of purpose is meant the general problem of giving shape
and meaning to the patterns of activity and resource allocation within the
organisation. It is not necessary to claim that purposes are stable over long
periods of time, for they may be formulated in ways which are contingent
upon changes, both within and outside an organisation.
If we are to define effectiveness as a measure of the achievement of purpose,
then without some notion of purpose it is impossible to conceive of any notion
of effectiveness. Usually an effective organisation is one which achieves a sub-
stantial number of its purposes in any given time period. Of course it is poss-
ible that an organisation's effectiveness could be weaker in the long run than
in the short run or, indeed, vice versa. It would appear, then, that the notion of
effectiveness may be rather inexact, in that some measure of the achievement
of a set of fuzzy purposes may itself be 'fuzzy'. Figure 1.1 illustrates this point.
Given that one might locate a boundary for an organisation and test, at that
boundary, whether purpose had been achieved to establish the notion of effect-
iveness, we might also create a boundary around the notion of effectiveness,
and only within that boundary may we discuss the questions of efficiency. Here
efficiency is the relationship of outputs to given sets of inputs. These relation-
6 The domain of organisational control

ships can be expressed in many ways. Often accountants, for example, find
themselves relating the value of outputs in the market-place to the value of
inputs in the factor market-place, and concluding that efficiency gains might
occur if either the value of the outputs rises per unit of input, or the cost of the
inputs fall per unit of output. It is helpful to notice that even these technical
efficiency gains measured in this way might be a confusion of relative price
changes and gains in the technical efficiency in the transformation process.
Thompson has argued, and the present authors agree with him, that technical
efficiency can only be discussed in a bounded system, where the boundaries are
closed for analysis, and that effectiveness can only be discussed in an equally,
but differently bounded, system (Figure 1.1). Purpose, however, by its very
nature, will tend to be unbounded and to be the product of social interaction.

FIGURE 1.1 Bounding domains, for purpose, effectiveness and efficiency

FEASIBLE REGIONS
Purposes

Effectiveness

Input - - . -'Output

\ /
Efficiency
 Anthony J. Berry, Jane Broadbent and David Otley 7

These notions, then, of purpose, effectiveness and efficiency, lie at the heart
of the task of control in an organisation. The regulation of the processes for
formulating purposes becomes an arena of considerable interest, as indeed
does the regulation of the processes of the achievement of effectiveness. It is a
commonplace observation that most accounting control has tended to focus on
the processes of attaining efficiency. However recent developments in, for
example, strategic accounting, have begun to move the focus of this work to
the more general problems of organisational management. In this sense the
control model of Robert Anthony (which we shall discuss in Chapter 2 in
some depth) which talks of strategic control, managerial control and opera-
tional control, is clearly a mirror of the layering of purpose, effectiveness and
efficiency. These three problems, then, will be with us as we pursue the
general puzzles of control, especially that of goals and their achievement.
However we can for the moment sidestep the complexities arising from the
problem of defining goals in a precise manner. For our present purposes we
can substitute 'accepted plan of action' for 'goal' in most control applications.
For example, many organisations seem to move between periods of relative
stability, while agreed plans of action are pursued, and more turbulent
periods when various interest groups engage in processes of negotiation and
bargaining to establish new agreements. Perhaps the minimum overall goal
we need to consider is that of survival. This goal is less problematic in that
most organisations seem to exhibit a fundamental commitment to remaining
in existence (Lowe and Chua, 1983). Beyond that, the importance assigned to
various subsidiary goals appears to be largely a function of the relative power
of the interest groups espousing particular concerns.
We are thus taking a stakeholder view of an organisation where various
interested parties exert their influence to ensure that the programmes of action
undertaken reflect their individual concerns to the greatest extent possible.
The overall plan of action finally settled upon will depend upon the relative
power and influence of the groups involved. Yet it is usually still in the inter-
est of most groups that the organisation continue to survive. From this per-
spective, it is the feasibility of a plan (that is, its acceptability to disparate
groups of interested parties, given their respective bargaining power) that is
the fundamental guiding principle on which subsequent control actions will
be based. Ideas of optimality are very much the icing on the cake, for identify-
ing and operating within such a feasible region is difficult enough. We are
therefore adopting a 'satisficing' (Simon, 1957) point of view where the attain-
ment of satisfactory results is regarded as adequate, rather than pursuing
some concept of the best possible result. However this is not to understate the
importance of goals as having a cultural and symbolic significance.
Individuals may feel commitment to an organisation because of its espoused
goals, even when these goals do not necessarily guide many of its actions.
Indeed it is probably more helpful to think of goals in this symbolic manner,
rather than as their being the guiding feature of a control process.
8 The domain of organisational control

However different organisations exhibit significant differences in the nature

and the use made of their goals. Etzioni (1961) developed a useful typology. He
distinguished between three ideal types of organisation; normative, utilitarian
and coercive. Normative organisations exist where most participants share the
same goals; here the concept of an overall organisational goal is helpful and
can be considered as the aggregate of those goals held in common. Utilitarian
organisations exist where the goals of participants are irrelevant to the activities
of the organisation; participants' involvement is on a contractual or instrumen-
tal basis. Here some form of inducement/contribution analysis (Barnard, 1938)
is appropriate, but the derivation of overall organisational goals is not possible
or helpful. It can be argued that most business organisations are predominantly
of this type. Finally, there are coercive organisations, where the values of partic-
ipants are opposed to those of the organisation (or perhaps, more precisely, to
those of the ruling coalition within it). Here it is only the power of this coalition
that enables it to impose its values upon unwilling participants.
Mechanisms of control differ significantly in each of these different sets of
circumstances. Thus the relationship that exists between individual and organ-
isational goals can be seen as one determinant of the processes of control that
will be used rather than the definition of the object of control. One task of the
management control system of an organisation may be seen as assisting the
organisation to identify a feasible set of activities that will provide acceptable
inducements to all participants to carry them out. In such a way, the organisa-
tion will continue to exist as a viable entity.
But in practice the issue is considerably more complicated than the preced-
ing ideas might suggest. Not only do different organisations exhibit different
characteristics, but different parts of the same organisation may also behave
differently. The norms and values prevalent in one part are often quite dissim-
ilar to those found in another. Different forms of involvement also occur at
different hierarchical levels, with senior managers exhibiting (or being
expected to exhibitl) normative involvement, and lower level workers exhibit-
ing instrumental or coercive involvement.
The study of organisational control therefore involves considerable com-
plexity and is subject to the vagaries of human behaviour. Nevertheless it is
open to analysis and has been approached in a number of different ways. In
Chapter 2 we will take up the approaches of organisation theorists. Here we
turn to a discussion of the approaches to organisation and control in the litera-
ture of cybernetics and systems theory.

Cybernetic and systems approaches to control

The purpose of this section is to outline and examine the concept of control
from the perspective of cybernetics and general systems theory. The term
 Anthony J. Berry, Jane Broadbent and David Otley 9

'cybernetics' was coined by Norbert Weiner in 1947 and was intended to

denote an area of study which covered 'the entire field of control and commu-
nication theory, whether in the machine or the animal' (Weiner, 1948). A more
modern definition, given by Pask (1961), extends it to the study of 'how
systems regulate themselves, reproduce themselves, evolve and learn'. Thus
cybernetics merges into the wider field of General Systems Theory (GST) and
no attempt is made here to limit the scope of the term ; the aim is to draw out
those concepts which are useful in understanding the process of management
control in organisations.
The cybernetic paradigm has underlain much work in management control.
Hofstede (1978) reports that 'a review of nearly 100 books and articles on man-
agement control theory issued between 1900 and 1972 reflects entirely the
cybernetic paradigm'. From the beginning, cyberneticians have been con-
cerned with the common processes of communication and control in people
and machines that were used in attempting to attain desirable objectives, and
have attempted to map the self-regulating principles found in human biologi-
cal systems onto systems of machines. Others have attempted to adapt the
self-regulating principles found in the human brain to organisations. Most
notable in this area is the work of Stafford Beer, in books such as Decision and
Control (1966) and Brain of the Firm (1972). Finally systems theorists have
attempted to develop approaches to the study of systems as wholes, rather
than merely as the sum of their parts. Such approaches have not always been
well-fitted to organisational systems, leading to the development of 'soft
systems' approaches, pioneered by researchers such as Checkland and
expounded in his Systems Thinking, Systems Practice (1981).
We will begin by outlining the cybernetic approach to the study of con-
trolled systems and then extend it by considering the contribution of GST in
general and the 'soft systems' theorists in particular. This section will conclude
with an evaluation of the contribution that these approaches can make to the
study of management control.

The cybernetic concept of control

Cybernetics is intentionally non-specific about the nature of the process being

controlled; in this way it hopes to derive general principles of control that can
be applied in different situations. The basis of controlled activity is seen as
reducing deviations between actual process outputs and those which are
desired; that is, it focuses on negative feedback. Although this may seem to be
a very restricted point of view, part of the contribution of the cybernetic
framework lies in its contention that this negative feedback mechanism is able
to explain much apparently purposive and adaptive behaviour.
This basic process of error reduction can be elaborated following the general
definition of control put forward by Tocher (1970,1976). From his work, a basic
10 The domain of organisational control

model of a cybernetic control process can be derived which indicates four

necessary conditions which must be satisfied before control can be said to exist.

1. the existence of an objective which is desired;

2. a means of measuring process outputs in terms of this objective;
3. the ability to predict the effect of potential control actions;
4. the ability to take actions to reduce deviations from the objective.

These conditions are schematically represented in Figure 1.2, taken from Otley
and Berry (1980). The contribution of this particular scheme over and above
similar models, often presented in the early pages of management control
texts in terms of a detector, a comparator and an effector, is that it emphasises
the central role of the predictive model. The role of this model is reinforced
when anticipatory (or feedforward) control is considered in addition to reac-
tive (or feedback) control. Whereas reactive control waits for the occurrence of
an error and then takes action to counteract it, anticipatory control predicts the
likely occurrence of an error and takes action to prevent it occurring. Thus
control is most effective when the process never deviates from its desired
state. In the context of business enterprises, anticipatory controls essentially
reflect the operation of planning systems. The more complex the system, the
more likely it is that reliance will be placed on anticipatory controls as
opposed to reactive controls. Ashby (1956) notes that a lesson that can be
drawn from biological systems is that it is advantageous to control not by
error but by what gives rise to that error.

FI GURE 1.2 Outline scheme of necessary conditions for a controlled process

I Inputs I >
Measures

(l) Change inputs

(first-order control)

(3) Amend model

of process
j Interrogation of model
(reality judgements)
(internal learning)

.....-1.;--'-::....,.,...----, Generation and evaluation Mismatch

of alternative courses of action signal

(4) Change process (systemic learning)

(2) Amend objectives (second-order control)

 Anthony J. Berry, Jane Broadbent and David Otley 11

Unfortunately these concepts of control do not apply in any straightforward

manner to the analysis of organisational control, an issue explored in some
depth by Otley and Berry (1980). Nevertheless the contribution of cybernetics
may well lie most importantly in the idea that error avoidance can explain
much apparently goal-seeking behaviour. This point of view is put forward
most strongly by Morgan (1979) when he states that: 'Organisms in nature do
not orient themselves towards the achievement of given purposes or ends;
they do not orient themselves towards the goal of survival. Rather they adopt
modes of behaviour and organizational forms which help them avoid certain
undesirable states.' Such an approach certainly seems applicable in explaining
much organisational and economic behaviour, but it must be recognized that
the feedback process is often highly imperfect. As one of the most cogent
administrative writers informed by a cybernetic perspective, Geoffrey Vickers
(1967) writes: 'In the management of human organizations, feedback is often
absent, ambiguous or uninformative and [the cybernetic concept of control]
points to the complementary process of mental simulation which enables
management to function in such conditions.'
This process of mental simulation is essentially that of attempting to predict
the possible outcomes of alternative courses of action. In this context it should
be noted that the cybernetic control model presented here allows the possibil-
ity of adaptation and learning. Indeed this is one of the most important fea-
tures of a viable control system operating in an open system, and will be
explored further when GST is considered. It is this cybernetic perspective that
informs Lowe (1971) when he defines a management control system as:

A system of organizational information seeking and gathering, accountability and

feedback designed to ensure that the enterprise adapts to changes in its substantive
environment and that the work behaviour of its employees is measured by reference
to a set of operational sub-goals (which conform to overall objectives) so that the dis-
crepancy between the two can be reconciled and corrected for.

The cybernetic approach thus represents a logical and abstract approach to

the study of management control systems. It can give some powerful insights
into the operation of control systems, but its very power means that it lacks
specificity. It also tends to assume that control is exercised from outside the
system; self-controlled systems may be better approached from the GST
framework.

The General Systems Theory Approach

The central concept of systems thinking is that of a system itself. That is, the
systems approach stresses the point of view that seeks to explain behaviour by
means of studying the interrelationship of parts rather than the nature of the
12 The domain of organisationalcontrol

parts themselves. It is thus essentially holistic in nature, in contrast to the

reductionist stance of much scientific activity. This approach to systems analy-
sis stresses the importance of emergent properties, that is to say, properties
which are characteristic of the level of complexity being studied and which
may not have meaning at lower levels of analysis. (An example from a physi-
cal system is the concept of temperature which is a property of an assembly of
molecules but which has no meaning in relation to a single molecule). Systems
may be arranged in a hierarchy, derived from an inherent distinction of com-
plexity (Boulding (1956):
1. static frameworks,
2. dynamic systems with predetermined motions,
3. closed loop control or cybernetic systems,
4. homeostatic systems, such as biological cells,
5. the living plant,
6. animals,
7. man,
8. organisations,
9. transcendental systems.
Most control systems theory is derived at a relatively low level of analysis,
and the attempt is then made to transfer it to a much higher level of analysis,
with probable adverse consequences in terms of applicability.
A major contributor to the application of cybernetics and GST to the man-
agement of organisations has been Stafford Beer. In his Brain of the Firm (1972)
he takes the human brain and nervous system as a model for organisational
control. Using this, he identifies five levels of control or systems, labelled
Systems 1 to 5. The lower levels are concerned with the transformation
processes required by the whole system and the maintenance of internal sta-
bility. The penultimate level is concerned with the maintenance of dynamic
equilibrium with the external world, and the final , fifth level, with the self-
conscious determination of goals . The interactions between these systems are
modelled directly from the neurophysiological analogy and are interpreted in
terms of managerial situations. The presentation of Beer's work is intuitive
rather than carefully argued, but, while it contains much stimulating material,
it is difficult to assess how much derives from the models propounded and
therefore its validity is not demonstrable.
GST is primarily a tool for dealing with very high levels of complexity, part-
icularly with reference to systems which display adaptive and apparently goal-
seeking behaviour. That this approach to complexity can be of value to the
study of management control can perhaps best be illustrated by a definition of
accounting put forward by Weick (1979): 'Accounting is the attempt to wrest
coherence and meaning out of more reality than we ordinarily deal with.'
Although borrowed from a definition of art, it is a graphic portrayal of the
central problem faced by accounting, and by management control more gener-
 Anthony J. Berry, Jane Broadbent and David Otley 13

ally. Indeed the approach to management control via accounting, as exemplified

by Anthony (1965), can be seen as an attempt to deal with the control of a
complex, interconnected human activity system by a systematic approach. The
problems in such an approach, summarised in Weick's definition, can perhaps
be best appreciated when one adopts a systemic viewpoint. Thus accounting
controls are the result of a great deal of effort being put into the development of
organisational controls by being systematic; it remains to be seen what will be
the result of a similar amount of effort being applied to being systemic.
An important distinction has been made between 'hard' and 'soft' systems
approaches. The former tend to relate to physical systems having relatively
clear objectives and decision processes, with quantitative measures of perfor-
mance. The latter tend to relate to systems which include human beings,
where objectives are vague and ambiguous, decision processes that are ill-
defined and possibly irrational, and where, at best, only qualitative measures
of performance exist. From this distinction it appears that the 'soft' systems
approaches will have most to offer the study of management control. The
leading proponent of soft systems approaches (Peter Checkland) claims that
systems ideas are used primarily in a process of inquiry, an exploration of the
meanings which actors attribute to that which they observe. Thus the soft
systems methodology copes with the central problem of objectives in a subjec-
tive manner (Checkland, 1981). First, there is a stage of analysis, deliberately
undertaken in non-systemic terms where the analyst becomes familiar with
the rich complexity of the system being studied. Second, a root definition-that
is a fundamental statement of purpose from first principles-of the basic nature
of the system thought to be relevant to the problem situation is sought. Third,
a conceptual model of the system is constructed using the minimum neces-
sary system that can achieve the root definition; this is validated by the data
gathered at the analysis stage. The crucial step in this process is the construc-
tion of the root definition, and this is evidently the most subjective part of the
process. Smyth and Checkland (1976) have attempted to build some safe-
guards into the process of root definition formulation, using checklists and
suggesting that definitions are exposed to participants in the situation. Both
these safeguards are designed to facilitate what Vickers (1965) would term the
process of appreciation, that is, the development of a rich and insightful way
of viewing a real-world situation. Such an appreciative judgement requires
both factual (reality) and value judgements, and therefore any assessment of
its validity is itself an appreciative judgement.
Despite such safeguards it is inescapable that the application of systems
methodology to organisations is dependent upon the subjective judgement of
the analyst. Whether this is considered to be a strength or a weakness depends
upon your point of view; however it signals a substantial shift from the
methodologies of the physical and biological sciences. The model of scientific
activity used by this type of systems approach is quite distinct from that used
in the natural sciences.
14 The domain of organisational control

It may therefore be argued that a systems approach carries within it a con-

servative ideology (Lilienfeld, 1978) as the analyst works within a framework
of co-operative people who co-operate with the ends of the system. It has also
been suggested that systems analysis is in the same philosophical tradition as
sociological structural functionalism. Although this is clearly true of cybernet-
ics and hard systems approaches, it is less true of soft systems approaches
which may be seen as lying more within the verstehen tradition of thought.
This tradition takes its name from the German verb 'to understand' and seeks
through detailed knowledge of the system to understand how the participants
within a system understand it. As Berry (1983) observes, in this school great
stress is laid upon the accuracy and honesty of observation, the sensitivity and
perception of the observer and on imaginative interpretations of observations.
Although there has been much academic criticism of the functionalist
approach, it is important to recognise that it is less limiting than is often sup-
posed. Burrell and Morgan (1979) point out that organisation theorists have
often mistakenly equated open systems theory with the use of an organismic
analogy. There are also wider perspectives available that have been much less
fully explored. These involve either taking a more subjective stance (moving
to an interpretive position) or being more concerned with radical change than
with regulation (regulation refers to the maintenance and continuity of the
system relationships which have been established or have emerged over time).
From the perspective of the study of control systems, the subjectivist position
poses no particular problems; however regulation is evidently of central
importance, although the study of regulatory processes does not necessarily
preclude the use of more radical perspectives.

The contribution of systems thinking and cybernetics

GST and cybernetics can contribute to the study of management control

systems (MCS) in various ways. First, a systems point of view can be adopted
in (MCS) analysis. This is the least controversial approach and may represent
little more than the adoption of an organisational level of analysis and a con-
scious attempt to be holistic rather than reductionist in approach. The most
insightful use of this approach is that of Vickers (1965,1967) who, as a practis-
ing administrator, has attempted to codify his experience in more general
terms by adopting a systems point of view and using cybernetic terms. In par-
ticular he argues for a systemic point of view which explains organizational
behaviour in terms of ongoing relationships rather than by the imputation of
objectives.
Second, there is the use of systems approaches to handling real-world MC
problems, perhaps best exemplified by the Checkland methodology. But
although such an approach may provide a means by which real-world prob-
lems can be dealt with, it does not provide any theoretical basis for the study
 Anthony J. Berry, Jane Broadbent and David Otley 15

of management control. Indeed it discounts the possibility of any such general

theories, arguing that each problem is unique and must be dealt with on its
own merits.
Third, there is the use of concepts developed in cybernetics and GST to
study MCSs and to develpp a theory of management control. However there
seems to be a gulf between the available concepts and their application to the
study of management control. For example, the introductory text by
Schoderbek et al. (1975) presents the basic concepts of systems theory and their
application to management systems in a particularly lucid manner, but imme-
diately runs into difficulty when applying these concepts to any particular
topic. Admittedly this is partly because of a hard systems orientation, but the
conclusions seem disappointingly diffuse and vague after the initial excite-
ment offered by the original concepts. Similarly Amey's (1979) analysis of bud-
getary planning and control systems misses out on the behavioural,
organisational, forensic, strategic and political elements of control, according
to one reviewer (McCosh, 1990). Perhaps most successful in this area is the text
by Maciariello (1984) who makes a consistent attempt to apply the cybernetic
paradigm to management control systems, but even so achieves only mixed
results. Perhaps we shall do no better; the reader is the best judge of that.
Finally, systems ideas have been used in other disciplines related to the
study of MCSs, such as the open systems and sociotechnical systems move-
ments in organisation theory. It is notable that all the open systems theories
considered here adopt, explicitly or implicity, an organismic analogy, utilising
concepts such as survival and functional attributes related to this end, such as
differentiation and integration, and purposive rationality.
In summary, the application of systems thinking and cybernetics to the
study of management control raises more questions than it answers. Yet the
very fact that these approaches raise questions may be valuable in a field that
is pre-paradigmatic in its development. In the West the discipline which has
probably had the greatest influence on management control to date has been
accounting, which is a clear application of a systematic approach. The
systemic approach of systems theorists may provide a useful countervailing
force in developing more comprehensive theories of management control.
A social system contains components, each of which is self-controlling, and
which contain models of the behaviour of the whole. Further, in the study of
organisational control, the system's environment is itself complex, as noted by
Buckley (1968): 'The environment of the enterprise is largely composed of
other equally groping, loose-limbed, more or less flexible, illusion ridden,
adaptive organizations.' If management is an activity which involves an
attempt to control situations involving greater amounts of complexity and
uncertainty than we have techniques adequately to cope with, then the
systems approach offers a method of studying management control that has
the potential to at least recognize some of the fundamental problems. For
example, it may indicate areas in which the wrong questions are being asked
16 The domain of organisational control

and in which inappropriate concepts are being adopted. It is in no sense a

complete theory, but it is the nearest to a method of developing a theory that
we currently have, and it deserves serious consideration for that reason alone.

Summary

In this chapter we have introduced the idea of organisational control and

some definitions of management control. We discussed the idea of the domain
in which control might be exercised, raising the interesting questions of pur-
poses and goals, using them, following J.D. Thompson, to connect organisa-
tional control to the ideas of purpose, effectiveness and efficiency. From a
review of the cybernetic and systems approaches to organisational control we
have argued that these approaches are not limited to mere description, but
offer a systemic frame of thought through which ongoing relationships
become foci of analysis. Further, as Checkland has demonstrated, there is
value in these approaches for the understanding of real-world problems.
While the step from the abstract idea to the concrete event is not trivial, these
approaches do offer both possible routes to practical and intelligent theory
development and also to helpful problem solving.
Far from replacing managers with general system theories, we argue that
the role of the manager is still central to the notion of organisational control.
Also it is possible, we saw, to relate theories of learning to the cybernetic con-
cepts of control to show that these approaches are not static.
 CHAPTER 2

Approaches to control in the

organisational literature
Anthony J. Berry, Jane Broadbent and David Otley

As we noted in Chapter 1, not all aproaches to control have adopted a systems

perspective. The aim of this chapter is to provide a broad outline of the work
of some of the authors and researchers who have considered the problem of
control using various social and organisational frameworks. Social and organ-
isational approaches seek to locate control in their context and, therefore, in
various different ways take into account the structures, the people connected
to them and the environment of the organisation. The approaches give differ-
ential emphasis to the various elements. Subsequent chapters will seek to look
in more depth at the impact on control of the structures of organisations and
the contexts in which they exist, as well as the roles of people within them.
The intention in this chapter is not to provide any prescription as to a 'best'
way to operate systems of control, but to continue the task we began in
Chapter 1, of sketching an outline of the huge diversity of approaches which
have been adopted in seeking to theorise control in organisations.
The 'mainstream' approach is perhaps best represented by the work of
Anthony et al. who can perhaps be credited with the first contemporary
attempts to formalise the subject area of management control (MC), through
the management control systems textbook which is now in its sixth edition
(Anthony et al., 1992). Working from an accounting base they claim to have
sought to broaden accounting's interest and to link it to other disciplines to
provide the frame of interest of Me. In so doing they seek to consider the
structure and design of relationships in the organisational system, as well as
the process - the set of activities that the system does .
More specifically Anthony et al. (1989) provide a framework of control
processes within the organisation in order to try to bound the notion of man-
agement control systems. The management control processes that they define
are threefold: strategic planning and control, task control and management
17
18 Control in the organisational literature

control. Strategic planning and control is concerned with the longer-term goals
and objectives, deciding what these are and evaluating the means by which
they are to be achieved. It is seen as oriented to the external environmental
issues. Task control is the more routine process of ensuring that tasks in the
organisation are carried out, in the terminology of Anthony et al. (1989, p. 11)
'effectively and efficiently'. Management control is the link between these two
elements; it is seen as the process that makes sure that the strategy of the
organisation is reflected in the tasks which are carried out.
The relationship between these three processes is claimed to be distinct and
hierarchical (with strategy setting the agenda of management control, which
in turn sets the agenda for task control) yet the boundaries are seen as overlap-
ping. By bounding the interest of Me in this fashion the authors arrive at their
definition of the subject:

Management control is primarily a process for motivating and inspiring people to

perform organization activities that will further the organization's goal. It is also a
process for detecting and correcting unintentional performance errors and inten-
tional irregularities, such as theft or misuse of resources.

The process they describe is one which includes organisational responsibilities

and authorities. It is a tool for managers and is argued to have psychological
considerations, needing the ability to communicate and inspire other employ-
ees. It is taken for granted that the process is goal-oriented, and efficient and
effective performance is also related to the organisation's goals. Management
control is argued by these authors to be built around a financial control system
and, as such, the budgetary control cycle is seen as central.
The inclusion of financial systems is important enough for the authors to
highlight the distinctions between management, task and accounting control.
Amongst these the authors highlight the people primarily involved in the differ-
ent functions, and their source disciplines. Thus management control is the
domain of managers and has a source discipline of psychology, accountants
deal with accounting control using the source discipline of economics, while
supervisors control tasks, using engineering as their source discipline.
Accounting control, however, as articulated by these authors is seen as very
much akin to the stewardship function of accounting and is not proactive (this
latter feature being seen as attributable to management control). When consider-
ation is given, however, to the descriptions of management control systems
elaborated by the authors, its structures are ones which relate to what are
usually seen as accounting based - a responsibility centre based model - and the
process is one of budgetary control. Thus, although formally the authors distin-
guish between accounting and management control, this rests on what might be
argued to be a very limiting view of accounting. Given the more usual view of
accounting, which includes more managerial accounting such as budgeting, it is
hard to see a distinction between accounting and management control.
 Anthony J. Berry, lane Broadbent and David Otley 19

Thus the exposition of management control that is provided does not do

much to broaden the interest of the topic. Psychology, the source discipline of
management control, does not form a substantive part of the thesis offered.
The actuality is that management control is located very firmly in the domain
of accounting, albeit more managerially based than that rather restricted stew-
ardship function described by the authors. This observation is one which has
been made elsewhere (Lowe and Puxty, 1989) in their critique of this approach
and it is to a critical evaluation of the classical model that we shall now turn.
Lowe and Puxty (1989) provide a critique of what they call the 'prevailing
orthodoxy in management control' and in so doing provide a recognition of
the influence and importance of the 'Anthony' approach. Specifically this
approach reflects a very narrow view of MC, despite its early promise to
broaden the subject. Other of the organisational approaches will show how
the boundaries of the domain can be usefully extended. Lowe and Puxty point
to other problems. The three-tier approach offered by Anthony et al. does
mean that the possibility of offering a holistic view of the organisation is lost.
This means that interlinkages between the different levels, which are interde-
pendent, are not shown, also the environment in which the organisation exists
is omitted from consideration. It is this concern with the internal processes
which leads to a tendency to see control as simply a feedback process, and to a
decoupling of planning and control, such that planning takes place, action
follows and finally there is feedback on that action. This may be contrasted
with a situation in which there is a more dynamic interplay between the plan-
ning and control processes, such that feedforward as well as feedback control
is possible.
Another crucial issue which Anthony et al. do not consider is the issue of
organisational goals. They assume their existence and the possibility of their
existence as completely unproblematic. In so doing, they reify the organisa-
tion; that is, they .treat it as an unambiguous, concretely existing object, and
forget the debate as to whether it is organisations or the people within the
organisations which have goals (Cyert and March, 1963; Pfeffer and Salanick,
1977; Lowe and Chua, 1983). Along with this is the issue of whether stated
goals can be used as an indicator of actual behaviour (Perrow, 1961). This
issue is well debated elsewhere (Lowe and Chua, 1983) but it is essential to
realise that the stated goals of an organisation may not be unproblematically
accepted as the actuality towards which organisational members strive.
These weaknesses can be seen in some of the more specific techniques of
MC which are widely used and can be seen as elements illustrative of this clas-
sical approach. Thus accounting-based controls such as budgeting, standard
costing and variance analysis are typical of the 'classical' controls which are
used in organisations. These approaches are ones which are geared to the role
of management controls which is concerned to operationalise the strategy of
the organisation. They assume organisational goals and indeed institutionalise
them in 'the budget'. The extent to which the environment of the organisation
20 Control in the organisationalliterature

is considered will depend on individual circumstances, but there may well be

a tendency to promote an inward-looking attitude and one in which the
means to control becomes more important than the control which is desired
(Drucker, 1964).

Moving towards different approaches to control

Hopwood (1974) focused on the notion of control, pointing out that control
of the enterprise involves administrative, social and self controls. The
accounting-type controls of the 'classical' model are ones which fit into the
administrative category, but now different dimensions are being added. In
defining three elements of control Hopwood's work provides a bridge be-
tween the 'classical' approach and the ideas in the anthropological literature.
He also provides the linkage to self controls, which will be discussed below.
Social controls are those which are reflected in the 'social perspectives and
the patterns of social interactions' (Hopwood, 1974, p. 26). In essence they are
the elements which define 'the way we do things here'. Hopwood points to
the fact that these factors are likely to intervene in the implementation of the
administrative controls and therefore organisational control cannot be
achieved without reference to these patterns of social interaction and norms.
In a similar vein, Merchant (1985) is concerned to emphasise the social or
behavioural side of control: 'Control is seen as having one basic function: to
help ensure the proper behaviours of people in the organisation.... Control, as
the word applies to the function of management, involves influencing human
behaviour' (p. 4). This he calls 'personnel control' and it includes both social
and self control; it is also linked to what he calls 'action controls' and 'results
controls' . Results control refers to an approach which holds individuals
accountable for achieving particular results, and then rewards them for their
achievement. In so doing it allows the possibility of managers having auton-
omy for detailed action, provided that they produce the desired outcomes.
The outcomes must therefore be quantifiable, and when this criterion is not
available action control might be used. In this situation the outcome might be
difficult to define, but the actions which are required can be specified, and so
the control process is geared towards seeing that the correct actions are
carried out. Neither of these is geared to the social aspects of control with
which this chapter is concerned, but they help to define the domain which
Merchant sees as being the area in which social controls are useful. Thus, in
areas where it is difficult to define and measure outputs and in areas where it
is not entirely clear what actions are required, social controls are seen as
important.
Social controls, for Merchant, include such items as 'getting the right person
for the job' or training and culture. The latter two items may well be linked, as
 Anthony J. Berry, Jane Broadbent and David Otley 21

periods of training may be seen as socialisation processes instilling into a new

recruit the culture of the organisation, 'the way we do things round here' .
Merchant provides a 'broad conceptualization of control' (p. 4) which he
focuses upon the control of behaviour, he also provides some discussion of
the implementation of the systems. While he sees control of behaviour as
important, he sees this as coming, not just from personnel control, but from its
combination with action and results controls. This leads to his conclusion that
multiple forms of control are desirable. However the drawbacks of this (cost
and the increased possibility of dysfunctional side effects, p. 131) are glossed
over by the author. Also, while some consideration of the advantages and dis-
advantages of the different approaches is given, the choice between them is
left to the individual in their particular circumstances, and as such is under-
conceptualised. The conflict between the different elements is not, therefore,
considered and developed. Moreover the 'broad' conceptualisation which
Merchant provides is one which is very inward-looking, and which does tend
to neglect the influence of the wider environment on the behaviour of people
in organisations.
The 'social' theme is also developed in the work of Etzioni and of Ouchi,
who also provides a three-part analysis of the approaches to control, again
linking the social and classical aspects, but providing a rather different
emphasis. Etzioni (as mentioned in Chapter 1) bases his work on the assump-
tion that there are three major sources of control: coercion, economic assets
and normative values. All would exist in organisations but some might pre-
dominate at different times or in different locations. He is centrally concerned
with the notion of power, and power is the means to make subjects comply.
This, if we see control as influencing others to carry out your intentions, is
therefore a control mechanism. Coercive power relies on the physical threat of
pain, or restriction of movement, and might be found in prisons.
Remunerative power is founded on the ability to provide material reward,
salaries and wages for example, and this can be seen to provide the main
element of the control which an employer has over the employee. Normative
power rests on the ability of an organisation to provide symbolic reward to its
members, in a 'pure' form it might be a religious organisation such as a
monastery.
The involvement that individuals have with organisations will then rest on
different bases, and the way in which control is exercised will depend on the
type of involvement. For example, if coercion was the influential mechanism it
would be pointless expecting an individual to act on the norms of the organ-
isation in a voluntary fashion. In most organisations there may well be differ-
ent combinations of all these types of power, those who work in the 'caring
services' might well forgo a larger remunerative reward for the normative
reward of 'doing a worthwhile job', but would still require a salary. This
signals the pluralistic nature of control and it raises the possibility that differ-
ent people may well respond differently to various approaches.
22 Control in theorganisational literature

The attention to the role of power and the use of the pluralist framework is
to be applauded, but the work does have its problems. While the different
notions of power are highlighted the implications of its existence are not dealt
with rigorously. Power is not a zero-sum issue and thus conflict is always a
possibility and this needs to be considered. The extent to which power is legit-
imate or illegitimate can be closely linked to this issue of conflict, as can the
question of positive and negative power (the ability to get things done or not
get things done). Thus the theory does not give much indication of the extent
to which its insights might be operationalised to achieve control in different
circumstances.

Contingency approaches

The different approaches considered above can be seen to provide a link

between the 'mainstream' approaches and those which take a rather broader
view of control. As well as dealing with internal matters such as the social and
self controls, acknowledgement is also given to the structures and the environ-
ment and the technologies of the organisations and the tasks performed in
them. In that respect a link to the area of contingency theory, which has devel-
oped in the organisational literature, can be made. Contingency theory argues
that there is no one 'best' way to approach organisations, but that the organ-
isational design should reflect the environment in which it is found . There is
much work in this area , one well-known study being that of Burns and Stalker
(1961) which differentiated between the 'mechanistic' organisation, operating
in stable conditions and with a known task technology, and the 'organic'
organisation operating in a more uncertain environment, with less defined
technologies. In the former a bureaucratic structure is seen as appropriate; in
the latter situation the laying down of such tight rules is seen as inappropriate
and impossible and thus a more flexible structure is required.
The more recent organisational work of Hannan and Freeman (1977) and
Aldrich (1979) sees the environment as more active in shaping the organisa-
tion . In the population-ecology approach the environment is seen as provid-
ing a 'natural' selection process for firms, those which survive being ones
which are suitable for the given environment. Thus the organisation is seen as
less able to adapt to the environment, survival depends upon fitness for the
changing environment.
There is, then, some tension between the notion that by adopting the 'right'
structure and control system an organisation can survive, and the notion that
the organisation has only a limited space within which to operate in a given
environment. This tension raises questions as to the ability of contingency
theory to provide a satisfactory basis for action. There is also a problem of
specifying the environment. Typically the environment is seen as one or two
 Anthony J. Berry, Jane Broadbent and David Otley 23

specific issues, and there is a danger that it is underspecified in many contin-

gency approaches. Nevertheless contingency theory does signal that managers
and controllers should give consideration to variables in the environment of
the organisation. Consequently it alerts us to a wider set of issues than does
the classical approach.
This rather broader approach can be illustrated by the work of Emmanuel et
al. (1990) which shows how this approach can be used, in the form of account-
ing techniques, as a way to achieve management control. Unlike the classical
approach it signals the different approaches which need to be adopted in
various circumstances. One particular difference which is highlighted is that
between programmed and non-programmed decisions: in programmed deci-
sions there is a clear and predictable link between action and outcome and the
outcome can be planned for; in non-programmed decisions there is a greater
degree of uncertainty making the potential outcome less predictable, and
planning correspondingly more problematic.

Anthropological approaches

Some writers, in moving away from the classical approach, focus exclusively
on the social domain. This type of work can take many guises, but has been
popularised by the work of Peters and Waterman (1982), which has brought
the notion of 'culture' to the fore in management thought. Culture is not easy
to define closely, but may be seen broadly as the norms, values and symbols
which enable members of a society or an organisation to make sense of what
happens in similar ways.
The reviews of Allaire and Firsirotu (984) and Smircich (983) both showed
that there are many different ways to approach the study of culture in organ-
isations. A broad difference emerges between the view that culture is a vari-
able which can be manipulated in order to achieve the 'correct' outcome for
the organisation, and the view that culture is a dynamic and symbolic element
which needs to be taken account of but which cannot be determined exter-
nally. The former situation is that portrayed by the Peters and Waterman
(1982) approach. Here the thesis is put forward that the role of management is
to promote the appropriate culture from which control will 'naturally' flow.
The implicit idea is that control will thereby come from internal self-control,
rather than being externally imposed by bureaucratic rules and regulations.
This assumes that it is possible to impose a culture, and this may not necessar-
ily be the case. Socialisation, through an apprenticeship period, staff training
and selection, may well promote this 'cultural' control, but it is likely that such
a process will be tenuous.
While being helpful in bringing the 'softer' cultural aspects to the fore, this
work has been criticised for its thesis that certain 'excellent' companies achieve
24 Control in the organisational literature

their position because of their culture. The categorisation of excellence has

been questioned, and the companies 'excellent' position has not always been
sustained, throwing some doubt on the thesis. If a more symbolic view of
culture is adopted then the use of culture as a tool of control must also be
questioned. In this view culture is not an independent variable which can be
adopted or imposed at will; it is constructed by social actors in their daily
lives. The Peters and Waterman approach takes a position which suggests that
cultures are things which are imposed or encouraged by managers, and does
not seem to position the ability of other groups to develop their own culture or
to resist cultural imposition.
Brunsson (1985) considers what is required to promote action and differenti-
ates between weak and strong ideologies. Action, he argues, is often promoted
by the ability to act according to predetermined ideas of what is appropriate -
ideologies. Strong ideologies avoid the need to follow long decision processes,
making action and reaction more a pre-programmed issue. Strong ideologies
mean that there is greater agreement about the way action should proceed,
weak ideologies make action less likely as there is room for dissent. It is easier
to promote action through strong ideologies, thus control via a strong ideol-
ogy is relatively easy. However by its very nature a strong ideology is difficult
to change and, should the ideology not be that which is needed (say in a
changing environment) then promoting new patterns of action will be very
difficult. While this provides an interesting insight into the possibility of
change, there is little here that provides guidance as to how a strong or weak
ideology might be recognised. Control in this schema is not so much proac-
tive, as reactive, depending on the strength of the ideology of the organisation
and the extent to which this id eology is aligned to that of the would-be
controller. It is, therefore, not an insight which gives an easy route to help to
operationalise a control system.
Pettigrew (1979) in a processual study looked at the emergence of an organ-
isational culture, the way that reality was constructed within a developing
organisation. This study considers culture much more as a symbolic artefact,
acknowledging its emergence and relating it to its context and history. In a
later study based on the NHS (Pettigrew, McKee and Ferlie, 1989), the implica-
tions of this emergent view of culture are examined. The view put forward
here sees the cultural dimension as fundamental to change, but is much more
tentative and sensitive to the possibilities of promoting change. They point out
that a universal recipe for change is not possible. Highlighting local history
politics and people, they are sceptical about the role of rational design, and
point to the paradox that stability may be necessary to promote effective
change at a later stage. Control in this context is therefore much more of a
political balancing act. It involves working within the cultural system and
moulding it rather than seeking to 'knock it into shape'. While again helping to
sens itise us to the role of cultural issues, this type of study does not provide
any overall answer to the issue of control, and neither does it seek to. This type
 Anthony J. Berry, Jane Broadbent and David Otley 25

of approach is useful in that it highlights the process which must be adopted in

seeking to develop control systems, but it does not specify the content.
This approach is one which has been adopted by Bourn and Ezzamel (1986)
who also consider the NHS and the role of culture in a situation of change.
This work uses that of Ouchi (1980), previously mentioned, and the notion of
clan control: that is, the social control exerted on individual members because
they are part of a group and identify with the group and the common values
of the group. It suggests that new moves for change in the NHS (intended to
change the control of the service) may not have the intended effect as they
promote a culture which is not in line with that which already exists in the
service. This study therefore underlines yet again the importance of consider-
ing culture when control within an organisation is desired. It also highlights
the difficulty of achieving a desired outcome by imposition.
That is not to say that change cannot be achieved. Dent in a study based on
the railway industry, shows a situation where a change has taken place in the
culture of an organisation. This study provides no generalisable conclusions:
again it is an interpretive piece of research and sees situations as uniquely con-
textualised. Laughlin (1991), however, uses the case to provide a rather more
structured argument about the role of the symbolic elements of an organisa-
tion in the process of change. Using a theoretical framework developed from
the Critical Theory of Jurgen Habermas, Laughlin emphasises the symbolic
elements, the interpretive schemes, suggesting possibilities which will help to
understand the processes of change in organisations. This theory is meant to
provide a framework of possibilities which will have to be fleshed out in the
context of particular situations. The point is that it holds as central the role of
the interpretive schemes in determining the pathway of change. This points to
the notion of the interpretive scheme in defining the level of control that is
possible in a situation where control is deemed to be reliant on changing
organisational practices and the organisational cultures which produce them.
The anthropological literature which is relevant to control is therefore some-
what diverse. While all those who have used the construct are equally con-
vinced about its importance, there are those who see it as a possible tool of the
controller, and those who see it instead as a constraining influence, particu-
larly in a changing situation. To this extent it can be implied that an existing
culture can be a useful control in a situation in which there is no change.
When change is needed this is only unproblematic when it does not seek to
alter the culture which already exists. The main problem with these types of
approaches is that they offer no guidance as to the problem of recognising cul-
tures, although in the case of those who see culture as a concrete entity their
manipulation is a substantive control tool. If the more symbolic approach to
culture is adopted then the possibility of controlling cultures disappears. In
different ways both these approaches fail to give direction as to the content of
control systems; they do, however, provide good advice as to the process
which needs to be followed to identify particular controls in particular
26 Control in theorganisational literature

situations. A further problem which is posed by critics of the approach is the

lack of consideration of the element of power.

Summary

In these first two chapters we have attempted to give a broad outline of the
various approaches to control which have been suggested by a number of
authors. In Chapter 2 we have been specifically concerned with sketching the
mainstream accounting-based approach to control and considering various
approaches which have sought to broaden the focus . In so doing we have used
as illustration the work of authors such as Hopwood and Merchant who have
introduced the idea that there are multiple dimensions to the controls in
organisations, some which are administrative and based on organisational
structures and procedures, others being based on social relationships and self
controls. Contingency theorists have suggested that different contexts require
different approaches to control and have sought to identify frameworks for
'matching' control approach to context. Anthropological approaches have
emphasised the importance of the values, ideologies and cultures of those who
are the subject of attempts to control (or who are controlling).
We have sought to explain, briefly, the strengths of the different approaches
as well as the weaknesses in order that managers may be aware, not only of
the diversity of possibilities, but also of their potential in different circum-
stances. We may seem to be somewhat pessimistic in our approach, as every
method we suggest appears to have weaknesses. This is not meant to be dis-
couraging, but is based on our belief that there are no clear indications of the
'right' or the 'best' way to achieve control. We are simply seeking to be honest
about our view that there are no perfect solutions. This does not mean to say
that we can ignore the problems and forget any attempt to control. The rest of
the book will move on to consider, in more depth, different issues of control
and some of the solutions which are offered.
 CHAPTER 3

Structures of control
Anthony J. Berry, Jane Broadbent and David Otley

As Chapters 1 and 2 have shown, there are lots of different approaches to con-
sidering control. Different approaches lead to the creation of diverse structures
within organisations. When we come to consider the structures of control then,
once more, we can look at the creation of these structures through a number of
different 'lenses'. The approaches adopted by organisational theorists and by
economists are the two with which we shall be concerned in this chapter. The
approaches adopted by organisational theorists are important because they
have provided ways of structuring organisations and the processes within
them. The approaches of economists are included because they provide a
framework for discussion of the question as to whether large organisations are
the best way of structuring industries and services . This then provides a frame-
work for the chapter. First, consideration will be given to the question of struc-
turing large organisations and the tasks within them. This will be done
through a reflection of the influence of classical management theorists, as well
as by giving some attention to more recent attempts to structure through divi-
sions or matrices. Next the contribution of the economists will be presented, in
particular the debate as to the reasons for the formation of markets or hierar-
chies in particular circumstances. Finally the debate as to the superiority of
either markets or hierarchies as a structure for organising the supply of goods
and services in contemporary society will be raised.

Organisational theorists and approaches to structures of control

The classical management theorists provide a useful starting-point to begin a

consideration of control structures. The work of F.W. Taylor provides an
important, and, even today, influential account of the way tasks can be

27
28 Structures of control

controlled in organisations. After studying the process of manual labour in a

particular task (shovelling pig iron) he defined the way in which the task
should be done most efficiently, thus increasing productivity. This definition
of the 'one best way' to do a job, coupled with the choice of 'the best man' is
called 'scientific management'. The focus for control was in the separation of
conception and execution of the tasks within a workplace. Managers had to
plan and then workers carried out their behest. This may seem unsurprising to
those of us born into present society, but in an era of craft control of the work
task this was a revolutionary idea. In craft control the task was under the
control of the craftsman who held the knowledge and expertise needed to
carry out the task. Taylor wished to minimise 'sold iering', the restriction of
work from both natural laziness and the restriction of workloads to group
norms. He also wished to make management more effective. He saw the task
of management as being to gain the knowledge of the task skills from the
workers and then to organise those skills in a 'scientific' fashion to ensure
maximum efficiency. Managers had to ensure that the 'best man' for the job
was hired and that the tools required were available. There was also an
element of payment by results, although this was designed to ensure that the
working man could not earn sufficient to engage in drunkenness!
Taylor's work was oriented to the problem of task control. Working in a
similar vein, but looking at the overall management of the firm, was Henri
Fayol, who has also been seen as highly influential. He provides a model
which structures the whole organisation as a unified hierarchy, with control
and authority coming from the highest level and filtering down the organisa-
tion . This familiar structure is illustrated in Figure 3.1.
Fayol, like Taylor, distinguished the role of managers from other functions
in the organisation. This was in line with his overall ideas of division of
labour and specialisation. Part of the manager's role was that of providing
discipline and being responsible for actions of subordinates. For this to
happen there has to be a structure which provided 'unity of command' -
workers being responsible to just one superior. The model supplied by Fayol
has a conception of the organisation as having a unified set of purposes (indi-

FIGURE 3.1 The functional hierarchy

Top Level ..

Management Levels ..

Bottom Levels ..
 Anthony J. Berry, Jane Broadbent and David Otley 29

vidual interests being subordinated to these) . Communication, through which

instructions to enable control are passed, basically follows the hierarchical 'up
and down' structure, although for efficiency horizontal communication is
enabled.
The structures of control which are typified in the work of Taylor and Fayol
are formally described by Weber. Weber was not an organisational theorist as
such. He did not seek to promote particular ways of organising, but he was
interested in understanding the society in which he lived and how it was
developing. He saw the ways of organising as typical of approaches in a
society in general, in which authority is given by the acceptance of rational-
legal principles. This can be contrasted with other societies in which authority
comes from either tradition (direct rule by elders or by the monarchy as a
matter of tradition and by reason of their position) or charisma (the acceptance
of control by those perceived to have special personal qualities, for example a
religious leader or a strong political leader).
Rational-legal authority is that which comes from following a set of rules
underpinned by rational calculation. Authority is given to people because they
have particular abilities and knowledge and the whole structure is based on
predetermined roles and impersonalised relationships. The aim is to provide
efficiency. In providing this description Weber formalised his model of a
bureaucratic structure. We sometimes use the term 'bureaucracy' in a pejora-
tive fashion, as a way of condemning 'red tape', but the term was originally
used positively as a means of improving efficiency. While we might some-
times be dismissive of the elements of bureaucracy, we might also do well to
consider the proposition that the bureaucratic structures of control contained
in the model are routinely employed in almost all organisations in our society,
be they large or small. The reader might also try to imagine an organisation
which does not use bureaucratic controls and what it might be like. This can
be a difficult task!
We cannot consider classical management approaches to the structure of
control without some mention of the Hawthorne experiments, which were
argued to show the importance of the social relations of the work group. These
experiments were carried out in the Hawthorne works of the Western
Electrical Company from 1924 for the next 20 years or so. Many experiments
were undertaken, but they included the study of changing physical conditions
on output. The 'Hawthorne Effect' was that changes in conditions, whether
making them 'worse' or 'better', all produced increased output. This was
argued to be as a result of the increased interest in the work group as a result
of the research. There has been some debate about the interpretation of the
data collected in this huge project (Rose, 1975) but the results are convention-
ally argued to show that the social composition of the work group is as impor-
tant as physical aspects of the working environment. Thus the psychological
effect of increased interest was argued to be as important as physical issues
and issues of 'human relations' became the concern of organisational theorists.
30 Structures of control

While the organisational structures of the whole organisation are not

affected by the adoption of Hawthorne-type approaches, the organisation of
task control is somewhat different. The human relations approach does move
away from the underlying assumption of bureaucratic ones, in which the
employee is seen as little more than a cog in a machine. It accepts that humans
are not just motivated by money. Control of the work process is undertaken
with more emphasis both on the creation of group solidarity and on the social
needs of the workers. One technique which has been influenced by this type of
approach to control is budgeting, where the use of participation is advocated
in setting targets.
The classical management theorists have, therefore, considered the struc-
tures of both organisation and of task control. The structure of task control
suggested is 'nested' in organisational structures which emphasise hierarchi-
cal and centralised structures, and in which authority lies with top manage-
ment. However, as organisations have increased in size, problems of a
centralised and unitary control structure have materialised. Thus division-
alised structures were developed to deal with the problems of allowing
sufficient specialisation and yet provide an integration of the tasks carried out.
Divisionalisation allows for a large organisation to be structured in a series
of smaller units which organise their own activities to an agreed extent, but
which are still accountable to the parent organisation. The amount of decen-
tralisation, that is the extent to which decisions are made at lower levels of
management, can differ between divisionalised organisations and even within
divisions of the same organisation. The divisionalised structure came into
prominence through its use in the General Motors Corporation, by AP. Sloan,
and provides various advantages. Using this structure, decisions can be dele-
gated to be made at or near its environment where there is argued to be better
knowledge of environmental conditions. The structure also provides the possi-
bility to give training and responsibility to a wider range of employees. This, it
is argued, will not only motivate them but also provide a pool of workers
from which to recruit top managers in the future. The structure is also claimed
to provide a means of filtering information flows to top management so that
they do not suffer overload. This is said to make it easier for top management
to identify strategic issues. The careful allocation of responsibility for decision
making can also ensure that decisions can be made faster, as they do not all
have to be made at one central point. The main disadvantage and a crucial
problem for an organisation which has adopted a divisionalised structure is
that decisions which are suboptimal for the organisation as a whole may be
adopted because they are optimal for a particular division.
Divisionalisation can be organised in a number of ways, but popular bases
are product differentiation or geographic location. Whatever the basis, the
extent of decentralisation of decision making can differ depending on
company policy, thus, while control structures of divisions may be similar, the
operation of control within them may differ. (Similarly, within centrally
 Anthony J. Berry, Jane Broadbent and David Otley 31

managed or unitary organisations there can be differences in the levels and

extent of decentralisation.) The operation of control within the divisionalised
company may still operate on the basis of bureaucratic form. To that extent
divisionalisation breaks down the unitary bureaucratic structure but does not
replace it completely.
The decision as to the formal structure of the organisation, be it unitary or
multidivisional, and the decision as to how much to centralise or decentralise
have to be taken and organisational theorists have also considered the extent
to which it is possible to provide general rules about the applicability of differ-
ent structures to different circumstances. This has led to the contingency
theories of organisations which imply that there are 'best structures' but that
these best structures are associated with (or are contingent on) other factors in
the organisation and its environment.
Burns and Stalker (1961) studied the structures of firms and suggested that
those which existed in a stable environment with well-recognised technolo-
gies (mechanistic systems) should be structured in the sort of way suggested
by the classical theorists. In more uncertain environments, where there is
rapid change and a need to respond to this change, organismic structures are
more appropriate. Organismic structures are much more fluid and dynamic,
adapting to their environment in order to survive, as do living organisms.
Here lateral communications are encouraged, roles may change and consulta-
tion rather than command is more appropriate.
Other contingent variables have been suggested; for example, Woodward
(1958) suggested the type of technology of the production process was impor-
tant, resulting in different spans of control. Where the technology is not complex,
the span of control can be wider. Hence the organisation structure must reflect
the need for diverse spans of control. Company size has also been claimed to be
important. Whatever the contingency, the claim is that once the important
contingent factors have been identified then an appropriate control structure can
be identified and used in order to provide the most efficient operation.
We shall finally consider the matrix structure which has been suggested by
organisational theorists to offer another approach to structure. Matrix manage-
ment is that which provides leadership on a functional basis and also on a
product basis. For example, in the educational world it is common to find a
structure which gives subject leadership (function) and also course leadership
(product). Thus a member of staff may be responsible to two different people
for different aspects of their work. This approach allows for development and
control of expertise within the function, but also allows some integration of the
different functions in the development of different products. In the latter case
faster responses to customers can be given. In this way matrix management
structures come nearer to meeting the needs of the organismic organisation
than do bureaucratic structures. The problem of having 'two bosses' can be a
significant disadvantage, and if the product teams are working on short term
projects then the frequent need to build new teams can also cause difficulties.
32 Structures of control

Organisational theory has thus provided a series of suggested structures

which can be used in organisations. All these structures provide different
solutions to control problems and in real-life situations are no doubt adapted
to deal with specific problems which arise. On the whole the solutions offered
by the organisational theorists we have considered are based on an a priori
view of a legal-rational system and on the nature of control resting on internal
organisational relationships.
In many ways economists have a similar view of the structures available for
control. However their existence is justified by a very different view of why
the firm exists . Also more consideration is given to the interrelationships of
different organisations within the economic environment. It is to these views
that we shall now turn.

Markets or hierarchies?

Economists have taken a rather different view of organisations from that dis-
cussed above; they have focused on the exchange relationships which are an
essential part of social life. In particular they have used ideas of economic ration-
ality as a basis to guide the decision as to how to structure exchange activities.
Their basic premise is that division of labour and specialisation which charac-
terises modern production leads to a need to exchange goods and services.
These exchanges or transactions must be co-ordinated and it is the cost of co-
ordination which leads to the decision as to the best structure in which to
organise them. The determining factor will be the cost of the information.
The line of argument can be traced back to the work of Coase (1937). The
starting-point of the discussion relates to the notion of ideal markets. It is
through the medium of the market that exchanges or transactions are sup-
posed to take place . Here is the platform where co-ordination of the transac-
tions takes place, buyers and sellers can be introduced, and allocation takes
place through the price system which balances supply and demand. The ideal
market is one in which this allocation process is efficiently carried out because
of the costless availability of perfect information to all those who might wish
to engage in the market-place. Coase pointed out that the allocation process is
not cost-free, and that the market is therefore not perfect. Therefore exchange
relationships may be better organised through the medium of an agreed con-
tract, especially where the conditions of the exchange are complex. The devel-
oping and agreeing of a contract has the possibility of incurring significant
costs. When these contractual arrangements are of a longer-term nature the
development of an organisational context in which to operate becomes an
alternative and possibly cheaper way of operating. Thus the cost of informa-
tion on which to co-ordinate the allocation process becomes a key variable in
the decision as to whether transactions are best organised through the market
 Anthony J. Berry,Jane Broadbent and David Otley 33

or in a hierarchical organisation. The issue is to decide whether the exchanges

are such that transaction cost can be minimised best by entering into the
longer-term contractual relationship of an organisation, or whether a market
relationship is appropriate.
Williamson (1975) has developed these ideas in the context of what is
referred to as 'transaction cost economics'. In doing so he is more specific
about both the nature of transaction costs and what gives rise to them. First, he
raises the issue of bounded rationality. Bounded rationality accepts that
human beings do not necessarily act to maximise the outcomes of a set of
actions. Because of the complexity of decisions and the limits of people to
process information, people will seek outcomes which are sufficient to satisfy.
March and Simon give the example of the difference between searching a
haystack for the sharpest needle (maximising) and searching for a needle
sharp enough to sew with (satisficing) (March and Simon, 1958, p. 141). Thus,
we have the person who intends to be rational but can only be rational within
the limits of his or her own capacity to process the information available
in a complex world. The consequences, in transaction cost terms, of bounded
rationality are increased by complexity.
A second element giving rise to transaction cost is opportunism. This arises
from the information asymmetry which might exist. Here one person will
have better access to information than the other and this allows a person to act
in their own interest. If this can happen, then the allocations in the market will
be affected and the exchanges in the market-place will not seem to be fair or
effective. This is enabled, in particular, when there are few transactions and
few trading partners. If there are large numbers involved in the market, then a
customer finding him or herself compromised by a supplier can move to
another trading partner; this is not as easily available in a limited market.
It is the bounded rationality implicit in a situation, along with the issue of
the possibility of opportunistic behaviour, which are seen as producing trans-
action costs. To these is added the issue of 'atmosphere'. This relates to the
preferences that individuals have for working in different types of organisa-
tion, and with groups who hold to different value sets. Thus human attributes
are considered along with economic transaction costs, and so different struc-
tures can be seen to be the result. In a situation where there is great uncer-
tainty, Williamson argues that the desire to minimise the costs resulting from
bounded rationality will lead to keeping the transactions within the organisa-
tion, thus favouring the organisational form over the market form of exchange.
Market exchange is seen to be favoured in the situation where there are large
numbers of trading partners, minimising the cost of opportunistic behaviour.
In relation to a consideration of the issue of opportunistic behaviour, the
cost of obtaining information is also significant. Agency theorists (see Baiman,
1982 and 1990, for a thorough overview of the area) who focus on the contrac-
tual relationship between the superior (manager) and subordinate are particu-
larly interested in the cost of the information needed by the superior to
34 Structures of control

monitor performance of the contract. This can be seen as complementary or

related to the markets and hierarchies approach. Spicer and Ballew (1983)
argue that agency theory is a special case of organisation failure. Organisation
failure (Williamson, 1975) is said to occur when the cost of organising in a par-
ticular fashion is greater than moving to another structure.
The two extremes of market and hierarchy of organisation can be argued to
be best seen as ends of a continuum, and various levels of combination of the
two extremes may be the best way to achieve efficiency. Thus structures may
be designed in such a fashion as to mix hierarchy and market in the form of a
divisional structure. Williamson discusses several forms of organisation oper-
ating at different levels of complexity and relating in different ways to co-
operating. Peer groups are seen as a simple non-hierarchical way of
co-operating in which economies of scale in the use of expertise, information
and physical assets may be achieved. As the groups get bigger the problem of
avoiding 'free-riders' means that more formal organisation is required and
simple hierarchies will be preferred. Here there is a recognised superior to co-
ordinate, make decisions and inform.
Because of the impact of bounded rationality and because the impetus to
specialise as a firm grows, the co-ordination problem increases; Williamson
posits a structure called the unitary or If-form organisation in which the co-
ordination process is such that several managers are needed. These managers
will be in control of functional areas, such as production or marketing, and
there will be at least two management layers. The need to communicate up
and down the hierarchy will lead to loss of information and of control. This is
exacerbated by the problem of bounded rationality in a situation which
requires the summarising of increasing amounts of information. Bounded
rationality also means that, as only limited amounts of information can be con-
sidered, some decisions may be unattended. Given the pressing needs of day-
to-day activities, the danger is that longer-term strategic activities may be
neglected.
Because of the growing complexity in larger organisations, and particularly
those which have moved into diverse product markets, Williamson suggests
that a multidivisional (or M-form) organisational structure would be advanta-
geous. This type of organisational structure is one in which there is a special-
isation of managerial decision making and a splitting of responsibility for
strategic and day-to-day decisions. In an M-form structure the day-to-day
operations are carried out in divisions which are quasi-autonomous, allowing
the division to be committed to its own operations. The group is controlled in
a strategic sense from a centre, which is committed to the future of the organ-
isation as a whole, and which has a monitoring role in relation to the per-
formance of the divisions. Organising in this fashion is seen to be reduce
information loss due to bounded rationality. This is the result of the divisions
having responsibility for their day-to-day decisions; this in turn reduces the
need to communicate information about these activities to so many people.
 Anthony J. Berry, Jane Broadbent and David Otley 35

The M-form of organisation can be seen to provide a mixture of markets and

hierarchical organisation. The M-form organisation allows the mixing of the
hierarchy, which controls the strategic issues, and the market, in which the day-
to-day activities are placed. Thus at divisional level some competition takes
place, not just between the divisions and their external competitors, but also
between the divisions for internal allocations of resources. This latter type of
competition may be for intermediate goods to be transferred between the divi-
sions (producing transfer pricing problems) or for basic resources such as
capital or human resources. It is the possibility of achieving the advantages of
the market alongside the strategic direction of the hierarchy which has perhaps
led to the movement towards the implementation of internal markets in some of
the bureaucratically and hierarchically organised public services. This phenom-
enon will be examined in more detail later, in conjunction with considerations of
more recent developments. First, in order to finish the overview of the basic the-
oretical work, consideration will be given to the work of Ouchi (1979, 1980),
who used the transaction cost approach to examine organisational control.
Ouchi alerted us to the fact that transaction costs are linked to the equity or
reciprocity in the terms of any exchange. Where it is difficult to value the
goods or services which are exchanged, transaction costs will be increased as
valuations need to be made and uncertainty must be borne. Ouchi is also con-
cerned with the ability to measure output and the understanding of the way in
which inputs are transformed into outputs. Where it is possible to measure
outputs easily, but the transformation process is not well understood, it makes
sense to use a control system which is concerned with output measurement,
and this Ouchi sees as being most economically dealt with in a market. This is
because there will be little need for internal mediation of the exchange - all
can be measured by the output. Thus the transaction does not need to be
organised through extensive contracts.
In the case where outputs are not easily measured but the transaction
process is well understood, the key element to control is the behaviour of
those involved in the transformation process. In this case the most economical
form of control is argued to be through the organisational structure of a hier-
archy. This is the result of the ongoing employment relationship, which does
not require constant renegotiation and which engenders a sense of belonging
and commitment to the organisation. The costs of monitoring employee
behaviour are therefore seen as less than those of monitoring them through
the market. Figure 3.2 illustrates these relationships.
Where it is possible to measure the outputs easily, and where transforma-
tion is well understood, either markets or hierarchies offer possibilities for
control. Where the opposite scenario exists and it is not easy to measure
outputs and the transformation process is not well understood, then neither
markets nor hierarchies are seen as offering a way of controlling. It is in this
situation that Ouchi posits the usefulness of the self and social controls of the
'clan'. Clan control offers a mutual monitoring through the norms and values
36 Structures of control

FI GURE 3.2 Approach to control of organisations

OUTPUTS Easily measu rabl e Not easily measurable

TRANSFORMATION PROCESS
Well understood Mark et /hierarchies Hierarchi es
Not well understood Mark ets Social contro l of the 'clan'

Source: W.G. Ouchi (1979) (adapted).

of the group. This type of control is more part of the process of action within
the organisations concerned than structure. In evaluating the market-
structured approach to control it is important to remember the existence of
approaches such as clan control. Often markets and hierarchies are considered
but the issue of clan control is not discussed.

The use of different structural configurations

The debate as to what structure is most appropriate is not just an academic

question, but relates to issues which are affecting most our lives in the 1990s.
The economists' approach to organisational structure has been adopted as a
frame of reference, and there is a distinct movement towards the adoption of
'market'-type answers to the economic problems of societies as different as the
UK and the 'emergent' economies of the countries of Eastern Europe.
Ha ving reviewed the main background literature in the area, the ne xt aim
will be to examine an example of the most recent attempts to approach control
issues through the market structure. The overriding str ength of this approach
is illustrated by the fact that it can be found in relation to both the public and
the private sectors. We shall first consider the use of this logic in the public
sector. The belief of the market approach is summed up by Enthoven (1985):
'When all alternatives have been considered, it becomes apparent that there is
nothing like a competitive market to motivate quality and economy of
service. ' (p. 42). This belief has been the basis for much of the recent legislation
in respect of the UK public sector. First implementation of these principles in
organisations such as the National Health Service (NHS) was in their periph-
eral acti vities. It was required that services such as catering, cleaning and
laundry be put out to competitive tender to test the efficiency of the internally
organised serv ices. Similar local government services ha ve also been required
to submit to competitive tendering processes. Having attempted to impose
market discipline on these services, legislation has now gone on to require that
professional services be subject to market discipline. Therefore in local author-
 Anthony J. Berry, Jane Broadbent and David Otley 37

ities, for example, services such as the legal department are planning for the
introduction of competitive tendering for some of their services and the provi-
sion of service level agreements in others. In the latter case a quasi-market dis-
cipline is being attempted, in which local government departments act as
purchasers and providers of each other's services, rather than as different
functional parts of the same organisation. This requires the formalisation of
the requirements of the purchaser and the agreement of the standards of
service. In essence exchange relationships are being created within the organi-
sation, and a type of M-form logic is being used.
This type of logic is being explicitly used to guide changes in the structure
of the UK health and education services. Delegation of responsibility to units
directly concerned with the service provision is being introduced. Thus those
who previously had responsibility for managing the service through direction
of the service units are now the delegated purchasers of the service . The man-
agement of the service provision is in the hands of the units themselves. In the
case of the NHS the units are, for example, hospitals and the creation of self-
governing trusts has given greater autonomy to some units. The intention is
that units must now compete to offer their services to the purchasers and that,
in this competition, efficiency gains will obtain. In the case of education, dele-
gation for service provision has been to school level and, given that funding is
largely related to pupil numbers and freer choice is available in school enroll-
ment, competition between schools is being promoted. There are other issues,
such as the publication of league tables of school examination and test results,
designed to promote this competition.
The question is whether the imposition of market-type controls is effective in
these situations. One of the problems which is being encountered in the NHS is
the provision of sufficient information to allow the market to carryon transac-
tions efficiently (Purdy, 1993a,1993b). Another problem is the lack of use of the
information that has been provided (Purdy, 1991, 1993a) Transaction costs
must be increasing, at least in the short term, in relation to this provision of
basic information. This seems on the face of things to be against to the logic of
transaction cost economics. Using Ouchi's (1979, 1980) framework, it could be
argued that, as the process of transformation of inputs to outputs is not well
understood, and as there is no easy measure of output, neither market nor
hierarchical control is appropriate, but clan control is needed.
In terms of the private sector, the need for flexibility in response to change
has led to an interest in and use of the market structure (Starkey et al., 1991).
Flexibility is seen as an important issue in the context of a rapidly changing
environment and in terms of strategic management. The new manufacturing
technologies and the approaches to management promoted by responses to
Japanese success have promoted flexibility at many different levels . Popular
management texts (for example, Peters and Waterman, 1982) have stressed the
need for flexibility in order to achieve excellence in a firm. The economic and
political failure of the bureaucratically organised regimes in Eastern Europe
38 Structures of control

has given impetus to the belief that the flexibility promoted by the market is
the key to economic success.
In this context, the organisation of exchanges within and between firms has
been examined. In the case of internal organisation, firms have been advised
to ensure that there is flexibility to cope with rapidly changing environments.
Perhaps the greatest effect of this has been in the way in which the boundaries
of firms have been redrawn in order that activities which were once internally
performed are now contracted out. Management buy-outs of sectors of firms
and the contracting out of peripheral activities are consistent with the advice
of Peters and Waterman (1982) to go back to basics and to divest other areas.
Joint ventures provide another means of co-operating in a shorter-term time
horizon. This approach can be seen as an attempt to negotiate and control a
complex environment through the use of contracts. This allows greater flex-
ibility to change the contract mix and also the activity mix of the firm . The
increased cost of renewing and renegotiating contracts may be offset by the
benefits of being able to react to change much more quickly. Yet in some situ-
ations, such as 'just-in-time' systems, the whole approach relies on very close
relationships organised through specific contracts. In this type of relationship
the extent of flexibility is questionable. The extent to which this kind of rela-
tionship is more aligned to market or to hierarchy can also be debated!
The ultimate question to be answered is whether this 'economic' approach
can help us to define the best structure for control in a given set of circum-
stances. The economic or market model is certainly achieving a good level of
support at present, hence its adoption in current thinking about the public and
private sector. However some questions need to be asked, based on the
insights of Ouchi's approach to the area. His model provides an indication
that in some circumstances it is not appropriate to adopt market-type controls,
and the key issue is the ability to measure outputs. There is some evidence
that the current legislation in the public sector is seeking to impose ways of
defining and measuring outputs as this is a fundamental requirement of the
market model. This is fine on one level, but if the situation is such that inad-
equate or inappropriate output controls are being developed, then problems
are likely to occur. There is some evidence (for example in the area of educa-
tion : Broadbent et al., 1992, 1993) of grave doubt about the ability of the output
measures being developed to measure the activity of the organisation in a
meaningful manner. This fear is most pronounced in the public sector in ser-
vices which are geared towards care and service to individuals, such as the
aforementioned case of education, or in health and welfare services. If this is
the case then thee market cannot control effectively. What is more, the
increased transaction costs expended to provide the increased information for
the market to develop cannot therefore be set off against any benefit.
The situation in the private sector also needs to be considered. Although the
divestment of activities and the downsizing of operations may be a good way
of achieving short-term flexibility, there is still a need for co-operation
 Anthony J. Berry, Jane Broadbent and David Otley 39

between producers. The relationships between suppliers and producers in the

'just-in-time' system must be close if the system is to work well. If contracts
are intermittent and all that happens is that the risks of a changing env iron-
ment are passed down to suppliers, the costs of the contract may increase. This
would be the result of producers seeing their risk profile increase and there-
fore increasing prices accordingly.
It needs to be recognised that the decision as to an appropriate structure is
not merely a question contingent on factors such as addressing complexity. It
must also attend to the need for long-term stability and investment (in people
as well as financial resources). This needs a consideration of the patterns of
working, the identity that employees have with their firms, and the scale of
operations, together with patterns of uncertainty and ambiguity. Unless these
things are taken into consideration the danger is that the decision to adopt a
particular control structure will be an ideological one. At the moment evi-
dence as to the success or otherwise of the implementation of the market
based approach in both the public and private sector is not yet available. More
research and exploration need to be carried out in order to inform an ongoing
discussion. This will enable us to broach the question as to whether the deci-
sion between markets and hierarchies is one which can be based on objective
assessment, or whether it is more an ideological choice. Indeed the issue as to
whether the choice between markets and hierarchies is culturally or ideologi-
cally based, rather than objectively directed, is an intriguing one. It is not ,
perhaps, possible to give an answer, but it might provide a useful considera-
tion to bring into play in deliberations about the structures in given situations.

Some final comments: do not forget the organisational theorists

The latter part of this chapter has considered the current debate as to whether
markets or hierarchies provide the better way of structuring the organisations
which provide the goods and services we need in our economy. The reader
might be forgiven for thinking that the suggestions of the organisational theor-
ists are forgotten and part of some intellectual history. This is far from the
case. The explication of the idea of a bureaucracy is one which is still
extremely influential and indeed is reflected in the whole idea of a hierarchy.
The two approaches are not mutually exclusive but provide many overlaps.
Structure is indeed an important element in providing control in organisa-
tions. Whether the structure adopted is theorised through organisational or
economic theory is perhaps less important than recognising that we have no
perfect structures of control available to us yet. It should also be recognised
that in the context of the contemporary debate as to whether markets or
bureaucracies provide the 'best' organisational structure the issue of clan
control has become somewhat forgotten. It should be remembered that clan
40 Structures ofcontrol

control is useful in situations where it is difficult to measure the outputs and,

indeed, where the process of achieving those outputs is difficult to define.
The issue of clan control can be explored in the context of a short considera-
tion of the work of Merchant (1985) who categorises three approaches to
control: action control, results control and personnel control. Action control
involves supervision which ensures that the correct processes are followed.
To use this approach those who seek to control must know the process
required. Results control requires that the outputs be measurable and moni-
tored to ensure they are as required. Where neither of these two possibilities
exists then personnel controls have to be used. This is comparable to clan
control and involves appointing those who can be trusted to do what the con-
troller would wish to do in the same situation.
Much of the recent legislation in respect of the public sector in the UK has
sought to adopt behaviour or results control and personnel controls have not
been pre-eminent. For example, medical practitioners are encouraged to prac-
tise particular approaches to medicine by linking actions such as immunization
programmes to pay . This is an approach based on ideas of action control. In
schools, examination results tables have been published as a way of measuring
outputs. Hence an output control has been developed and adopted. In both
these examples, assumptions have been made about the relevant processes and
outputs. Previously the professionals who undertake these tasks would have
been controlled by personnel controls. Professionals would have been trusted
and subject to much less scrutiny in the day-to-day operation of their tasks.
Now the tasks have been redefined in such a way as to minimise the need for
personnel control, and to allow the control process to operate through preferred
structures, in the case of the UK public sector by the introduction of the market.
The issue of structure, it must be recognised, is not the only issue for consid-
eration when control is considered. It is, however, in the context of contempo-
rary debates, a central consideration. The aim of this chapter has not,
therefore, been to present a picture of what control structure should be. It has
been to present an overview of the ideas available from the organisational and
the economic field. It has aimed to show the diversity of approaches available
and to present an overview of where they are suggested to be of most use.
The strengths and weaknesses of the different approaches have been consid-
ered. Finally some illustration of the current issues and debates of control has
been provided. The fact that structures of control are a current and controver-
sial element in the world in which we live has been highlighted, and some
doubt about the solution of markets as a universal panacea has been voiced.

Notes

1. The gender of the 'best man' is that which was used by Taylor, not our own! We
retain this original gender throughout our discussion of Taylor's work
 CHAPTER 4

Procedures for control

Anthony J. Berry, Jane Broadbent and David Otley

In previous chapters we have addressed the issues of control and of control

structures. In this chapter we address the question of procedures for control
primarily through consideration of planning. We work from the idea that a
plan constitutes a model of the expected or intended future activities of the
organisation and can be a basis for control actions; however we will also argue
that the procedure of planning contains within it more of the requisite variety
for control than is contained in the plan itself.
We can commonly talk of planning as consisting of different elements in the
simple control loop . First, there is a process of establishing the pattern of activ-
ities to be undertaken in the future; it is the representation of that pattern of
activities in physical or financial terms, or perhaps both, that constitutes the
plan. The second part of the cycle is undertaking the work that is contained in
the activity programme and measuring the work that is done. The third part of
the cycle is constructing reports and statements about the work done in relation
to the plan. The fourth element in the cycle is reviewing the relationship
between the expected pattern of activities and the actual pattern of activities and
deciding what, if anything, should be done about it. This is, of course, the simple
control or planning cycle which is best described as a first-order feedback.
In the simple feedback control loop (see Chapter 2) there is a requirement
for a controller or decision maker to make decisions that would change either
the input mix or the transformation process to achieve the desired ends. The
desired ends are usually expressed in the plan which an organisation may
have created. The plan is not the control model; for that, the controller needs
models of the 'transformation process in their environmental context' to
enable him or her to establish a plan or a modification to that plan. Hence
there is a need to consider the control models in use; that is, we must consider
the kind of planning processes which an organisation might create. In this
chapter we set out to discuss, first, planning; second, managing complexity;

41
42 Procedures for control

third, managing with uncertainty and the question of whether it is possible to

conceive of some overall way of fitting these issues together, and what
problems emerge.

Planning

At the heart of a great deal of the literature on planning lie the structural func-
tional ideas of Talcott Parsons (1964). Planning may be seen as a process of
choosing and setting in train patterns of activities in order to achieve certain
goals. The pattern of activities in any sizeable organisation is a matter of estab-
lishing activities and differentiating them into sub-activities. This differentia-
tion necessarily leads to the need for processes of integration. As all of these
activities take place in a changing world, there is a need for adaptation and
change in the pattern of activities and, indeed, the goals. All of this takes place
in the context of a social structure; that is, in a human organisation where
people mayor may not share values, histories, traditions, languages, skills, or
knowledge. These four ideas from the structural functional literature, that is of
goal attainment, integration, adaptation and the social structure, underlie a
great deal of what is actually written about planning and budgeting.
Note that we are distinguishing between the concepts of organisation struc-
ture (see Chapter 3) and social structure. By the latter we mean the patterns of
emergent values, and beliefs, identity and identification, of persons within the
organisation. By the former we mean the pattern of roles, authority, respons-
ibility and accountabilities formally established. Hence, it will be noted, our
interest, in most parts of this text, in the interplay of social structure and
organisational structures and procedures.
In the example of the vehicle rebuild (Box 4.1), drawn from the real experi-
ence of one of the authors, we see how the imperatives of the social structure
can dominate the apparent rationality of an accounting system which was
reflecting the goals differentiation and integration procedures of this part of
the enterprise. In effect the managers were being highly intelligent They were
more prepared to set aside an accounting system which did not reflect the full
circumstances and they were not prepared to force unnecessary changes on
the patterns of work, effort and use of skill of their labour force . Here we see
that task activities in human organisations, whether they are of the physical
work or the accounting representation or payment, take place within social
structures. While these social structures may exhibit multiple values and mul-
tiple characteristics and may rarely operate with any kind of unitary consen-
sus or, indeed, unitary authority, it is visible from this example that the social
structure enabled managers to pursue organisational goals when the organisa-
tion procedures of accounting would cause dysfunctional behaviour.
 Anthony J. Berry, Jane Broadbent and David Otley 43

Box 4.1 Accounting costs and social structures

A company owned a fleet of vehicles, and the life of each vehicle, with
regular maintenance and irregular but well-programmed major rebuild-
ing, was about 30 years. The company knew, five years ahead, which
vehicles would appear when for the rebuild programme. In this pro-
gramme they had organised a system of budgeting and were attempting
to move to a system of flexible budgeting in relation to the work to be
done. Clearly, the work was actually very predictable and was very well
predicted. The company was puzzled because it had apparently lost
control of the cost of the rebuild.
Upon examination it was clear that the standard costs which had been
developed for the rebuild were being largely abused. The company
faced problems in its local labour market in recruiting skilled manpower
without apparently changing the pay rates that it was prepared to offer .
It was necessary then to use the bonus and incentive system as a vehicle
for increasing the take-home pay of the skilled workers without chang-
ing the daily and hourly rate. The method chosen for this purpose was to
book some of the skilled workers' time to a category called 'idle time' .
This then shortened the amount of time booked to do the specific pieces
of work, and this in turn increased the bonus payable on each job. As
booked idle time was held to be the responsibility of the management
and not of the workers, it was paid at the rate of the average bonus
earned that week in the whole of the rebuild system. So the allocation of
time to booked id le time meant that there was a ratchet effect in increas-
ing the bonus and thereby increasing the take-home pay. Further it was
difficult to find the original standards that had been set up by work
study engineers for the maintenance task.

It is sometimes assumed that there is an unequivocally determined time

horizon for organisation plans. In the early days of corporate planning, as in
national planning, it was common to find injunctions to have five-year plans.
However the parts of real organisations work with very varied time horizons -
from a few days or months to many years. So different parts of complex organ-
isations plan over different time horizons, with the financial plan horizon
being a compromise. For example, in the electricity supply industry the time
span of decision making, implementation and consequences for a nuclear
power station total some 45 years (and perhaps much longer). For a fossil
power station they span anything between 20 and 40 years. Laying a new
power supply to your house (a supply which perhaps lasts 30 years) may only
44 Procedures for control

require a few weeks. The very long time scales of implementation and conse-
quences, characteristic of major projects like roads, tunnels, very large bridges
and so on, seem to create problems all of their own, in that it is very difficult to
predict either accurately or with any degree of confidence costs and revenues
over such extended periods. Even in businesses with much shorter time hori-
zons it is fairly clear that the time span of thinking and the time span of control
vary considerably in different parts of complicated organisations. The follow-
ing example illustrates this point. In an organisation which was supplying a
commodity product, the accuracy of estimates or the accuracy of plans that
was established by the organisations was as represented in Table 4.1.
Now it may be seen from the table that the senior managers could claim to
be more effective managers than their subordinates in that they were able to
have a more accurate short-term forecast of revenues and costs at their level of
the organisation than the subordinate managers were able to attain at theirs.
However we might note that the senior managers' major focus of concern was
not on short-term control but on longer-term issues of effectiveness. It can be
seen from the table that there was a curious similarity between some admit-
tedly imprecise notion of time span of control and the error level which was
apparently tolerable or, indeed, merely existing within this organisation (see
Berry and Otley,1986, for a further examination of these issues). Senior man-
agers, though, were not above using their better short term performance as a
means of exerting claims of the greater competence in relation to their subor-
dinates. In addition to the issue of time there are the problems of complexity
and uncertainty, and we will take these up in the following sections.

Managing complexity

There are many studies and approaches to the structures which modern
organisations have developed in order to manage their inherent complexity.
In Chapter 3 we saw that Williamson discusses the difference between unitary
and multidivisional forms of organisation. Another writer, Harrison, has sug-
gested that there are four basic forms of managing complexity: the first of

TABLE 4.1 Estimation accuracy of plans at different managerial levels

Management levels Planning years horizon

1 2 3

Top 0.5 0.11 0.15

Middle 0.11 0.15
Lower 0.15
 Anthony J. Berry, Jane Broadbent and David Otley 45

these is organisational bureaucracy; the second is the organisational matrix;

the third form is a person-centred organisation; and the fourth is a power-
centred organisation. Obviously Williamson and Harrison (1972) and other
authors (for example, Morgan 1986) have different pictures in their minds and
different ways of thinking about organisations. If we take the first model of
the functional bureaucracy, represented in Figure 4.1, it is fairly clear that this
is a model of an organisation with a head office function, where each of the
organisational arms represents some major activity of the organisation; for
example, purchasing, personnel, manufacturing, marketing, finance, and so
on. The character of this sort of organisation, as in all bureaucracies, is that it
will operate with rules, structures and procedures, and certainly its planning
and budgeting will be clearly structured via established organisational proce-
dures. The principal criticism of functional bureaucracies is that, while in
stable states they produce efficient goal attainment, in changing worlds they
tend to be maladaptive. Clearly it is the functional model of Parsons being
replayed here: the bureaucracy is essentially an organisation dedicated to the
pursuit of goals through structures of integration to cope with its internal dif-
ferentiation. It has the control modes of hierarchies and rules, from which are
derived the processes of planning and plans to ensure compliance and goal
attainment.
The next step in organisational form is to move to the horizontal linkages,
shown in Figure 4.2, of the matrix organisation, where it is necessary to cope
with adaptation by linking across the functions because of the interdepend-
ence of the operations. This linking via plans produces the capacity to relate
the inner interdependent parts to the outside world; that is, it creates the
capacity to think about and create adaptive processes. In the person-centred
world, the third type suggested by Harrison, we have an organisation which is
essentially rooted in its social rather than organisational structure, and the

FiGURE 4.1 The Organisational Bureaucracy

1. Funtions such as production, marketing, etc the.

2. Work is delegated from the top .
3. Co-ordination is by rules and procedures, up and
down the hierarchy.
46 Procedures for control

FIGURE 4.2 The Organisational Matrix

Horizontiallinkages across functions modifies the process

of co-ordination via hierarchy to include co-ordination
across the functions, providing scope for faster
adaptive behaviours.

notion of goals and differentiation, integration and adaptation flow from the
interests of the individuals as they go about their work. Some professional
partnerships are like this .
The fourth kind of organisation which Harrison addresses, the power-
centred organisation, is a kind of despotism where all power flows from the
one central figure. It is a picture of monarchy, and it describes quite nicely
some kinds of entrepreneur-created organisations, where the entrepreneur
cannot delegate and cannot trust anybody to undertake work on his behalf or,
indeed, on their own behalf. Most of the rest of this chapter will consider pri-
marily the organisational forms of functional bureaucracy and of the matrix,
as these are the most common core forms of managerial organisations.
Financial planning is, then, a control procedure in most organisations. A
financial plan is usually represented by the four key financial statements; that
is, the profit and loss account, which expresses the expected patterns of activ-
ity and the consequential revenues and costs; the balance sheet, which repre-
sents the pattern of asset structure to support those activities; sources and
application of funds statements, which show the patterning in the way rev-
enues and asset use are intermingled; and the cash flow, which is a simplified
representation of money movements in and out of the organisation. Probably
the most useful of these four for financial planning and control would be a
sophisticated sources and application of funds statement, because that state-
ment encapsulates the changing patterns of financial flows in relation to activ-
ity through time. Overall these four statements, taken together, are
observations of resource flow and the consequences of activities.
The financial plan so created becomes a model for control. However, as it has
been socially constructed and is used in the context of both an organisational
structure and a social structure, it would be unwise to assume (see Chapter 3)
that the 'plan' is uncritically accepted as the basis for all future actions. This
 Anthony J. Berry, Jane Broadbent and David Otley 47

idea of a financial plan fits well with the bureaucratic form of organisation,
expressing integration of the programmes of activity in financial terms .
One critique arises from consideration of the source or the rationales under-
lying the plan. In earlier chapters we referred to the work of Etzioni and his
typology of normative, instrumental and coercive organisations. If, for
example, an organisation has a normative form, then clearly the question of
goal differences does not arise, even though, of course, there are other prob-
lems. If it is a coercive organisation, the goals of those being coerced do not
matter. If it is an instrumental organisation (as most commercial organisations
appear to be) then there must be some mediation of the differences of values
and goals, and so on, within an organisation. Organisations, in pursuit of econ-
omic rationality, use their need to satisfy capital markets, to justify entering
into varieties of control processes of either motivation or manipulation to try
to align organisation members to such institutional goals in a process of goal
congruence. Otley has commented that it might be easier to recognise and
work with behavioural congruence; that is, people can actually work together
while having disparate views on why it is they are working together. The
response of the authoritarian is, in the face of apparent deviance from institu-
tional rationality, to propose more controls. It is also evident that the greater
the control, or the number of controls that are invented, then the greater the
deviance that apparently takes place. This endless game was charted elegantly
by Hofstede (1967) and more sharply and more clearly by Van Gunsteren
(1976) who pointed out the infinite regress of rule-based control (if there is a
rule A, then we need a rule B, to define when A shall be applied. If there is
now a rule B, we need a rule C, and so on). It seems that we cannot solve the
issues in the social structure by creating additional control procedures in the
organisation structure.
Simon (1957) and other researchers noted that managers in organisations
behave as though they are not pursuing the same goals; as though the organ-
isational plan and control mode did not wholly provide the criteria for their
actions. He offered the following explanation. Given that large organisations
deal with complexity via differentiated structures and procedures, they create
the existence of sub-goals (and a need for integration). As the differentiation is
not a simple decomposition of the work of the board, but is an elaboration of it
- in products, technologies, markets and so on - then local rationalities are
created, in relation to which local goals emerge. Hence managers working on
such local goals are constrained by corporate goals rather than merely subor-
dinated to them.
This brings us to what we might term the complexity critique in financial
planning and control. Given the nature of complexity in organisations and the
nature of differentiation, we clearly create a need for integration (a need that is
powerfully served by accounting processes) and this is integration both of
thought and action. In the Anglo-Saxon cultures, great store is set by a notion
of autonomy and autonomous, personally responsible action. In the Asian cul-
48 Procedures for control

tures, a greater store is set by the notion of dependence and joint action. It
would seem that, as in so many things, the truth of managing complex organ-
isations lies between these two polar positions of autonomy and dependence.
We find ourselves struggling to handle the interdependence between the
internal and external worlds of organisations, meeting highly varied environ-
mental demands, opportunities and constraints, while trying to hold to the
stability of a transformation process and deliver goods and services into the
product market-place. Given that statement, the question arises as to what can
be differentiated. It would appear that the simple notion of the institutional
goal and the institutional decision maker merely delegating different parts of
the tasks is too simple. It would appear that, as organisations grow in size and
complexity, it becomes logically necessary to differentiate work and the deci-
sion authority about what that work actually is. Having had that level of dif-
ferentiation it would seem that we must rethink our notion of financial
planning to create a control model, because it becomes impossible to see this
merely as a process of disaggregation and aggregation of well understood and
completely rational processes.
In more formal terms, we are arguing that the financial plan - because it is
created in a complex organisation through highly differentiated procedures -
is a variety-reduced descriptive model of the future activities of the organisa-
tion . Such a plan as model will lack the requisite variety for control which
Ashby (1956) demonstrated was necessary for determining the state of the
system. However it is the case that the model/plan and the processes which
create it would contain substantially more variety and might approach the
requisite variety for control.
[alland (1989) introduced an elegant notion of plans between parts of organi-
sations forming contracts or compacts, multidimensioned in nature, rooted in
patterns of history and notions of incomplete understanding. These compacts,
then, were spaces or domains within which there was an agreement to work, an
agreement to behave in a certain way, through which skill, experience, knowl-
edge and competence were mobilised at, for example, lower levels, as part of a
bargaining process for resources to enable the organisation to function (with
variety in the compact and in the process). It is a way of construing organisa-
tion as a rather late-twentieth-century form of Rousseau's social contract, that
the planner's compact creates sufficiently common frames of understanding
through which people will have the confidence to allocate major amounts of
resource and to enable managers of differentiated units to handle both the dif-
ferentiated tasks and the differentiated decision authorities that, perforce, must
go with them. Through this process, then, the problem of adaptation can be
handled, for it follows that adaptation must also be handled as a differentiated
process. Here we see adaptation taking place on two levels: firstly, the macro-
institutional level, which concerns the basic configuration of enterprise, and
secondly, at a micro level, which concerns specific elements of the environment
to the organisation and finding the appropriate configuration of the internal
 Anthony J. Berry, Jane Broadbent and David Otley 49

micro world for effective relationship with these specific elements.

Clearly one response to the problem of complexity is to seek control through
vertical co-ordination. This process of establishing goal programmes and sub-
goals and so on is characteristic of the functional bureaucracy which presup-
poses that decomposition is an appropriate algorithm for its internal world. To
some degree the technique of management by objectives follows that particular
approach. We suppose that in some limited circumstances that might work.
More likely, though, is that the debate about goals would be on something of a
dialectic or a process of negotiation between institutional and internal sub-
systems. The dialectic would have to recognise the location of knowledge and
expertise and competence to undertake work in the complex organisation.
Depending upon the nature of work, complexity may be severe at 12 or more
people, and almost certainly becomes so at numbers from 20 to 40 people. It is
actually impossible for one person to know everything, therefore the contribu-
tions of others are required, and these contributions are unique. So we see
again that the procedures of creating a plan or control model are richer and
probably more important than the plan itself. In the dialogue about the cre-
ation of the plan we would expect to find the explanatory and predictive
models from which managers argue their case for the plan itself.
What seems to arise here is the principle of subsidiarity, by which is meant
that, as a means of coping with complexity, tasks and decision authority
should be delegated as close to the point of enactment as one could possibly
get. This principle of subsidiarity, which is a part of the evolution of the
European Economic Community and its relations with its member states and
the regions and local structures of those member states, constitutes a very
interesting model for commercial organisations. The principle of subsidiarity
is a response to the problem of complexity and internal differentiation which
would appear to be appropriate if the differentiated units could effectively
work for the most part separately from each other; for example, where the
interdependence between the units is low we can create the classic multidivi-
sional form of organisation. Where the interdependence between parts is very
high, one has to find a means of managing that interdependence. It is clear
that major problems arise at this point for goals, co-ordination and control.
An extreme notion of subsidiarity essentially says that an institution is an
aggregation of parts; in this sense the differentiated parts may negotiate
between themselves as to what form of institution is actually needed. In the
reverse of the case of an institution dominating its differentiated parts,
co-ordination procedures would emerge to constrain the managers' actions.
Those of our readers with a sense of history will recognise very nicely the
conflict between the king and the barons which led the barons to force the
king to sign the Magna Carta. This problem has also been played out in
the evolution of the political structures of the United States of America and is
unlikely to rest easily in the emerging structures of the European Economic
Community. It is also a problem that emerges time and time again in the
50 Procedures for control

debates about the nature of multidivisional organisations. It is possible to see a

rather simple-minded dynamic between integration and adaptation when
problems arise in a highly complex multidivisional organisation, for if prob-
lems emerge about managing interdependence hierarchically and between the
divisions, then the solution proposed is almost inevitably in forms of central-
isation. If problems arise in centralised or unitary organisations as a product
of complexity and ambiguity, the solution normally proposed is to decen-
tralise or divisionalise. Of course these are polarities and, as Peters and
Waterman (1982) found, there is much scope for acknowledging the paradox
when they observed organisations with what they termed tight or loose
control structures, by which they meant that decision authority on some
matters was held centrally (tight co-ordination) while on other matters it was
delegated (loose co-ordination to permit adaptation).
Perhaps more formally it is possible to argue that the principle of subsidiar-
ity does not imply either a multidivisional form or a unitary form. What the
principle of subsidiarity might lead us to in the design of organisation control
structures is discerning the relevant and appropriate location of task and deci-
sion authority. Note, though, that these are all processes of control through
vertical co-ordination, where the task is to hold the integrity of the institution
together in the interests of the owners, or other principal stakeholders, such as
the managers. The problems that arise here are the problems of aggregation
across complex differentiated units which insist on working with different
technologies and different time horizons. In the face of such complexity there
is much sense in central or institutional control structures being construed
with minimum variety. Hence financial models and financial targets, using the
four financial statements, constitute one reduction of the complexity, delegat-
ing the processes of interpretation (or holding uncertainty) to unit managers.

Programme planning and budgeting systems and zero-based budgeting

In this section we turn to a discussion of two approaches to managing com-

plexity through financial planning procedures: the first , programme planning
and budgetary systems (PPBS), refers to the continuing integration of pro-
grammes with plans and budgets; the second purports to be more radical,
zero-based budgeting, which is control through a process of requiring future
rejustifying of all actions.
Where there was high complexity and interdependence in organisations, it
became important to try to create some institutional means of handling it. The
story is often told of how Robert McNamara used his industrial experience as
a basis for his attempt to manage complexities and interdependence in the US
Department of Defense (see Box 4.2) by the introduction of programme plan-
ning and budgeting systems. The essential point of programme planning and
budgeting systems may be seen by referring to the models of organisation as
 Anthony J. Berry, Jane Broadbent and David Otley 51

functional bureaucracy and as a matrix. In a government there may be activ-

ities which contribute to some overarching policy; for example, the depart-
ments concerned with education, health, industry, welfare and housing may
all be contributing to a policy of raising the level of social and economic well-
being of a section of the community. Should the government organise its work
only on a functional basis then the recipients of these efforts may experience a
measure of discontinuity of policy and a lack of effectiveness. Further,

Box 4.2 An example of changes in the distribution of power through a

change from a functional to a matrix organisation

It is said that in the bad old days there were four armed forces in the
United States, each of which contributed one member to the Joint Chiefs
of Staff Committee which reported to the Secretary of State for Defense.
These four units were the Navy, the Air Force, the Army and the
Marines. It was observed that in any serious fighting these four units
mostly had to co-operate in a highly interdependent way to provide an
effective military force in any given theatre of operations. It was also
observable to the public that the United States Air Force maintained a
small army and a rather small navy, the United States Army had some
seaborne capability and a great deal of airborne attack capability, the
United States Navy had some major aircraft carriers and battle fleets
which had enormous numbers of strike aircraft as part of their military
capability, and the United States Marines were always designed as a
flexible force with flexible resources. Interesting questions arose as to
which of the three major branches of the armed forces should have
control of intercontinental ballistic missiles. In practice they each had
their own systems. It was therefore felt that the growth of these four
enterprises had led to a certain measure of redundancy. However, given
the structure of the Joint Chiefs of Staff Committee, it was also observed
that the Secretary of State for Defense would always be presented by the
Joint Chiefs of Staff with their own solution to whatever problems
emerged. Essentially the power of decision making rested not with the
Secretary of State but with the Joint Chiefs of Staff. Readers may have
entirely different views about what they regard as appropriate.
However, Robert McNamara, as Secretary of State for Defense, it is
reported, created civilian directors of theatre operations reporting
directly to him. The Joint Chiefs of Staff were then told that if they
wanted any military resource or any money to provide that resource
they would have to negotiate with the directors of theatre operations,
who were the budget holders.
52 Procedures for control

departments may use the existence of such a policy to bid for extra funds but
find it expedient to allocate some of them to equally worthy but competing
activities. In short the very thrust for efficiency which functional specialism
enables may render the policies less effective.
The idea of creating programme managers, across the functions and hence
along the matrix, to co-ordinate such a policy initiative seems simple enough.
However if the co-ordinators have less authority or micropolitical force than
the functional managers then their co-ordination role will become one of mes-
sengers rather than mediators. A second-stage solution to this problem is to
provide a programme or policy manager with the budgetary allocation and
invite the functions or departments to bid for resources from the programme
manager so that he or she can hold them to account and in turn be accountable
for the policy initiative. Here the matrix managers have more resource power
than the function heads. Hence it is clear that the idea of PPBS is the creation
of a means of integration to ensure the adaptation to programme needs
becomes crucial to the behaviour of departments. It is also clear that the idea
of an internal market can be used here, for that is how functions obtain
resources to do work.
Of course in any ongoing organisation as complex as a government it is
unlikely that such internal market power will be the dominant mode of behav-
iour. Long-lasting organisations develop all manner of social structures,
customs and histories which would mediate the crude application of these inter-
nal market arrangements. A slight insight into such social structures may be
gained by noting the routes for promotion. For example, do the programme
managers aspire to be departmental heads, or vice versa? Further consideration
of the fact that government policy initiatives do not get created in a vacuum but
are the outcome of considerable external and internal debate should give the
sceptical good cause to doubt that simple structural changes would give rise to
major changes. However PPBS stands as an interesting contribution to the
pursuit of the rational in the allocation of resources in pursuit of organisational
goals. Its very introduction might create some interesting political realignments.
A further technique used in pursuit of rational resource allocation in the
achievement of goals is zero-based budgeting (ZBB). This requires all bidders
for resources to assume no history or precedent and to justify from a zero base
all resources sought in the budget. This somewhat unlikely approach, espe-
cially in organisations with ongoing commitments, might cause a deal of irri-
tation and time wasting in order to satisfy the 'new manager' that previous
behaviour was not perverse. It makes a good if dangerous weapon in the
hands of a new manager, for in such an approach is an invitation to reassert
authority and perhaps power. ZBB is susceptible to the same critiques as PPBS
and all other prescriptions for the narrowly rational pursuit of resource
control in the pursuit of goals . In simple terms, PPBS may be represented as in
Figure 4.3. Note that in a PPBS system resources are allocated to functions by
programme managers.
 Anthony J. Berry, Jane Broadbent and David Otley 53

FIGURE 4.3 ResourceFlows in PPBS

Top Management
PPBS I !TraditionalRoule
Route I /

Function Function Function

Manager A ManagerB ManagerC

-.. Programme 1
(Manager)
10.., -,
J'L... - '
10..,
"L... - '
-, ... ..- Co
J'L... I--'

-. Programme 2
(Manager)
10..
J'L
i\
~
10..
J'L. ~
/1\ 10..
J'L. ~
/1\

~
Programme 3
(Manager)
...
"'"~ ~
" r-, Io../' r-,
J'\.
~
... /' r-,
"'"\.~

Note that in PPBS the programme managers are the primary conduct
of resources for their functions.

In functional goal disaggregated planning and budgeting, the function

managers bid for resources in the context of a prior plan. The programme
managers negotiate for resources from the function managers. In PPBS the
resources are given to the programme managers. The function managers have
to bid for resources from the programme managers. The programme man-
agers could be held to have a more systemic view of the operations to product
outputs, and are likely to encourage further processes of adaptation. Clearly,
though, the pathways of differentiation and integration are significantly
different.
It is not at all clear that McNamara's attempt actually worked well, or that
the civilians ever did manage to get control of the military. Aaron
Wildavsky's (1975) study of the introduction of programme planning and
budgeting systems in the US Department of Agriculture suggested that its
impact as a means of changing resource allocation from that of the prior
bureaucracy was very modest indeed. The combination of traditional separa-
tion and major programmes seemed to create a life of their own, for, while
attempts might have been made to organise apparent management in a
54 Procedures for control

different way by having programme managers, the substantive reality was

not much affected. This could be seen as further support for the behavioural
critique that social structures both endure and are of great significance. It is
easier to change the symbolic representation of what is happening than it is
to change the social structure and the grounded reality of the way people
actually engage with their work and with each other. In any event,
Wildavski's memorable conclusion to his study was that 'some butterflies
were caught, but no elephants were stopped'. He found that the depart-
ments were the progenitors of programmes, they possessed the expertise
to analyse and create; they were a major source of knowledge for the
programme managers.
Of course, programme planning and budgeting systems seem to be the
classic solution to the problem of matrix organisations. PPBS offers (see Figure
4.3) the product managers of matrix organisation the capacity to make the
crucial decisions about the resource flows that create the goods and services
for which they are responsible. If properly developed, PPBS gives them the
data to enable them to manage the planning and control loops around their
particular responsibility. We can see how internal markets can be developed
(see Chapter 7).
The question arises as to the efficacy of these approaches. One of the more
interesting studies of budgeting in the UK government observed that, in any
given financial year, the most any government could affect the overall level of
budgets was of the order of 1 per cent and the most they could affect the distri-
bution or redistribution of resources amongst the programmes and ministries
was probably not much more than that. This means that, as in any large organi-
sation, the impetus of past commitments and the limited rationality, together
with the political processes of bargaining and negotiation, suggest that incre-
mentalism is not merely inevitable but also, according to Braybrooke and
Lindblom (1963), a desirable process for managing major institutions. The argu-
ments of the incrementalists contained in Braybrook and Lindblom essentially
state that, because of plurality, the pluralism of values, limitation of knowledge,
the weight of past decisions and the continuity of programmes, the ambiguity
of goals, the shifting relationship of means to ends, the bargaining between one
sector of activity and another and so on, the processes of mutual adjustment
and incrementalism are inevitable if an organisation is not to fragment.
As noted earlier, the rationalist attack on incrementalism was to create the
notion of zero-based budgeting; that is, managers were supposed to justify
anew all resources to be required for the following planning period. The basic
idea is clearly difficult in operation yet, as a kind of mental discipline, it offers
some opportunities for reflective rigour. It suggests that it might be possible to
look at an organisation in action and develop a critique from a zero base and
then ask, firstly, how the particular current configuration of resources came to
be and, secondly, whether that is the most effective configuration of resources
that can be used to achieve what the organisation and differentiated units now
 Anthony J. Berry, Jane Broadbent and David Otley 55

wish to achieve. So ZBB has a helpful critical contribution to make, providing

it is more than an analytical exercise and is connected to the operating, day-to-
day reality of the managers concerned, where future activities are the result of
past decisions.

The rediscovery of programme planning and budgeting systems

Johnson and Kaplan (1987), in their critique of management accounting and

control, argued that the systematic structures of cost accounting did not make
much of a connection with the systemic problems of managing an enterprise.
To recapitulate a little, the structures of traditional cost accounting, especially
such things as full absorption costing, fit beautifully with the notion of the
functional bureaucracy, in that they are a classic case of decomposition and dis-
aggregation of hierarchic elements to lower-order activities. If it is important
that the bureaucratic functions work together, then this programme of system-
atic accounting, taking no cognisance of the need to manage interdependence,
is likely to produce an accounting structure with limited usefulness.
The manager of the fish-finger factory (see Box 4.3) had, through the use of
systematic accounting, become a rather poorly paid servant of a retail opera-
tion. Had the factory managers had a systemic notion of accounting they
would never have got themselves into this predicament. They needed such an
analysis to enable them to see the relationship between the way in which they
were using resources and the way in which they were allocating cost, so as to
adopt a different business behaviour. Even if we accept the economic mar-
ginal resource - marginal cost analysis, it will be noted that the use of full
absorption costing to reflect opportunity cost was unhelpful and inadequate.
Kaplan and Johnson's critique is pretty much that story told over and over
again. What Kaplan argued was that it was important to break this systematic
notion of full absorption costing, and to actually create the idea of activity
pools to which overheads could be allocated. Then these activity pools would
be rigorously criticised (in other words, there would be the emergence of the
ideas of zero-based budgeting in a new form) to decide upon appropriate cost
structures, then the costs in the activity pools would be allocated to products
on some reasoned basis. Now, clearly, this model of activity-based costing has
some useful advantages in that it does introduce a critical review. With cost in
its application it is, however, an alternative model of systematic analysis and
is not necessarily a matter of systemic analysis. The reason why this is so is
that it is using fundamentally the observation model of accounting rather than
an explanatory model which might lie in economic theory or organisation
theory for the behaviour of the firm.
In the late 1980s and early 1990s activity-based costing has become rela-
tively widespread. It is being followed, at the time this chapter is being
written, by the rise of activity-based management (ABM). Of course activity-
56 Procedures for control

Box 4.3 Competing with oneself

In a fish-finger factory in Europe the company discovered that they had

some excess production capacity. Their current brands were selling well
and the product was acceptable in many markets. The company was
approached by a major retail operation asking them to manufacture,
with their spare capacity, some fish-fingers which would be slightly dif-
ferent and wrapped in the retail company's packaging, as an own-label
(as it is called) brand. The fish-finger company, believing, rightly at that
time, that there was no problem with competition with their own
market, readily agreed and through the negotiation of volume contracts
offered a very competitive price to the retailer. In the ensuing five to ten
years the structure of cost accounting and its full absorption model were
pursued until the company accountant noticed that the own-label busi-
ness was now more than half of the total output, and that the own-label
business was costed with a lower level of cost absorption than their own
brand business because of the systematic structure of the accounting.
The company had moved itself into a rather unpleasant corner, for its
brand prices were under pressure in the market-place. It was a problem
for the management to understand how they got themselves into this
predicament and how they could get themselves out of it.

based management is nothing other than programme planning and budgeting

systems rediscovered in what is recognisably a more practical and more
grounded way. Activity-based management is the horizontal integration
process which addresses the systemic issues of any enterprise, and with the
reconfiguring of accounting data flows in support of ABM we can begin to see
an encouraging development of accounting to support a systemic mode of
management control and decision making rather than supporting a systematic
mode. It is important to note that the problems of differentiation of task and
decision authority, the issue of autonomy and the management of interdepend-
ence, are not fully resolved; they are only reconfigured. Inevitably the recogni-
tion of an activity constitutes the recognition of an interdependence which is
essentially task-based. There would also be powerful and helpful interdepend-
encies between different activity systems and the same functional bureaucratic
system. An illustration of this might be that the existence of one activity might
well make it possible to have a related activity (such as a similar product or
service) and have a cost for do ing both of them in the same enterprise rather
lower than do ing them in two separate enterprises . Some authors regard this
as strategic cost management - finding cost savings as a result of the available
interdependencies within a very complex system.
 Anthony J. Berry, Jane Broadbent and David Otley 57

In the 1950s and 1960s Boeing developed the family of Boeing 707, 727 and
737 civil jet airliners. The commonality of the technology, design of some
parts, and of the operating characteristics of aircraft in airline use enabled very
considerable cost savings to be made by having the family of aircraft products
made within the same organisation. That is, given the development of the 707,
it was cheaper for Boeing to build the 727 and 737 than it would have been for
another manufacturer to build similar aircraft. Of course this is to replicate the
familiar problems of economics of product range and market scope.

Managing with uncertainty

The previous discussions of procedures for dealing with complexity are them-
selves a matter of order and ordering. We hope that the reader is not too
caught up with the approach of order and stability to forget that what one
sees in a financial plan, especially one which is projected on a spreadsheet as
single-point estimates over ten years, might be one where the problem of
uncertainty has been set aside. For we live in a state of uncertainty about the
future of the universe. We cannot predict anything accurately; the best we can
do is to create some understanding of probability distributions of distant
events. The fact that nearly all financial plans are couched in single-point esti-
mates (presumably drawn from unknown probability distributions and there-
fore unknown likelihoods of being achieved) provides us with something of a
puzzle, for why would that be so?
The behavioural critique of the economic theory of the firm from Cyert and
March (1963) included the observation that managers set out to avoid uncer-
tainty. We suspect that we all have a great deal more sympathy with avoiding
uncertainty than engaging with it, to which task both Donald Michael (1973)
and Donald Schon (1983) encourage us. Schon wrote that the denial of uncer-
tainty merely leads one to live in a disconnected way and it was imperative to
have thought structures and behaviours which did connect parts of our expe-
rience and understandings. Schon and Michael agreed, though, that managers
must learn to live and work with uncertainty and might, in particular, explore
the way in which the anxieties that seem to stem from uncertainty bump into
the defences that managers and organisations mobilise within themselves.
Michael argues that only if future responsive planning systems are created can
organisations come to terms with uncertainty and create the essentially intelli-
gent, goal-seeking, adaptive organisations which are necessary. So in terms of
our functional quartet of goals, integration, adaptation and social structure,
Michael is recognising clearly that the social structure has to bear the costs
and difficulties of managing uncertainty through personal and organisational
adaptation. He argues that this can only happen if individuals within the
social structure learn to give each other support and, further, that in an uncer-
58 Procedures for control

tain world the only likely outcome of quasi-certain financial planning, indeed
any other human activity, is that we will get it wrong. If we close down the
notion of getting it wrong, either by punishing wrong-doers or by creating a
climate within which getting it wrong is unacceptable, we will merely repli-
cate conservative, limited, inward-looking and life-denying organisations (you
may well work in one of those!). In order not to do that the social structures,
that is the people and the way they work together, have to develop the compe-
tence and the capacity to learn. It is in this arena, this awareness of the self and
behaviour and the awareness of organisational processes through which
defence structures are mobilised against anxiety, that Michael makes his
sharpest contribution. He argues that managers need considerably more
knowledge of themselves, much greater interpersonal sensitivity and a much
greater capacity for giving each other support if we are ever able to enter a
future responsive learning organisation that can cope with the ambiguities
and uncertainties which face it.
It might be thought that what is being argued for here is something that
looks rather like what is believed to have been the management style of some
Japanese corporations in the 1980s. In these organisations, goals and pro-
grammes were formulated after lengthy processes of discourse and dialogue,
examining a wide range of possibilities and ensuring, as best they can, that
they create understanding of the world they work in, how they work together
in it, how to cope with surprises and opportunities, and how to reflect and
then learn anew about the world as they encounter it. To some extent that may
be a reasonable description. In cultures of high dependence, where people are
not primarily anxious about whether they belong, these processes appear to be
possible. In cultures of egocentric ambition, competition and fight , which char-
acterise the Anglo-Saxon world, it would appear to be more difficult to do
that. What seems to be needed in the Anglo-Saxon world is a process of social
adaptation to enable the social structures to function effectively in handling
the problems of high uncertainty of markets and technologies. It is important
to note that the techno-structure is inanimate and the anxieties that are pro-
voked by uncertainty are provoked inside people and do not float in some
kind of institutional miasma where they can be located, packaged and locked
in a spreadsheet.
The problem of uncertainty can lead organisations to create and mirror the
varieties of chaos which they face. They do not deal with the uncertainty,
rather they let it tumble in across their defences and randomly disturb what
were believed to be orderly patterns. An example of this would be a financial
plan subject to a multivariate sensitivity analysis with no exploration of man-
aging consequences. Another would be the presentation of the future of the
enterprise as a probability distribution of the net present value of future
expected cash flows, with no understanding of the loss function of the deci-
sion makers. The problem of uncertainty disturbs the apparent stability of
budgeting and the patterns of expectation. It certainly radically disturbs what
 Anthony J. Berry, Jane Broadbent and David Otley 59

one thinks one is doing when one measures events. Even more significantly,
the accounting measurement is difficult, for the nature of observation of the
past as a basis for future predictions in an uncertain world is highly problem-
atic . What becomes clear here is that the nature of observation of the past to
infer exact predictions in an uncertain world is a highly difficult business.
Statisticians lead us to think about variability, measures of variability and
issues of sampling. So if the problems of observation and measurement are
difficult then the elicitation of meaning is even more difficult. What the
problem of uncertainty does is focus the requirement for interpretative analy-
sis as part of the control loop, rather than mere resort to observation and com-
parison of what is observed by what was expected. The issue of uncertainty
leads us to have disturbed pictures of expectations and possibly inadequate
measures of phenomena as they actually occur. Therefore some much richer
frame of interpretative analysis is required to enable those in the position of
controller to actually come to some judgement as to the antecedents of
observed activity and what might be done to shape future activities.
This takes us back to the beginning of this chapter, where we touched upon
the difference between decision making and control. What seems to emerge
here is support for Geoffrey Vickers' (1965) general proposition that managers,
in using procedures for control, do not, and perhaps should not, make deci-
sions, and nor should they believe that they do. Rather, he argues, and we
agree, that managers need to form appreciations of what they are setting out
to manage. The appreciation is the multidimensioned, multitheoried under-
standing of the complexities and uncertainties of the domains in which they
find themselves in relation to the environment. This notion of an appreciation
then creates the context within which action might be considered and resource
patterns shifted. It provides the basis for [alland's notion of a planning
compact. What this does, then, is relate the techno-structure of control to the
social structure of organisations, recognising that the techno-structure can
contain the complexity, while the social structure contains the uncertainty.

Conclusions

We are essentially arguing here that the techno-structure of planning and

control procedures is necessary, valid and very important, because it is the
only process that we actually have in an adequate variety-reduced form that
enables us at least to relate analytically the factor markets, transformation
process and product/service markets of any enterprise. The limitations of the
techno-structure of such control procedures is that it provides a category of
observations, largely without an explanatory theory; it cannot deal well with
uncertainty and it cannot, apparently, survive unscathed either the behav-
ioural critique, the complexity critique or the uncertainty critique which we
60 Procedures for control

have offered. We are arguing that the way out of this is a linked frame of
thinking which would enable us to locate the control procedures to the social
structure of the organisation.
Note also that, in our view, the plan as created (with its attention to com-
plexity - variety - and to uncertainty) is a description of the outcome of the
control procedure which creates it. Hence the control procedure of planning
provides the activity within which explanatory and predictive models are
brought into use to provide the substantive arguments for the plan as selected.
Thus the variety model for control of the enterprise as system is that of the
planning process, and not the plan, while the uncertainty is contained by the
people in the social structures.
 CHAPTERS

The context of control

Anthony J. Berry, Jane Broadbent and David Otley

Introduction: some general issues

In order to consider the context of control this chapter will examine issues
outside the organisation, that is to say, the environment in which the organisa-
tion exists. This endeavour should not be seen as unproblematically imposing
a strict delineation of the organisation and its environment, nor as a one-way
relationship between the two. Thus, in order to examine the context of control,
two prior points about the nature of control itself need to be raised. The first of
these relates to the extent to which the locus of control is within or outside
organisations, the second relates to the extent to which it is possible to differ-
entiate the organisation from its environment. These two issues will be exam-
ined in turn.

The locus of control

Controls are conceived and can operate at different levels : internal to the
organisation, conceived and operated within them and relatively concrete, for
example, the management accounting systems. Other control systems are con-
ceived externally, on a societal level, sometimes by governments, but operated
within an organisation. Again these systems are relatively concrete, for
example systems to cope with issues such as health and safety, or the require-
ments of the companies acts. On a more abstract level are the values and ethics
of a society which influence the laws and controls which are possible as well
as separating deviant from acceptable behaviour. An extreme example is the
use of the death sentence as a tool of control. The death sentence is used far
more in some societies than in others, and this, it is argued, is related to the

61
62 The context ofcontrol

values of the different societies. So the context of control is argued to relate,

not just to concrete external issues, but crucially to the values and the ethics of
the society in question. As members of organisations are also citizens in their
societies this context is not just 'something out there', but is also related to the
abstract and embedded controls which spring from inner values gained from
membership of that society. This will be a central issue in the current chapter.
There are interactions between the values of a given society and the structures
which are produced within it. It must be recognised that values affect the
structures which are created, but also these structures, in turn, provide con-
texts which might be constraining in themselves and which may not easily be
changed. These relationships between the structures of society and the value
systems have been studied in different ways. Some of the ideas of Anthony
Giddens and Jurgen Habermas will be explored later in the chapter.

The organisation and its environment

If we wish to study the context of control and are focusing upon an examina-
tion of control within organisations, then we need a means whereby we can
separate the outside from the inside and hence define what is the organisation
and what is its context. It has to be recognised that the relationship between
the organisation and its environment is likely to be complex. Contingency and
functional theorists argue that an organisation should adapt to the environ-
ment, applying what might be seen as akin to a Darwinian logic. However it is
argued that this gives too little emphasis to the actors in the environment to
have an impact upon and shape an organisation. The population ecology view
(Hannan and Freeman, 1977) argues that there is a natural selection process in
which the organisations which are fittest for their environment will survive,
whilst others will fail. Organisational ecology theory (Trist, 1976), on the other
hand, sees the environment and the many organisations within it operating as
a complex ecosystem, evolving together. The environment of anyone organ-
isation therefore comprises many other organisations and each one, in this
situation, is part of the environment to the others. Any organisation can, there-
fore, influence that environment, and the latter perhaps is a negotiated context
and is not independently and externally imposed.
One way in which the boundary between the inside and the outside of the
organisation is defined is by legislation. In this chapter, the organisation will
be generally assumed to be that defined by legal governance and the analysis
will look closely at controls which are imposed (or are attempted to be
imposed) by external bodies of varying kinds. External regulation of the
legally constituted body will, therefore, be the main focus of interest.
However, in a final analysis, we will return to the problem of defining the
organisation and its environment and their interrelationships through con-
sideration of the extent to which external controls can be said to be effective.
 Anthony J. Berry, Jane Broadbent and David Otley 63

External regulation and its impact

We can now turn to the controls themselves. One influential source of control
is the machinery of the law, developed through the regulatory and enforce-
ment powers of government. This is not the only source of control, for trade
and professional groupings may also provide regulation and enforcement,
although this is most commonly viewed as self-regulation. The extent to
which the regulations which are developed by both government and other
associations accord with the values of society in general will be discussed
later. The next section of this chapter will be primarily concerned with provid-
ing a discussion of the controls which exist.

Governmental legislation

As has been argued above, government legislation provides an important

element in the context of control of many organisations. It provides a direct
control over the actions of organisations in many different ways. It can create
and bound the organisation. The UK Companies Act 1985 provides the regula-
tive framework for the creation of public and private limited companies. This
Act allows for the creation of a limitation of the liability which can be borne by
the owners and defines the difference between public and private limited
companies. The latter, in contrast to the former, have shares which cannot be
traded publicly. They provide a framework in which companies can, through
their articles and memoranda, define their relationships with the outside
world as well as internal rules of procedure for the relationships between the
shareholders.
UK legislation allows for the existence of other organisational forms: the
Partnership Act, 1890, governs the relationships between partners in a busi-
ness situation, the Building Societies Act, 1986 regulates building societies, the
Charity Commissioners regulate the operation of charities, local authorities
operate within a legal framework, as do health authorities. Without the requis-
ite legislation the organisations in question could not exist in the form they do .
The legislation provides broad control over the governance of all the organisa-
tions as well as defining certain levels of public and private accountability
which must be met. The provision of annual accounts to shareholders in a
limited company is a good example of how one level of accountability is pro-
moted between the managers and owners of a company. It is interesting to
note how much of the accountability which is imposed on companies is
expressed in financial terms and is directed to a selected audience (Laughlin
and Gray, 1988, p. 296). Thus external accountability and control have a
tendency to be identified with accounting-based information.
Not only does this type of legislation impose organisational boundaries and
therefore control the existence of organisational types, it also creates opportu-
nities and possibilities for the control of and within the organisations which
64 Thecontextof control

are created. The markets and hierarchies framework has been used in earlier
chapters as a way of focusing on the controls within organisations. It must be
recognised that, to allow for different types of organisational control, the envir-
onment in which either markets or hierarchies can exist has to be created. This
point can be illustrated by the example of the National Health Service in the
UK. Until recently the NHS has been organised as a hierarchy in which
administrators through resource provision and professionals through a system
of shared values, a clan culture, have been enabled to control the organisa-
tion . The National Health and Community Care Act 1990 has legislated for
change, which, by creating structures for provision and purchase of services,
requires a more market-based approach to the interrelationships in the service.
This structure was not adopted spontaneously; it required the provision of a
legislative framework before its implementation. Here we see that the external
regulation which is provided by government is essential to the enablement of
the particular form of control which has been adopted.
The actual structure which the government legislates to impose can, by its
very nature, have profound implications for the control structures which are
possible. The difference between the privatisation of British Telecom as a
whole unit and the separation of the different water authorities into compet-
ing units illustrates the variety of possible structures.
Lindblom (1977), in the context of a critique of the competence of markets,
shows the extent to which government has provided privileges which enable
business to function in a market environment. The provision of limited liabil-
ity confers business with a crucial privilege; other examples include tax incen-
tives, influence in policy-making circles and legislation to control the labour
force. Without this support business would not be able to function as it does
and this raises questions, which cannot be pursued here in any depth, as to the
extent to which the market really functions as an invisible hand. Imai and
Itami (1984) provide a discussion as to the interpenetration of organisation
ana market in different settings. They examine the US and Japanese
economies, and show how market ideas and organisational forms of resource
allocation impinge upon each other in different ways in the two situations.
Their work emphasises that, while they are analytically separable, in a practi-
cal situation the distinction between the market and organisational forms of
resource allocation is not a clear one .
Other regulations are related to the relationships between the organisation
and both the general community and its workers. Health and safety regula-
tions, employment regulations and pollution controls are examples of this
type of control. These define the obligations of the organisation in very specific
ways and through inspection and enforcement they control particular activi-
ties directly. These regulations are therefore much more direct than the former
controls (which relate to boundary definitions and the creation of a supportive
environment); this type of regulation provides a framework of specific actions
and requirements.
 Anthony J. Berry, Jane Broadbent and David Otley 65

Given British membership of the European Community (EC), regulation of

organisations from external parties will increasingly have an international
dimension. Regulation from the EC is concerned with many areas of organisa-
tion and may well impinge much more in the future as transnational activities
expand and as new treaties are enacted. The influence of the EC will be felt
both on the issue of boundary regulation and in terms of direct intervention to
define the actions of organisations in particular circumstances. Harmonisation
of practices in the European Community is the aim of this institution and this
is desired in the name of maintaining conditions for a free market within the
community. As well as the more obvious ventures in, for example, harmonis-
ing accounting practices, the EC also provides requirements in wider issues
such as the environment. This emerging power of the EC is an example of the
use of transnational regulation to provide the context in which organisations
and markets can function.
It has to be recognised that the legislative framework provided by govern-
ment (national and the EC) is just a framework. Any formal control system
cannot be comprehensive unless the context in which it operates is known
completely and is predictable; this is likely only to be possible in very simple
situations and the context of the social world is not simple. Because of this the
legislation needs fleshing out and this may be done in a number of ways,
sometimes through the legal processes in which case law is developed. Case
law is an important vehicle by which the law is clarified in relation to specific
situations, which then can be applied to related areas. Sometimes this fleshing
out is through processes of administration. The delegation of powers to make
decisions to responsible officials is one way in which this is achieved. For
example, in the UK, the Education Reform Act and the National Health and
Community Care Act both provide for the respective Secretaries of State to be
able to rule on particular issues. Within the health service there are well used
channels in which the policies of government can be communicated to health
authorities, the bureaucratic structure providing for lines of authority from
the Secretary of State to the operating units. Direct instruction in the frame-
work provided by the general legislation is therefore possible.

Other external regulation

Not all regulation flows from government; some regulations may even be seen
as the means by which government regulation is avoided. This type of control
is introduced by groups working together on a voluntary level; adherence to
them might be seen as the price of joining a 'club' or exclusive group. For
example, the UK stock market has regulations which are amongst the most
comprehensive controls which are imposed upon companies and these are
administered by the body itself and not through the companies acts. If a
company wishes to have a quotation on the exchange then it must comply
with the requirements. These regulations can also be used to define
66 Thecontextof control

boundaries. An example is provided by the concept of the 'Chinese wall' in

stockbroking firms, created on the inception of the 'Big Bang' in the stock
exchange. Broking and dealing can now be undertaken by the same firm, but
confidentiality between and separation of the two areas is demanded. There is
a requirement in the Stock Exchange Regulations to provide and presumably
maintain an internal boundary.
Not all associations are so visible. Cartels, the association of firms in the
same market acting together to control the market, are illegal in the UK but
have existed in the past and operate elsewhere. OPEC is a well-known current
example of a group of producers who work together to control both oil pro-
duction and oil prices . There are arguments as to how long and how effect-
ively any cartel can exercise control, but some seem to flourish as examples of
predatory monopolies.
Professional associations are another source of control. These bodies act to
control entry into professions through education and examination pro-
grammes and have procedures for expulsion upon proof of unacceptable
behaviour. They control the rules of conduct for their members and act as a
lobbying body on their behalf in many cases. The UK accounting professional
bodies fulfil all these functions. They control the educational process by
setting minimum standards for registration as students and they have their
own examination systems. Membership of these bodies is usually the result of
both passing examinations and demonstrating a minimum of practical experi-
ence . The professional accounting bodies also have a great input into the
setting of standards for the practice of accounting. They are closely involved
in the processes of generating Financial Reporting Standards (FRSs), which
outline the way in which accountants should approach their compilation of
financial data; again there is a sense of harmonisation of practice. This is a way
of achieving data which will help the functioning of the capital markets and
allow a fairer assessment of the results of companies by the general public.
Again this regulation can be seen as a way of achieving the conditions in
which a market can function.
On another level it could be that the professional associations exist to
protect the values which are implicit in membership of those groups. The
control of the profession by the profession is argued for in the context of the
expertise of the particular group. Professional judgement and expertise are
argued to be such that they cannot be legislated for; implicit in professional
control is a concern for the protection of the 'culture' of the profession. The
medical profession, with their central value of medical autonomy, and the aca-
demic community who value their academic freedom are both communities
who work hard to maintain their professional autonomy and (as they see it)
their integrity. The issue, as far as we are concerned in this chapter, is that the
values of the professional grouping act as a strong control on the behaviour of
its members. The regulations which are formally produced stem from those
 Anthony J. Berry, Jane Broadbentand David Otley 67

values and are produced in some cases as a means of maintaining the auton-
omy of the profession to regulate itself.
Trade associations might agree codes of practice on a voluntary basis as a
way of protecting both their independence of behaviour and also the
confidence of the general public. The scheme operated by the Association of
British Travel Agents (ABTA) to protect holiday-makers when travel firms get
into financial difficulties can be seen as controlling the trade and protecting
public confidence. Trade unions, similarly, impose controls upon their
members as well as trying to control, through national negotiations of wages
and conditions, the employers of their members. They seek to control the
employer-employee relationship for the benefit of their members. In so doing
the trade unions become part of the environment of employing organisations
as well as being organisations in their own right. This type of interrelationship
is illustrative of those in the organisational ecology model of Trist (1976).

QUANGOS

Quasi-autonomous non-government organisations (QUANG OS) can also be

used as a means of control. These are set up by government action but are
argued to have autonomy from them, acting to control other organisations
'objectively' w ithin a policy framework, rather than 'politically'. For example,
regulatory boards have been set up to control the prices and profits of the util-
ities which have recently been transferred to private ownership and which,
because of their near monopoly, could possibly exploit the general public
through their pricing policies. The Office of Water Regulation (OFWAT), the
Office for Electricity Regulation (OFER), OFGAS, concerned with the gas
industry, and OFTEL, the regulator of the Telecommunications industry are
examples of such organisations.

Overview

This overview of some sources of regulation is not exhaustive, but is illustra-

tive. It does provide some indication of the nature of regulation and also the
difference in scope. A regulation which provides a boundary for an organisa-
tion or one which provides an environment should be differentiated from a
regulation which defines actions more closely . The different sources of exter-
nal control are also worthy of note. It can be seen that there are multiple
sources of regulation. The question which will be asked in the next section of
this chapter is how this regulation actually works to achieve control. The exis-
tence of regulation does not mean that it necessarily achieves what it intends,
neither does it explain why organisations take notice of regulation. The
answers to these questions are crucial if we are to understand the context of
control and it is to these we shall turn in the next section.
68 The contextof control

Regulation: how does it work, why does it work?

Legislation does not always 'work'. People do not always do as they are
legally required. A drive along any of our roads will illustrate the extent to
which speed limits, for example, are interpreted and indeed broken. Police
officers have publicly admitted that they will not prosecute those exceeding
the speed limit on UK motorways by 10 miles per hour or less . This effectively
means that one can travel down a motorway at 80mph with little fear of prose-
cution for speeding. The short history of the Community Charge gives another
example of the manner in which legislation can fail to control the behaviour of
large numbers of the community; it also illustrates the danger to the legislators
of implementing laws which are not acceptable to large numbers of the com-
munity. It could be argued that the fall of the UK prime minister, Margaret
Thatcher, from power in 1991, was to some extent the result of the reception of
the 'poll tax' and its lack of compatibility with the values of great numbers of
the community. These examples provide a base from which to discuss the
relationship between the regulations which exist as formal and concrete struc-
tures in society and the less tangible elements, the prevalent societal values or
culture.
Many social theorists have given precedence to either the structures in
society or the subjective role of the individual. The work of Marx, for example,
stresses the importance of the structures of society, in particular the centrality
of the capitalist mode of production. In contrast to this the work of Goffman is
much more concerned with the way in which individuals shape their own
social world, but does not concern itself with structural issues. Two theorists
have attempted to link the two elements together to provide what they believe
is a rather richer picture of society. We shall consider the work of these
authors (although in a modest and limited fashion) because they provide some
rationale for recognising the importance of both the tangible and non-tangible
organisational elements in systems of control.
Anthony Giddens has developed a theory of structuration which aims to
address the relationship between the structures of society and the systems
made up by the interdependent actions of the individuals which make up that
society . This approach highlights the relationship between very different ele-
ments; intangible elements such as meanings and values are linked to the
more tangible structures. A brief introduction to the theory (and an applica-
tion of it) is provided by Capps et al. (1989). Capps et al. use structuration
theory as a way of exploring and understanding the role of culture in organi-
sational control. The structures provide the rules and resources which people
can use in social life and it is through their use that the structures are main-
tained and developed. This is what Giddens calls the duality of structure.
Three forms of structure are suggested: meaning, morality and domination.
Together these provide the rules which are linked to the communication of
 Anthony J. Berry, Jane Broadbent and David Otley 69

meaning, the norms which guide relationships and the power relationships
between parties. Capps and her colleagues use this theory in the context of a
case study in the NCB showing the set of practices which create and sustain
the structures alongside the production of meaning. The study aims to show
how interpretive studies of culture which look at meanings can be extended to
include considerations of structure and power.
Jurgen Habermas also recognises the importance of intangible elements in
the social world in which we exist, although he adopts a rather different
approach. He differentiates the life world, the stock of taken-for-granted
definitions and understanding we have of the world, from the systems of
society. The systems of society are seen to be ones which are the product of
the values and beliefs of the life world. There are mechanisms which inter-
vene to 'steer' the systems in a way which is commensurate with the life
world values. This is an abstract model of societal development which has
been adapted for use on an organisational level by Laughlin, who argues
that the organisation can be seen as a microcosm of society and has a life
world, steering media and systems. Any process of change which is not led
by evolution of the life world and which therefore is at odds with the values
of the organisation is likely to be disruptive to the organisation and may not
achieve what it set out to do . (Laughlin, 1987, 1991, provides a discussion of
the issues which give much more detail for readers who would like to
pursue the issues in more depth.) The value of referring to the work of
Habermas is that it provides another articulation of the argument that the
intangible aspects of organisational and social life are at least as important as
the tangible ones.
Using ideas developed from the work of Habermas outlined above, along
with those of some critical lawyers, Laughlin and Broadbent (1993) have
examined the implications of legislation which is not in line with the life
world of organisations or societies. In particular they have looked at the UK
Education Reform Act of 1986 and the UK National Health and Community
Care Act 1990. Their aim has been to see whether and to what extent these
Acts are overreaching the life world demands and the self-reproductive
processes of the organisations with which they are concerned. The implica-
tions of this are important in a consideration of the context of control,
especially in the context of a discussion of the extent to which external legis-
lation can control. This brings us back to the questions raised at the begin-
ning of this section. Why do people obey laws and how is control achieved
through external legislation? We argue that the law can only achieve control
when it is in harmony with the life world of the context in which it is to be
applied. If this is not the case, then, in the long term, either the law is
ignored or the downfall of those who seek to impose the law or the down
fall of the law itself are possibilities. Thus Sunday trading carries on in the
UK and the prohibition of alcohol earlier this century in the USA was
unsuccessful.
70 The context of control

One set of answers to our original questions, how does regulation work and
why does it work, can now be suggested. Regulation and control will succeed
when there are in accordance with the values and beliefs which predominate
in a society or an organisation. They 'work' because they are imposing con-
trols which reflect the values and beliefs of most of society. This does not
mean that all of society will conform - there will always be deviants and dif-
ferent shades of opinion in a society - we refer here to the acceptance and
implicit acceptance by the majority.
It has to be recognized that the context of control is not simply the result of
regulation. If values and beliefs are important components of control then they
may be so deeply ingrained that they do not require formal regulation. This is
apparent in the normative controls (Etzioni, 1961) and clan controls (Ouchi,
1980) mentioned in Chapter 2. Using a rather wider perspective, the question
of why controls succeed can be related to the notion of their legitimacy
(Weber, 1948). Authority is gained through the legitimacy which is accorded
to those who seek to control and legitimacy is achieved in a number of ways.
Modern society is seen as being oriented towards systems of control which
achieve their legitimacy through democratic, legal, rational structures such as
those we have discussed earlier. On the other hand, legitimacy might be
achieved through tradition or charisma. These possibilities will be considered
in turn.
Traditional authority lies with those who have traditionally been accorded
that right. This is a situation in which the values of the society are such that
particular relationships are the norm and to some extent this type of authority
is similar to that encompassed in clan control or normative control. For
example, in some societies age traditionally receives deference from youth and
authority therefore relates to the role or position held by particular people.
Weber uses patriarchy as an example of an important type of domination
which rests on tradition. In this situation, control can be exercised by a person
who holds a given role because he or she is the holder of that role and for no
other reason.
Charismatic authority, on the other hand, is achieved by those with (or per-
ceived to have) extraordinary qualities, and authority therefore lies with that
particular person. Religious leaders or national leaders such as Mahatma
Ghandi, Mother Theresa or Adolf Hitler might be seen as examples of this
type of leader and control can be achieved by these people through the
authority of their persona. This type of control is substantially different from
that based upon either legal- rational or traditional authority as it is personal-
ity-led and may appear at random. For charismatic authority to be achieved
the person in question must have the support of substantial numbers in an
organisation or in a society. If this is achieved, the person in question will have
a great capacity to control and this will indeed be an important element in the
context of control.
 Anthony J. Berry, Jane Broadbent and David Otley 71

The possibilities of achieving control through external regulation

This last section of the chapter will seek to reflect on what the context of
control might be seen to be and the extent to which external regulation is an
important component of that environment. The importance of various kinds of
regulation in providing a context for the control of individual organisations
cannot be denied. The extent to which market or bureaucratic control is
enabled by the external regulatory framework in existence is clear, and actors
in the environment can be powerful determinants of control. Legal and exter-
nal regulation have also sought to try to define the boundaries of organisa-
tions, having been more successful in some cases than others. The intertwining
of the affairs of the companies which were controlled by Robert Maxwell
shows how the attempts to provide boundaries and maintain them can be
breached by those who wish to do so.
The extent to which the external environment is constitutive of the organisa-
tion or is actually constituted by the organisation perhaps remains open. In
some cases there is little doubt that the organisation is constrained by the envi-
ronmental context in which it exists: on the whole people and organisations do
not go out and break the law. Also people and organisations do react to cul-
tural values; issues of social welfare are now common in Western societies .
However it is equally possible for organisations to have a great influence on
their own environment either directly or indirectly. The main employer in a
small market town in the south of England was seen by local residents to exert
an 'unfair' pressure on the local planners because of its threat to relocate if
new offices did not receive planning permission. It seems likely that there is a
two-way interchange and that there will be variability in the direction in
which the power flows . For those who are designing controls an awareness of
the possible environmental influences is essential; the judgement as to what
that context is and what its effect may be cannot be specified away from the
actual situation.
The other central issue for those who seek to develop controls systems is
whether the systems they develop will produce the control required. External
regulation is imposed with the expectation that desired effects can be achieved
in this fashion. Members of parliament, for example, who promote legislation,
do so on the premise that desired consequences will follow, that they can
change things! The argument that has been presented in this chapter is that
controls can only work when there is some alignment between the spirit of
the control and the values of the organisation or the society in which the
control is placed. One exception to this has been ignored and that is where
there is the presence of direct coercion or physical restraint. Thus controls in a
situation of slavery or in a prison may work in a rather different fashion
(although even in situations like this conflict is possible).
72 The context ofcontrol

If we restrict ourselves to a discussion of situations where direct or physical

control is not the case, then the question has to be asked as to how we can
develop external controls which are effective. In this case the argument for
law which reflects the values of the organisation or society as a whole is essen-
tial. It is also perhaps sensible to use what has been called 'reflexive law'
(Teubner, 1983) that is, law which deals with broad frameworks within which
practical problems can be resolved. It is a law which defines processes within
which answers are enabled. It does not seek to define answers directly. This
type of law recognizes that controls which are developed with one particular
situation may not transpose either spatially or temporally. This would seem to
be a good model for developing systems of control on both a micro and a
macro level. It allows for the possibility for controls to develop and mutate as
need require. If this type of law was developed, regulation from external
sources would not only have the chance of being accepted and therefore
effective, but also it would have the potential for longevity.
There is also an ethical dimension to the existence of regulation. Regulation
is needed in many cases not only to control, but also, to provide a framework
of sanctions for those who contravene the law . These are essential because
people are sometimes naughty! There are always people who will break the
laws which we set. Thus there needs, morally, to be some way by which the
bounds of acceptable behaviour are set and the penalties for breaking them
defined. For this reason it is important that external regulation can work.
However it is important to underline once again that the regulation will be
stronger when it relates to a firm value foundation, held by the community
which has imposed the control. One further question (which this chapter will
not seek to answer) can be raised as a point for discussion. Is there any law or
regulation which can be stated unequivocally? Is there a point about which,
morally, there is no debate, and external regulation can and should be
imposed unilaterally?

Summary

This chapter has been concerned with the context of control and has focused
on the effect of external regulation in defining the context. We have examined
the role that external regulation has in defining the boundaries of the organ-
isations and the relationships between them. Regulation has been seen to
come through the rule of law and through the voluntary association of groups
of individuals and organisations. It has been argued that for regulation to be
effective it must have a strong association with the values of those it seeks to
regulate, or the law will either be ignored or will have the potential to under-
mine the law-makers. The use of reflexive law, a law which defines the
processes by which agreement as to the nature of particular controls is
 Anthony J. Berry, Jane Broadbent and David Otley 73

achieved, is suggested as this law will have the mutability to deal with chang-
ing values and circumstances.
We cannot pretend that this is the whole story of the context of control; we
have addressed areas which give rise to both passion and conflict and these
issues can be illuminated by people from many other disciplines, for example,
political scientists, sociologists or novelists. Deeper insights can be achieved
by extending the boundary of our considerations and we encourage the inter-
ested reader to do this . On a practical level, it is clear it is essential to consider
the context of control if any attempt to achieve control is to be made. The
extent of the success of any attempt will relate to the interplay of the actions
and their context.
 PART II

Issues of Control
 CHAPTER 6

Accounting systems and control

Anthony J. Berry, Jane Broadbent and David Otley

Introduction

The aim of this chapter is to consider the role of accounting systems in organ-
isational control. Every organisation, be it small or large, business or family
unit, has some type of accounting system. Accounting systems provide a fun-
damental way of handling high levels of complexity by the imposition of a set
of standard operating procedures; this provides a major strength of account-
ing as a control system. However weaknesses also stem from the imposition of
standard systems in complex situations because inventive human minds find
ways of reporting desired results through the manipulation of the system
rather than by behaving in expected ways.
Simon et al. (1957) suggested that accounting information serves three major
functions: attention directing, problem solving and scorecard keeping. Control
involves all of these functions. Attention must be directed to the process being
controlled when results are not as expected. The idea of management by
exception follows directly from this approach. Accounting provides data for
problem solving which may be proactive or reactive, in this latter case dealing
with issues highlighted by the attention-directing function. Accounts for
scorecard keeping are the result of a process which examines the extent to
which the organisation as a whole as well as the individuals within it meet the
performance targets set. Performance evaluation of managers and the busi-
ness units they direct is heavily dependent upon accounting-based measures,
such as profit, return on investment, residual income or value added.
In this chapter we shall be concerned with the attention-directing and score-
card keeping functions of accounting-based control systems as they are central
to the formal and routine control of the organisation. Problem solving, whilst
a familiar feature of organisational life, tends to be ad hoc, reactive and rarely

77
78 Accounting systems andcontrol

a simple routine. However routinely collected data are often the basis for
problem recognition and are used extensively in finding solutions, so there
are close linkages between attention-directing routines and decisions taken in
respect of unique problems. Hence we will also consider the use of accounting
data in problem solving. A final note of caution should be sounded before
moving on to look at these areas in more detail; as the earlier part of this text
has illustrated, control is related to many issues, other than accounting.
Accounting is just one element used in systems of control, albeit the central
concern of this chapter.

Problem solving: techniques for gathering and organising data for

accounting-based control

The generation of data and information about the internal operations of a firm
is the focus of management rather than financial accounting, the latter being
geared to provide information for formal and external purposes and for legal
controls. Management accounting is often defined as providing information
for decision making or problem solving. These management accounting
approaches will be outlined in this chapter. For a more detailed examination,
the reader is referred to standard management accounting texts (for example,
Emmanuel et al., 1990, provide an overview of the area; detailed accounts can
be found in Arnold and Hope, 1990; Drury, 1992; Wilson and Chua, 1992). The
approaches which we shall outline are concerned with providing information
for short-term and long-term decisions about activity and with the ascertain-
ment of total product costs. However, before considering specific techniques, a
more general point about the limited nature of the models which underlie
management accounting should be noted.
Many management accounting techniques are based on a simple first order
control loop. This model suggests that control requires the formation of object-
ives, the development of plans (using predictive models) to achieve these
objectives, measurement of the extent of such achievement and, finally, the
monitoring of variances from plan to see if any corrective action is required
(Otley and Berry, 1980). Management accounting models tend not to give full
consideration to all the stages of this process, mainly focusing on the imple-
mentation of plans and the monitoring of variances. The formation of object-
ives is often taken as unproblematic; the debate as to whether we can talk
about the goals of the organisation, or whether we really should refer to the
goals of coalitions or individuals, is often neglected; and the micropolitics of the
organisation are thus ignored. The process of scanning of the environment for
opportunities has also received scant attention (King, 1975). Much of the
investment appraisal literature is based upon the assumption that suitable
projects will present themselves without any effort on the part of managers.
 Anthony J. Berry, Jane Broadbent and David Otley 79

Furthermore the development of predictive models has been glossed over,

despite the fact that the use of appropriate predictive models enables pro-
active, feed forward control to be used. Feedforward control may be less ex-
pensive than reactive feedback control, which can only rectify errors
already made. By planning to avoid predicted consequences which are unde-
sirable, considerable savings can be made. Most of the predictive models
in management accounting are rooted in economic theories which see 'profit
maximisation' as the main business objective. The maximisation of the pres-
ent values of cash flows has been used as a surrogate for profit maximisation.
However this is not always the reason for which people enter business, as
Box 6.1 illustrates.
We now turn to a short overview of the main techniques of management
accounting.

Short-term decisions

An important issue for managers is the effect of short term changes in activity
on the financial results. Techniques have been developed to provide an analy-
sis of the effect on costs of levels of activity in the short term. These techniques
also enable the examination of the profitability of operating at different activ-
ity levels. The interested reader should examine chapters concerned with cost
structure, marginal costing and cost-profit-volume analysis in management
accounting texts. This approach is based on the fact that some costs can, in the
short term, be seen to vary in proportion to volume of output (variable cost)

Box 6.1 Comparisons for performance?

A comparison scheme for hotel businesses sought to provide hoteliers

with an analysis of their financial results. Comparison with the regional
average, it was suggested, would enable them to identify areas where
they were stronger or weaker than this average and so indicate areas
where performance might improve. The scheme was not well supported.
Many hoteliers argued that they operated in a pleasant south coast resort
because they enjoyed the lifestyle. For them improving financial results
was not a prime issue as long as a standard of living satisfactory to them
was being achieved. This illustrates that to focus solely on objectives of
profit maximisation can give an impoverished account of why people
enter business and what they seek to achieve. We need to be sensitive to
the variety of motivations managers may have when considering the use
of techniques based on a narrow, economic perspective.
80 Accounting systemsand control

whilst other costs (fixed or period costs) do not change with variations in
volume. However the categorisation of costs into fixed and variable compo-
nents depends upon the time horizon of the decision being considered. In the
very short term, most costs are fixed; in the long term, all costs become vari-
able. It is therefore vital to consider the time span of the decision being taken
before embarking upon a cost analysis of this nature.
Of key importance is the 'contribution' each unit of production makes.
Contribution is defined as the difference between the variable cost of produc-
tion and the selling price of each unit. Total contribution can therefore be
found for any level of activity (within the relevant range) by multiplying the
contribution per unit and the activity level. Contribution accumulates, first, to
cover fixed costs, then to provide profits. Profits are not, therefore, earned
incrementally with each unit of production sold. A given level of activity is
required to reach 'break-even point', the point at which contribution is
sufficient to cover fixed costs . This approach allows the calculation of costs
and profits at different levels of production and can be used as a feedforward
control device to aid decisions as to whether particular activities will be
profitable. Where there are limiting resources or production bottlenecks, the
idea of measuring the contribution per unit of limiting resource is a widely
applicable heuristic, popularised in texts such as Eli Goldratt's The Goal
(Goldratt, 1984).
The contribution approach can also be used as a basis for decisions about
whether a firm should make components internally or buy them from outside
suppliers. If spare capacity exists (and only then), the relevant comparison will
be between the internal variable cost and the external purchase price . The
fixed costs which will be incurred whether the new production goes ahead or
not can be argued to be irrelevant in this situation. Thus the focus of interest is
the marginal cost of the decision, which may be obscured by the widespread
use of full-cost absorption accounting systems which allocate fixed costs to
units of production.
The contribution technique can also be used to aid some pricing decisions,
particularly in secondary pricing situations, where there is spare capacity.
Here we may choose to adopt the principle that any contribution, however
small, is better than no contribution. Thus a lower price may be accepted than
that usually charged, provided that variable costs are covered. It must be
remembered that short term decisions do, however, have long-term conse-
quences. It is to these long term decisions we now turn.

Long-term decisions

In the long term all costs are variable and one of the costs which assumes
significant importance in long-term decisions is the time value of money. We
are generally concerned with a situation in which an initial investment is
 Anthony J. Berry, Jane Broadbent and David Otley 81

made now which will produce expected benefits over a number of future
years. The estimation of the magnitude of these future benefits is a significant
difficulty faced by all investment appraisal techniques, and most tend to
assume the existence of good predictive models. In practice we usually have
only rather poor models and are forced into performing sensitivity analyses to
examine how our decision might be affected by variations in estimates.
There are several different approaches to the appraisal of long term invest-
ments. Three approaches will be examined in this section: net present value
techniques, payback and accounting rate of return. The mainstream manage-
ment accounting texts deal with these techniques under the heading of capital
investment appraisal and provide detailed discussion of the benefits of the
different approaches.
The theoretically favoured techniques to appraise long-term decisions focus
on the incremental cash flows which are expected to be generated. The reason
for this preference lies in the assumptions about the objectives of the firm
which are linked to the maximisation of shareholder wealth. Evaluating new
activities on the basis of maximising the net present value of the expected
future cash flows provides a surrogate for the maximisation of shareholder
wealth.
Net Present Value (NPV) methods take estimated incremental cash flows
over the life of a project and discount them to their present value. Discounting
is undertaken to account for the fact that cash received in the future has an
opportunity cost when compared with cash received now. To be acceptable, an
investment must have a positive net present value at an appropriate discount
rate. Another derivative of this approach is the calculation of an internal rate of
return (IRR). This rate is simply the discount rate at which the NPV of an
investment becomes zero. Here the decision rule is that an acceptable project
must have an IRR better than some predetermined hurdle rate. Whilst these
approaches are not technically difficult to implement, there are problems in
both deciding the discount rate to be applied and forecasting future cash flows.
The problem of estimating an appropriate discount rate to be used in NPV
calculations has engaged academics for some time . The discount rate can be
argued to depend on both the cost of capital to the firm and the specific risk
associated with the project. In practice the weighted average cost of capital to
the firm as a whole may be used to give a discount factor to be used in all
capital appraisals. Another refinement is to attempt to relate the discount
factor to the risk of the particular project, and its interrelationship with other
investments. As the discount factor determines whether the NPV is positive or
negative it is vital to ensure this is an appropriate estimate. It is also important
to ensure that inflation is properly dealt with. If cash flows are estimated in
real terms, then a real cost of capital must be used; if estimated in nominal
terms, then a nominal rate is appropriate. Although this seems straightfor-
ward, all too often it is observed that the discount rate used is inappropriate to
the basis of cash flow estimation used.
82 Accounting systems and control

The forecasting of the project life and its associated future cash flows is also
problematic, as the application of the most sophisticated techniques to wrong
estimates will not give good results - Garbage In, Garbage Out! Predicting the
future is always difficult and cash flow prediction is no different. Hertz (1964)
offered one solution when he argued that the use of Monte Carlo simulation
on probabilistic cash flows to produce a probability distribution of NPV esti-
mates was the best we might achieve. In his favour it can be seen that this at
least provides the decision maker with an assessment of the probabilities of
loss and gain, should the project be undertaken.
It should be noted that surveys designed to detect whether the adoption of
these 'sophisticated' techniques do provide 'better' results as measured by
increased earnings per share have not been able to show significant relation-
ships (Haka et al., 1985). There are various reasons for this and these may have a
relationship to the context of their use as well as practical difficulties in their
application (for an overview, see Northcott, 1991, 1992). Indeed some studies
have shown an inverse relationship between the use of sophisticated evaluation
methods and subsequent performance. However this may well be an example
of reverse causation: that is, poorly performing firms are more likely to use
sophisticated appraisal methods in an attempt to improve their performance.
In the face of such problems of estimation of both cost of capital and cash flows,
many firms adopt the less sophisticated technique of payback, which is the length of
time the initial cash investment takes to be recuperated by the cash flows generated by
the scheme. This is a much used and well understood technique, with the advantage of
simplicity. If the cost of capital is about 20 per cent, then a three year payback rule will
provide a fairly robust criterion for project acceptance. Some firms use it only for
smaller schemes, some in conjunction with other approaches, but others use it in isola-
tion. A more sophisticated variant is to calculate discounted payback period, which at
least provides one estimate of the time period over which the firm is at risk from its
investment.
Debates as to the strengths and weaknesses of the diverse techniques are
recounted in the texts. Suffice it to say at this stage that the discounting tech-
niques are theoretically superior, but are not always used. However their
adoption is spreading (Pike, 1983, 1988), perhaps as the result of business edu-
cation. There is a danger that the techniques may be used to justify decisions
which have in effect already been made. If, for example, an NPV calculation
does not achieve satisfactory results, then the cash flows or the discount rate
might be changed to provide the result which is required. It seems unlikely
that a manager would put forward any plan which does not appear to meet a
firm's current criterion of positive NPV or required payback, as it would be
rejected at first sight. However managers may well be influenced to adjust the
estimates in order to promote projects which they believe to be desirable for
their own purposes.
The choice of the appraisal technique is complicated by the existence of
another approach, the accounting rate of return (that is the ratio of profit gen-
 Anthony J. Berry, Jane Broadbent and David Otley 83

erated by the scheme to the initial- (or average) - investment). This provides
another way of appraising investments and is important, because it is also
often used to appraise managers' performance. It is seen as theoretically infe-
rior to NPV because it deals with profit flows rather than cash flows, and these
can be manipulated by the application of different accounting policies and
techniques. NPV deals with cash flows, which are argued to be more objective
and less subject to manipulation than profit flows, which are seen as more
subjective. (Interestingly this argument ignores the manipulation which can
also bias cash flow estimates!) Different accounting rates of return can be
obtained from similar schemes which are accounted for using different
accounting policies. In particular, ARR raises problems of asset valuation.
Despite these problems it is used, probably because it produces results similar
to the familiar measurement of return on capital employed (ROCE), used to
appraise results post hoc.
A fundamental conflict can arise between capital investment appraisal
methods, which are based on discounted cash flows, and subsequent manage-
rial performance assessment, which is based on accounting profit and return
on investment measures. A 'good' project (on NPV criteria) may have poor
accounting returns in its early years; conversely, a 'poor' project can show
good accounting returns initially. Thus a manager may be reluctant to propose
a project which is clearly in the firm's best interest because it would adversely
affect his or her performance reports in the short term. This conflict between
future-oriented investment appraisal techniques and historically oriented
accounting measurement techniques is difficult to resolve. The most straight-
forward suggestion is that the financial impact of a new capital investment
should be incorporated into future budgets when it is accepted. The respons-
ible manager's performance should then be monitored in terms of achieving
the budgeted figures, rather than any preset ROCE target.
As the application of each of the techniques has different aims, and the pre-
scription for action might be different under each, appropriate action cannot
be easily decided. The forecasting of future cash flows always provides the
opportunity for biasing and manipulation in the interests of individuals and
against the interest of the organisation as a whole. If appraisal of managers'
performance is linked to the results produced, manipulation will always be a
problem and can only be minimised (for example, by the use of post-audit
techniques) rather than eliminated.

Data for routine control

Information about the costs of existing products or parts of an organisation's

operation is also needed for decision making and control. In particular, total
costs for each product are needed for the valuation of inventories in the
financial accounts. The technique of absorption costing traces to the cost object
84 Accounting systems and control

those costs which can be directly identified with that object. It then allocates the
other indirect costs between the units produced. The total cost of each cost unit
is thus the direct cost of each unit plus the share of indirect costs apportioned to
it. Absorption costing is, therefore, a systematic approach to the construction of
total cost information. As well as providing estimates of total cost for the
financial accounts, this approach gives information which can form a basis for
comparison between different operating units or provide information which
might be useful in pricing cost objects. The basis of 'sharing out' the indirect
costs can be problematic as it is an arbitrary act. This is not necessarily a
problem in itself but, when the figures produced are not recognised as arbitrary
and then become the basis of decisions, problems can arise. Johnson and Kaplan
(1987) have offered a critique of the relevance of management accounting
because of issues such as this. Full-cost information can give only a general indi-
cation of long-term viability of a product (if appropriately constructed), whereas
variable cost information provides a sound basis for short-run decision making.
Activity-based costing (ABC), latterly developed into activity based man-
agement, has been devised as a more defensible approach to the problem of
calculating the cost of a product, department or service. ABC is a systematic
approach but, unlike absorption costing, it seeks to relate costs to the cost
object in alignment with the actual processes of transformation it undergoes.
This requires an analysis of the 'cost drivers' within a business. For example,
the cost of filling an order for £100 of goods is likely to be similar to the cost of
filling an order for £10 000. The cost driver in this case is therefore the number
of invoices rather than the value of the invoice. Allocating costs by the use of
cost drivers is argued to produce a more 'realistic' cost, although it must be
recognised that it still involves arbitrary allocations. Identifying and examin-
ing the cost drivers can focus management attention on the activities which
cause costs to be incurred.
The above techniques provide the foundation for many of management
accounting's virtues as well as the problems which arise from inherent short-
comings. Some of these problems relate to the issue of observing and classify-
ing costs in the ways required to use the techniques, but other problems relate
to the difficulty of using the techniques in a context which requires predic-
tion. However these basic techniques provide the elements of the 'tool-kit'
which can be used to approach the main problems which management
accounting is asked to answer.

Attention directing: budgetary control

The implementation of a system of accounting-based control needs more than

just the techniques for calculating costs noted above. Systems for co-
 Anthony J. Berry, Jane Broadbent and David Otley 85

ordinating the information generated and integrating it with the other busi-
ness information in functional areas such as production and marketing are
also required. Thus, whilst accounting-based control of the business could not
be enacted without the tools to calculate relevant costs, the process of control
is carried out through a different set of activities, those of budgeting.
Budgetary control forms the foundation for the other two functions of
accounting information systems, attention directing and scorecard keeping. It
involves the development of plans for action, expressed in financial terms, and
the monitoring of subsequent activity to monitor the extent to which plans
have been achieved.
Emmanuel et al. (I990) alert us to the fact that budgets fulfil many different
purposes within an organisation. Beside being an important element in
enabling the process of decision making, some possible roles that budgets may
serve include:

1. authorisation of actions,
2. a means of forecasting and planning,
3. a channel of communication and co-ordination,
4. a means of motivating organisational members,
5. a vehicle for performance evaluation and control.

The first four roles are oriented more to attention directing; the fifth is more
related to scorecard keeping. Different elements of control are to be found in
the different roles. In different organisations, each of the above roles may be
given a different emphasis; a budget which is designed to serve one purpose
well will probably be less effective at serving other purposes. Thus the design
of a budgetary control system is essentially a set of compromises and will
result in each purpose being served to a greater or lesser degree.

Budgets as authorisation

Formal authorisation of spending limits is one way of controlling the actions of

subordinates. Thus school governing bodies now have the responsibility, del-
egated from the Local Education Authority, to spend within a preset budget.
They can spend that budget as they wish (within the requirement to provide
education), but they cannot spend more than the set amount. In contrast to this
idea of a global sum, a budget may be much more detailed and used as a
means of authorising spending only on particular line items, such as salaries or
building maintenance. The extent to which a budget is delegated as a total
figure, rather than item by item, defines the extent of the control which is being
attempted. A line-item budget is very much a behaviour (action) control.
The scope for controlling resourceful human beings by simple devices such
as budgets is limited by creative attempts to modify the control which is being
86 Accounting systems and control

sought. For example, cash-limited Health Authorities, it is alleged, hold back

invoices for payment at the year end in order to stay within their authorised
cash budget. Thus activity exceeds that which is recorded and budgeted for,
yet the cash budget is met. In other circumstances activities may be halted
even though profitable production could be continued, because budgetary
spending limits on direct materials have been reached. This latter situation
may be the perverse achievement when a budget is given line by line to the
manager of a cost centre. A less detailed situation where managers are given
profit targets may produce better results, but inevitably the detailed controls
are fewer. Indeed greater control may be achieved by the reduction of the
number of individual controls (Drucker, 1964).
There is a tension between retaining centralised control and delegating deci-
sions to an operational manager. A local manager should have greater knowl-
edge of that environment and delegation to him or her will mean that better
decisions can be taken more rapidly. However the local manager may make
decisions which are good for that local operation but which do not accord
with the overall needs of the organisation. This tension between knowledge of
what is feasible (generally concentrated at lower levels) and knowledge of
what is desirable (generally concentrated at higher levels) is a notable feature
of all budgeting systems. Responsibility for either a cost centre (in which the
decision of the extent of each budget item is centrally controlled) or a profit
centre (where there is greater delegation of decisions as to the budgetary ele-
ments) is therefore qualitatively different, as is the control achieved. Decisions
about the type of control to be exercised must recognise these tensions.

Forecasting, planning and communication

The second set of roles, forecasting and planning, come into play when devel-
oping a budget. Horngren (1981) defined a budget as 'A quantitative expres-
sion of a plan of action and an aid to co-ordination and implementation. In
most cases the budget is the best practical approximation to a formal model of
the whole organisation: its objective, its inputs and its outputs.' An alternative
definition is 'a forecast of the financial implications of an operating plan',
which also indicates the predictive role inherent in the budgeting process.
The third role of budgeting suggested, that of communicating and co-
ordinating, is also suggested by this definition. The way in which these func-
tions are served can best be illustrated by examining the stages of budget
preparation. The nature of the control process can also be illustrated. The
interested reader can find more detail of these issues in the mainstream
accounting texts (Arnold and Hope, 1990; Drury, 1992; Wilson and Chua,
1992);here, yet again, there is only space for an outline to be presented.
Preparation of any budget first requires the forecasting of future activity
from all parts of the organisation. Activity levels for the diverse areas must be
 Anthony J. Berry, Jane Broadbent and David Otley 87

decided upon and then expressed in monetary terms using the forecasts of
economic activity price changes and so on . These individual plans are then
amalgamated into an organisational budget (the master budget), which co-
ordinates all the different functions, balancing, for example, production
and selling plans.
This process of budget preparation will involve communication between
different functional areas in the organisation. The type of communication
engaged in is important in the formation of perceptions about the budgeting
process by the different organisational members. A 'top-down' approach, in
which plans are imposed from above, may well lead organisational members
to consider the budget as a severe constraint. Suspicion and subversion of the
budget and the budgeting process can result. The active participation of
budget holders and the possibility of their negotiating the final budget
outcome, that is, a 'bottom-up' approach make it less likely that the budget
will be the focus of organisational conflict. But it should be recognised that all
budgeting is both top-down and bottom-up. The budget circulates up and
down the organisation in several iterations involving negotiation and bargain-
ing. 'Top-down' and 'bottom-up' refer only to the starting-point, and there
may often be little to choose between them as approaches. Much more im-
portant is the way in which changes are negotiated and communicated.
The accounting literature suggests that the generation of budgets should
follow the organisational structure of responsibility and accountability. It
should also be as detailed as is needed to ensure co-ordination. Organisational
structure does not necessarily determine the type of control system to be oper-
ated, for it can be altered in order to achieve a configuration which allows the
desired control, including budgetary control, to be applied. In particular, the
alignment of cost centre or profit centre responsibilities may be changed in
line with a desire to change the type of control desired.
A budget may be based upon standard costs. The forecasting, planning and
co-ordination requirements are particularly stringent in this case. The stand-
ard costing approach defines standards for material and labour usage as well
as standard costs for each of the quantified elements. Based on the philosophy
of scientific management which sees 'one best way' to achieve a result, the
'best way' is decided upon and costed. The process therefore finds a predeter-
mined standard cost for the production of a unit and results will be analysed
to see if these standards have been achieved. The use of this technique allows
a detailed analysis of results to find out exactly where standards are not being
achieved, and represents strong behaviour control. Explanations for non-
achievement can then be sought and remedies may be applied. Decisions may
also have to be taken about the level of variation to be tolerated before investi-
gations are instigated. It would be misguided to expect complete accuracy in
developing, or adhering to, standards and judgements as to the action taken in
response to variances which have to be made. Variances may not always be
the result of poor performance; it may be that the standards set were inappro-
88 Accounting systems and control

priate. The development of a standard costing system can be expensive and

the use of a poor set of standards means that much time may be spent investi-
gating deviation from standards which are the result of poor standard setting
and not poor control. Because of the expense of generating good standards,
standard costing is best applied in situations where there is a standard
product that is produced in large numbers.
Because the system of standard costing provides such a tight system of
control, products and services are now often deliberately 'standardised' in
order to achieve such control. Although common in mass production, this
idea is being carried across into service organisations. One example of this is
the generation of Diagnosis Related Groups (DRGs) in the National Health
Service . DRGs are groups of medical problems which have similar diagnoses
and are seen to require similar treatment, generating similar costs. They
were developed as the basis of reimbursing medical insurers in the USA,
but are now being seen as a way of distributing resources and controlling
expenditure.
Thus, in different ways, the process of budget preparation achieves the
focus for forecasting and planning and provides the channel for communica-
tion and co-ordination. However budgetary control requires that a further
stage of activity be followed.

Motivating and evaluating performance

After the budget is produced and the activity to which it relates is undertaken,
then information on actual results must be collected. This is in line with the
fifth role of budgeting, that of feedback control. Such control will be sought by
action to both remedy the situation where budgets are not being achieved and
to feed back to update the planning process. This information about the align-
ment of actual and budgeted performance can also be used as the basis for
performance evaluation. Further the fact that performance is evaluated in this
way means that achievement of budget targets should also provide a source of
motivation for organisational members.
In the context of motivation we can investigate the effect of budgets when
they are used to provide targets for individual performance. Tosi (1975) pro-
vided evidence that performance is better when clear, defined, quantitative
targets are provided. While a difficult target tends to motivate higher perform-
ance, if a target is so high as to be perceived as unattainable (and is not
accepted by the individual who is responsible for its achievement) results are
likely to be worse than if a lower, but accepted, goal was set (Locke, 1968).
This phenomenon was also demonstrated by Hofstede (1967). This book
carried on the investigations started by Argyris (1952) into the effects of
budgets and targets on human behaviour. The conclusions which Hofstede
reached were as follows .
 Anthony J. Berry, Jane Broadbent and David Otley 89

- Budgets only motivate if they are 'owned' by the manager concerned.

- Provided the budget does not exceed the most demanding target which an
individual accepts, the results will increase in line with increasing
difficulty.
- Participation in budget creation facilitates acceptance of budgets.
- Cultural and organisational norms as well as individual personality affect
managers' reactions to particular budget targets.

The title of Hofstede's book is an apt one as there is much gamesmanship in

the setting of budgets. Participation in budget setting, for example, can be
helpful in ensuring the acceptance of budgets, lead to improved communica-
tion and can decrease the likelihood of distortion and manipulation of infor-
mation. However it can also provide the opportunity to bias the budget. For
example, slack may be built into targets (Lowe and Shaw, 1968) and managers
may affect budgets (Schiff and Lewin, 1970) more than budgets affect man-
agers (Argyris, 1952). This is particularly likely where the achievement of
budget is linked to remuneration.
Cyert et al. (1961) showed that the forecast projections of the same set of
figures differed according to the label attached to them. Thus a lower rate of
increase was estimated when the figure was labelled 'sales' than when it was
labelled 'costs'. It is likely that the forecast is made with regard to the conse-
quences of error and a conservative estimate of sales or a generous estimate of
costs provides less threat of non-achievement of the target. The work of both
Lowe and Shaw (1968) and Otley (1978) showed that the biased estimates
might not always be in the direction of setting easier targets. In a situation
where current performance was poor, managers might feel obliged to promise
better for the future and overestimate future performance. This could provide
them with the time and opportunity to continue in their jobs in the hope of
solving the problems.
The usefulness of participation is also contingent on the culture and the per-
sonality of different managers. Those who feel that they are in control of their
own destiny react more positively to participation than those who feel they
are the victims of destiny (Brownell, 1981). The latter group exhibited poorer
performance when participating in budget formation, possibly as a result of
increased stress from an environment they perceived as uncontrollable and
uncertain. Bruns and Waterhouse (1975) suggested that participation was
most useful in decentralised organisations which were engaged in tasks of a
well-defined and structured nature.
The findings relating to the use of budgets as motivators suggest that there
is likely to be some tension when budgets designed to motivate are used as
vehicles for performance evaluation. This is because a good motivational
budget will always be slightly more difficult to attain than the performance
which can be achieved. Managers evaluating performance need to bear this in
mind and not treat small adverse variance too harshly. Those using budgets as
90 Accounting systems and control

a basis for planning must also remember that the estimates may contain bias
for reasons which are entirely rational from the point of view of the manager
who has prepared the budget, but which may undermine some of the useful-
ness of the budget from the point of view of the organisation as a whole.
While there will be some intrinsic satisfaction in achieving a set target, a
major source of motivation lies in the rewards which are often contingent on
achieving target performance. In this sense the budget can be seen as an
important tool of performance evaluation, and may be used as such by senior
management. If achievement of targets is necessary for achieving reward, then
an important catalyst for bias has been constructed. This bias will be quite
difficult to detect in the processes of negotiation which precede the setting up
and agreement of targets.
Thus budget-based incentive schemes should be designed in such a way as
to try to ensure that the individual's behaviour in achieving rewards is in line
with the company objectives (Hopwood, 1972). This may mean that good per-
formance is not necessarily the same thing as successfully meeting budget
targets. Indeed there are organisational structures where the control and
reward of performance are not best achieved by such detailed control.
Accounting measures may still be used in such circumstances, and it is to a
consideration of these controls that we will now move.

Scorecard keeping: performance evaluation and accounting controls

The requirement to adhere to detailed budgets is perhaps most useful as a

control in a situation where responsibility for costs alone is delegated to the
manager. The reporting of line-by-line results against budget allows for prob-
lems to be pinpointed by the delegator. If the responsibility delegated is for the
achievement of given levels of profit, the line-by-line detail of the budget is of
central interest to the manager to whom responsibility is delegated, but not to
the delegator. In this case, a profit target is often used as a means of control.
Performance is evaluated by the achievement of these targets. The logic of this
type of control is that those who have responsibility are closest to the condi-
tions in which operations are carried out and know best how to achieve results.
This type of logic achieves its fullest expression in the context of the divi-
sionalised company. Here the control of the company is delegated to those in
operating divisions who are considered to be closest to the environment and
who can therefore make the 'best' decisions because of this local knowledge.
In these situations the use of return on investment (ROD or residual income
(RI) measurements can be used to measure the performance of managers. ROI
looks at the profit which is generated in comparison with the asset base used
to generate that profit, expressing it in percentage terms. RI adjusts the profit
figure attained, introducing an imputed interest charge for the financing of the
 Anthony J. Berry, Jane Broadbent and David Otley 91

assets used. Managers will be given specific targets and, while detailed control
of day-to-day operations is not required, the overall target result must be
achieved. Bonus payments are often used to enhance the motivation of man-
agers to achieve the required targets. The use and development of perform-
ance measures are not straightforward and problems can arise for many
reasons, as suggested by Emmanuel et al. (1990, p . 176):

- Organisations have many objectives and purposes which cannot be meas-

ured easily or effectively by single measures of performance.
- Organisations often require that co-operative action be taken. Trying to
measure individual performance will not necessarily reflect the co-
operative aspects of tasks and may be dysfunctional in that individuals may
pursue actions which enhance their individual ratings at the expense of the
organisation's best interests.
- The specification of tasks and targets in advance may be a fruitless task
because of the ambiguous nature of managerial responsibilities.
- Not all aspects of performance, especially such issues as quality or morale
of staff, which are increasingly being stressed as important in the modern
organisation can be measured in quantitative terms.
- The measurement of results may not be adequate to reward effort, espe-
cially when the environment has not been as was expected when targets
were set.

The overall problem is that reward systems reward report results and not
behaviour. This leads to the situation where managerial behaviour is geared
towards the achievement of reported results. In a situation where the environ-
ment is rapidly changing the standards against which results are assessed may
be inappropriate; The retention rather than renewal of fully depreciated assets
may be the result of their beneficial effect on achieving ROI targets (Dearden,
1962). The purchase of a new asset will increase the capital base upon which
the profits generated are to be assessed, as well as generating depreciation
charges against profit. This may be dysfunctional where, for example, new
investment which would improve the competitive advantage of the firms
products by achieving better quality is being postponed.
The research findings on the issues of participation in budget setting must
be borne in mind when performance is being assessed. One finding by Kenis
(1979) is worthy of inclusion here as it extends the discussion beyond that of
budget achievement. In his study, Kenis notes that, while budget achievement
was improved by participation, there was no relation to other measures of
overall job performance. Studies by Ivanevitch (1976), Milani (1975) and Steers
(1975) show relatively insignificant connections between budgetary character-
istics and job performance. This must alert us to the fact that quantitative evalu-
ation might have little effect on overall job performance. This may be the
result of the fact that managers may have no control over some important
92 Accounting systems and control

variables, no matter whether they are involved in the construction of the

budget or not. Evaluation styles can vary however. Merchant (1981) found
that large, decentralised firms with diversified activities tended to operate
administrative rather than personnel controls. Greater stress was given to
formal controls such as budgeting and lower levels of managers participated
in developing budgets.
Hopwood (1972) studied cost centre managers in a sequentially independ-
ent situation and argued that three diverse orientations to evaluation could
be identified. In each case there are different approaches to the linkage of
performance and rewards.

1. A budget-constrained style: here a rigid insistence upon the short term

achievement of the budget is the central feature.
2. A profit-conscious style: here the general effectiveness of the unit's opera-
tions is the central feature. The approach is more flexible than that in the
budget-constrained style, and budgetary information is supplemented by
other information. Thus a manager who has a good explanation for over-
spending may still be evaluated favourably.
3. A non-accounting style: in which budgetary data are seen as relatively
unimportant and other measures of managerial performance are used.

Managers who were evaluated in a non-accounting style were less cost-

conscious than those evaluated using accounting-based styles. Those man-
agers who were evaluated using the budget-constrained style reported higher
levels of stress, poorer relationships with colleagues and a greater tendency to
manipulate financial reports than managers evaluated under a profit-
conscious style.
Otley (1978) repeated some of Hopwood's work in the setting of independ-
ent profit centres. He thought that, in the chosen setting, budgetary informa-
tion would represent a more adequate measure of managerial performance
than in Hopwood's study. He found style of evaluation had little impact on
job-related tension or on manipulation of data. Performance evaluation based
on budget achievement led to an emphasis on the short term, but, apart from
this, provided an effective management style. Otley also found that per-
formance affected the choice of management style, with better performing
managers being more likely to be evaluated in the more flexible style. Thus
the choice of the way to use budgetary information in performance evaluation
is a complex matter, with the potential to generate unexpected side-effects.
Hirst (1981) studied the effect of different environmental conditions. Where
there is a high level of uncertainty, accounting measures are seen as providing
a less complete description of performance than in a more stable environment.
Govindarajan (1984) supports this position, suggesting that in a highly uncer-
tain environment more subjective evaluation procedures are likely to be
adopted. These studies suggest there are two central issues that need to be
 Anthony J. Berry, Jane Broadbent and David Otley 93

resolved in each specific situation. First, the impact of the control system
depends on the way in which accounting information is used by managers
and the rewards that are contingent upon its achievement. Secondly, the effect
of high reliance on budgetary measures of performance is contingent upon the
extent to which we can be certain of the linkage between managerial behav-
iour and desired results. This is often quite low and is made more ambiguous
in an uncertain environment. Reward systems must be designed with this in
mind and must be tolerant of occasional failure, especially in situations where
innovation may be needed.
Individual differences between managers must also be borne in mind, and
there is some indication that the cognitive style of individuals will affect the
structure of the information which will be of most use to them. Highly aggre-
gated data, such as balance sheets and formal accounting reports, are of most
use to high-analytic types (those who use conceptual models to understand
situations). Low-analytic types (who see situations more holistically) are better
served by disaggregated raw data (Benbassat and Dexter, 1979). Macintosh
(1985) also suggested that information should be supplied in a manner which
is consistent with a manager's cognitive style.
Another important individual difference relates to the extent to which indi-
viduals accept personal responsibility for what happens to them. Those with
an external locus of control see events as being outside their control, those
with an internal locus of control see events as the result of their own behav-
iour. The latter group performed best when involved in budget setting, the
former group when targets were imposed.
This short overview of the literature relating to performance evaluation sug-
gests that there are inherent difficulties in providing suitable measures of per-
formance, that the style by which individuals are evaluated by their superiors
will affect performance and that individual differences and the state of the
external environment are all important factors for consideration in designing
systems. All these complexities suggest that the design of performance evalu-
ation systems, as part of a control system, is complex and difficult to deter-
mine away from the particular situation and the individuals within it. What
works in one place for a certain set of people may not work in another situa-
tion or even in the same situation when different individuals are involved.

Closing comments

An overview of the uses of accounting-based controls shows how diverse the

roles served by such controls are, as well as the various interlinkages and ten-
sions which exist within and between the roles. Despite this complexity,
accounting is clearly an important element of control and is generally per-
ceived to be so. It is now being introduced into areas which have previously
94 Accounting systems and control

had little financial control at an operational level - schools, probation offices,

the arts, hospitals, universities and voluntary organisations. Because of this
level of influence the synergies and tensions associated with the use of
accounting based controls need to be clearly understood if we are to exploit
the virtues of accounting and not be prey to the weaknesses and dysfunctions
which can be created by its misguided use.
If we return to the simple control model upon which management account-
ing is based (that is, where an objective exists for an activity in a quasi-stable
environment which can be described by a predictive model, and where it is
possible to measure what has been achieved and allow monitoring to take
place) we can see the type of environment in which accounting based control
is likely to work best. It is a situation in which certainty and well defined rela-
tionships exist . The irony is that it is in precisely these circumstances that any
approach to control will be most easily applied. Where there is a more uncer-
tain environment, where interdependencies are least easily defined and where
the future is most difficult to predict we are most in need of some way to help
us achieve control. Here any accounting-based approach to control is of least
use because it relies on the ability to predict and translate predictions into
financial values. Despite the variability of the circumstances in which account-
ing information is appropriate there is little variability in its application in
Western societies. But accounting still provides the articulation of the one
measure which can serve to aggregate or contrast information from all aspects
of a diversified business, monetary value. However flawed it may be, it can be
argued that it is still the best control we have available to us.
Nevertheless there is some disquiet about the central role of management
accounting information in management control. Johnson and Kaplan (1987)
have raised the issue of the relevance of the type of information which man-
agement accounting produces. While there is some contention as to the ade-
quacy of the account these authors offer (Ezzarnel et al., 1990, Hopper and
Armstrong, 1991) their book has provided the impetus for a debate on the
direction which management accounting should take. Linked to this debate
has been a closer inspection of the techniques adopted by Japanese compa-
nies, which are less financially oriented, and consideration has been given to
the question of whether Western companies can learn from their Eastern coun-
terparts. The outcome of these deliberations will be important for the develop-
ment of accounting-based control and for management control in general.
There are unlikely to be simple solutions, but the complexity of the task
should not discourage us from undertaking an attempt to seek the resolution.
 CHAPTER 7

Divisional control
M. Broadbent and J. Cullen

Introduction

Chapter 3 has already considered both how an increase in an organisation's

size may lead to problems if a centralised and unitary control structure is
maintained and how divisional structures were developed to balance the
problems of allowing sufficient specialisation and yet providing an integra-
tion of the tasks carried out. That chapter further stated that divisional struc-
tures may lead to suboptimal decisions being taken by divisional
management. The work of Williamson (1970, 1975) was used to support the
notion of multidivisional (M-form organisations) as a medium for control to
cope with the growing complexity of larger organisations and particularly
those operating in 'diverse' product markets. M-form organisations are
argued to integrate the concept of markets and hierarchies within one organi-
sation, at divisional level through the market considering day-to-day activit-
ies, and at the group or head office level, through hierarchies setting strategies
for the divisions. Put more distinctly, strategic decisions are vested in top
management whereas operating decisions are the prerogative of divisional
managers (Ezzamel, 1992, p . 5).
Contingency theory approaches to control, introduced in Chapter 2, argue
there is no one 'optimal' way to structure an organisation, but that structure
should reflect the environment in which it is found . The contingency literature
suggests that environmental complexity is significantly correlated to the extent
of decentralisation and that a strategy of product diversification has caused
many organisations to divisionalise their structures (Chandler, 1962). Lorsch
and Allen (1973) linked greater market differentiation and independence
between products to divisionalisation along product lines, while organisations
operating in integrated markets tend to remain non-divisionalised.

95
96 Divisional control

Chapter 4 introduced the notion of subsidiarity as a means of coping with

complexity; in this case task and decision authority should be delegated as
close to the point of enactment as possible. The principle of subsidiarity does
not imply either a divisional or unitary form but leads to a consideration of
the relevant and appropriate location of task and decision authority which, as
argued above, leads to a separation of strategy and operational decision
making and control for divisionalised companies.
Having linked back to the threads developed in the initial chapters of this
text, we aim in this chapter to explore the issues raised in greater detail, ini-
tially returning to the arguments put forward by Chandler and others, before
moving towards a control model for divisional companies and consideration
of the ensuing problems that creates. The traditional approaches to control,
through an accounting information system, will be explored, as will non-
accounting controls (for example, Kaplan and Norton, 1992, 1993). The asser-
tion of the separation of strategies and operational decision making will be
questioned using, inter alia, the work of Goold and Campbell (1987a, 1987b).

The M-model arguments revised

Dermer (1977) emphasises that an understanding of the wayan organisation

structures itself through differentiation and the allocation of responsibilities
through decentralisation is necessary in order to study control. Organisations
differentiate their activities into subunits so each faces a more homogeneous
and manageable environment. Differentiation can take many forms, the two
major ones being differentiation by function and by product. Decentralisation
determines the organisational structure, the scope of each unit and its poten-
tial range of activities. Differentiation determines what each unit is capable of
doing, while it is decentralisation that determines what decisions it is respons-
ible for and determines the range of discretion allowed to each subunit within
the organisation.
Ezzamel and Hilton (1980a, 1980b) bring these two concepts together by
classifying decentralised structures as 'functional' or 'federal' . Functional
decentralisation is the delegation of decision-making power to lower levels of
management on the basis of functional specification, for example production
or marketing, while federal decentralisation involves the partitioning of the
firm into two or more quasi-autonomous subunits, divisions or business units,
whose activities are co-ordinated primarily through price mechanisms such as
product lines, customers or geographical location.
Loft (1991), writing about the history of management accounting, suggests
that at the turn of the nineteenth century the US economy had several huge
vertically integrated firms, successfully adopting a centralised (or U-form)
organisation structure by employing decentralised decision making through
 M. Broadbent and I. Cullen 97

differentiation into highly specialist units. Hence top management could co-
ordinate activities and direct strategy and policy. Johnson and Kaplan (1987),
using the Williamson (1975) argument, maintained such large companies
recognised the benefits of a well managed hierarchy over those of market
exchangers). The firms were not at this stage divisionalised: they were decen-
tralised and it was not until after the First World War that divisionalised
companies started to emerge, well known examples including Du Pont and
General Motors.
The links between decentralisation and environmental uncertainty are well
stated in the contingency theory literature. The work of Burns and Stalker
(1961) who identified the categories mechanistic and organic organisational
structure lead to research suggesting that environmental complexity is
significantly linked to the extent of decentralisation within an organisation.
The work by Chandler (1962) linked product diversification with decentralised
structures and was supported by Lorsch and Allen (1973). This led to the con-
clusion that, as firms differentiate their products, divisionalisation will arise as
a consequence of the resulting independence of different lines. This argument
can be augmented by the analysis of technology as a contingent variable. This
can be illustrated by the work of Woodward (1958a) and Perrow (1967, 1970)
who maintain different organisational structures are a reflection of their tech-
nologies. This work suggests that, if a large company which operated in
several diverse product markets and which used several different production
technologies were to decentralise, divisionalisation would be the likely result.
Such a conclusion is supported by the general systems theory literature,
which argues that organisations, like individuals, seek to cope with their envir-
onments. If the environment is uncertain and highly complex then the creation
of subunits or divisions within an organisation as a 'coping' mechanism
would seem appropriate. Thompson (1967) summarised this approach, stating
that organisations cope with uncertainty by 'creating certain parts specifically
to deal with it' .
The arguments put forward so far within this chapter follow those of
Chandler (1962), supported by Johnson and Kaplan (1987), summarised by
Loft (1991) as follows: 'that the environment caused large companies to follow
certain strategies, which in turn 'caused' the multi divisional company'. This is
not to suggest there is certainty about this. Loft questions this assertion, sug-
gesting an opposite line of causation that may be equally as valid, 'that large
companies' strategy of trying to dominate the market by driving competitors
out, or swallowing them up, lead to monopoly and oligopoly and in turn to
bureaucracy and wastefulness'.
Returning to the mainstream literature, Scott (1971) summarised the devel-
opment of a firm into four stages, linking these with organisational structure
and their respective attributes (see Figure 7.1] Scott (1971). The Scott matrix
views the development of organisational structures through growth in size
and complexity. Initial growth in size is seen as leading to the development of
98 Divisional control

FiGURE 7.1 The Scott Matrix

Owner Growth in size Growth in Highly diverse,

manager complexity but no central
central strategy strategy: set by
business units

Stage II III IV
Product line Single Single Multiple Multiple
Distribution Single Single Multiple Multiple
Organisational Little or no Functionally Product market Product market
structure formal based and based and quasi- based and
structure integrated autonomous largely
autonomous

Source: Four strages of Organisational Development adapted from Scott (971).

decentralised, functionally based organisations. With increased complexity the

structure becomes product based and quasi-autonomous, with strategy held
centrally - the classic multidivisional company. The final stage suggests that,
as organisations become so diverse, the holding of a central strategy is
replaced by the delegation of individual strategies to divisions or strategic
business units. This matrix presents a logical progression of structure, but any
large diversified company may be adopting different control mechanisms for
different divisions at anyone point of time - so control for a core business
division will be different from that of a peripheral business unit. Such argu-
ments will be developed later in this chapter using the work of Goold and
Campbell (1987(a), 1987(b» and Miles and Snow (1978).

Control with multidivisional organisations

The major advantage of decentralised and multidivisional organisation in par-

ticular is that it allows different layers of management to 'concentrate on those
issues with which they are best placed to deal' (Emmanuel et al., 1990a, b). The
M-form organisation, if applied appropriately within organisations, should
provide for effective performance and profit maximisation (Williamson, 1975)
through the following three attributes. First is the efficient allocation of
resources within the organisation. If the organisation is compared to a capital
market, where lenders with surplus units are seeking out borrowers requiring
those units for investment through a risk-return trade-off, then headquarters
or central management is allocating resources between divisions on an equally
competitive basis. Funds raised externally are obtained at a lower cost than
 M. Broadbent and J. Cullen 99

any individual division could obtain, they are then allocated on a competitive
basis between divisions and the efficiency of these funds is monitored through
the use of performance indicators, so providing a control loop for the alloca-
tion of these resources.
Secondly, by the delegation of day-to-day and operational decision making
to largely autonomous divisions, the communications channels to higher
levels of management are not overburdened by the detailed information
filtered out at lower levels. Such information channels can be used for sum-
marised information flows and strategy dissemination. The concentration by
divisions in particular areas of activity also reduces information flows, as
interdivisional transactions are reduced.
Thirdly, the M-form minimises suboptimal behaviour. The vesting of strate-
gic decisions with top management and the day-to-day operating decisions at
divisional level reduces the risk and complexity of the decision making, as
each is specialising at different levels . Because the organisation is a collection
of interdependent units, transactions between these units will reduce the
potential for opportunistic behaviour, which would be encouraged in transac-
tions external to the organisation. The use of control systems and performance-
monitoring devices, linked to a flexible reward system, may ensure optimal
behaviour. Headquarters can, if required, mediate in disputes between divi-
sions expediently without recourse to legal intermediaries.
According to this theory, hierarchies are only liable to replace markets if it is
more economical to organise transactions through such hierarchies, so any
managerial costs of divisionalisation must be less than those incurred by oper-
ating through the markets. The delegation of decision making to divisional
managers will automatically increase the managerial costs through increased
monitoring to ensure that delegated authority is not misused. Scapens and
Roberts (1993) argue that decentralisation is often ambiguous because of a mis-
match between the commercial unity of a division and the arbitrary divisions
created by the system of profit centres. This ambiguity is supported by Otley
(1990) in a case study of British Coal. As divisional level managers are given
greater decision responsibility, they in turn expect greater rewards for that
increased responsibility and a share in the success of the divisional activities.
It is possible for the performance of one division to be enhanced at the
expense of the total organisation and the sheer competitive nature of division-
alisation may breed separatist attitudes and overcompetitive behaviour
amongst divisions. Since the loss of central control through decentralisation
and divisionalisation means a higher incidence of non-programmed decisions
(Emmanuel et al., 1990) for divisional managers, it is essential that these deci-
sions are taken using judgement and intuition which is organisationally goal
based, rather than solely divisionally based.
The use of divisionalised structures may be counterproductive where there
are excessive interdependencies between divisions, for example where the
organisation goal is indivisible amongst divisions, or where organisational
100 Divisional control

resources are so complementary that separation between divisions reduces the

scope for scale economics. The interdependencies of production functions may
require high management costs to manage the relationships and may negate
any benefits obtained from divisionalisation. Hopper (1988) argued that such
costs arose from the requirement to have corporate controls over the manager-
iallabour process, and the growth of management itself necessitated its man-
agement through monitoring processes.
Williamson (1970, 1975) argued that the divisionalised organisational struc-
ture possesses the requisite control apparatus required for profit maximisation:

1. an incentive machinery, including pecuniary and non pecuniary rewards,

which can be adapted to monitor the behaviour of divisional managers
towards the goals of central management;
2. an internal audit system which reviews and evaluates the performance of
divisional managers;
3. an allocation system which assigns resources (cash flows) to the highest
yielding projects after evaluation of divisional investment proposals.

He further puts forward the procedures for the implementation of divisional

structure, as follows:

- identification of divisional boundaries;

- assignment of quasi-autonomous status for each division, so determining
the extent of division autonomy;
- allocation of company resources to divisions;
- the use of performance measures and reward schemes to monitor divisional
activities;
- the performance of strategic planning wherever possible.

Ezzamel and Hart (1987) suggest the management accounting literature has
devoted little attention to the identification of divisional boundaries, the
assignment of quasi autonomous status to divisions, and the performing of
strategic planning. We hope to redress this balance by considering just these
issues in the following section before considering the traditional accounting
approaches to divisional control, to be followed by a broadening of the perfor-
mance measures and the development of the strategic planning issues.

Controls in divisionalised organisations

This discussion will draw heavily on the work of Ezzamel and Hart (1987)
who maintain that divisional control emphasises both financial and structural
control devices, that these interact and, to some extent, constrain one another.
 M. Broadbent and J. Cullen 101

They state the overall control apparatus can encompass a variety of individual
controls which may relate to:

- defining divisional environment,

determining divisional size,
- defining and co-ordinating divisional interdependencies,
- determining the extent of divisional decision-making autonomy,
- characteristics of information and information flow,
- designing an internal audit system, and
- designing appropriate reward systems.

The above act as interacting decision variables which may constrain manager-
ial action and which may be subject to other managers' manipulations. Each
will be considered in turn.
The divisional environment will be a subset of the environment faced by
the total organisation. The degree of differentiation required by each division
to meet the demands of its environment will create internal diversity. The
greater the differentiation the greater will be the problem of integration
between different divisions. Divisions operating in relatively certain environ-
ments will rely on more formalised organisational practices, whilst those oper-
ating in environments reflecting greater complexity will have less formalised
systems. The greater the environmental variety the greater the potential for
disunity and dysfunctional behaviour amongst divisions, yet each division is
unlikely to be fully independent of other divisions. The organisation needs to
achieve integration between units to remain a whole. Lorsch and Allen (1973)
maintained that the greater the differentiation between divisions the harder
the task of integration, with a resulting complexity of integrative devices ; task
forces, committees, divisional specialists working at head office and so on.
Within conglomerate companies, by definition highly diverse and operating in
different product markets, Lorsch and Allen (1973) found performance evalua-
tion systems that had direct linkages between performance in a financial sense
and monetary rewards, while, in vertically integrated firms, having greater
interdependencies and facing a similar environment, the performance evalua-
tion systems were more formally administered with less emphasis on financial
results and more on operating criteria.
Divisional size may reflect a compromise between economic efficiency and
administrative efficiency. While classical economics would have us believe
there are optimum sizes for particular units, divisional managers may have
valid personal reasons to extend plant size beyond this norm for their own
rewards. Similarly head office staff may limit plant size below this norm to
ensure that the power base of individual divisional managers is not enhanced.
The issue of divisional interdependence includes the relationships between
central management and divisional management, between divisional
managers themselves, divisional managers and their operating units, and
102 Divisional control

between operating units within anyone division. The interdependencies

(Ezzamel and Hart, 1987) could take several forms, and relate to demand and
transfer pricing issues as well as issues of technological interdependency as
well as behavioural effects. Each must be co-ordinated to avoid managers pur-
suing behaviour which is dysfunctional for the organisation as a whole but
which may further their own interests. It is suggested that, in order to improve
the effectiveness and usefulness of various accounting procedures, accoun-
tants have to team up with organisation designers and behavioural scientists.
The level of autonomy granted to a division manager is stated clearly in the
work of Williamson (1970, 1975): central managers take strategic decisions,
whilst operating decisions are vested with divisional managers. Williamson
argues that the mixing of these decision levels may lead to suboptimal deci-
sion making and a reduction in organisational effectiveness. The evidence
cited is a product of studies within traditional US divisionalised companies,
yet more recent work by Goold and Campbell (1987a, 1987b) illustrates the
sometimes clouded nature of the two levels of decision making, particularly
for strategic business units (see Scott, 1971, Figure 7.1).
Even within the strategic business unit and the more traditional division it
is essential that central management maintain the decision process which allo-
cates resources amongst competing units. It is this process which can regulate
the growth of anyone division and its ultimate size and importance within the
whole organisation. The rationing of resources to divisions that do not
conform with central strategies may be a powerful incentive to conform.
Ezzamel and Hart (1987) maintain a commonsense view about the amount
of decision discretion granted to divisional managers in that it would 'pre-
sumably imply that division discretion should be limited in decision areas
characterised by excessive interdependencies, as well as in areas of critical
importance to the organisation as a whole' . Lorsch and Allen (1973), in the
study already quoted, found div isional managers within conglomerates had
greater autonomy in their decision making than vertically integrated organisa-
tions . Other academic studies seem inconclusive and conflicting. The logical
view would suggest that divisional managers facing environmental uncer-
tainty which is different to that of the organisation as a whole, operating with
few interdependencies, would have a greater range of decision responsibilities
than the divisional manager working in contrary circumstances.
Information flows between levels of management reduce the level of uncer-
tainty in which they operate. If decentralisation and divisionalisation is to work
through the notion of subsidiarity, then ex ante information gathering for deci-
sion making should be done at the appropriate level; equally ex post informa-
tion should be reliable for performance evaluation. Information characteristic
and information flows will vary according to the reliability of the data. When
output is difficult to observe and verify then the focus of information is likely
to shift to 'softer' information flows. There has been much criticism of the use
of 'hard' accounting information for the evaluation of divisional performance
 M. Broadbent and J. Cullen 103

measures: see, for example, Ezzamel et al. (1990), who argue for the use of soft
information for performance evaluation. Other proponents of this approach
include Kaplan and Norton (1992, 1993) whose work will be discussed later.
Ezzamel and Hart (1987) considered that the internal audit of divisional
organisations can operate at three different levels: advanced, contemporane-
ous and ex post. The advanced level involves reviewing a divisional proposed
course of action; contemporaneous evaluation entails continuous monitoring
of divisional performance; ex post evaluation involves end of period perfor-
mance evaluation. Williamson (1970, 1975) views audit as a necessary cost of
using a hierarchy as opposed to a market. The audit function, he argues, is
necessary to attain the greater efficiencies generated by internal transactions.

'Traditional' accounting methods for the measurement of divisional

performance

Next we provide an overview of the issues involved in providing accounting

measures of divisional performance. The bulk of the literature is contained
within management accounting texts, many of which assume that manage-
ment accounting exists in splendid isolation rather than being a subset of
management control!
Measures of divisional performance are usually financially based. There is
substantial research evidence (Reece and Cool, 1978; Vancil, 1979; Tomkins,
1973) demonstrating the continued use of profit measures and return on
investment ratios as the main measures of divisional performance. Solomons
(1965) gave three reasons why some sort of index of divisional profitability
would be useful.

1. as a guide to central management in assessing the efficiency of each divi-

sion as an economic entity;
2. as aid to central managers in assessing the efficiency with which divi-
sional managers discharge their duties;
3. to guide divisional managers in making decisions in respect of the daily
activities of their own divisions.

While the three reasons provide a basic rationale for the use of divisional per-
formance indexes within one organisation, Skinner (1990) maintains that a
further economic justification for the use of performance indexes is inter-firm
comparisons, where results of one division can be compared with those of the
results of a whole company in a similar industry. Recent developments in
company segmental reporting also highlight this external focus. Solomon con-
centrates on central managers and divisional managers to the exclusion of
104 Divisional control

operating managers whose efficiency has a direct link to profitability. The

three reasons stated by Solomon do agree with the Williamson 0970,1975)
arguments that, through effective resource allocation and performance evalua-
tion by central management, divisionalised organisations can pursue the
notion of profit maximisation.
A major challenge facing this approach is to devise performance measures
which do not encourage divisional managers to pursue policies that are
beneficial to their own division but detrimental to the organisation as a whole.
Shillinglaw (1982) suggests three rules with which divisional profit measures
must comply before they can be regarded as acceptable; summarised by Drury
(1985) they are:

1. divisional profit should not be increased by any action which reduces total
organisational profit;
2. each division's profit should be as independent as possible of performance
efficiency and managerial decisions elsewhere in the company;
3. each division's profit should reflect items which are subject to any
substantial degree of control by the divisional management or his (or her)
subordinates.

Traditional measures of divisional performance include profit, return on

investment (ROD, residual income (RD and discounted cash flow (DCF). Each
will be considered in turn.

Profit

Profit is an appropriate measure of performance if the divisional manager has

decision responsibility for all the variables within its calculation. There are,
however, problems concerning profit measurement, profit definition and
profit calculation, and the issue of common costs, common revenues and
transfer prices. Absolute profit can be defined in many ways using different
accounting bases. Which level of profit is applicable may also present prob-
lems - contribution margin, direct divisional profit, controllable divisional
profit, income before taxation or net income.

Return on investment and residual income

Return on investment (ROD or return on capital employed (ROCE) are very

commonly used indexes of managerial and divisional performance levels.
Skinner (1990) indicated that surveys from a variety of countries have shown
that about 70 per cent of divisionalised companies use profit, most often in
the form of return on investment ratios, as their main measure of divisional
 M. Broadbent and J. Cullen 105

Box 7.1 How to reject an acceptable investment

A division within a large conglomerate has achieved a fairly consistent

ROI of about 20 per cent for the last five years. This is above the return
for the company as a whole which has varied from 11 per cent to 15 per
cent over the same period with a cost of capital estimated to be 10 per
cent. The divisional management team are considering a new investment
in advanced manufacturing technology. The new investment ROI is esti-
mated at 18 per cent with a positive net present value of £68 500 and an
internal rate of return of 14 per cent. The divisional management team
have decided not to submit the new investment to headquarters for
approval as it would reduce the ROI of the division in the short term.

financial performance. Return on investment has many advantages: it is a true

efficiency ratio focusing attention on both assets and profits; it is a measure
which is widely used (and abused) by managers; it provides a basis for com-
parison of performance between divisions, between divisions and outside
companies and between divisions and alternative investment of funds. ROI is
a ratio, which divisional managers will be required to maintain or improve
over time. A divisional manager may be able to maintain the ratio by replac-
ing assets as necessary and making progress improvements, but there is no
strong encouragement to make the division grow in size, hence the use of
absolute profit in parallel with ROI. Horngren (1962) has suggested that the
change in ROI is often more significant than its absolute level, whilst Amey
(1969) argues that profit maximisation must be the true objective.
The use of ROI can lead to the suboptimisation of resources. Divisional
managers may reject new capital investment projects which are in the
company's best long-term interests by having a positive net present value,
because of the effect such a project would have on the short-term divisional
return on investment. (See the example in Box 7.1 above.) Because new pro-
jects usually originate at operational levels within companies (King, 1975;
Bower, 1986) it is possible for divisional management to filter out such projects
before central management know of their existence. While the bulk of acade-
mic literature supports this view, Lillis (1992) repudiates this generalist con-
clusion in her work on three large divisionalised firms .
Solomons (1965) argued that the use of residual income overcomes these
difficulties. Residual income imputes a cost of capital on the investment base
which, when deducted from profits (or net earnings) gives an absolute figure
rather than a percentage, like ROI. The charge for capital would be useful for
evaluating performance and for guiding investment decisions. Divisional
managers, he argues, would have an incentive to pursue all projects that have
106 Divisional control

estimated internal rates of return greater than the cost of capital as this could
increase absolute residual income. In the short run, Solomons also argued that
residual income is the long run counterpart of discounted cash flow, so
making it consistent with the wealth maximisation of model classical econ-
omics. This latter assertion has met with much criticism from Amey (1969)
who argued, amongst other things, that the cost of capital may be inappropri-
ate. ROI, regardless of the definitional problems of its two components; profit
and investment, does provide a measure of efficiency. Such a measure,
however, may encourage managers to pursue short-term goals and have a
very narrow view of the organisation objectives. 5wieringa and Weick (1987)
maintain that a measure like ROI at least initiates action within the organisa-
tion, whilst Covaleski et al (1987) links this internal measure with the external
pressures from the financial markets.
Financial measures of performance tend to concentrate on the short term,
that is the effect of actions on the current financial year. In fact the time scale
may be even less: Tomkins (1973) found that in British divisionalised compa-
nies a very large proportion (46 out of 53) appraised their divisions against
detailed financial budgets on a time scale of one month or less. 50 projects
which reduce short-term financial performance are likely to be held back by
divisional managers unless specific provision is made for them in the capital
budgeting process.
The use of financial performance targets results in divisional managers
making decisions which are consistent with the achievement of a corporate
finance objective, but this ignores the fact that organisations have multiple
objectives, many of which cannot be easily measured. This narrowness of
objectives will be considered in some detail after the next section on the use of
discounted cash flow techniques and the issue of the divisional cost of capital.

Discounted cash flow

The discounted cash flow approach to divisional performance attempts to

compare the expected net present value of the division at the beginning and at
the end of a given period. It has clear links to the economic value concepts of
financial accounting which have not been adopted for corporate financial
reporting. While future cash flow may be difficult to estimate, the academic
and practical problem is the establishment of a discount rate or cost of capital
for application to the estimated cash flows .
The establishment of such a discount rate is highly problematic. Less formal
approaches may include the use of past divisional returns, returns achieved by a
company comparable to the division, budget returns of the division, the average
return of industry in which the division operates or management judgement.
More formal approaches may include the weighted average cost of capital
adjusted for risk levels, or the use of the capital asset pricing model (CAPM) to
 M. Broadbent and J. Cullen 107

derive covariances for divisional returns. The establishment of a weighted

average cost of capital for a division may be computationally difficult - there are
problems with gearing levels for divisions, the issue of the holding company not
holding securities issued by divisions and traded in capital markets, although
Williamson (1970,1975) likens the relationship to a capital market.
The use of a CAPM approach has several computation problems, including
the following.

1. Thode (1986) argued that four conditions need to exist before the CAPM
could be adopted: the selection of a company of equal business risk to the
division, the proxy company's business risk can be quantified; no scale
economies or other synergies exist; and the growth opportunities of the
division exactly match those of the proxy companies. Thode further
argued that it is highly unlikely that such conditions will exist jointly.
2. There is a potential problem which may arise because of the changing
expectations of managers having applied a single corporate rate in the
past to using different rates for each division (Welch and Kainen, 1983).
Some divisions may have projects rejected which otherwise might have
been accepted, so having implications for divisional growth, profitability
and management rewards. The change may be culturally unacceptable to
divisional managers.
3. The use of divisional costs of capital to evaluate all projects in a division is
likely to bias divisional managers in favour of riskier projects. A portfolio
of projects of high, medium and low returns is normally present.
Adherence to one cost of capital to evaluate all projects in a division may
mean rejection of low-return but low-risk projects.

Grinyer (1986) has argued that the CAPM approach which emphasises sys-
tematic risk will not be adopted by managers who are subject to total risk.
The discounted cash flow model needs careful consideration but currently a
profitability index like ROJ is the dominant divisional measure of perfor-
mance. This emphasis on accounting measurement and asset valuation has
resulted in broad-based criticism of performance appraisal techniques which
will be reviewed within the next section, which considers recent developments
in divisional performance measurement.

Developments in divisional performance measures

Traditionally there has been a tendency for management accounting control

systems to concentrate solely on internal issues. As indicated earlier in the
chapter, there has also been a tendency for these control systems to focus
solely on financial measures of performance. It has been well documented that
108 Divisional control

financial measures of performance tend to concentrate on the short term (that

is, the effect of actions on one year's profits, return on investment or residual
income) . Projects which harm short-term financial performance are likely to be
held back by a divisional management team unless specific provision is made
for them in the capital budgeting process. We have argued earlier that, if too
much stress is placed on the corporate financial objectives, this may mean
ignoring the fact that organisations have multiple objectives, many of which
are not easily measured in financial terms. To use one key measure of financial
performance is almost sure to produce dysfunctional behaviour as regards one
or more other objectives. Other objectives may relate to sales growth, market
share, employee relations, quality or social responsibility. It is to the need to
link financial measures of performance to non-financial measures of perform-
ance that we now turn.
Many readers will be familiar with the work of Bromwich and Bhimani
(1989), who pointed out that there was mounting empirical evidence support-
ing the need for management accounting to become more externally focused
to enable the organisation to look outwards to the final goods market. This
followed a call for strategic management accounting by Simmonds (1981), to
be further developed by Bromwich (1990). While discussing the joint use of
financial and non -financial measures of performance, we must simultaneously
discuss the link between performance measures and the strategy being
followed by the organisation.
As early as 1979, Parker produced a paper which put forward an argument
for a balanced assessment of divisional performance to be measured by a com-
posite mix of quantitative and qualitative indices. Parker (1979) argues that
much of the traditional divisional performance literature appears to be based
upon a simplistic and unrealistic view of corporate and divisional goals:
'Given the existing range of changing corporate and divisional goals, the divi-
sional profit test taken by itself is inadequate as a measure of any division's
progress towards the attainment of the corporate goal set' (p, 316). He went on
to suggest an alternative approach to divisional performance measurement
which accountants could adopt for the benefit of the decentralised company as
a whole. The suggestion (p. 317) was that accountants should:

- discard the belief that accounting measures should be used to promote goal
congruence among divisional managers;
- recognise the need to preserve some degree of autonomy in divisional oper-
ations;
- review the possible methods of assessing divisional performance with a
view to accounting for the needs and objectives of all levels of management
above and within each division;
- move beyond the single divisional profit-based index to provide an
expanded number of measures of divisional performance which account for
a broader range of success criteria.
 M. Broadbent and J. Cullen 109

Parker suggested the following as possible additions to the traditional

profit/ROI measure:

- financial management ability - stock and asset turnover, gearing ratio,

sources and application of funds, fixed asset statistics such as age, main-
tenance expenditures, depreciation policies;
- productivity - profit before interest and tax.
- marketing - sales volume, market share, sales effort indicators (for
example, visits per customer);
- research and development - research and development costs to sales, re-
search and development cost per employee, project performance indicators;
- social responsibility - social budget, narrative report;
- employee relations -lost time accidents, hours lost as a percentage of hours
worked.

Other writers (Dearden, 1968; Kaplan, 1984; Johnson and Kaplan, 1987) have
suggested before and since that profit/ROI is too narrow a view. However
Parker's main argument is that there is a plurality of objectives for a company
and its divisions and that there must therefore be a more balanced view of
performance and the indicators used to appraise it.
This idea of a balanced view has been further developed by Kaplan and
Norton 0992, 1993). They introduced the idea of a balanced scorecard (see
Figure 7.2) which represents a set of measures aimed at giving managers a fast
but comprehensive view of the business.

The balanced scorecard includes financial measures that tell the results of actions
already taken. And it complements the financial measures with operational meas-
ures on customer satisfaction, internal processes, and the organisation's innovation
and improvement activities - operational measures that are the drivers of future
financial performance. (Kaplan and Norton, 1992, p. 71)

An important feature of this approach is that it is looking at both internal and

external matters concerning the organisation. The balanced scorecard intro-
duces the idea of competitor benchmarking in relation to new product intro-
ductions and technology capability. Another important feature is that it is
related to the key elements of a company's strategy. Different divisions may be
following different strategies and the items in the balanced scorecard would
therefore need to reflect these differences. A final important feature is the fact
that financial and non-financial measures are linked together. There is no sug-
gestion that financial performance measures should be discarded altogether.
Periodic financial statements remind managers that improved quality,
response time, productivity or new products only benefit the company when
they are translated into improved sales and market share, reduced expenses or
higher asset turnover. An illustration of the way the balanced scorecard could
110 Divisional control

FIGURE 7.2 The Balanced Scorecard

Financial perspective Customer perspective

GOALS MEASURES GOALS MEASURES

Survive Cash flow New Percent of sales from

products new products
Succeed Quarterly sales growth
and operating Percent of sales from
income by division proprietary products

Prosper Increased market Responsive On-time delivery

share and ROI supply (defined by
customer)

Preferred Share of key accounts'

supplier purchases

Ranking by key
accounts

Customer Number of cooperative

partnership engineering efforts

Internal Innovation
business perspective and learning perspective

GOALS MEASURES GOALS MEASURES

Technology Manufacturing Technology Time to develop next

capability geometry vs leadership generation
competition Manufacturing Process time to
learning maturity
Manufacturing Cycle time Product focus Percent of products
excellence Unit cost that equal 80% sales
Yield Time to market New product
introduction vs
Design Silicon efficiency competition
productivity Engineering
efficiency

New product Actual introduction

introduction schedule vs plan

Source: Kaplan and Norton (1992), P: 76.

 M. Broadbent and J. Cullen 111

be developed within a divisionalised organisation is given below in Box 7.2

taken from Kaplan and Norton (1993) (see also Figure 7.2).

Box 7.2

FMC Corporation is one of the most diversified companies in the United States,
producing more than 300 product lines in 21 divisions organised into five busi-
ness segments: industrial chemicals, performance chemicals, precious metals,
defence systems, and machinery and equipment. Based in Chicago, FMC has
worldwide revenues in excess of $4 billion.

If we were going to create value by managing a group of diversified companies,

we had to understand and provide strategic focus to their operations. We had to
be sure that each division had a strategy that would give it sustainable competi-
tive advantage. In addition, through measurement of their operations, whether
or not the divisions were meeting their strategic objectives.

If you are going to ask a division or the corporation to change its strategy, you
had better change the system of measurement to be consistent with the new
strategy. We acknowledged that the company may have become too short-term
and too internally focused in its business measures. Defining what should
replace the financial focus was more difficult. We wanted managers to sustain
their search for continuous improvement, but we also wanted them to identify
the opportunities for breakthrough performance.

A new measurement system was needed to lead operating managers beyond

achieving internal goals to searching for competitive breakthroughs in the global
market-place. The system would have to focus on measures of customer service,
market position and new products that could generate long-term value for the
business. We used the scorecard as the focal point for the discussion. It forced
division managers to answer these questions: How do we become our customers'
most valued supplier? How do we become more externally focused? What is my
division's competitive advantage? What is my competitive vulnerability?

We decided to try a pilot programme. We selected six division managers to

develop prototype scorecards for their operations. Each division had to perform
a strategic analysis to identify its sources of competitive advantage. The 15 to 20
measures in the balanced scorecard had to be organisation-specific and had to
communicate clearly what short-term measures of operating performance were
consistent with a long term trajectory of strategic success.

We definitely wanted the division managers to perform their own strategic

analysis and to develop their own measures. That was an essential part of creat-
ing a consensus between senior and divisional management on operating objec-
tives. Senior management did, however, place some conditions on the output.
112 Divisional control

The extract describes an organisation where the divisional managers were

involved in the process of strategic analysis and it is to this area that we now
turn our attention. As mentioned earlier in the chapter, the work of
Williamson (1970, 1975) is based on the premise that central managers take
strategic decisions, whilst divisional managers are vested with operational
decisions. We will now look at a range of literature which challenges this par-
ticular assumption. The main thrust of this literature is the need to accommo-
date, within the divisional performance measurement system, the fact that
many divisional managers are responsible for the development of their own
business strategy. Simons (1990) explains that business strategy refers to the
way a company competes in a given business and positions itself among its
competitors. This contrasts with corporate strategy which Simons (1990)
explains as being concerned with the determination of which businessles) the
organisation chooses to compete in and the most effective way of allocating
scarce resources among business units.
As part of the discussion on the work of Kaplan and Norton (1992, 1993),
the need to link the performance measurement system to the key elements of a
company's strategy was highlighted. The point was also made that different
divisions within the organisation may be following different strategies and
consequently the items in the balanced scorecard would need to reflect these
differences. Hall (1978) described how General Electric , recognising the need
to differentiate between divisions, introduced the concept of Strategic Business
Units (SBUs). Hall outlined (p. 394) that the SBU concept of planning is based
on the following principles:

- the diversified firm should be managed as a 'portfolio' of businesses, with

each business unit serving a dearly defined product-market segment with a
dearly defined strategy;
- each business unit in the portfolio should develop a strategy tailored to its
capabilities and competitive needs, but consistent with the overall capabili-
ties and needs.

Hall (p. 396) identified four steps which are required to operationalise the
concept of SBUs:

1. identification of strategic business elements, or units;

2. strategic analysis of these units to ascertain their competitive position and
long-term product market attractiveness;
3. strategic management of these units, given their overall positioning;
4. strategic follow-up and reappraisal of SBU and corporate performance.

Using the concept of strategy as a position, Hall emphasised the need for dif-
ferent performance measurement and appraisal systems for business units
holding positions in the different strategic classifications of dogs, question
 M. Broadbent and J. Cullen 113

marks, stars and cash cows. However, as part of a critical evaluation of the
SBU process, he suggested that organisations had generally failed to develop
the managerial control aspects of the process. He found it illogical to go
through the process of SBU analysis and then to continue to measure and
reward SBU management on annual performance against a profit target.

Most firms have gone only half way with the SHU concept - they position the
product market segments and then go right on rewarding and promoting managers
on traditional criteria. In the end the companies which make the SHU concept work
will be those which change all management systems; developing and rewarding SHU
managers differentially depending on their SHU position and the strategic handling
which is appropriate for their element of the portfolio . (Hall, 1978, p. 402)

This link between strategy and control systems formed the basis of further
work by Govindarajan and Gupta (1985) and Simons (1987a). Govindarajan
and Gupta (1985) examined the linkage between strategy, incentive bonus
systems and effectiveness at the SBU level within diversified firms. The results
of the study can be summarised as follows: (a) greater reliance on long-run cri-
teria (for example, product development, market development, personnel
development, political/public affairs) as well as greater reliance on subjective
(non-formula) approaches for determining the SBU general manager's bonus
contributions to effectiveness and (b) the relationships between the extent of
the bonus system's reliance on short-run criteria and SBU effectiveness is
effectively independent of SBU strategy.
The study by Govindarajan and Gupta therefore provides empirical support
for the idea that, in terms of SBU effectiveness, the utility of any particular
incentive bonus scheme employed in an attempt to influence the SBU general
manager's behaviour is contingent upon the strategy of the focal SBU. Simons
(1987), again using the concept of strategy as a position, carried out an empiri-
cal study which investigated the relationship between business strategy and
accounting-based control systems. Building on the Miles and Snow (1978)
typology for identifying generic strategies (prospector, defender, analyser),
Simons (1987a) studied firms classified as either prospectors or defenders to
determine whether management control systems differ between the two
groups. He found that successful prospector firms seem to attach a great deal
of importance to the use of forecast data in control reports, set tight budget
goals and monitor outputs carefully. For prospectors cost control is reduced. In
addition, large prospector firms appear to emphasise frequent reporting and
the use of uniform control systems which are modified when necessary.
Defenders, particularly large firms, appear to use their management control
systems less intensively. In fact negative correlations were noted between
profit performance and attributes such as tight budget goals and performance
monitoring. Defenders emphasised bonus remuneration based on achievement
of budget targets and tended to have little change in their control systems.
114 Divisional control

Goold and Campbell (1987a, 1987b) took this debate a stage further by
carrying out a major piece of empirical work looking at the management
of diversification in 16 large UK companies (1987a, b):

For senior managers at the corporate headquarters of most large companies, diver-
sity is a fact of life. The portfolio of these companies frequently includes businesses
from several industries, at different stages of maturity, with different growth option
and different financial performance. Understanding and controlling each of the busi-
nesses in a portfolio of this sort is a severe test of corporate management. (Goold
and Campbell, 1987, p. 42)

Goold and Campbell identified three main different central management

styles (strategic planning, strategic control and financial control) and three
main philosophies for building and managing a diverse portfolio (core busi-
ness, diverse business and manageable business). A brief description of the
different philosophies and management styles is shown in Figure 7.3.

An important finding from the Goold and Campbell work is that companies
tend to employ a uniform style across most of their businesses, and that
changes in style rarely occur. Companies with a core business philosophy tend
to adopt a strategic planning style . Those with a manageable business philoso-
phy adopt a financial control style. Companies operating a diverse business
philosophy use a strategic control style. This relationship is highlighted in
Figure 7.4. Comparing Figure 7.4 with the Scott (1971) matrix (see Figure 7.1)
the Goold and Campbell analysis discusses specific control issues within
stages III and IV in more depth by considering different management philoso-
phies and styles associated with these different stages.

The idea of strategic control is of particular interest to us in terms of the

degree of decentralisation given to business unit managers. This links into the
concept of Strategic Business Units, discussed earlier in the chapter, intro-
duced by Hall (1978). The essence of this decentralisation is the location of the
primary responsibility for proposing strategy and achieving results with the
business unit manager and not with central management. Companies, in this
situation, decentralise strategic responsibility in order to ensure that strategies
are based on a detailed knowledge of specific products markets, to increase
business level 'ownership' of strategy, and to reduce the overload on the chief
executive and senior management. Goold (1991) agrees with these objectives
but emphasises that such decentralisation can only work well if two basic con-
ditions are fulfilled :

- The centre must be able to determine whether the business is on track with
its own strategy. Unless the centre knows when to intervene, decentralisa-
tion becomes abdication of responsibility.
 M. Broadbent and J. Cullen 115

FIGURE 7.3 Different Philosophies and Management Styles

PHILOSOPHIES

Core Business

Company commits itself to a few industries and sets out to win big in those industries.

Manageable Business

The emphasis is on selecting businesses for the portfolio which can be effectively managed
using short-term financial controls. There may be extensive diversity in terms of indus-
tries, but there is homogeneity in the nature of the businesses.

Diverse Business

The emphasis is on diversity rather than focus. The centre seeks to build a portfolio that
spreads risk across industries and geographic areas, as well as ensuring that the portfolio is
balanced in terms of growth, profitability and cash flow.

CENTRAL MANAGEMENT STYLE

Strategic Planning

Corporate management in Strategic Planning companies believe that the centre should
participate in and influence the development of business unit strategies. Their influence
takes two forms : establishing a planning process and contributing to strategic thinking. In
general they place less emphasis on financial controls.

Financial Control

Financial Control companies focus on annual profit targets. There are no long-term
planning systems and no strategy documents. The centre limits its role to approving
investments and budgets, and monitoring performance. Targets are expected to be
stretching and once they are agreed they become part of a contract between the business
unit and the centre. Failure to deliver the promised figures can lead to management
changes.

Strategic Control

The centre of Strategic Control companies is concerned with the plans of its business units.
But it believes in autonomy for business unit managers. Plans are reviewed in a formal
planning process with the objective being to upgrade the quality of the thinking. However
the centre does not want to avocate strategies or interfere with the major decisions. Control
is maintained by the use of financial targets and strategic objectives .

Source: Goold and Campbell (l987b), P: 27.

 ....
....
0\

FiGURE 7.4 Management of a diverse portfolio

tl
~.

Philosophies (j) '

0-
;::t
Core business Manageable businesses Diverse Businesses ::..
Diversity across Low Very high High
8;::t
industries
.....
~
Diversity across Fairly low Low High
types of
businesses
Style at centre Strategic Planning Financial Control Strategic control
How mismatches avoided Core business mainly Portfolio selection Structure into
responsive to Strategic Planning and retention of 'manageable' homogeneous groups
businesses
Growth Mainly organic Mainly acquisition Organic and acquisition
Drawbacks - Limited industry diversity - Limits to acquisition-based growth - Low key centre
- Maturation of core businesses - Vulnerable to aggressive competition - Limited central added value
- Non core businesses - Does not build - Cultural complexity

Source: (Goold and Campbell (1987), p. 50)

 M. Broadbent and J. Cullen 117

- The business heads must know what will be counted as good performance
by the centre. Without clear goals, the whole concept of decentralised
responsibility suffers, since the conditions under which a business head can
expect to operate free from central intervention are ill-defined (p. 69).

In the companies using a strategic control style, formal strategic control

processes typically had the following key stages (Goold, 1991):

- Periodic strategy reviews for each business - the business proposes a

number of specific strategic objectives or milestones and negotiates with the
centre until they reach agreement. These objectives are non-financial indica-
tors of underlying competitive position, and provide a longer-term, more
strategic balance to financial objectives.
- Annual operating plans - explicit non-financial objectives derived from the
strategic plan are often included alongside financial budget objectives in
annual operating plans.
- Formal monitoring of strategic results - some companies combine this with
budget or operating plan monitoring. In others it is a separate process.
- Personal rewards and central intervention - explicit strategic objectives are
built into personal reward schemes and complement financial performance
in guiding central intervention. Most companies avoid a formula-based link
between achievement of strategic objectives and personal rewards, and
prefer a more direct and flexible link.

The last two points in this strategic control process can be related to the work
of Simons (1990)/ who introduces the idea of interactive management control
as opposed to programmed control:

Management controls become interactive when business managers use planning and
control procedures to actively monitor and intervene in on-going decision activities
of subordinates. Since this intervention provides an opportunity for top management
to debate and challenge underlying data, assumptions and action plans, interactive
management controls demand regular attention from operating subordinates at all
levels of the company. Programmed controls, by contrast, rely heavily on staff spe-
cialists in preparing and interpreting information. Data are transmitted through
formal reporting procedures and operating managers are involved infrequently and
on an exception basis. (Simons, 1990, p. 136)

The focus of attention in Simons' work is business strategy and the way in
which top managers use interactive control systems to monitor the progress of
a business unit towards its business goals. Simons argues that the top managers
make selected control systems interactive to monitor personally the strategic
uncertainties that they believe are critical to achieving the organisational goals.
It is also suggested, in a similar way to Goold (1991)/ that the use of interactive
control systems can be linked to the use of subjective reward systems which are
118 Divisional control

not formula based. Company B, for instance, in Simons (1990) has a reward
system based on effort. As a result of the debate and dialogue that surrounds
the interactive management control process, new strategies and tactics emerge
over time. Simons (1990) therefore illustrates how management control systems
are important not only for strategy implementation but also for strategy for-
mulation. The work illustrates how interactive control systems can be used to
manage emergent strategy (Mintzberg, 1978). Similar points are made by Dent
(1990) who illustrates how embryonic management notions can become mani-
fest through new systems of planning, accountability and performance mea-
surement. These in tum can provide conditions of possibility for organisational
reform and the emergence of new strategies.

Conclusion

The later parts of the chapter have discussed the way in which management
control systems, particularly the use of a combination of financial and non-
financial measures as part of a strategic control process, need to be understood
in their organisational context. The overall issues of control within division-
alised companies is a highly complex mechanism, the linkages between strat-
egy/ structure and organisational content provide an opaque picture of
control, which is not reflected in the clarity of the Williamson (1970/ 1975)
model. Hopwood (1987/ 1990) and Dermer (1990) provide both theoretical and
empirical debate about the way in which accounting needs to be seen in terms
of shaping, rather than just enabling, organisational affairs. The approach
taken in the development of strategic control in this chapter has been uncriti-
cal. A critique of strategic control is included within the next chapter.
 CHAPTER 8

Strategic control
AlanCoad

Introduction

The study of management during the last three decades has been dominated
by two concepts: strategy and cybernetics (Dermer, 1988). The resulting
conventional wisdom has tended to be teleological. Strategy is usually
described as management's definition of what the organisation should be
doing. Control is then effected cybernetically, through goal-related feedback
and adjustment. However there is another body of literature, more recent and
rooted in a more empirical research tradition, which regards strategy and
control as something other than the creation of top management. According to
this perspective, organisations are made up of a variety of stakeholders
attempting to satisfy their wants amidst a host of conflicts and constraints.
Strategy in this light is not the prerogative of top management, but rather the
outcome of organisational struggles. It is what the organisation actually does
rather than what top management intends it to do. The resulting order is not
the creation of an overall designer; rather it emerges from the collectivity of
interactions of organisational actors (Mintzberg and Waters, 1985; Johnson,
1987; Dermer, 1988, 1990).
Research has indicated that strategic change is shaped by such things as
administrative planning and control systems (Ansoff, 1979; Anthony, 1965;
Goold and Quinn, 1990a); the perceptions and biases of key decision makers
(Anderson and Paine, 1975; Schwenk, 1984) and political factors (Mintzberg,
1983; Pettigrew, 1985). Much of the conventional wisdom regarding the rela-
tionship of strategy and control has focused on the first of these, thereby,
perhaps inadvertently, denying important aspects of organisational reality.
This chapter attempts to redress the balance somewhat. Initially it examines
the concept of strategy and the conventional wisdom of the relationship of

119
120 Strategic control

strategy and control. It then explores issues such as cognition, learning, ideol-
ogy and politics to gain some understanding of the way they affect this rela-
tionship. Finally it attempts a synthesis of the ideas and considers the
implications for strategic management.

Strategy

The term 'strategy' is probably one of the most ill-defined in the business
vocabulary, having a wide range of connotations. Published definitions vary,
with each writer adding his or her own ideas and emphasis. To some, strategy
refers to a plan, the end product of strategy formulation (Newman and Logan,
1971). Some include objectives as part of the strategy (Andrews, 1971; Quinn,
1980) whilst others see objectives as what the strategy is to achieve (Ackoff,
1974). Many mention the allocation of resources as a critical aspect of strategy
(Hofer and Schendel, 1978). Some prescribe a review of the market and
specifically mention competitive position (Porter, 1985). To further complicate
matters, some writers suggest that strategy is reserved for a formal logic,
explicitly stated, that links together the activities of business (Ansoff, 1979),
while others indicate that a strategy can emerge from a set of decisions and
need not be explicitly stated (Mintzberg and Waters, 1985).
Clearly the word 'strategy' is used in different ways. However within this
variety two major themes are evident. Firstly, there is the idea of strategy as a
position. This concept involves identifying where an organisation locates itself
in its environment. By this definition, strategy becomes the mediating force, or
'match' (Hofer and Schendel, 1978), between the internal and external context.
In ecological terms, strategy in this sense is an environmental 'niche'; in econ-
omic terms, a place that generates 'rent'; in management terms, a product-
market 'domain', the place in the environment where resources are
concentrated. Strategy, by this definition, is a version of positional analysis,
concerned with the status of an organisation relative to competition and other
aspects of its environment. A feature of this concept of strategy is that all
organisations can be said to have a strategy. Thus, while the match between an
organisation's resources and its environment mayor may not be explicitly
developed and while it mayor may not be a good match, the characteristics of
this match can be described for all organisations (Hofer and Schendel, 1978).
The second major theme is the idea of strategy as a process. Landmark
works by Chandler (1962), Ansoff (1965) and Andrews (1971) were among the
first to propose the distinction between the process of strategy and its content.
The distinction has tended to divide research ever since. Strategy process
research has been concerned with the way strategy is formed and imple-
mented, whereas 'what is being decided' has been claimed as the province of
position-oriented research. Questions of 'who is involved in strategy' and
 Alan Coad 121

'why strategy arises' have been addressed by both groups, but in different
ways. The business unit, the company and populations of organisations have
been the focus of position research, whereas process research is largely preoc-
cupied with the individual and the group. The question 'why' has been seen
primarily as one of economic performance by position researchers, whereas
process researchers have looked either to logical or behavioural rationales.
However this distinction between process and position research leaves both
streams weakened, theoretically. Most scholars would agree that the content
of a strategy is likely to be affected by the processes whereby the strategy is
developed and implemented. Also it is possible that the processes are affected
by the content and outcomes of previous strategic decisions. Close examina-
tion of positioning strategies (for example, Miles and Snow's prospectors,
analysers, defenders and reactors) infers a variety of organisational processes.
Moreover many of the prescriptions for the processes of strategy (such as
Andrew's search of the environment to identify opportunities which may be
exploited by company strengths) imply a search for environmental niches, that
is positions.
This overlap of two key themes in the strategy literature is accommodated
by Stacey's (1993) model of the strategy concept. He emphasises the import-
ance of thinking of strategy as a game people play. This overcomes a tendency
to depersonalise strategy. Much of the strategy literature infers an objective
reality in which 'the organisation' moves in response to changes in 'the envi-
ronment' (Glaister and Thwaites, 1993). Unconsciously we begin to see strat-
egy in mechanical terms where one 'thing' moves in predetermined ways in
relation to another 'thing' . This oversimplifies strategic management because
organisations and their environments are not things, but rather groupings of
people interacting with one another. By focusing on strategy as a game, we
remind ourselves to examine the moves, countermoves and further responses
by inter- and intra-organisational players which are the dynamics governing
success or failure.
Stacey's model of this gaming process is shown in Figure 8.1. Starting at the
left-hand side, we see that people in the environment of organisation A do
something that people in that organisation discover. People in organisation A
choose how to respond and then act upon that choice. These actions then have
consequences for people in the environment, for example competitors, cus-
tomers, suppliers, distributors and legislators. These then choose how to
respond to their perceptions of the actions of people in organisation A. And so
the game goes on. People both within and outside an organisation interact
through a series of feedback loops resulting in discovery, decisions and action
that control and develop their organisations.
Figure 8.1 illustrates that we can look back in time from the present to, and
if we are able to perceive a pattern in past actions, we call this pattern the past
strategy of the organisation. Similarly, if we are able to look forward to tf, we
may discern a pattern, or strategy, in future actions. For Stacey, then, strategy
F IGURE 8.1 The Strategy Concept ....
N
N

/;'" - - - J - - - ....
- ~ - I
, \
Choose I \
I Discover \ ' _
\ I I
..... - -It- ;'" I

Act
t -,/ --. .... / I ~
I ' - -\ / Oth er orga nisa tions
I \ and peopl e:
the env ironme nt
Con sequ ences
r r
Conse quences
\ \ Organi sation
\ \ A
\ \
\,n~
\ \
\ \
\ ;' .... \
r r -- \
\,.
~ ,. !, Act .... ~, -_ ~
/
--t ,
: Discover \ ,
, I I
\ Choose \
, I

-- ,_ ...... - -

Sna pshot of Snaps hot of Snapshot of

Organi sat ion A's Organisation A's Organisation A's
Posture Posture Posture
Position Position Position
Performance Performance Perform ance
,
,------ ~ -- ~~-
,
I~ - - ~ - - -- --
,
, - -----".. ...
~ ~ if TI~

Source: Stacey ( 993).

 Alan Coad 123

is a perceived pattern in actions past or yet to come. The emphasis on percep-

tions in this definition is important because strategy does not have an objec-
tive reality independent of the observer: we cannot touch or feel a strategy.
Strategy is a label that people apply to patterns in action. Moreover different
people might apply different labels, focusing on different attributes of the
unfolding action. This leads us to the concept of multiple realities, whereby
different organisational actors perceive the patterns of the strategy game in
different ways, which may lead to different prescriptions regarding how to
act in the future. This concept of multiple realities and the attendant multiple
rationalities will be examined further in the section on political factors, below.
Having explored the concept of strategy, it is possible to offer a working
definition of what the present author understands by the term 'strategic
control':

Strategic control is concerned with the decisions and actions undertaken by organ-
isational actors in response to perceived environmental patterns in organisational
action, past or yet to come.

In particular, this chapter focuses on the way in which organisational actors

make sense of their environment and how this, in turn, contributes to the
development of strategies in the cultural, political and social milieux of
organisations.
It is acknowledged that the definition above is much broader than that
which is normally found in the literature of strategic management and man-
agement control. The conventional wisdom of strategic control tends to place
emphasis on formal planning and monitoring systems. This conventional
wisdom will be described in the next section of this chapter, and subsequent
sections will examine some of the theories of organisational dynamics which
imply a need for an expanded concept of strategic control.

The conventional wisdom

The term 'conventional wisdom' refers to those general prescriptions for

success that command most attention in best-selling management textbooks
and among practising managers. In respect of Stacey's three elements of the
strategy process - discovery, choice and action - the conventional wisdom
prescribes that the discovery phase should comprise some systematic
appraisal of an organisation's strengths and weaknesses, its environmental
threats and opportunities, together with a recognition of the values and expec-
tations of influential stakeholders (usually assumed to be top management).
This knowledge is then summarised in a form of position statement (such as a
SWOT analysis) which gives an indication of where current strategies are
124 Strategic control

likely to lead the organisation into the predicted future. If a gap exists between
predicted performance and managerial expectations then the gap should be
closed by selecting appropriate strategies. The 'choice' phase of the conven-
tional wisdom, then, comprises the identification of possible strategies and
their evaluation according to a favoured set of criteria. The criteria prescribed
for assessing strategies are often summarised under classifications such as
tests of 'fit' with environmental opportunities and threats, 'feasibility' in terms
of the resource capability of the focal organisation, and 'acceptability' to the
influential stakeholders. Together, these stages of discovery and choice are
referred to as strategy formulation. In the conventional wisdom, the action
phase involves a formal implementation of chosen strategies by designing an
appropriate organisation structure, adopting a management style that will
influence organisation culture in a way regarded as instrumental by top man-
agement, and establishing the appropriate control systems to check that
actions align with managerial expectations.
It is acknowledged that this description is a simplification of the conven-
tional wisdom of strategic management. However it presents a picture that
will be familiar to readers of popular textbooks on the subject (see Dess and
Miller, 1993;Johnson and Scholes, 1993; Morden, 1993). Moreover it should be
emphasised that the conventional wisdom is not some monolithic, uniform
body of explanation and prescription. On the contrary, it consists of a number
of approaches that appear to be very different. For example, some writers
(such as Andrews, 1971) regard strategy formation as a creative, conscious
intellectual process of organisational design, whereas others (such as Ansoff,
1979) regard strategy as the outcome of systematic, formal planning and
control. Nevertheless the coventional wisdom tends to adopt a teleological
perspective in that perceived organisational order is assumed to have been
put in place by a designer. It is assumed that organisations follow particular
strategies because that is what their senior management have decided.
Consistent with this belief in the designing influence of top management is
the belief that the circular process of discovery, choice and action that is strate-
gic management is a deliberate and intentional one (Stacey, 1993). The ques-
tion of the validity of these beliefs will be dealt with later in this chapter. For
now it is time to concentrate on what has conventionally been understood to
be the relationship between strategy and control.
From the mid-1960s to the mid-1980s the dominant view of the connection
of strategy to management control was based on the work of R.N. Anthony,
who defined management control as 'the process by which managers assure
that resources are obtained and used effectively in the accomplishment of the
organisation's objectives' (Anthony, 1965). This definition distinguished man-
agement control from both strategic planning and operational control and led
to the development of formal, systematic, organisation-wide, data-handling
systems which are designed to facilitate management control (Machin, 1983).
It is a perspective that has been reinforced by a tendency in the strategy litera-
 Alan Coad 125

ture conceptually to separate strategy formulation from strategy implementa-

tion. Hence strategic planning sets the strategies, management control checks
they are being pursued appropriately. Control is effected through negative
feedback. The management control system measures the performance of a
process and compares it with a standard. If the performance fails to meet the
standard, subsequent performance is adjusted to ensure compliance. The
model is, therefore, essentially cybernetic, although a very simple kind of
cybernetics is involved (Lowe and Puxty, 1989).
It was noted in Chapter 2 that Anthony's perspective brings into close align-
ment the concepts of management control systems and management account-
ing systems. For him, management control is built around financial
information about revenues, costs and resources. The information covers both
planned and actual levels, such as are found in a budget. In addition to
financial information, the system may include data about volume of resources
and number of personnel, as well as the quantity and quality of output. Plans
are presented and approved at certain dates each year. Reports on actual out-
comes are submitted, reviewed and evaluated in a prearranged sequence.
Overall the management control process is depicted as an integrated one,
encompassing programming, budgeting, accounting, reporting, analysis and
corrective action.
Despite its plausibility and widespread acceptance, Anthony's model of
management control has a number of serious limitations. Some of these have
been referred to in Chapter 2. What we seek to emphasise here is the fact that
a number of limitations came to be acknowledged in the strategy literature
during the 1980s and led to a variety of prescriptions for systems of strategic
control (see Hurst, 1982; Lorange et al., 1986). In particular it was noted that
Anthony's model stresses financial objectives and, via the medium of the
budgetary control system, usually concentrates only on the coming 12 months.
It pays little explicit attention to longer-term goals and objectives. Moreover,
by virtually ignoring issues such as organisations' environments and their
positioning relative to other entities, it fails to accommodate constructs that
are fundamental to the concepts of strategy and control. Furthermore it plays
down the potential of management control systems to influence strategy for-
mation and draws unnecessarily restrictive boundaries around concepts such
as management control and strategic planning.
The prescriptions for strategic control systems that emerged in the strategy
literature during the 1980s broadly fall into two categories. Firstly, there are
those elements of the system which are concerned with ensuring that the
firm's strategy is on track. These may be referred to as implementation con-
trols. Secondly, there are recommendations for systematic approaches to check
that the present strategy remains relevant in changing circumstances. This is
the role of strategic control systems which enable top management to respond
to unforeseen external and internal developments with changes in strategy.
These may be referred to as relevance controls.
126 Strategic control

Strategy implementation control is concerned with whether or not a firm's

strategy is being implemented as planned. By analogy with budgetary control
systems, this requires that objectives be established and performance meas-
ured against the objectives to assess whether strategy is going according to
plan. However there is a fundamental problem here, in that strategic object-
ives tend to be relatively long-term. Hrebiniak and Joyce (1986) point to a ten-
dency for managers to focus more on short-term than on long-term objectives.
Controls set against objectives that are several years ahead will not be as
influential as controls set against this year's targets. Therefore they suggest
that strategic control systems should specify short-term goals (milestones)
which need to be achieved in order that the strategy is implemented in the
long term. The strategic milestones are not targets in themselves, but rather
things which are realised on the way (for example, completion of a new
factory) . To strike a proper balance between long-term and short-term object-
ives, Lorange (1988) argues in favour of developing separate strategic and
operational budgets which should be controlled against independently. His
strategic budget is based on two elements. Firstly, it includes an allocation of
resources necessary to change the direction of a particular business, that is,
those resources necessary to implement a major change in business strategy.
Secondly, the resources should be identified with specific strategic projects or
programmes. Lorange suggests that the primary justification for a particular
allocation of strategic resources is that it should enhance the firm or business
unit's long-term competitive position. The operational budget, on the other
hand, should include those resources necessary to maintain current operations
at a successful level. For each organisational unit, Lorange suggests establish-
ing strategic objectives (the eventual position competitively); strategic pro-
grammes and milestones (specific tasks en route to the strategic objectives);
strategic budgets (the resources to be spent on strategic programmes); and
operating budgets. Performance should be monitored against each heading
and the reward system should, at least in part, be geared to the implementa-
tion of strategy.
Recommendations in the literature regarding the use of relevance controls
are concerned with a much broader question: should the firm's overall strat-
egy be changed in the light of unfolding events and trends? Hurst (1982)
reasons that there are differences between strategic control systems and
budgetary control. Strategic control requires more data from more sources,
particularly external sources, and may be concerned with competitive bench-
marks and with non-financial measures, as well as with long-term outcomes.
Strategic control may therefore be less precise and less formal than budgetary
control and is concerned more with the accuracy of the premises on which a
managerial strategy has been based and much less with quantitative devi-
ations from a standard. Premise control requires systematic and continuous
strategic survelliance of internal and environmental factors to determine if the
 Alan Coad 127

assumptions on which a managerial strategy has been based remain valid. It is

argued that control processes such as these provide top management with the
information they need to decide when to intervene to adjust strategy
(Schreyogg and Steinmann, 1987).
We see from the above sections that there is no shortage of literature recom-
mending the establishment of strategic controls. However, paradoxically, very
few companies identify formal and explicit strategic control measures and build
them into their control systems (Horovitz, 1979; Goold and Campbell, 1987a,
1987b). A recent survey (Goold and Quinn, 1990a) shows that only 15 per cent
of those large companies in the UK which have formal planning systems use
them as a standard against which to monitor actions. Goold and Quinn (1990b)
question why this apparent paradox should exist. Is it because there is a
significant lag between theory and practice? Or is it that the benefits of strategic
control systems have been greatly overstated in the literature of the 1980s?
We may go some way towards answering these questions by returning to
the idea that the conventional wisdom is teleological in nature. It pictures the
firm as being guided by a comprehensive and explicit strategy which is sys-
tematically planned and co-operatively executed. It considers strategic deci-
sions to be the domain of senior management. As management's definition of
what the organisation is and what it should be doing, strategy articulates,
shapes and justifies the context of organisational activity for all concerned
(Dermer, 1988). The conventional wisdom admits that there are barriers to
strategic control arising from systems complexity and behavioural and politi-
cal factors (Lorange and Murphy, 1984). However these are mere stumbling-
blocks; they do not negate the definition of success which is still measured in
terms of managements' ability to deal with the issues it faces (Dermer, 1990).
The conventional wisdom defines issues as events in the organisation's en-
vironment to which management must respond. Accordingly prescriptions
are made to sharpen management's response by developing systematic
approaches to environmental scanning.
However, as we have seen, there is a theory/practice gap regarding strate-
gic control. The position taken here is that the teleological model of strategic
management, whilst of value as a logical model in structuring debate and
informing learning, does not describe strategic management in practice.
Specifically it attaches too much importance to the role of top management in
strategy formation. It is too optimistic about the possibility of synoptic, ra-
tional analysis by top management and co-operative execution of strategies. It
fails to address such issues as the cognitive limitations on analytical and ra-
tional behaviour, the interaction of a variety of stakeholders in the location
and definition of strategic issues or the exercise of choice at the interpersonal
level. In short, it is a depersonalised model of management which fails to
recognise the emergent nature of strategy. It is to these issues that we turn in
the next two sections.
128 Strategic control

Cognition and shared mental models

The material in this section deals with organisational members' cognitive

models of strategic problems and the factors which affect the ways they
develop these models. This, in turn, affects strategic issue diagnosis, problem
formulation and choice of action. Much of this perspective derives from work
in the 1940s and 1950s by Nobel prizewinner H.A Simon, who laid the
groundwork for the treatment of cognitive simplification in his discussion of
'bounded rationality' by suggesting that choice of action in the real world
departs from the requirements of objective rationality because of human cog-
nitive limitations.
Within the field of cognitive psychology, the construct theory of Kelly (1955)
and the 'field expectations' of Tolman (1949) are both models of cognition
which suggest that individuals have a coded perception of their world to
which they relate new experiences. Their models indicate that managers will
process environmental and organisational events through pre-existing knowl-
edge systems known as schemata. Schemata are mental models which repre-
sent beliefs, theories and propositions that have developed over time, based
on individuals' personal experiences and which allow them to categorise
events, assess consequences and consider appropriate actions (Fiske and
Taylor, 1991).
A number of researchers have argued that the brain's capacity to store infor-
mation vastly exceeds its capacity to process that information (for example,
Hyman and Anderson, 1982). As a result, there is a need for some sort of
patterning of the information into larger units so that it can be managed.
Furthermore, when faced with new situations, individuals seek to allocate the
experience to an existing pattern. Schemata thereby provide frames for prob-
lems which make it unnecessary for organisational decision makers to expend
the mental effort necessary to diagnose completely each element of a new
strategic problem (Taylor and Crocker, 1983). Schon (1983) reasons that
schemata provide managers with a repertoire of images, examples and actions
such that when a new situation arises it is perceived as something already
present in the repertoire. There is a tendency then to filter new information so
as to focus only on those data which fit in with established mental models.
In a similar vein, Weick (1979) argues that schemata - he refers to them as
'perceptual sets' - are ready-made explanations that the actor carries from situ-
ation to situation. Futhermore he suggests that human actors do not react to
an environment, rather they enact it; that is, they create meaning from envir-
onmental stimuli on the basis of attention to what has already occurred.
Moreover what they understand to be occurring currently affects their percep-
tion of past occurrences. They thereby continually amend their mental models
through their ongoing experiences or enactments. In this sense they invent
their own environment. For Weick, environments are enactments by managers
 Alan Coad 129

of the world in which they live; their 'reality' in terms of perceptions about the
interactions of people inside and outside an organisation is not to be under-
stood as an empirical reality but rather as a function of management cogni-
tion. Indeed it may be a mistake to think of managerial thought and
managerial action as separable. When managers act, their thinking occurs con-
currently; there is a presumption of logic in meeting a situation, so action is
natural and thinking (in action) in turn endows the action itself with greater
meaning (Weick, 1983). It may, then, only be possible for managers to make
sense of what they are doing after they have done it. This reasoning has impli-
cations for our understanding of strategic control. In highly complex and tur-
bulent situations, explanations of strategic change need to take account of the
view that environments are invented or created in managers' minds and that
they can often only make sense of what they are doing with hindsight.
One further consequence of using schemata to process information deserves
to be highlighted. Their use allows individuals to reason by means of analogy.
This enables people to speed up the process of recall and application to a new
situation. We use analogies to make the novel seem familiar by relating it to
prior experience. A basic form of learning uses some form of repetition to
push schemata into the subconscious where they can be recalled and used
very rapidly. The richer the stock of schemata, the more expert the person.
This form of learning is referred to as 'single-loop learning'. A stimulus results
in a reaction based on our previous experience. No attempt is made to retrieve
and examine the subconsious schemata to test it for continuing validity. This
carries with it a number of dangers in the strategic setting. The fact that the
schemata being used to design actions are subconscious means that we are not
questioning them. The more expert people are, the quicker they use subcon-
sious models, and therefore the more readily they take for granted the
simplifications and assumptions on which the models have been based. This is
highly efficient in stable conditions, but in changing conditions it may lead to
skilled incompetence (Argyris, 1990). This gives rise to the need for double
loop learning (Argyris and Schon, 1978) where we learn by questioning and
adjusting subconscious schemata.
However, so far the discussion has been at an individual level. Strategies are
rarely formed by individuals alone. More usually they are the product of
people choosing and acting in groups. It is worth considering whether the
lessons from cognitive psychology have relevance at the group level. Sims and
Gioia (1986) suggest that the schemata concept can be applied to management
groups and used to describe similarities in assumptions and ways of process-
ing information within the group. According to the authors, common types of
schemata include implicit theories about relationships among variables within
a company and its environment, and scripts, which provide directions for
behaviour.
Beyer (1981) argues that there exist organisational ideologies that can be
defined as a relatively coherent set of beliefs that bind some people together
130 Strategic control

and that explain their worlds in terms of cause and effect relations. Other
writers refer to similar sets of beliefs as 'paradigms' (Sheldon, 1980; Pfeffer,
1981). Aparadigm is a set of beliefs and assumptions people have about the
world. It is usually subconscious and therefore rarely questioned. As people live
and work in a group they come to share ways of looking at the world. The para-
digm flows from shared past experience and reduces the communication and
information flows necessary to secure cohesive group action. This sharing of
implicit models is often referred to as the culture of the group or organisation.
Of importance to our discussion about strategic change are the insights from
researchers who have observed the tendency of paradigms to dominate organ-
isations in a relatively conservative way. Sheldon (1980) argues that organisa-
tions may enter a 'paradigmatic state' in which they cease to adapt to their
environment, or may simply adjust marginally within their paradigm. The
paradigm tends to be quite robust, even in the face of potentially
disconfirming evidence. Grinyer and Spender (1974) describe how managers
will defend their paradigm by cycling problems through routines of adjust-
ment and interpretation (single loop learning). Confronted by a challenge, for
example declining performance, management are first likely to seek means of
improving the implementation of existing strategy, perhaps by tightening con-
trols . If this is ineffective, they may seek to change strategy incrementally, but
still in a manner in line with the existing paradigm. Only if there is persistent
disconfirming evidence will management consider abandoning the old recipe
and shifting to a new paradigm. Similar observations are made by Argyris and
Schon (1978) who provide illustrations of the difficulty managers have in
moving from single loop learning, where they are making responses within
established cognitive frameworks, to double loop learning, where they inter-
nalise divergent views which do not correspond to these frameworks.
These observations suggest that paradigms and their attendant strategies
may not change readily in response to environmental change. This provides
the phenomenon of strategic momentum (Miller and Friesen, 1980) where
managers resist changes that conflict with their way of understanding their
organisation and its environment until some crisis makes such resistance
unsustainable. Johnson (1987) describes a typical pattern of strategic change in
which the organisation is driven down a particular path by its own momen-
tum, becoming more and more out of line with the requirements of its envir-
onment. When the resulting strategic drift has taken the organisation too far
from these requirements, a revolutionary adjustment is necessary which
involves breaking the old paradigm and establishing a new one.
In this section it has been suggested that central to the process of forming
strategy are the cognitive systems of the people involved. In the previous
section we examined the conventional wisdom of strategic management. This
represents a particular paradigm, the assumptions of which suggest that
strategy making is the domain of top management who through a systematic
assessment of an organisation and its environment can identify and imple-
 Alan Coad 131

ment strategies that maintain an organisation in a state of dynamic equilib-

rium in which it is adapted to its environment. The implementation and
control of strategies is assumed to be effected by an appropriate organisation
structure, sophisticated control systems to sense environmental and organisa-
tional change, and the establishment of a strong organisational culture. It
follows that the behaviour patterns of successful organisations are assumed to
be ones of order, regularity and consensus around a stable belief system.
However the empirical research of Miller (1990) and Pascale (1990) suggests
that such harmony, consistency and fit achieved by organisations may be at
one and the same time their strength and their downfall. Reinforcing their par-
adigm, the tendency is for companies in this position to concentrate on doing
better what they already do well. This results in what Miller refers to as the
Icarus paradox. As organisations keep modifying and building upon what ini -
tially made them successful, a momentum results which leads to disaster.
Pascale emphasises the need to change today's dominant paradigm. The con-
ventional wisdom which focuses on order, objectives, planning and control
systems tends to emphasise doing the same things better. Consequently the
existing paradigm blocks alternatives. Pascale argues for a new paradigm that
recognises the non-equilibrium nature of innovative organisations and the
connection between tension and creativity. For him, successful organisations
demonstrate a paradox. On the one hand, they must achieve 'fit', that is coher-
ence among units, central control and synergies; on the other hand, they also
require 'split', that is decentralisation, differentiation, variety and rivalry. Split
is required in order to develop new perspectives and innovative actions. This
leads to the destruction of old paradigms. This need for fit and split inevitably
causes tension, but Pascale regards this as creative tension that provokes
inquiry and questioning and leads to a learning organisation which supports a
continual dialogue between contradicting points of view.
Overall, then, a picture begins to emerge that innovative strategies will
rarely flow from centralised systematic strategic planning. Instead most innov-
ative strategies emerge from a complex organisational learning process. We
will return to this theme in a later section. First it is necessary to consider
another important facet of group dynamics: political factors.

Political factors

Underlying the conventional wisdom is what is referred to as a unitary view

of organisations. This pictures organisations as being integrated wholes where
the interests of the individual and the organisation are synonymous. It empha-
sises the importance of individuals subordinating themselves in the service of
an organisation as a means of realising their interests and the common good
(Morgan, 1986). The unitary perspective reifies organisations by regarding
132 Strategiccontrol

participants in organisational life as being united under the umbrella of

'organisational goals' which 'it' attempts to achieve in a team-like manner
(Fox, 1974). Consequently organisations are pictured as harmonious, consen-
sual phenomena existing for the pursuit of common ends. Conflict tends to be
explained in terms of pathological causes (such as ignorance or antisocial
behaviour), and it is seen as an unwanted intrusion. It is the role of the team
leaders (top management) to eliminate or suppress conflict wherever possible
so as to reinforce the team spirit. Power is regarded in such a way that it is
limited to management; formal authority is the only legitimate source of
power, and the rights and abilities of others to influence organisational
processes are rarely acknowledged (Morgan, 1986).
It will be argued here that the unitary perspective not only inaccurately
describes organisational life but also may lead to a suppression of dissenting
views which are necessary for organisations to achieve creatively and innova-
tively a strategic advantage. An alternative to the unitary perspective is the
pluralist view. This sees organisations as comprising diverse socioeconomic
groups whose pursuit of sectional interests inevitably creates some conflict
(Fox, 1974). Hence conflict between organisational stakeholders (management,
workers and so on) is not abnormal: it should be expected. Morgan (1986)
argues that conflict can serve both positive and negative functions. On the
positive side, conflict can energise an organisation. It may counter tendencies
towards lethargy, staleness and apathetic compliance. It may encourage forms
of self-evaluation that challenge the conventional wisdom and established
organisational paradigms. Consequently, it may help stimulate learning by
keeping the organisation in touch with what is happening in the environment
and by being an important source of innovation. In particular, a healthy atti-
tude towards conflict may overcome pathological conformity arising from
'groupthink' and 'organisational defensive routines'.
Groupthink (Janis, 1972) refers to a tendency of some groups who work
together over a period of time to produce poorly reasoned decisions. Group
cohesiveness and a desire for consensus lead to a number of symptoms which
include self-censorship by members who avoid speaking up against the major-
ity opinion; a shared illusion of invulnerability; and erroneous stereotyping
of people outside the group. These phenomena can then lead to serious
deficiencies in the decision-making process. These include incomplete survey
of alternatives, failure to examine the risks of the preferred choice; failure to
reappraise initially rejected alternatives; and selective bias in processing infor-
mation (Janis and Mann, 1977). Overall too few alternatives are examined,
with the policy alternative first tabled often being the one adopted, whether it
was sound or not.
Organisational defensive routines (Argyris, 1985, 1990) are patterns of
behaviour deployed by people in an organisation to protect themselves from
embarrassment and anxiety. A key defensive routine is to make matters undis-
cussable and to make the undiscussability itself undiscussable. So, for
 Alan Caad 133

example, subordinates may refrain from telling their superiors the truth if it is
likely that the truth may be something they will dislike. Of course the subordi-
nates do not publicly admit they are doing this. Nevertheless their superiors
know it is going on, because they do it themselves. The result is an undis-
cussed game of pretence in which all indulge and all know it is going on. The
pay-off from the game is that it avoids confronting potentially embarrassing
or dissonant situations. Argyris describes such defensive routines as barriers
to organisational learning. They may become so entrenched that managers
actually avoid discussing contentious open-ended issues altogether. This is
particularly dangerous in circumstances requiring strategic control.
Organisational participants may become desensitised to gradually accumulat-
ing changes in the environment and the prevailing paradigm will become
reinforced by default.
The existence of rival points of view, recognised by a pluralist perspective,
can do much to overcome such pathologies and improve the quality of strate-
gic decision making. There are of course negative aspects of conflict. Too much
of it can immobilise an organisation by channelling the efforts of its members
into unproductive activities. Nevertheless it is a matter of balance because, as
we have seen, too little conflict may result in complacency and lethargy.
The basic unit of analysis in the pluralist perspective is the interest group.
Each unit is presumed to have its own perspective, its own rationality, and to
operate consistently within the rationality (Cooper, 1983; Dermer and Lucas,
1986). The multirational nature of organisations generates conflict wherein
stakeholders contend for the control of strategy. There is no presumption of
managerial prerogative. Managerial plans do not necessarily determine the
contents of the strategic agenda (Dermer, 1990). Strategic concerns may origi-
nate anywhere in the organisation and the role of 'change entrepreneur' can
be assumed by anyone with the necessary motivation and ability : 'the strategic
agenda emerges from the reconciliation of conflicting rationalities, not from
the imposition of management's global scheme. Because no grand design is
assumed to exist, organisational strategy does not necessarily need to make
sense from any single perspective' (Dermer, 1990).
March and Simon (1958) suggest that each interest group recognises, defines
and attempts to resolve into limited and manageable problems the uncertainty
it faces. It then seeks to articulate and legitimise those problems for which it
feels it has implementable solutions. Dermer and Lucas (1986) reason that it is
probable that there is more than one model of controllable reality operating at
anyone time. Moreover multiple models imply the necessity for multiple
control systems (even though only one system, that of top management, is
acknowledged as legitimate).
For pluralists, sources of power are the media through which conflicts of
interest are resolved. If we accept the assumptions of shared power, multiple
realities and multirationality, it is likely that the formal, authoritative control
systems cannot fully prescribe or predetermine behaviour. Because of
134 Strategic control

mutual dependency and the need for mutual survival, differences between
organisational units must be harmonised through processes such as negotia-
tion and compromise. The role of each organisational unit is regulated, not
only by a system of (strategic) control imposed by top management, but also
as a result of negotiation with other stakeholder units. Thus Dermer and
Lucas (1986) contend that actual control systems are, in fact, the emergent
results of prior political negotiations between organisational units. An impli-
cation of this reasoning is that the conventional wisdom of strategy and
control as it is taught (and practised?) does not reflect the whole of
organisational reality.

Towards a broader concept of strategic control

In recent years, the strategy literature has begun to reflect a perspective that
coalesces in the notion of emergent strategy (Mintzberg and Waters, 1985).
Strategy in this light is not the prerogative of top management, but rather the
outcome of organisational struggles. It is what the organisation actually does
rather than what top management intends it to do .
By considering the ways in which cognitive factors interact with political
factors, we may develop a clearer understanding of the ways in which strate-
gic change occurs. This will have implications for what we understand by the
term 'strategic control'. Below we present a simplified model of the strategic
change process, which adopts a pluralist perspective. In this model, the
process of strategic change begins with an event in the internal or external
environment of an organisation. The event becomes known to individuals
within the organisation either because they have direct contact with it or
because it is communicated to them through the formal and informal systems
of the company. Research suggests that the awareness of strategically
significant events often exists in the organisation long before any action is
taken to deal with them (Pettigrew, 1985; Schwenk, 1988). Indeed the event
may be dealt with by local subsystem adaptation, rather than precipitating an
organisation-wide response (Quinn, 1980).
According to the pluralist view, the organisation will comprise interest
groups, each of which has a paradigm which will be used to interpret the
event and define its significance. The paradigm will also suggest a possible
response to the event on the part of the organisation. This response will tend
to be consistent with the political interests of the group proposing it. Given
sufficient motivation, the interest group will push for organisational attention
to be paid to their issue and proposed solution. A complex process of advo-
cacy and coalition building is required before the issue can be said to have
gained organisational attention. Once an issue has gained sufficient support,
 Alan Coad 135

in the sense that it is being discussed by people with sufficient power to do

something about it, then the issue becomes part of the strategic agenda
(Dermer, 1990). This process is often characterised by the interaction of multi-
ple participants advocating alternative solutions to multiple problems that are
being dealt with simultaneously (Cohen et al., 1972; Pettigrew, 1985). Through
this interaction, new understandings of strategic issues emerge (March and
Olsen, 1976; Quinn, 1980). There is no overall framework to which organisat-
ional participants may refer when tackling issues on the strategic agenda
because each issue is unique. Potential solutions become apparent through
discussion with other members of the organisation, with customers, suppliers
and even competitors. This progress of the issue is likely to have implications
for participants' paradigms. They are in effect modifying old mental models
and existing company recipes to come up with new ways of doing things.
Thus the people involved with issues on the strategic agenda experience a
real-time learning activity which results in new paradigms. These then form
the basis for a new strategic orientation. This, in turn, may necessitate changes
in organisational structure and information processing, which will, of course,
take time. The new paradigm and its associated organisational implications
inevitably create conditions for changes in political alliances and the relative
power of interest groups. This may create a resistance to change which
reinforces the notions of strategic momentum and strategic drift described
earlier.
This simplification of the political and cognitive factors involved in strategic
change is sometimes referred to as self-organisation. This is because no one
centrally organises event recognition, issue framing, support building and
choice of action. Self-organisation is a process of political interaction and
group learning from which innovations and new strategic directions may
emerge. It occurs when people form a group that produces patterns of behav-
iour, despite the absence of formal hierarchy within the group or authority
imposed from outside it.
Concepts of emergent strategies and self organisation lead us to reconsider
what is meant by strategic control. It is most usually portrayed as something
to do with ensuring conformity with the central organisation-wide intent of
top management. However Dermer (1988) argues that cognitive and political
models view organisations as non-goal-oriented, non-instrumental social
systems which are susceptible to change in infinite ways because their strate-
gies can be affected to some extent by any stakeholder. For Dermer, control
must therefore be redefined as that which causes activities and outcomes to
happen. In this chapter, we have attempted to follow his advice, and hope that
this is reflected in the definition of strategic control proposed on page 123. In
this way, we broaden the concept so as to explicitly recognise that forces other
than senior management can and do shape strategic evolution. Organisational
effectiveness is no longer goal-related but can be evaluated only in terms of
adaptability and, ultimately, survival.
136 Strategic control

Implications for management

We appear to have painted a pessimistic picture regarding mangements' abilit-

ies to effect strategic control. This is unintentional. What we have attempted is
of a more realistic portrayal of the context in which strategic mangement takes
place . It is a complex context in which simultaneously organisations require
both stability to effectively implement current strategies via approximations to
planning/monitoring forms of control; and instability to encourage innova-
tion via political and group learning modes of decision making and self-
organising forms of control. In this context, an important role for strategic
management is one of effectively using their positions of power to influence
the structural arrangements and dynamics of organisations so as to provide an
appropriate balance between the stability of planning and monitoring systems
and the discontinuities of learning and political behaviour.
Assuredly scholars are a long way from becoming prescriptive regarding
the management of the inevitable tensions. The legacy of Burns and Stalker
(1961) was an 'either/or' way of thinking about the design of organisations.
There became a widely held belief that formal bureaucratic controls are inap-
propriate for organisations facing high levels of uncertainty. Changing con-
texts apparently require more organic practices: co-ordination is achieved
through the use of informal, personal communications, rather than by rules
and standard operating procedures. The implication is that organisations face
choices between designs for order or for disorder, for consistency or for dis -
turbance, and for continuity or for change.
We would argue that this dichotomy has been overstated. Katz and Kahn
(1978) observe that organisations possess both 'maintaining systems', which
insulate them from change and uncertainty and perpetuate the status quo, and
'adaptive systems', which stimulate innovation and experimentation. An issue
which remains largely unaddressed in the literature is the extent to which the
balance between these types of system may be designed through the conscious
intervention of senior management, or left to spontaneous self-organisation.
Some indirect evidence is provided by comparing and contrasting the
approaches to control taken by two large successful organisations described in
case studies in the same text (Bruns and Kaplan, 1987). Dent (1987) describes
the structure and processes of control systems at Eurocorp, a company which
develops, manufactures and distributes a wide range of computing products.
Simons (1987a, 1987b) describes the control processes at Johnson & Johnson, a
company with business interests in pharmaceutical, cosmetic and health care
products. Both companies are large organisations. Contingency studies
suggest that increased size leads to task specialisation, especially where scale
economies may be obtained. This in turn creates task interdependence and
pressures for the development of bureaucratic planning and control proce-
dures (Mintzberg, 1979). In addition, both companies face high levels of
 Alan Coad 137

environmental uncertainty. Such uncertainty tends to produce pressures for

decentralised decision making, giving discretion to those with specialised
market knowledge, and may be best managed through organic arrangements
(Mintzberg, 1979). What is especially striking is a comparison of the different
ways in which the two companies cope with their challenging circumstances.
At a superficial level, Eurocorp's control systems appear to be designed
according to traditional principles. Formal planning procedures exist.
However they are not continuous activities. In the rapid rate of change in the
company's product-markets, planning details quickly become obsolete.
Planning creates corporate direction but, in practice, operational co-ordination
is achieved through a complex pattern of spontaneous interaction. Dent
argues that the structure of the control system is significant in supporting this
interaction. Responsibility exceeds authority; managers depend on others to
achieve their own unit's performance targets. This creates tensions in the
organisation, encouraging unit managers to think beyond their functional
tasks and to negotiate with managers of other units to act on their behalf.
In contrast, Johnson & Johnson cope with the learning requirements of an
uncertain environment by formalising frequent superior / subordinate interac-
tion. Simons distinguishes two types of control process. The first he terms
'interactive control' to describe situations in which business managers actively
use planning and control procedures to monitor and intervene in ongoing
decision activities. At this company, long-range and financial planning
systems are used interactively: superiors are highly involved in negotiations
with subordinates regarding revisions to plans and actual outcomes. The
involvement is formally programmed and is an ongoing activity. However
this interaction is set against a background of a second type of control process,
which Simons refers to as 'programmed control' . This is used in areas of the
business which are less exposed to uncertainty. Here managers direct their
attention primarily to ensuring that predetermined control procedures are
established and maintained by designated subordinates. For programmed
control, managers intervene only if outcomes are not in accordance with
predetermined standards.
Obviously, the richness of the case studies by Dent and Simons cannot be
conveyed in two short paragraphs. The reader is encouraged to refer to the
source material. Nevertheless we observe here two significantly different ways
of coming to terms with some of the issues explored in this chapter. Both cases
demonstrate the paradox of 'fit' and 'split' (Pascale, 1990). At Eurocorp, stabil-
ity, central control and synergies (fit), are encouraged through the annual
planning cycle, which examines broad competitive issues, allocates resources
and establishes an attitude of profit consciousness through the specification of
unit financial objectives. However new perspectives and innovative actions
(split) are encouraged through decentralised decision making and overlapping
responsibilities, which results in informal negotiations between managers of
different units. Hence needs for double loop learning are primarily accommo-
138 Strategic control

dated by horizontal interaction and self-organisation. In contrast, at Johnson &

Johnson, stable activities are subject to programmed controls. The need con-
tinually to re-examine organisational paradigms is formally designed into
the routine and frequent interaction between superior and subordinates.
Both approaches encourage organisational learning and the identification of
strategic issues by people at lower levels of the organisational hierarchy. It
may be that, to some extent, the approaches are substitutable, and involve
certain trade-offs. Dent observes that the processes at Eurocorp are inefficient.
They consume a lot of management time and energy. Without formal author-
ity over other units, managers expend a lot of effort in persuasion of others.
Sometimes there are suspicions that agreements may not be implemented.
Levels of stress are high, and frictions emerge. Nevertheless it is far from clear
that alternative approaches would be less costly. At Johnson & Johnson, inter-
active controls necessitate continual replanning and frequent communication
between superiors and subordinates. These procedures are costly to operate.
If the approaches are substitutable, they call into question simple contin-
gency frameworks which tend to infer that mutually exclusive modes of
strategic control may be appropriate in different circumstances. We really
need far more detailed case studies of the nature of those provided by Dent
and Simons before we are in a position to describe more accurately strategic
control in practice. The provision of these descriptions, together with consid-
eration of the generalisability of findings, and the extent to which strategic
control processes may be intentionally designed by top management, are some
of the more important challenges facing researchers and practitioners in the
years to come.
 CHAPTER 9

Control of embedded operations

spanning traditional boundaries
Anthony J. Berry

As so often in discussions of management control, we return to the seminal

idea of Robert Anthony's three levels of control: strategic, managerial and
operational. Earlier chapters of this book have been critical of this taxonomy,
but it very serviceable as an ideal type from which we may extend our analy-
sis . In this chapter we leave aside the issues of strategic and managerial
aspects and turn to the consideration of operational control.
Typically, in management accounting and control texts, control of opera-
tions is handled through the establishment of budgets of various kinds (see
Chapter 6). In this chapter we wish to take a broader view of operations, fol-
lowed firstly by consideration of modes of control for networks, and secondly
by consideration of financial control and clan control issues in what we shall
term 'embed d ed networks' or chains of operations (Grabner, 1993).
Embeddedness means that such a firm is linked by medium-and longer-term
contracts to both suppliers and to purchasers in a complex chain of organisa-
tions, which themselves are similarily connected. In the first section of this
chapter some of the characteristics of the embedded firm are developed. This is
followed by a discussion of the problems of control in and of such chains. The
issues for financial control are examined prior to an exploration of the social
and organisational processes within and between partners in such chains.

Operations and organisations

We will consider operations from two standpoini:s: firstly, those which are
wholly within a given organisation, and secondly, those that lie within and

139
140 Control ofembedded operations

outside the boundaries of a number of participating organisations. In the first

case we will be considering operations which are wholly under the control of
the owners or managers of an enterprise. Such an operation might be simple
and only inside one part of a complex organisation but it might also be an
operation which requires the co-operation of many different organisation sub
units. In Figure 9.l(a) we give a simplified example of an operation to produce
products or services which require a series of steps from inputs to outputs
through a convergent procedure. In the figure the transfers into and out of the
organisation involve transactions at the boundary of the organisation.
The idea of boundary to organisation is used in two different ways: firstly,
there is the legal and institutional boundary of the organisation, which is
clearly marked by the usual transactions of purchase and sale of goods and
services; secondly, there are the social and cultural boundaries of the organisa-
tion, which may be considerably fuzzier than the legal or transactional ones.
These social and cultural boundaries are marked by individuals and sub-
groups in an organisation.
In this first example the operating system is located wholly within the
legal boundary of the organisation and the legal boundary is itself wholly
within a country boundary. What is presented in this example is a firm oper-
ating in markets and acting as an independent trader. In Figure 9.Hb). the

FIGURE 9.1(a) A Firm and its Boundaries within a Country

Country boundary
Firm-legal boundary

Operating system boundary

..
~

D-
•
Resource

y-
o.. o..
inputs Assembly Selling products/

o-U 9
~

services

noJ

Inthisexample, theoperating system is within thefirm andthefirm is within a country.

 Anthony J. Berry 141

FIGURE 9.1(b) A Firm and its Boundaries across Countries

Firm-level bounda

Operating system

B
Country boundaries

In this example, the firm operates in five countries, takes in resources from three countries
and sells in three countries. The firm is essentially independent.

example is complicated by relaxing the country constraint and we can see

that the steps of the operation still lie within the legal boundary but now
take place in a number of different countries. Here a third idea of boundary
is introduced, that of the country. An example of this is the Ford Motor
Company in Europe, which operates on an integrated strategy, marketing
and manufacturing process, so that engines might be made in Germany,
Spain or England; the assembly plants might be in Belgium, England or
Spain, or other countries; parts suppliers could be located in any Western
European country or other part of the world, and the product marketed in
all Western European countries and outside it. Indeed it is said that the auto-
motive industry is one of the first truly global networks of provisioning and
selling. Here there are complex issues in the legal control of Ford that cross
the social, cultural, national boundaries, but the system that they are man-
aging , as the final suppliers of vehicles to consumers,includes controlling
the system which produces all of the parts for the assembly lines that they
have established.
142 Control ofembedded operations

In Figure 9.2 the situation is made more complex by showing how the oper-
ating system to produce products and services has been modified by having
the steps in the operation take part in legally separate organisations. In this
latter case the firms or organisations are, to a greater or lesser extent, depend-
ent upon one another for the functioning of the operating system. In this case,
following the terminology of Grabner (1993), we refer to these organisations or
firms as being 'embedded'. So we can see that this simplified example leads us
to consider the problems of the control of operations within independent firms
and within firms embedded in an operating system which includes other
embedded firms . Some writers refer to these kinds of systems as networks,
others as chains. Note that the boundaries which are applied to denote what is
inside the system, network or chain are an arbitrary choice, perhaps based
upon considerations of the relative significance of the contributions of the
actors to the operation in question.
In the case of the embedded firms the operations of focus are inside and
outside the ownership of the boundaries of the specific enterprises. An
example here would be an operation to supply clothes at a retail shop which
involves textile technology firms, dye makers, fashion designers, textile manu-
facturers, makers of trims, suppliers of computer-controlled production
machinery, garment makers and assemblers, merchants, retail houses and so
on working together to design, manufacture, merchandise, market and sell the

FI GURE 9.2 The Embedded Firms

Operating system

--+---1~I--D-~- - - - - - - - - - - ~ ®
1
1 o
r----------- -----------,
I
I-t---.---, 1
I 1
1
'" I
-=C l.-'-If--c}----, 1----- --
I~

-10 ·E
1 v
.5 I I r - -
OJ
~
I Assembly I 1 ~
::s
Q
I I ~
1 I-- -~ ~
'"
OJ
"0
£
~ ...-1----i:J---..J I
1
I
1
1L 1
..J
@

:®
 Anthony J. Berry 143

garments. So, in addition to the legal, cultural and national boundaries, we

note that a fourth idea of boundary arises, that of technology. The relationship
of different technological approaches in organisations would create consider-
able problems for the establishment of embedded chains. To overcome such
difficulties would imply that the organisation was not technologically separate
from other enterprises in the network or chain : for example, the imperatives of
technological integration of production equipment and procedures and, of
great interest for control, electronic data interchange will enter the operations
of anyone enterprise. Clearly, then, in any complex operational chain to
produce goods and services which is typical of most advanced manufacturing
countries, most of the operating systems that are established and managed
involve transactions across all of these four boundaries.

Control in embedded networks

Two primary modes of control, direct control, that is direct personal supervi-
sion of those doing the work, and indirect control, that is supervision of work
through mediating layers of organisation, are familiar. However in both,
quite simple, organisations (see Box 9.1) direct control is difficult and, in
complex organisations, direct supervision and oversight is hardly possible.
Furthermore, rule-based control is difficult to establish and sustain in organ-
isations which have high degrees of variety, both in terms of required skills
and knowledge, geographical locations and technologies, and hence is clearly
problematic in a firm in an embedded chain . Thus the control of the embed-
ded networks of complex organisations becomes of interest.

Box 9.1 The African bus driver

An African bus driver in what is now Zambia had a bus route that
meant his driving a bus away from his depot one day, and driving it
back the following day. Sometimes a problem arose in collecting the
ticket money. On a particular day, the inspector who had been given by
the driver the proceeds of one of these two-day trips was rather suspi-
cious of the small amount of money that had been handed in. He asked
the driver, 'Is that all?' and the driver said, 'Yes'. The very large inspec-
tor then picked the driver up, turned him upside down and shook him,
and a great deal more money rattled out of the driver's pockets, clothes
and so on, over the floor . When all this was scooped up they came to an
equable agreement about how much of this belonged to the bus
company and how much belonged to the driver.
144 Control ofembedded operations

It appears that three different problems arise, both in the case of the operat-
ing systems in the independent firms and within the chain of embedded firms .
Firstly, there is the problem of deciding which organisations will enter the
chain. Secondly, there is the problem of establishing the operating system
from which role, work demands, the nature of work and work organisation
are derived (and which in turn serve as an experience from which changes
might be managed). A common approach to managing this process can be
found in the literature of project management, where the actors negotiate and
agree to a programme for establishing the new system. Thirdly, there is the
ongoing management of the now established operating system.
In order to begin exploring these problems we first consider a process of evo-
lution from a 'firm' operation through a procedure of decentralisation (alterna-
tive hierarchic control) to one which is part of an embedded chain either
through contracting out or through franchising. Operational control of complex
and wide-ranging activities engender difficulties in reporting action and imple-
menting decisions both because of the time lags and because of the degree of
unexpected variety which local managers must handle. This latter point is
often used as an argument for decentralisation, in order to make a contract for
responsibility and accountability within the organisation, a form of control
through hierarchies and rules, in order to minimise the variety that senior man-
agers need to handle. Of course decentralisation as a form of differentiation
designed to cope with external complexity and uncertainty leads to a need for
new modes of integration between the decentralised units; for example, new
rules emerge for transfer of products and services. When this becomes too
difficult to manage (because of significance, the nature of interdependence and
of complexity and uncertainty) then quasi-market related controls are allowed,
in various forms, to enter the internal world of the organisation. Should these
still leave control problems, the troublesome activities can be exported from
the organisation in two primary ways: subcontracting for products or services
instead of providing them internally, (and perhaps permitting a management
buy-out to create the subcontractor) and franchising (for example, retail
outlets) or selling off parts of operations instead of owning them.
Here we note that complexity and uncertainty of operations within the inde-
pendent firm are seen to lead to variations of the control structures and proce-
dures through changes in hierarchy and rules and forms of markets. These are
not only matters from the technical system, for the issues and problems associ-
ated with motivation, identity, belonging and loyalty are used as arguments
for franchising, where there is a procedure to regain control at the boundary
by placing the ex-employee, now franchisee, firmly outside the organisational
boundary. Here the problems of motivation are apparently solved by ensuring
that the franchisee's self-interest is now aligned with the economic goals of
the franchiser. Such a practice has occurred within the domestic bread and
milk delivery business in the UK. In general we can see that organisational
and environmental complexity and uncertainty can lead independent firms to
 Anthony J. Berry 145

develop approaches to control which can create our second case of the embed-
ded firms in a 'chain' operating system. Of course these two examples - that of
the independent firm and that of the embedded firm - are two ideal types
which mark a continuum. It is probable that firms will operate in a close
approximation to these. It is likely however that some firms will operate in a
more hybrid mode, with some operations largely within their control and
some embedded in networks or chains.
The imperatives which create such very complex chains of systems making
up an operating system include the need to cope with behaviour of competi-
tive actors in product markets. The need to cope with market uncertainty and
volatility requires the establishment of operating systems which can be
flexible in response to external environmental changes (particularly product
market changes). However it is important that this whole operating system
exhibits some characteristics of stability and predictability; that is, that the
manner in which the linked organisations can work to actually produce goods
and services for the market-place has a reasonable degree of likelihood of
operating within the parameters that managers and actors in the system have
established. Here we rediscover one of the eternal tensions in the design of
control procedures, that which lies between the wish for stability and the wish
to cope with change.
In the literature we notice the prevalence of this theme. For example, the
system dynamics work of Forester (1961) demonstrated that if systems in a
chain operate independently then dynamic instabilities may arise. This point
is nicely illustrated by the brewery example in Senge (1993) where the inter-
play of retailers, responding on their own time scales to perceived growth in
demand and perceived supply shortages by overordering in the expectation of
rationing, led to huge overproduction. One solution to this kind of predica-
ment when operating systems are open to discrete decisions is to include
significant amounts of redundancy to prevent instabilities. This solution,
which is to have excess operating capacity and to hold stocks of goods and
services at intermediate stages in the chain, may be of lower cost than that
delivered by large instabilities but is more expensive than the embedded firms
operating the whole system or network in concert to minimise capacity
throughout the whole system and stockholding at all stages, or perhaps
persuading the consumers to hold stocks instead.
The broad question arises as to what modes of control might be applied to
the three stages identified earlier, that is the membership, establishment and
operation of these networks or chains. Here we suggest three 'ideal' types as a
basis for discussion. These are control through (a) dominance, (b) collabora-
tion and (c) competition.
By dominance we mean that the most powerful player in some part of the
operating system is enabled to ensure compliance to his or her requirements
by the relative weakness of the others. It is characteristic of many advanced
manufacturing systems with many embedded firms that such complex opera-
146 Control ofembedded operations

tions are dominated by one of the actors. This domination exists in different
forms, depending upon what power may be exercised by the various actors.
For example, the 1980s saw system domination being exercised by retailers, as
in the case of much clothing and foodstuffs in the UK. The domination was
created by capture of retail channels through the creation of supermarkets,
with lower costs of operation (reducing employment in retailing), provision of
a wide range of goods in one place (customer convenience) and cost efficiency
in purchasing and distribution. But domination elsewhere is in the hands of
manufacturers, as in the case of motor cars, where manufacturers have consid-
erable control over the retail operations through direct ownership but more
commonly through tight franchising arrangements. In this industry such dom-
inance has led to embedded suppliers being required to 'open' their internal
accounts to the dominant actor. Further power may be the property of compo-
nents suppliers, as perhaps in the case of certain computers where microchip
technology and supply constrain downstream activities.
By collaboration we mean the joint control of transorganisational networks.
An example here would be the development of aircraft, flight systems and
engine technologies, where close collaboration between aircraft, component,
engine manufacturers and air forces and airline companies is essential in order
not only to design but also to operate embedded systems.
By competition, we mean that, by the day-to-day, week-to-week interplay of
markets and competition, suppliers are able to supply and buyers to buy. In
order to explore these types we need to consider the nature of the interdepend-
ence between the firms in the embedded chain. Before taking up that point it is
necessary to consider the fact of differing time scales. It seems reasonable to
argue that the time scales for the operation of an embedded system would
depend upon the critical path of the system and would also be relatively short,
of the order of months. In contrast it is argued that the time scales for negotia-
tion of membership and establishment of such systems are relatively long and
of the order of years, because of issues of materials, design and production
technologies. By extension it is argued that the time scales for modification of
such systems would be of intermediate duration. Following Miller (1976) we
adopt a taxonomy of interdependence as follows: serial, that is activity A pre-
cedes activity B and so on; mutual, where activities A and B take place jointly
and perhaps lead to activity C; reciprocal, where activity A and B are interac-
tively related. In relation to an established embedded operating system we
expect to find the interdependence of the embedded firms to be serial or
mutual while, in relation to the process of negotiating and establishing such
an operation, we expect the interdependence to be mutual or reciprocal.
Therefore, applying these ideas to an established embedded operating
system, we expect the time scales to be short and the dependence to be serial or
mutual and the form of control to lie in either dominance or collaboration. The
competition mode of control is the least likely, for it supposes not only that the
rationale for setting up the embedded chain was incorrect but that short-term
 Anthony J. Berry 147

switching of suppliers in such complex systems is feasible. Indeed it is possible

to claim that for such a competitive mode to exist the operating systems must
be simple and made up of available parts from alternative suppliers.
Extending these considerations to the processes of negotiating membership
of complex chains we expect the time scales to be long and the interdepend-
ence mutual, and hence we expect the first mode of control to be competition;
that is, organisations would bid and negotiate to take a role in a systemic
chain. Then, after the partners have come together, the likely mode of control
in the process of establishing such chains would shift to collaboration. Even so
the most powerful actor in the chain could exercise a decisive influence in
managing the market competition, perhaps by setting the criteria (such as
quality or resources) by which others might be permitted to enter the competi-
tion. Where the time scales are long and interdependence in the design or cre-
ation of the chain is reciprocal, we would expect to find a collaborative mode
of control. Here we expect to see joint working groups to ensure that the best
compromises are reached. It is unlikely that such joint groups of embedded
firms would then set up a competitive mode of control for part of this work.
Of course in this argument we posit tendencies of control modes following a
certain type. It is possible that, in any such chain of systems, all three modes of
control will actually be observed to a greater or lesser extent in both the estab-
lishment and operation of the chains. The argument so far is to try to under-
stand how the logic of the tasks of the embedded systems leads to mode of
control. We need to also ask how the social and cultural processes in these
embedded systems affect the control procedures, a question which will be
taken up in the fourth section of this chapter.
A counterpoint to what some might see as the despotism of the powerful
actor in the chain, effectively using power to gain the compliant behaviour of
other embedded firms by essentially tying them in to a specified plan for the
operation, is to work from the idea that three operations may be being
managed at the same time. These of course are those referred to earlier in this
chapter: negotiating a role, establishing an operating system and running it.
The logic of the needs of the 'establishing' system are for cooperation and
exchange such that the system established is the best overall compromise; the
'partners' in the chain need each other's energy and creativity and a willing-
ness to understand the consequences of their actions for each other. In contrast
the running of the operating system requires the embedded firms to stay with
the plan. The paradox here of partnership in one mode and power in another
leads to the idea of loosely coupled systems, which enables alternative logics
to work in the network while the overall system has the requisite degree of
stability and potential flexibility and renewal. Much writing on manufacturing
operations is based on such a systemic logic; that is, the procedures for mater-
ials requirement programming and 'just-in-time' delivery arrangements have
been established as modes of ensuring the operating integrity and efficiency of
the system chains created. It is now clear that as several modes of control can
148 Control of embedded operations

be observed in anyone embedded network. This explains why some authors

write about control as being 'tight-loose', as they note the paradox of appar-
ently conflicting control modes used around an operating system.

Financial control

Normally financial control is fitted into this world as a way of constructing

statements of the resource inputs and tracing the way in which those resource
inputs are transformed into outputs; much of the earlier discussion in Chapter
6 has addressed these questions. Within the system as established, the cost
management models of either full-absorption costing or activity-based costing
might be used. However it is clear that the idea of activity-based accounting
aims to achieve a closer approximation (because they are descriptive models) to
modelling the operations themselves. One might note the similarity of the idea
of activity-based accounting to the ideas of programming, planning and bud-
geting, popular in accounting and management literature of the early 1970s.
The financial management and control of such embedded chains replicate the
problems found in divisionalised organisations which have adopted investment
centre control and varieties of transfer pricing policies for aiding the manage-
ment of the interdependent operations across the divisions. All of the familiar
problems of optimisation and suboptimisation reappear, with the added com-
plexity of legal independence and perhaps a struggle to ensure that the eco-
nomic returns to the participating firms are appropriate, however that may be
determined. In the nature of collaboration of embedded firms one would expect
to find specificity of assets for particular purpose, product or service specificity
and probably knowledge and skill specificity in relation to the particular oper-
ating system . (This is not to argue that each firm is only involved in one such
operating system, though such multiplicity does add complexity.)
If the appropriate control mode is that of dominance we expect the dom-
inant actor to require openness of costing and the adoption of costing systems
that ensure that the dominant actor can assess the cost structure, through rel-
evant future time periods, of the whole operating system. To manage such a
system the dominant actor must have some knowledge of the relationship of
fixed and marginal costs as a function of time and activity volume of the
whole system as well as an estimate of the way cost reduction programmes
(all other matters being equal) can be designed and implemented by all of the
partners. In this way strategic direction can be given, provided there is ad-
equate understanding of market conditions over the requisite time scales. Of
course the detailed arrangements would need to reflect the likely volatility in
markets, so one would expect, as indeed there exists in the automotive supply
system, a series of contract arrangements which permit variations in call off of
volume and category of elements over appropriate time scales. In short it is
 Anthony J. Berry 149

possible to have flexible arrangements in the embedded chain even though

these are established through asymmetric power relations.
Even if the appropriate control mode is collaboration we expect to find exactly
the same management problems and the same need for cost data to be shared
among the partners (which could include parallel providers of similar elements,
that is multiple sourcing) to enable them to manage the operating system in the
light of expected market conditions. Of course the likelihood of stable symmet-
ric power relations is low, but there is no particular reason to doubt that a
shared process can be applied, though skill in negotiation may create ripples.
The particular accounting arrangements for an embedded network are a
matter of great concern. There is evidence from recent research that the
accounting arrangements in large divisionalised organisations are designed to
support financial reporting rather than management accounting and control.
While matrix organisation structures have been developed within organisa-
tion, Jones and his colleagues (1993) reported:
information overload, domination by the need for consolidation, the irrelevance of
group information for operating subsidiaries, e.g. needs of financial accounts dom-
inate the managerial accounting needs and control information dominates decision
and action information.

Such research suggests that, if the problems of intra organisational networks

are not readily addressed, those of interorganisational networks would be
even more difficult. This point leads to the suggestion that a derivative of
activity-based costing could be the basis for establishing the financial control
structure for an embedded operating system, for it would follow clearly the
character of the system itself. Whether the partners would be able to reconcile
ABC for the system with their other accounting procedures would raise
difficult technical questions. Note, however, that ABC is suitable for planning
and reporting, but less appropriate for the decision issues, noted earlier, in the
management of the system.
From this brief discussion of cost management we note that there are many
other issues in the provision of elements of such operating systems. These
would include quality, delivery agreements and compliance. Equally it is clear
enough that the 'operation' to establish an operating system would need a dif-
ferent cost management procedure (more like a project management process
focusing on effectiveness) from that of the operating system itself (more of a
focus upon efficiency).

Embedded operations and clan control

The control modes of dominance, collaboration and competition can be related

to Ouchi's constructs of hierarchies and rules and market-based control, which
150 Control of embedded operations

have been used in other chapters of this book as an heuristic framework. So far
in this chapter collaboration has been addressed in a technical manner. Little
has been said about the issue of clan based control. To illustrate the significance
of these factors we consider the findings of a research project in the American
automotive component supply industry, where alliances were made between
US and Japanese companies. The researchers reported that the Americans con-
cluded that they had learned nothing from the collaboration, while the
Japanese had different views. However the Japanese were puzzled at the atti-
tudes and behaviour of the US participants. It seems that the clan-based control
with which the Japanese were familiar, based upon open discussion and
sharing of problems, knowledge and data, was simply unavailable to the US
participants, who expected a rules- or market-based (competition) control.
Hence we observe the significance of the dynamics of competition and collabo-
ration, for it would appear that the embedded firms of the Anglo-Saxon world
expect power and dominance, rules and markets while the Asian organisations
naturally seek to create a significant degree of clan based control where collabo-
ration would exist both in the tasks and in the social and cultural systems.
Before jumping to familiar conclusions about differences in cultures, we
may note that potential actors in the embedded firms must have the ability to
contribute to the required reciprocal or mutual dependence. If potential firms
cannot do so, then they are treated as accidental contractors and subject to
market-based control. This idea expla ins why observers of Japanese manufac-
turing systems find complex embedded chains existing with some sweat shop
contractors on their edges.
In these complex manufacturing chains, for example aircraft, cat food and
cars, it may be necessary to create transorganisational workgroups so that the
effective modes of collaboration can take place. So the paradox of the complex
capitalist system manufacturing chains emerges. Capitalism is supposed to be
about competition and we have shown how competition enters into the
processes of negotiating for membership of embedded chains. However the
establishment and operation of the embedded chains requires high degrees of
collaboration through the working together of people to create the context for
other participants in the organisations. This implies that there is an imperative
to work at issues of mutualities and interdependencies both through contracts
and through the notion of trust.
What comes into focus then is the need for apparently independent firms to
work together in the operating system, or what some have called the 'virtual
organisation' or, less helpfully, the 'boundaryless organisation'. Particular
problems arise when we consider the issue of control across boundaries,
whether they are internal or external to an organisation. We can begin to
examine some of the difficulties with the example of a travelling salesman
employed in an independent firm. The salesman spends most of his time at the
physical and geographic boundaries of the operating systems of the enterprise.
Such a person is in effect continually negotiating organisational boundaries.
 Anthony J. Berry 151

The problem that arises is that the task boundaries of the operating systems
may become changed by the salesman's negotiation of special terms and deliv-
ery schedules disturbing the previous programmes and requiring consequent
ad hoc adjustments through the personal networks of the organisation.
Perhaps more importantly this simple example cause us to acknowledge the
significance of the social and cultural boundaries of the operating systems.
Note how these are modified or even blurred by the fact that such salesper-
sons often find a greater sense of identity with the clients than they do with
major parts of their employing organisation. As a consequence organisations
have to introduce procedures to ensure that the salesperson's identity as an
employee of the firm is continually reinforced. This problem is also encoun-
tered in governmental services, especially in the foreign services, where
ambassadors might come to be more sympathetic to their accredited govern-
ments than to their own state. These problems arise with equal sharpness for
the people who work in such embedded firms , for they now have to handle
the conflicts arising from alternative modes, collaboration and competition, at
the boundary of their firms. Of course this problem is as old as human experi-
ence but that marks it as difficult rather than trivial. A full discussion of the
manner in which we might understand how to handle these issues is to be
found in the extensive literature on boundary spanning roles.
In addition to the personal issues of boundary spanning it is important to
note that the culture of partners in the chains may clash in subtle as well as
more blatant ways. To assume that, because the technical characteristics of
operating systems can be created and run satisfactorily, such systems will
deliver expected benefits is to ignore the clan-based problems. Note how
matrix forms of organisation were invented for divisionalised independent
organisations as a means of creating coherence in the operating systems and
dampening conflict among the divisions, yet it is clear that multinational firms
still find that cultural issues pose major problems of organisational functioning,
as the extensive literature on international management demonstrates. The cul-
tural problems found in the example of the Japanese-American alliances
reported earlier have been repeated many times as is shown by the extensive
literature on joint ventures and strategic alliances. Indeed it is arguable that the
Anglo-Saxons and the Asians are equally able, from an assumption of long-
term self-interests, to see that an embedded system is appropriate, but it is
more likely that the Asians, with their greater disposition to co-operation, will
actually work the micro adjustment processes to realise the long-term benefits.

Systems do not always work

We have examined the control problems that arise in what we have called an
embedded operating system. We have shown how the nature of the interde-
152 Control of embedded operations

pendence, together with consideration of time spans, affects the choice of

control mode. For operations which are dedicated to the negotiation and
establishing of embedded operating systems we suggest that the control
modes would be competition and collaboration. For the operation so created
we suggest that the likely control modes are collaboration and dominance.
In the case of the embedded operating system it appears that the control of
the system must lie within it, for, although there may be asymmetric power
relations or a dominant partner, there is no superordinate authority. This
interesting and difficult state leads to a number of questions. Firstly, how do
we create a control system that actually has the capacity to serve the partner's
needs to learn about the inside of the system in relation to the outside? Neither
the traditional ideas of financial control nor activity-based costing approaches
can solve this problem. Neither have we done more than show how the
broader approach to control in this chapter and in this book may provide a
means of thinking about the parameters of the issues of control of embedded
networks. A second question arises as to how control procedures can be
designed to prevent the embedded operating systems becoming sclerotic
when the actors cannot find the energy to recreate such complexity. Can the
partners find ways of working away from the defensive routines that exist
within organisations (for example, secretiveness, personal fiefdoms, the lust
for power, conflict of ambitions or narcissistic leadership) and across organisa-
tions (fear and anxieties, failure of generosity, blaming the partner, using one
system operation as a bargaining counter for another, failure in understanding
the otherness of the partners, and so on)? The significance of the unaddressed
and unmodified defensive routines is that they lead to an immobilisation of
needed change and then to an inevitable shattering of trust, which in turn will
shatter the embedded system operations.

Summary

In this chapter we have shown how the control problems of complex embed-
ded systems follow from the tasks interdependence and time frames involved.
We have shown that different control modes come into play at different stages
in the negotiation, establishment and operation of such embedded chains. In
drawing attention to the implications for the financial control of such chains
we noted that we had recognised the limitations of current practice and
suggested a broader framework for taking matters further . Finally, having
developed the discussion from a technical standpoint, we argued that the
major control problems for embedded chains would lie in what Ouchi termed
'clan control', that is in the social and cultural processes within and between
the partners in such chains.
 CHAPTER 10

Economics and control

Willie Seal

Introduction

In Chapter 3, the 'market' approach to organisation was discussed in the

context of its pervasive influence on policy making in areas which were hith-
erto regarded as sancrosanct parts of the 'public sector' and thus organised
via a combination of political, bureaucratic and professional controls. Indeed
the policy influence of the 'subcontracting model' has become so dominant
that it is sometimes hard to see how the drive towards marketisation and con-
tracting out can be resisted! The aim of this chapter is to set out the economics
approach to control from a critical perspective. It is suggested that the superi-
ority of the market over hierarchy as the mode of allocation is not as clear-cut
as some policy makers seem to think. Furthermore it is quite possible that
forms of internal organisation with subtle forms of control are being replaced
by transactionally inferior governance structures.

Control in economics

'Control' in mainstream economics is most usually seen as something which is

exercised from 'outside' the economic system, generally by the government.
Thus there are discussions on the impact of, for example, price and wage con-
trols on inflation and unemployment, and on the effect of attempts to control
public expenditure. Significantly 'control' is rarely seen as being an endoge-
nous feature of economic systems. There are a number of plausible explana-
tions for this neglect. From a technical point of view, mainstream analysis has
focused on exchanges between individuals acting through equilibriating
markets. Thus, under strict assumptions, general equilibrium theory has

153
154 Economics and control

shown that this can lead to an optimal allocation of resources (Bator, 1957)
with the consequence that other controls are unnecessary.
Other explanations of this neglect of control recognise historical and ideo-
logical factors. Historically, the beginnings of the subject (at least in the
English-speaking world) are associated with Adam Smith's famous doctrine of
the 'invisible hand'. Thus the problematic of classical through to neoclassical
economics was the analysis of the way goods and factors of production which
were allocated through markets were guided by the 'invisible hand' of the
price system. The 'visible hand' which allocated resources within organisa-
tions such as firms was largely ignored. In as much as there was an 'economics
of the firm', it focused on the way the profit-maximising firm's price and
output were affected by the interplay between external market factors (such as
the competitive structure of the industry) and a technologically derived pro-
duction function (which determined costs) .
The internal allocation of the firm's resources (the main focus of the man-
agement control and accounting literatures) was reduced to a number of
simple decision rules such as equating marginal revenue and marginal cost or
choosing projects with a positive net present value. Somewhat paradoxically
these optimising rules have greatly influenced both management accounting
and, as was argued in Chapter 2, management control literatures. This
paradox of neglect and influence can be resolved by drawing on the distinc-
tion made in Chapter 1 between systematic and systemic approaches.
Economics offers through its normative, microeconomic paradigm, a number
of enticing decision rules which aid the development of systematic control.
However, in its systemic version, it emphasises the superiority of decentralisa-
tion allocation via markets rather than internal organisation. Thus (with the
obvious exception of the markets and hierarchies paradigm) the word 'control'
is bracketed with other essentially 'non-economic' words such as 'organisa-
tion', 'hierarchy', 'authority' and 'process' dealt with by other disciplines.
The preoccupation with atomistic, self-equilibriating systems has been criti-
cised at both macro and micro levels. At the macro level, the basis of Keynes's
critique of the 'classical model' was that long-term disequilibrium could lead
to 'involuntary unemployment'. Interestingly, in view of the earlier discussion
of systems and cybernetics in the control literature, more modern reinterpreta-
tions emphasise 'disequilibrium' - the breakdown of a price system when
decisions are based on non-equilibrium prices - in other words, the essence of
the Keynesian critique was not that unemployment could exist in equilibrium
but rather that the price system could send the wrong signals (Leijonhufvud,
1968). Even this somewhat limited perspective on control has been on the
retreat at both theoretical and policy-making levels. The theoretical riposte to
Keynes from economists has rested on 'rational expectations' and the 'new
classical' model. At the policy level, these ideas have spawned various forms
of monetarism.
 Willie Seal 155

Many modern Keynesians would have little difficulty in perceiving the

economy as a 'cybernetic system' with all the associated feedback processes.
However, as we shall see below, this perception of 'economics as a process'
has fewer adherents at the micro level. Perhaps surprising in the light of their
laissez-faire predilections, much of the recent criticism of the orthodox micro-
economics, especially the so-called 'theory of the firm,' has come from the so-
called 'new institutionalist' or 'contractual' economists. Contractualists reject
what they would term the 'engineering' model of the firm which seems to
emerge from the industrial economics literature. In their view, the production
function is not technologically determined. Indeed contractual perspectives
see it as being imperfectly specified as the result of informational asymme-
tries, bounded rationality, opportunism and other lacunae of the contractual
world introduced in Chapter 3.
There are two main versions of contractual theory - agency theory and
transaction cost economics.' From a systemic perspective, the agency vision is
of an economy as a network of interlocking contracts (Jensen and Meckling,
1976). As we have seen in Chapter 3, the transaction cost paradigm sees the
economy as a mixture of 'markets and hierarchies' . We may compare the two
approaches with simple examples. To illustrate the markets and hierarchies
(or transaction cost economics) approach consider a common managerial
problem, the 'make-or-buy' decision on a component. In management
accounting terms, the choice may be based on a simple comparison between
the unit cost of manufacturing the component 'in-house' compared with the
unit cost of buying the component from a subcontractor. The transaction cost
economics (TCE) approach would put this decision in a wider and more com-
prehensive framework. For example, does the production of the component
require investment in specialised machinery or specialised training for
workers? If it does, then we have a situation of asset specificity. This raises the
possibility that an external contract may be difficult to draw up and enforce
because small numbers bargaining may be hindered by opportunistic behav-
iour by one or both parties. For this reason, it may be 'cheaper' to produce the
component 'in-house' because of contractual rather than engineering problems.
In other words, it is less costly to use a hierarchy than a market.
The agency problem can be illustrated by the situation facing the owner of a
holiday bungalow who lives a long way away. The owner only uses the bun-
galow herself for a few weeks a year. For the rest of the time she rents it out to
other tenants or holiday makers. The obvious solution is for her (the principal)
to appoint a local agent to handle these sublets. The question is now, how can
the agent be motivated and monitored? Probably the owner will decide to
allow the agent a percentage of the rents. She will also expect the agent to send
her regular accounts recording income and expenditure. She is unlikely to
request that these accounts be independently audited but it is easy to see how
agency theorists have used this line of reasoning to explain the spontaneous
156 Economics andcontrol

development of financial accounting and auditing in more complex agency

relationshipsf
These two examples are admittedly in the 'Robinson Crusoe' tradition of
economic reasoning. What they do not capture are some of the underlying
behavioural assumptions and the sometimes contentious policy implications
which have been drawn from these two approaches. We therefore need to take
a more detailed and critical look at these issues.

Some problems with the 'economics' approach to control

One theoretical interpretation of the push for marketisation is that the agency,
rather than the transactions cost, approach has triumphed! To agency theor-
ists, the aim is to find the 'best' contractual arrangement - internal organisa-
tion is not an option because, in theory, it does not exist. The firm is a 'nexus
of contracts' and 'a legal myth' (Jensen and Meckling, 1976). In the context of
the issues discussed in Chapter 3, a control structure designed by an agency
theory-influenced consultant would be different from a system designed by a
transaction cost economics-influenced consultant, particularly one familiar
with the works of Ouchi and others.
The equivalent issue in the private sector is the long running debate begun
by Coase (1937) on the reason for the existence and boundary of the firm . Once
again, as the discussion in Chapter 3 illustrates, this debate has influenced
much of the recent management literature, whether it is 'sticking to the knit-
ting' (Peters and Waterman, 1982) or divestment and management buyouts
(Thompson and Wright, 1988). Indeed, rather than putting the policy issue in
terms of 'private sector versus public sector' - the common currency of the
popular debate - the question may be couched in the more general one of
'market versus non- market forms' of organisation. This problem has to be
answered in the private sector as well as the public sector but, as we shall see
below, the forces influencing the 'choice' of contractual arrangements are quite
different from those operating in the political sphere.

The limits to contracting

Williamson has always argued that a prescriptive application of transaction

cost economics involves a comparative institutional methodology (see, for
example, Williamson, 1985). Secondly, in their explanatory version of institu-
tional change and development, both the transactions cost and agency para-
digms draw on versions of natural selection theory. Thirdly, in contrast to
agency theorists, Williamson has acknowledged and defended the role of the
 Willie Seal 157

authority relationship in effecting transactional efficiency. We now turn to a crit-

ical examination of each of these issues, after which we will see how contrac-
tual theory may be applied to organisational choice problems.

Alternative types of organisation with different transaction cost characteristics

Transaction cost economics suggest a number of alternative governance struc-

tures, with spot market exchange at one end of the spectrum and hierarchy at
the other. Between these extremes lie other forms such as franchising, long-
term contracts, 'inside contracting' and quasi-vertical integration. We have
already seen that there are other organisational forms such as the 'clan'. We
may add various forms of participatory models. These may range from the
pure labour-managed firm to the informal forms of participation associated
with the Japanese firm (the J-firm - see especially Aoki, 1984, 1986, 1990a,
1990b).
Labour-managed firms are usually dismissed by both transaction cost and
agency theorists on theoretical and empirical grounds (Jensen and Meckling,
1979). In view of the immense success of Japanese industry and the numerous
attempts to copy features of Japanese organisation, the J-firm cannot be so
easily ignored. A variety of governance structures is also characteristic of most
'public sector' institutions. If we consider the forms of organisation in the
public sector, it would be a gross error to characterise them as being simple
hierarchies. Indeed the danger of contracting out is that the subtle forms of
social and professional control are replaced by contractors who actually do use
primitive hierarchical forms of organisation.
The comparative advantage of different forms of non-market governance
structures depends on contingency factors such as the task or the level of
uncertainty in the environment. For example, in terms of technological devel-
opment, the clan plays a crucial role since it is in activities such as research
and development that performance ambiguity is likely to be high and formal
output controls are not only ineffective but actually dysfunctional.
Furthermore if, as is often argued, technological progress is as much reliant on
the ideas of ordinary workers as it is the responsibility of the designated
researcher then the innovatory advantages of social arrangements are further
augmented.
Other models of the firm offer a superior adaptive response to environmen-
tal uncertainty compared with the simple hierarchy. For example, Aoki's
models of the firm emphasise that good horizontal communication between
functional areas is more important than the vertical communication up and
down the conventional hierarchy in dealing with emergent problems (Aoki,
1986). His approach emphasises participatory or bargaining models of the firm
which seem to make fewer normative demands than the clan. This organisa-
tional form is more explicable in terms of the self-interest rather than self-
158 Economics and control

abnegation of the worker (Aoki, 1984, 1990a, 1990b). Indeed, in contrast with
the classical capitalist firm, co-operation is based on a more equal constitu-
tional footing between shareholders and employees, with the management
acting as a sort of mediator. The firm is seen as a 'nexus of treaties' rather than
a 'nexus of contracts' (Aoki, 1990a).
Non-economists often criticise economists for basing all their theories on a
narrow version of self-interest. A less common criticism is that they tend to
argue that people are the same, whatever the organisational environment.
Thus, if behaviour does vary between different institutional settings (such as
between the public and private sectors), it is because the cost and reward struc-
ture differs, not because people are 'different' and/or 'better or worse'.
Although there are good methodological reasons for this assumption (without
it, any behaviour could be explained by the 'people are different' argument), it
ignores the potential benefits generated by governance structures because they
actually seek to 'make people different/better', For example, a 'clan' would be
relatively uninteresting if it was simply a by-product of a specifically Japanese
culture and not, as Ouchi argues, a type of firm or organisation that can be
created through socialisation (Ouchi, 1980, 1981).

Can we be sure that the best contractual modes are selected?

A second criticism of the new 'institutionalist economics', although it is gener-

ally more applicable to agency that to transaction cost theory (Putterman,
1986) is the assumption that 'efficient' organisational forms will be selected
through a form of natural selection. The use of biological analogies in social
science has attracted a number of general and specific criticisms.
Methodologically there is a danger that survival becomes its own
justification - that the 'survival of the fittest' is a tautology (see Langlois, 1986).
In response to this criticism, Jensen (1983) argues tautologies may be useful
heuristic devices in the development of refutable theories. A second charge
against natural selection theories is that they are 'Panglossian' in that they
tend to support the status quo as the 'best of all possible worlds' (Tinker,
1988).This charge may be rebutted on both theoretical and empirical grounds.
In the specific context of management control, a firm may face an environment
where conflicting pressures induce contradictory practices (Seal, 1993). For
example, firms such as Hanson or GEC that have been successful predators in
the market for control do not develop management styles which enhance
success in the product market (Goold and Campbell, 1987a, b). Organisational,
as with biological, adaptation is a compromise meaning that survival traits in
one part of the environment may weaken the firm's ability to compete in
another area. Thus survival does not imply optimal design. As well as
conflicting pressures, institutional development is a product of path dependen-
cies. Thus organisational forms which might be optimal in one historical
 Willie Seal 159

period may have been wiped out in a previous period when conditions were
different (Nelson and Winter, 1982).
In his critique of competitive selection, Dow (1987) raises the issues of
power and appropriability. We will be looking at power in the next section.
The issue of appropriability is that 'selection can .. . be expected to operate on
firms as a function of their profitability in a given market environment' (Dow,
1987, pp. 31-2). Dow makes the point that the survival of a firm in a capitalist
economy depends upon the ability of the entrepreneur to capture the benefits
of a particular governance structure in the form of profits. Thus , while a non-
capitalist governance structure such as a labour-managed firm may generate
larger transaction cost savings than the classical capitalist firm, these benefits
will be diffused throughout the workforce. The advantage (from a survival
point of view) of the capitalist firm is that the '[smaller] benefits of control by
capital can be concentrated quite easily in the hands of a single agent, by
bringing all physical capital under the umbrella of common ownership' (Dow,
1987, p. 32).
Although it is possible to rebut general criticisms of selection theory, it is
easy to understand why it has such a tainted reputation in the social sciences.
Indeed the problem with so much of the 'new institutionalist' literature is that,
although writers such as Williamson, Jensen, Meckling and Fama may be able
to defend their use of selection theory against general criticisms, it is relatively
easy to find specific examples of tautological and Panglossian reasoning in
their work. More damagingly at the policy level, crass invocation of selection
theory by politicians and consultants may be used to provide support for the
fashion for the application of 'business methods' in the public sector, arguing
that, if they have withstood competitive pressures in the private sector, they
must possess some universally optimal properties.

Authority, power and efficiency: the labour process debate

The transaction cost paradigm unambiguously espouses the efficiency-

enhancing properties of the authority relationship and is quite prepared to
engage in debate with the labour process school. It recognises that the need for
management control is a direct consequence of the assumptions of costly
information and bounded rationality. In contrast the neoclassical tradition is
coy about the role of authority and power. For example, Alchian and Demsetz
(1972) even suggest in a renowned passage that the employer has no more
power over the employee than the grocer has over a customer. Perhaps more
surprisingly, neoclassical theory also lacks an explicit theory of ownership.
Ownership, at least in popular interpretations of capitalism, is usually in-
extricably linked to both power and authority.
Putterman (1988) argues that a theory of ownership emerges once the sim-
plicities of the neoclassical world are left behind. The neoclassical firm is a
160 Economicsand control

'production function to which a maximand profit has been assigned' and

argues that the 'owner' of such a firm possesses the right to select the produc-
tion programme and appropriate the residual (p. 247). There is no need for
this person to actually own a resource such as capital. In the new institutional
economics, the owner of the firm 'must do more than simply determine a pro-
duction programme and collect the profits; he must also manage'. Putterman
criticises much of the economics literture because it tends to 'obscure a central
aspect of the nature of production organisation in the capitalist economy; pre-
cisely, the ownability of the entity that undertakes the task of organisation,
and the exclusion of factors hired by that entity from control and residual
rights' (Putterman, 1988, p . 256).
The role of authority is an area of transaction cost economics that has
attracted much critical attention from radical economists. Some have argued
that hierarchies are more to do with power than with economic efficiency
(Marglin, 1974). The well-known 'putting-out' debate revolves around
whether factory production replaced cottage industry for efficiency or power
reasons. Rather than being a transaction cost solution to problems of cheating,
embezzlement and high stocks, Marglin argues that 'the success of the factory,
as well as its inspiration, was the substitution of capitalists' for workers'
control of the production process' 0974, p. 46).
Dow, another critic of the transaction cost paradigm, focuses on the function
of authority in transaction cost economics, although its function is supposedly
to 'restrain the opportunism which would otherwise infect market exchange'
(Dow, 1987, p . 19). What is there to control the opportunism of the higher
levels of authority? Rather than restricting themselves to controlling oppor-
tunism among subordinates, Dow observes that 'agents holding positions of
authority might use data obtained through internal audits to gain strategic
advantages over lower level parties, use fiat to settle disputes in ways to suit
themselves, or impose self-serving incentive systems' 0987, p . 20).

Some policy implications

In the light of recent megafrauds in the UK and the increasing divergence

between corporate performance and director remuneration, Dow's observa-
tions on managerial opportunism seem particularly pertinent. In other areas,
the bias of the transaction cost paradigm is less clear-cut. In view of the repu-
tation that transaction cost economics has in the minds of radicals, it is some-
what ironic that it may actually be used to defend hierarchies against the
contracting out model. The contemporary relevance of this debate is that,
paradoxically, most of the modern debate about the boundary of the firm (or
the scope of the 'public sector') is concerned with institutional changes in the
opposite direction - away from hierarchy towards 'contracting out'!
 Willie Seal 161

There are differences, of course. The contracts are not generally going to
independent artisans, as in the pre-industrial era, but to firms with their own
characteristics of internal organisation. Thus the relevant comparison is not
simply between market and hierarchy but between different types of internal
organisation which are now linked by market rather than hierarchical links.
Whilst transaction cost economics may not actually support hierarchical gov-
ernance structures, the ambiguity of its prescriptions may actually produce
more caution in policy makers' minds. Dow suggests two major difficulties.
Firstly, we cannot always separate the transaction from the existing govern-
ance structure; secondly, bounded rationality and opportunism means we
cannot identify an efficient governance structure ex ante.
The first problem is a serious obstacle to making transaction cost logic fully
operational. In order to choose an appropriate governance structure, we need
to be able to identify independently a transaction. This might be relatively
easy in the production of widgets but it is far more difficult in the service tasks
which are typical of the public sector. For example, if one of the dimensions
for identifying a transaction is asset specificity, then the level of human asset
specificity seems to be an endogenous function of the governance structure
rather than an exogenous factor helping to determine it. For example, the UK
prison service has developed high levels of human asset specificity based
upon a 'relational team', yet recent decisions to contract out many of its activ-
ities suggest a different judgement about the nature of 'transactions' in the
prison service!
This example is also relevant to illustrate the second of Dow's criticisms.
Much subcontracting of former public sector services is based on a sort of
franchise process. This requires forecasts about ex post behaviour and con-
tractual safeguards against problems such as 'opportunistic quality degrada-
tion' . However 'perfect' contractual design itself conflicts with the transaction
cost assumption of bounded rationality. As Dow (1987) puts it, 'the mere
existence of positive transactions costs may suffice to prevent transaction cost
minimisation' (p. 28).
Contractual safeguards rely on accurate measurement of both quantities
and quality of output, yet one of the difficulties in managing so much of the
public sector is the lack of agreed output indicators.As was discussed in
Chapter 3, if output cannot be satisfactorily monitored then control must be
based on 'behaviour control' rather than 'output controls' ala Ouchi (1977).
In conclusion, contrary to the suspicions of many non-economists, transac-
tion cost economics is either extremely difficult to operationalise or biased
towards internal organisation rather than market organisation. Thus, if
governments are choosing to subcontract to the private sector (ignoring, for
the sake of argument, any venal or corrupt motives), it is because they feel
that the nature of transactions will not be changed by changes in contractual
arrangements. Secondly, they are confident that there will be sufficient ex ante
competition between contractors. Thirdly, they can design contracts which
162 Economics and control

monitor and punish any ex post contract opportunism by the franchisees. In

short, they are operating in a world of unbounded rationality and zero trans-
action costs - a world familiar to the more formal versions of agency theory
but alien by assumption to the world of the transaction cost and incomplete
contracting paradigms.

Notes

1. See Seal (1993) for an extended critical comparison of the various forms of agency
theory and transaction cost economics, with specific reference to issues of account-
ing and management control.
2. For example, a more complex agency relationship exists between the managers and
shareholders of a mature corporation where typically there has been a separation of
ownership and control. Jensen and Meckling (1976) and Watts and Zimmerman
(1979) have used the potential conflict of interest between shareholders and man-
agers (as well as other parties such as debtholders) to produce a voluntaristic
theory of accounting and auditing. For criticisms of this 'Positive Accounting'
school, see Christenson (1983) and Lowe et al. (1983) .
 CHAPTER 11

Performance indicators and control

in the public sector
Peter Smith

The public sector poses particularly challenging problems of control. Almost

by definition, there are no competitive markets for its products, and it is not
subjected to the traditional discipline of financial markets. Instead strategic
control must be secured through political processes, which can take many
forms. With the rapid reduction in the cost of data collection, governments are
increasingly relying on various types of performance indicator to inform the
political process, and thereby secure control of public sector management. The
purpose of this chapter is to examine the implications of the use of perfor-
mance indicators (PIs) for managerial control of public sector resources.
The first section sets the scene by tracing the recent history of performance
indicators in the UK. PIs are expected to fulfil many roles in securing control
of public sector organisations, and these are examined next. It is argued that
the principal criterion for introducing a PI system is that its benefits should
outweigh its costs. The third sections therefore investigates the principal
benefits that PI schemes can be expected to generate, while the forth examines
their associated costs. Next issues relating to the implementation of PI
schemes are discussed and the chapter ends with an appraisal of the strengths
and weaknesses of PIs in securing control of the public sector.

Introduction

During the 1980s, fuelled by the ready availability of cheap information

technology, the practice of publishing information about the performance of
the public sector became widespread. The principal purpose of these

163
164 Control in the public sector

developments was to enable various interested parties to secure control of

public sector resources. In the UK, no part of the public sector escaped the
scrutiny of the performance auditor. This section describes some of the more
important developments.
The first scheme receiving high-profile promotion by the UK government
was the 'Local Government Comparative Statistics' initiative (Department of
the Environment, 1981). This formed part of a code of practice that local gov-
ernments were urged to follow when publishing their annual reports. Every
report was to include the authority's performance on a range of about 50 indi-
cators, set beside data from other 'similar' authorities. The indicators were
principally related to costs and manpower. The similar authorities were
selected by the local government itself, and a wide variety of practices was
adopted (Ashley Smith and Smith, 1987). The Chartered Institute of Public
Finance and Accountancy published an annual volume listing the indicators
for every authority (CIPFA, 1982). However the initiative appeared to have
very little tangible effect on public, elected representatives or local govern-
ment (Chandler and Cook, 1986).
The National Health Service (NHS) in the UK is a central government pro-
gramme, but is administered by local health authorities. The central govern-
ment uses PIs as a central part of its mechanism for controlling the activities of
these local administrations. In 1983 a series of reports was published docu-
menting for each Local Health Authority 123 performance indicators relating
mainly to costs (Department of Health and Social Security, 1983). These thick
volumes had little immediate impact. However they were soon replaced by an
enhanced package of over 400 indicators, made available in machine-readable
form (Department of Health and Social Security, 1985). The government
rapidly developed computer software to give rudimentary analysis of the
data. In 1987 it published an expert system to help the performance auditor
make sense of the vast volume of data (Department of Heath and Social
Security, 1987), and commissioned an introductory guide to the PI system
(Department of Health and Social Security, 1988). The performance indicators
now form a central part of the system of performance review in the NHS,
whereby District Health Authorities are held to account by Regional Health
Authorities, which are in turn accountable to the central government. Thus
this set of PIs is central to securing control of the NHS.
Many other PI schemes have been put in place. For example, the Committee
of Vice Chancellors and Principals (1987) initiated a series of performance
indicators for UK universities. The process of public sector performance
measurement then reached its apotheosis with the publication in 1991 of the
'Citizen's Charter' (UK Government, 1991). Intrinsic to this initiative was the
notion that only by making available information about the activities and
achievements of public services could those services be held properly to
account. Accordingly a series of charters for individual services was
produced, amongst the first being a 'Parent's Charter' for schools (Department
 Peter Smith 165

of Education and Science, 1991) and a 'Patient's Charter' in the NHS

(Department of Health, 1991).
The Parent's Charter contains a number of measures intended to make
schools more sensitive to the demands of parents. Amongst them is the inten-
tion to publish 'league tables' of public examination results, truancy rates and
destinations of pupils leaving school. Coupled with this reporting initiative
was the establishment of a right of parents to express a preference for the
school their child should attend, offering parents some sanction against poor
performance.
The Patient's Charter also includes provisions for the publication of perform-
ance data, including waiting times for outpatient appointments, waiting times
for inpatient treatment, and waiting times for ambulance services. Associated
with these data are a series of patient rights, including guaranteed admission
for inpatient treatment within two years of being placed on a waiting list.
Perhaps the most ambitious PI initiative associated with the Citizen's
Charter has been the set of performance indicators developed for local govern-
ment by the Audit Commission (1992). Development of this package was com-
plicated by the fact that UK local government still enjoys some autonomy
from central government, and so can within limits specify its own priorities
and targets. As a result, a national unified system of performance reporting
may be inappropriate. Nevertheless the Audit Commission has devised a
system of over 200 PIs which seeks to allow some local autonomy, yet forces
local authorities to present comparative performance data.
Thus the performance indicator philosophy has permeated most of the UK
public sector. The first developments had little impact, principally because
there were few sanctions attached to continued poor reported performance.
However, especially since the introduction of the Citizen's Charter, real sanc-
tions, for example in the form of client choice, have forced public sector organ-
isations to pay increased attention to their activities as reported in PIs.

Roles of performance indicators

The publication of performance indicators is intended to serve a number of

objectives. Most fundamental is the desire to secure accountability of the public
sector organisation (the agent) to its principals. Stewart (1984) suggests that
accountability can be secured only if two conditions obtain: first, the agent
must have to give an account of performance to the principal; and second, the
principal must be able to hold the agent to account. The title of this chapter
reflects the need both for a report of performance and for the associated ability
to act on the report, and thereby to control the public sector organisation.
Agency theory has been developed principally in the context of the corpor-
ate sector (Baiman, 1990) in which the identification of principal (owner) and
166 Control in the public sector

agent (manager) is relatively straightforward. The pattern of accountability in

the public sector is far more complex. Consider, for example, the case of a local
government housing department. There are various 'stakeholders' in such a
department, each of which might consider itself a principal. Examples include
the chief executive of the local government; the local elected representatives;
the central government; the local taxpayers; the housing department tenants;
the broader public, including potential future tenants; and auditors acting on
behalf of one or more of these parties. To a greater or lesser extent, each of
these parties has a legitimate interest in the performance of the department,
and might seek to exercise some control over its operations. Thus the simple
model of a single principal is hopelessly inad equate as a reflection of the rich
pattern of accountability that obtains in the public sector.
Notice in particular that the users - or potential users - of public sector ser-
vices should be able to hold those services to account. In the corporate sector,
when product markets are competitive, customers do not expect to exercise
direct control over management because they have the strong sanction of
taking their custom elsewhere. In contrast, most public services are local
monopolies, and there are few real sanctions available to users. In addition,
many potential users of services or other interested parties do not have direct
experience of the quality of service delivered. For example, many citizens have
legitimate interests in the performance of police or personal social services,
even though they cannot be explicitly identified as clients of those services. To
secure accountability it therefore becomes desirable to furnish such parties
with some mechanism for controlling the performance of the services.
A strong framework of accountability is thought to be necessary for the
pursuit of the accountant's notions of efficiency and effectiveness. If the
pattern of accountability is faulty, it can be argued, managers may have no
incentive to use their resources efficiently. Such considerations are likely to be
the principal concern of the funders of the organisation, such as taxpayers and
the central government. Similarly, without accountability, there may be no
incentive for a public sector organisation to pursue the objectives of its princi-
pals, and so effectiveness might suffer. While most stakeholders will have an
interest in effectiveness, it is perhaps service users who have the most immedi-
ate interest in it.
However, once PI schemes have become operational, it soon becomes clear
that they are addressing issues broader than efficiency and effectiveness. For
example, in assessing the NHS scheme after it had been in place for a number
of years, Baroness Trumpington (1987) suggested that its main role was the
promotion of equity between individuals and between geographical areas. The
extension to equity considerations emphasises why it is important to consider
the broader concept of control as the principal purpose of PI schemes. In the
corporate sector, control might be considered synonymous with the promo-
tion of efficiency and effectiveness. The interest of owners and customers is in
persuading the company to produce goods required by product markets and
 Peter Smith 167

securing rates of return that satisfy financial markets. Any company failing to
do so runs the risk of bankruptcy. However in the public sector control might
have wider connotations, first because the principals are more diverse, and
second because there is much less consensus about what the public sector
should be doing.
The emphasis on control raises the issue of who is exercising the control on
behalf of whom. The mechanism for securing control is the prevailing pattern
of accountability. In practice most accountability structures have developed
in an arbitrary manner, and are the outcome of the vagaries of historical acci-
dent. There is no ineluctable model of accountability. As a result it is difficult
to generalise about how control over a public sector organization is exercised.
Thus, for example, UK local government is controlled in part by local electors,
who are periodically able to pass judgement on their elected representatives in
local elections. In addition, the central government feels it too has a legitimate
interest in local government, and has adopted the right to control expenditure
and implement a wide range of more detailed controls. On the other hand, the
NHS is directly run by the central government, albeit on an agency model,
with local Health Authorities having no elected members, and very weak local
patient representation, in the form of Community Health Councils. The citizen
wishing to effect a change may therefore have to seek influence via a very cir-
cuitous route, perhaps needing to put pressure on the secretary of state by lob-
bying a local member of parliament.
Thus the citizen with a legitimate interest in public services often has very
imperfect sanctions with which to affect the management of local services . The
principal mechanisms are various forms of political pressure, including
voting, and migration to areas with a preferred pattern of public services.
These alternatives may be either ineffectual (as in the case of voting) or very
costly to the individual. There is therefore a difficulty in persuading the citizen
to take an interest in the performance of public sector organisations. This
insight has informed many of the recent PI developments in the UK. For
example, the Citizen's Charter is intended to make it easier for ordinary
people to assess the performance of the public sector, and to give them an
incentive to do so.
A parallel, alternative strategy to overcome the problem of lack of public
interest is the creation of expert audit offices, which are intended to act on
behalf of citizens in scrutinising public sector performance. In England and
Wales, this role is undertaken by the National Audit Office (for central govern-
ment) and the Audit Commission (for local government and the NHS).
As the above examples have shown, PI schemes have traditionally been con-
sidered under two headings: those that are designed for external stakeholders
and those that serve internal management control purposes. These external
and internal purposes can be considered analogous to the external and inter-
nal accounting functions in the corporate sector. The two types of PI scheme
can be thought of as serving respectively the political and managerial control of
168 Control in the public sector

the organisation. An example of the former is the set of PIs which English local
governments are advised to publish in their annual reports. The explicit objec-
tives were 'to give [local taxpayers] clear information about local govern-
ment's activities; to make it easier for electors, [taxpayers] and other interested
parties to make comparisons of performance of their authorities; and to help
[elected representatives] form judgements about the performance of their own
authority' (Department of the Environment, 1981). On the other hand, the set
of PIs developed for the NHS address managerial control, were intended 'to
help [managers] assess the efficiency of the services for which they are respons-
ible' (Department of Health and Social Security, 1983).
Increasingly, however, there is emerging a third type of control, in which
the principal is a purchasing organisation, disbursing public funds, and the
agent a quasi-autonomous provider in receipt of those funds (Carter and
Greer, 1993). The relationship between purchaser and provider is guided by a
formal contract, through the terms of which the principal seeks to control the
agent. PIs might form an intrinsic part of the contractual arrangement. This
third type of control might be called agency control. Here an example from the
UK is the Benefits Agency, the organisation which is responsible for assessing
and paying social security benefits claims. This is held to account by the
central government through the medium of a range of performance indicators
and associated targets, such as 'to pay the correct amount [of income support]
in 92% of cases' (Benefits Agency, 1993).
At the interface between the external world and the public sector organisa-
tion is the board of governors. Anthony and Young (1984, p. 649) consider the
most important stimulus to improved management control in non-profit organ-
isations to be 'more active interest in the effective and efficient functioning of
the organization by its governing board'. In general, governing boards are
held accountable to external stakeholders by various mechanisms, such as
elections or patronage. Thus, when PI schemes are aimed at external stake-
holders, if they are at all effective, they are likely to influence the decisions of
the governing board. And so, in turn, they will influence the internal manage-
ment of the organisation. The external PI scheme will thereby be internalized.
Consequently this chapter makes no distinction between the various forms of
scheme described above, and seeks instead to examine the influence of all
types of PIon the internal control of the public sector organisation.
Two models of management implicitly underlie the traditional public sector
PI scheme: a model of production, and a model of control. These two models
indicate respectively what the PIs are seeking to represent, and how they are
subsequently to be used to effect change. Architects of PI schemes usually
appear to have in mind a standard neoclassical production model, in which
inputs are consumed by an organisation. Some production process takes place
and a set of valued outputs emerge. The outputs eventually have an outcome
in terms of their impact on society. Within this framework there are usually
four types of PI, seeking to measure different aspects of activities: environ-
 Peter Smith 169

ment, inputs, processes and outputs. Some authors distinguish between inter-
mediate measures of output and eventual outcome (Jackson and Palmer,
1989). However it is very rare for any PI scheme to be able to address outcome
issues, so for practical purposes it is reasonable to ignore the outcome aspect
of performance.
The model of control is based on the notion of managerial cybernetics dis-
cussed in Chapter 1. PIs serve the various stages of the cybernetic model by
enabling principals to set quantitative targets for management. Performance
can then be assessed in relation to these targets. Any deviation can be exam-
ined, remedial action proposed and revised targets set. Thus the principal
objective of the PI scheme is to facilitate estimation of the production function
confronting the organisation, so that best practice can be identified and
achievable targets set. The criterion for judging the PI scheme should therefore
be the extent to which its benefits - in terms of more sensitive control- exceed
its costs. The following sections set out the principal benefits and costs associ-
ated with a PI scheme.

Benefits of performance indicators

The relevant benefits and costs arising from PI initiatives are those impinging
on the external stakeholders. As discussed above, the putative benefits might
arise in the form of the greater efficiency, effectiveness, equity and probity of
the public sector. Costs include the direct costs of collecting and disseminating
data, and might also emerge in the shape of unintended dysfunctional conse-
quences of PI schemes. This section seeks to categorize the various benefits.
Costs are treated in the next section.
The principal benefits claimed for PI schemes can be considered under six
headings:

1. clarifying the objectives of the organisation;

2. developing agreed measures of activity;
3. gaining a greater understanding of the production process;
4. facilitating comparison of performance in different organizations;
5. facilitating the setting of targets for organisations and managers;
6. promoting the accountability of the organisation to its stakeholders.

Although forming an important component of most PI schemes, the measure-

ment of inputs - as expressed in costs - is usually relatively straightforward.
Indeed, many early schemes, such as the Local Government Comparative
Statistics initiative, concentrated almost exclusively on inputs, in the form of
personnel and costs. Increasingly, however, schemes are seeking to address
much more interesting questions, namely: what is going on inside the black
170 Control in the public sector

box known as the production process; how are the valued outputs (not to
mention eventual outcomes) to be measured; and what light do the PIs cast on
the performance of the organization? The first step in devising a PI scheme is
therefore to identify the processes and objectives of the organisation. In itself,
this procedure might prove valuable. It requires explicit statements of
methods and goals which hitherto may have been muddled and ambiguous. If
it proves possible to make such statements, the organisation may be given a
clearer sense of purpose, and the efforts of staff at all levels can be focused to
common objectives.
However one of the aspects of public sector management which distin-
guishes it very clearly from management in the corporate sector is that
identification of objectives may be difficult or even futile. Different stakehold-
ers often hold different expectations with regard to a public sector organisa-
tion . For example, parents, employers, the community at large and the central
government might wish to emphasise very different outputs of the secondary
education sector. Even within a group of stakeholders (say parents), there
might be a great mix of requirements. And the priorities of all groups might
change over time . In the end, of course, schools must reconcile the possibly
conflicting demands made on them. But to make explicit statements of object-
ives might show that they are favouring one group of stakeholders at the
expense of another, and preclude the flexibility needed to adapt to changing
demands. It is therefore important to emphasise that identifying objectives is
not a trivial process, and may in some circumstances even be dysfunctional.
Having identified processes and objectives, the next requirement is to devise
measures of the phenomena of interest. Only by identifying unambiguous and
consistent measures is it possible to compare the activities and achievements
of the organisation, either with its own performance in previous periods or
with that of other organisations in the same period. These two types of com-
parison - dynamic and cross-sectional - are the only realistic methods of
gaining some idea of the production possibilities of the organisation, as there
rarely exist ideal, engineering benchmarks of performance. It is therefore
imperative that the definitions of any measures used are clear and not open to
'creative' misinterpretation by different management teams. They should
therefore be readily audited, to minimise misinterpretation or even fraud .
In practice, many aspects of public sector performance are very difficult to
measure, and the measurement process is often dependent on data provided
directly by 'front line' workers, the very people whose performance is to be
assessed. For example, the measurement of medical caseload requires infor-
mation on the diagnosis of patients, which is provided by doctors. If such data
are to be used in judging the efficiency of those same doctors, then they have a
very strong incentive to maximize their apparent workload by assigning their
patients to diagnosis groups with a high workload 'tariff'. It is almost imposs-
ible to control such activity by external audit, as there is considerable scope for
discretion in matters of clinical judgement. In any case, any large-scale audit
 Peter Smith 171

would be infeasible. As a result, much of the public sector relies on the good-
will of staff to extract useful performance data, and so obtaining their support
in implementing any PI scheme is imperative.
Measurement is only the beginning of the comparison process. The phe-
nomena measured must be incorporated into a model of the production
process before any useful comparisons can be made. The modelling stage is
often particularly difficult because our understanding of the means whereby
public sector inputs are converted into outputs is very imperfect, particularly
when environmental circumstances have a significant influence on outcome.
Smith (1990) gives five reasons why variations in PIs might be observed
between organisations: they might be pursuing different objectives; they
might have different environments; they might face different resource costs;
they might report their performance differently; and they might have different
levels of efficiency. The purpose of modelling is to disentangle these causes of
variability.
Modelling is therefore an essential aspect of almost all comparison. In the
corporate sector, businesses have the option of closing down operations if the
environment is adverse - for example if local pay rates are very high, or local
transport infrastructure is poor. Indeed, if the business is operating in a com-
petitive market, such closure may be essential if the business is to remain com-
petitive. The need for modelling the production process sensitively is
therefore relatively modest in most competitive product markets.
In contrast, most public sector organisations have to continue operating
regardless of how adverse the external environment might be. There is what
one might call a 'geographical imperative' regarding the location of services
such as schools, hospitals and refuse collection. As a result, the modelling
stage is crucial, and the performance analyst must seek to take full account of
the external environment when coming to judgements on public sector
performance. In the schools sector, for example, there has emerged the notion
of the 'value added' by schools when judging their examination results (Gray
et al., 1990). That is, performance is judged, not on crude examination results,
but on the results secured in relation to the capabilities of the children as
measured when they entered the school. The intricacy of the statistical
models developed in the education sector is evidence of the complex process
involved in coming to informed judgements about performance once it
becomes necessary to take account of environment. It is often best to think of
uncontrollable environmental determinants of performance as additional
inputs into the production process, augmenting the set of more conventional
physical inputs.
The next stage in analysis of performance is to come to a judgement on the
performance of the organisation of interest. Dynamic or cross-sectional com-
parison yields benchmarks against which its achievements can be measured.
From this analysis, some statement can be made about the desired level of per-
formance, and so targets can be set in terms of the dimensions of performance
172 Control in the public sector

encompassed by the PI scheme. In itself, the publication of comparative data

might introduce pseudo-competition between public sector organisations.
Implicit in the comparison process there might be rewards and punishments
for management boards and their staff which lead the organisations under
scrutiny to pursue measures which improve their performance, as reflected in
the PIs. However it may be that publication alone is not enough, and that
targets must be associated with an explicit incentive scheme for organisations
or the managers within them. Thus, as well as facilitating retrospective
appraisal of past performance, the PIs can playa central prospective role in
servicing the system of targets and rewards - whether implicit or explicit -
adopted by the external controllers of the organisation.
For example, in the UK Benefits Agency, a key organisational objective is to
clear 67 per cent of child benefit claims within 10 days (Benefits Agency, 1993).
The senior management of the organisation can therefore be held to account
on clearance rates such as these, and they in turn can hold to account the man-
agement of individual offices using the same PIs. However, as noted above, in
doing so they may need to take into account the different environments
(reflected in the complexity of claims) in which different offices must operate.
In summary, therefore, the PI philosophy entails identifying objectives,
measuring progress towards those objectives, assessing performance, setting
targets and ensuring that incentives are compatible with progress towards
targets. The cybernetic model of management control underlying the PI phi-
losophy is intended to bring numerous benefits relating to management
control. The identification of objectives might clarify the organisation's
mission, and enable it to nurture a shared sense of purpose amongst staff.
Quantifying objectives enables targets to be set at all levels in the organisation,
to which incentives can be attached, and forces managers to address trade-offs
between conflicting objectives. The performance assessment process can gen-
erate an enhanced understanding of the production process underlying the
organisation's activities. And it is possible that the entire PI culture might
enhance the organisation's accountability to outside stakeholders, enabling
them to secure better control.

Costs of performance indicators

Although many of the potential benefits listed above are palpable, and can be
observed in many public sector organisations adopting the PI philosophy, no
PI scheme is without associated costs. As the Audit Commission noted when
developing its PI scheme for UK local government, the direct costs of collect-
ing, analysing, auditing and disseminating data can be considerable (Audit
Commission, 1992). And there are a large number of less obvious but possibly
 Peter Smith 173

more pernicious potential costs, in the form of unexpected behavioural conse-

quences of publishing performance data. These are examined in this section.
The possibly dysfunctional consequences arise for two fundamental
reasons. First, it is very difficult to implement the principles set out in the pre-
vious section. To a greater or lesser extent, most PI schemes fail to reflect all
the nuances of organisations' purposes and activities. This might be because
some of the processes and outputs are elusive and difficult to identify or
because - although identifiable - they are difficult to capture in simple PIs.
And second, controllers have only a limited ability to process the data emerg-
ing from a PI scheme, and so may be misled into making faulty inferences
about performance. These two limitations can be thought of as respectively
malfunctions of the PI sensing mechanism and malfunctions of the response
mechanism.
Following Smith (1993), it is possible to examine the unintended conse-
quences of PI schemes under eight headings: (l) tunnel vision, (2) sub-
optimisation, (3) myopia, (4) measure fixation, (5) misrepresentation, (6)
misinterpretation, (7) gaming, (8) ossification. Tunnel vision, suboptimisation
and myopia reflect different aspects of incompleteness in the PI scheme. If
managers are given incentives to pursue targets reflected in PIs, then they are
implicitly being given incentives to ignore aspects of performance not encom-
passed by the scheme. Thus if, as is usually the case, some valued aspects of
performance are not captured by PIs, there is a risk that they will be ignored
by management. The scheme induces tunnel vision.
Suboptimisation is the pursuit by managers of narrow local objectives at the
expense of the objectives of the organization as a whole. It is a very real
danger in the public sector, in which many of the outputs are the result of co-
operation between a number of agencies. Such co-operation is endangered if
each of the agencies is asked to pursue its own agenda. The enhancement of
joint outputs is in general not attractive to a manager given explicit perform-
ance incentives, because it entails the co-operation of individuals beyond his
or her direct control. Thus, even if joint objectives can be incorporated into a PI
scheme, it is likely that managers will give them less priority than objectives
more within their direct control.
Myopia is the neglect of long-term objectives. This phenomenon is endemic
to any PI scheme. Either the controller waits for all the long-term conse-
quences of activity to unfold before coming to a judgement on performance -
in which case the judgement may be too late to be useful - or a judgement is
made before the long-term consequences have become evident - in which case
the evaluation is incomplete. Long-term issues are very important in many
public sector programmes, particularly health and education.
Most measures of performance are only imperfect reflections of activity or
progress towards an objective. Therefore, if managers are held to account with
such imperfect measures, dysfunctional consequences may result because it is
in their interest to enhance performance as measured, and not necessarily to
174 Control in the public sector

maximise valued outputs related to the underlying objective (Kerr,1975). This

phenomenon - measure fixation - was endemic to the Soviet Union (Kornai,
1992) and results from the impossibility of capturing all the subtleties of per-
formance in a limited number of indices. For example, in the UK National
Health Service, managers have been set the target of ensuring that no patient
waits more than two years for an operation (Department of Health, 1991).
Managers have pursued this target with zeal , yet the outcome may have been
longer waiting times for patients awaiting serious surgery, as hospitals have
concentrated on reducing the longer waiting times of those in need of
relatively minor surgery. This outcome may not have been intended by the
controller (the UK government).
We have already noted the danger of excessive reliance on PI schemes
leading to misrepresentation, which can take the form of 'creative' reporting
or fraud . Creative reporting is legitimate, but entails the deliberate use of
judgement to show the organisation in the best possible light, and may lead to
biased perceptions of activity. Examples include smoothing - the deliberate
choice of the time period to which to assign an event (Ronen and Sadan, 1981).
Fraud is illegitimate manipulation of data, and is an ever-present danger
when auditing is costly and complex, as in the public sector. Mis-
representation of any type is likely to be dysfunctional because it leads to the
controller receiving misleading messages about processes and outputs.
The problem of misinterpretation is brought about principally by the need
to take account of environment, and the intrinsically complex transformation
processes inherent in much of the public sector. In addition, notwithstanding
the efforts of bodies such as the UK Audit Commission, there is a great short-
age of expert intermediaries able to cast dispassionate light on performance
data. Often the only experts available to interpret such data are employees of
the very organisation the performance of which is under scrutiny. Such
observers clearly have an incentive to put their own efforts in the best possible
light, and so their interpretation may lead to biased perceptions about the
organisation's performance.
Gaming is the deliberate distortion of behaviour by management to secure
strategic advantage (Jaworski and Young, 1992). For example, managers might
deliberately underperform year after year to avoid being set more demanding
targets. They can blame such underperformance on external factors and the
controller is often in no position to gainsay such judgement. The incentive for
gaming arises because, although managers might be given a reward for
outperforming their targets, the subsequent penalty is that all future targets
will be 'ratcheted' up by the controller in the knowledge that enhanced perform-
ance is possible. As a result, unless they are given an incentive to improve
targets as well as simply achieve them, managers may opt to achieve chronic-
ally mediocre performance (Weitzman, 1976). The scope for gaming therefore
arises because of shortcomings in the incentive scheme and long time hori-
zons amongst managers.
 Peter Smith 175

Setting up any PI scheme is a costly and complex process. It is tempting for

controllers to assume that, once designed and implemented, the scheme can be
left unaltered. However, if it is not to ossify, it must be kept under constant
review. New challenges and opportunities constantly arise, and should be
incorporated into the PI scheme when relevant. In addition, one of the means
of insuring against gaming behaviour is to lead managers to expect that the
criteria of success will be periodically changed.
Thus there exists a wide range of potentially serious adverse consequences
arising from the implementation of PI schemes. The next section considers
ways in which these might be mitigated.

Mitigating dysfunctional consequences

Numerous strategies exist to counter the unintended and possibly dysfunc-

tional consequences of PI schemes. Amongst the most important are the
following:

1. involving staff at all levels in the development and implementation of PI

schemes;
2. retaining flexibility in the use of PIs, and not relying on them exclusively
for control purposes;
3. keeping the PI system under constant review;
4. seeking to quantify every objective, however elusive;
5. measuring client satisfaction;
6. seeking expert interpretation of the PI scheme;
7. maintaining careful audit of the data;
8. nurturing long term career perspectives amongst staff;
9. keeping the number of indicators small;
10. developing performance benchmarks independent of past activity.

Involving staff at all levels 0), retaining flexibility (2) and keeping the PI
scheme under constant review (3) are strategies which acknowledge that no PI
scheme can offer a perfect representation of the production process, and that
control mechanisms are inevitably imprecise. The controller must recognize
that, unless applied with some sensitivity, PI schemes are doomed to fail in all
but the simplest circumstances. And it is important to recognize that many
aspects of performance in the public sector are resistant to quantification.
Rather than try to force the organisation into a rigid straitjacket of PI control, it
may be more appropriate to adopt a mixed strategy, in which other modes of
control are also adopted. For example, as discussed in Chapter 3, Ouchi (979)
notes that production processes in which outcomes are difficult to measure
and the transformation process imperfectly understood - characteristics of
176 Control in the public sector

much of the public sector - may be better suited to the 'clan' control of peer
pressure and shared objectives amongst staff. It is noteworthy that the culture
of the UK National Health Service has changed dramatically on the introduc-
tion of an internal market for health care in 1990, from that of clan control to
that of agency control, in which performance data playa much more import-
ant role than hitherto. Attention to (1), (2) and (3) might alleviate many of the
unintended consequences of PI schemes.
Measuring progress towards every objective (4) and measuring client satis-
faction (5) are different strategies for seeking to encompass all salient dimen-
sions of performance into the PI scheme. The measurement of every objective,
however roughly done, is increasingly being recognized as a key component
in non-financial performance measurement in the corporate sector (Kaplan
and Norton, 1992) and is essential if important dimensions of performance are
not to be neglected by managers. Yet the prospect of capturing all aspects of
public sector performance in a manageable number of measures is daunting.
The strategy of seeking measures of client satisfaction is one way of sidestep-
ping this massive measurement problem. Measures of outputs are after all
only proxies for their impact on clients. The problems involved in measuring
satisfaction are, however, very considerable. More fundamentally, the
identification of 'clients' is very difficult, perhaps including taxpayers, non-
users, future generations and governments, as well as the more obvious class
of clients - service users.
Expert interpretation of data (6) is needed because of the complexity of
assessing performance, and the lack of incentive for individual citizens to
undertake such analysis. Clearly public audit offices fulfil this role, and are
also able to undertake audit of data (7). There remains the problem, of course,
of who audits the auditors, and how their effectiveness is judged.
Strategies 0) to (7) are likely to be applicable to most PI schemes. The
remaining strategies are designed to address specific dysfunctional phenom-
ena which may arise in particular circumstances. Nurturing long-term career
perspectives (8) is designed to encourage staff to address long-term objectives.
The difficulty of capturing long-term issues in PI schemes is reduced if staff
expect to be in post to experience the consequences of their attention to long-
term issues. It should be noted, however, that long-term career perspectives
may encourage gaming amongst staff, as they seek to maintain modest perfor-
mance targets into the future.
Keeping the number of indicators small (9) reduces the scope for misinter-
pretation and increases the ability of the principal to exercise meaningful
control. However it may compromise the objective of developing a compre-
hensive PI scheme, and may encourage gaming. It is noteworthy that, in
seeking to develop a PI scheme for English local government, the Audit
Commission (992) came up with a package which did not 'cover every
service' and did not deal with 'individual services in excessive detail' . They
 Peter Smith 177

recognized that the limitations of the scheme would leave 'some commenta-
tors dissatisfied'. Yet the package includes over 200 indicators.
The need to develop independent benchmarks (0) is brought about by the
ability of individual managers or organisations to appeal to uncontrollable
external circumstances as the reason for their apparently poor performance.
There will always be organisations for which no comparator is available - for
example, schools operating in uniquely disadvantaged areas. And some public
services are natural monopolies with which few comparisons are possible. It is
possible that, with careful use of modelling techniques such as data envelop-
ment analysis (Ganley and Cubbin, 1992), benchmarking might be possible in
public sector services with simple production processes where a number of
comparable organisations exist. However such circumstances are relatively
rare, so the scope for benchmarking is limited.
Not all of these strategies for averting dysfunctional consequences are rel-
evant in every situation and some, such as nurturing long-term career per-
spectives, may be helpful in minimising some adverse consequences but
aggravate others. Nevertheless the list indicates that there do exist means of
addressing many of the unintended costs of performance measurement. The
wise controller must seek to balance the costs of implementing these strategies
against the benefits they might bring in terms of control.

Conclusions

Performance indicators are assuming increased importance in most public

sectors for a number of reasons. One is simply that over a short period the
costs of collecting and disseminating data have reduced dramatically. Another
is that public sectors are increasingly being reorganised away from monolithic
structures towards agency arrangements in which small policy units purchase
public sector services from quasi-autonomous agencies. The resulting arms'
length relationship between purchasers and providers must be serviced by
performance data, where before internal control could be by more informal
means. It is noteworthy, therefore, that no discussion of PIs can be made in
isolation from the organisational structure they are intended to service.
This chapter has viewed the use of PIs within a cost-benefit framework and
asserted that the acid test for a PI scheme should be that its benefits should
exceed its costs. A number of benefits claimed for PI schemes have been
identified, and it has also been noted that a number of unexpected costs might
arise when implementing PI schemes. Strategies exist to mitigate some of the
adverse consequences associated with PIs, and again the benefits of imple-
menting these strategies must be weighed against their costs.
178 Control in the public sector

The discussion suggests that, if handled carefully, the provision of more

performance data may be beneficial to the performance and accountability of
the public sector. However one cannot ignore the manifest dysfunctional
consequences that arose in the former Soviet Union when an attempt was
made to run an entire economy through the medium of such data. Only by
careful attention to the considerations set out here, and by keeping a sense of
proportion about the limitations of PIs in securing control, can the worst
consequences of excessive reliance on performance indicators be avoided.
 CHAPTER 12

Organisational culture and control

Kim Langfield-Smith

The term 'culture' has been used to describe many forms of human collectives
- nations, ethnic or regional groups, organisations, professions or occupations
and families. In this chapter the focus will be on cultures that characterise an
organisation. The study of human groups as cultures owes its origins to the
early writing in anthropology of Margaret Mead (1928), Malinowski (1922),
Radcliffe-Brown (1952) and, more recently, to Levi-Strauss (1967) and Geertz
(1973). Aspects that are now associated with organisational culture have been
studied in the organisational behaviour literature by writers such as Whyte
(1948), Dalton (1959), Blau (1955) and Roy (1960). However it was not until
the 1970s that cultural perspectives began to emerge in organisational research
and the terms 'organisational culture' or 'corporate culture' came into popular
and academic use. For academic researchers, culture became a framework for
understanding and attributing meaning to the structures, systems, events,
interactions and other phenomena that take place in organisations. In the
popular press, the interest in culture has focused on managing (or changing)
culture, or creating an awareness of the influence of an organisation's culture
on organisational success. Several popular management writers have sug-
gested that an effective culture can be planned and implemented rationally. It
is only very recently that academic researchers have begun to investigate the
significance of culture for control systems.
In this chapter we will discuss the concept of organisational culture as used
within organisational research, and its relevance to the design and operation of
control systems. This area of research is complex and fraught with competing
definitions and viewpoints about the nature of organisational culture and of
control. When organisational culture is viewed as embedded in the wider
sociocultural system, it becomes a metaphor for understanding the organisa-
tion (and its control systems). Alternatively organisational culture can be
viewed as an organisational variable, in the same way that we consider organ-
isational structure or technology. Thus culture can be viewed as part of the
179
180 Organisational culture and control

organisational context to be considered in the design of control systems, or as a

tool to be managed to achieve better control and organisational effectiveness.

Clans and clan control

We can begin our discussion of organisational culture by considering the

closely related idea of clans. Within a transactions cost framework, Ouchi
(1979) described three different organisational control mechanisms: markets,
bureaucracies and clans. Each of these mechanisms needs certain minimum
social and information requirements to operate. These are shown in Table 12.1.
Market mechanisms require there to be a shared norm of reciprocity, and
control is achieved through price mechanisms. Reciprocity refers to the notion
of trust. It is necessary for each party to a transaction to be assured that the
other is acting honestly, and this reduces transaction costs. All information rel-
evant to the transaction is captured in the price. In a bureaucracy there is the
additional social requirement of agreement on the legitimate source of author-
ity. The norm of reciprocity takes the form of 'an honest day's work for an
honest day's pay'. Control is achieved through formal rules. The most demand-
ing mechanisms are clan controls which require that the members of an organ-
isation share a range of values and beliefs on the forms of proper behaviours,
and have a high level of commitment to those behaviours. Thus, within a clan,
control is achieved implicitly. It will be noted that Ouchi's description of the
clan bears some similarity to Etzioni's normative organisations.
Clan controls are most effective in certain types, or parts, of organisations
where the outputs are ambiguous and difficult to measure, and where behav-
iour controls are inappropriate because the transformation process is
unknown (Ouchi, 1980). The imposition of output or behaviour controls,
which might be used within a bureaucracy, can actually impede effective
control within a clan by encouraging behaviour that is not consistent with

TABLE 12.1 Social and informational requirements for

organisational control mechanisms

Type of Social Informational

control requirements requirements

Market Norm of reciprocity Prices

Bureaucracy Norm of reciprocity Rules
Legitimate authority
Clan Norm of reciprocity Traditions
Legitimate authority
Shared values, beliefs
 Kim Langfield-Smith 181

organisational goals and objectives, or by directly intruding on the day-to-day

smooth functioning of the organisation. A research laboratory in a pharmaceu-
tical company, or a medical centre, might place reliance on clan controls. In
both situations formal and inflexible accounting controls may impede
workflow and redirect attention to aspects that do not lead to desired outputs.
Clan controls are subtle and not visible, particularly to the casual outside
observer. Key aspects of the control system are rewards that attach to display-
ing the correct attitudes and values. These may take the form of some ritual or
ceremony that serves to reinforce those same attitudes and values. For
example, in a university the awarding of degrees is signified by the ritual of
graduation, which emphasises the rewards obtained through scholarship.
Ouchi's (1980) model of markets, bureaucracies and clans is dominated by
the nature of the economic relationships between parties within an organisa-
tion . In particular, the condition of trust, or reciprocity, in a clan is based on
members' beliefs that co-operative behaviour is a good way to achieve higher
output, which can ultimately result in long-term financial equity . However
Alvesson and Lindkvist (1993) suggested that there are clans where there are
other bases of trust. They distinguished between the economic clan of Ouchi,
and the social clan and blood relationship clan. In a social clan, commitment and
reciprocity are based on the individuals' needs to belong and communicate
with other members as part of an organisation. Thus these organisations
satisfy the emotional and social needs of its members. A blood relationship
clan is connected to a 'biological imperative'. That is, clan relationships and
trust are based on family relationships that dominate that organisation.
While organisations that can be described as pure clans are rarely found,
many organisations rely to some extent on clan control. Examples include hos-
pitals, schools, research groups, legal firms and public accounting firms. A
characteristic of these types of organisations is the high level of commitment
that can be found among employees, the careful selection of individuals by
the organisation, stability of employment and extensive socialisation processes
(often through formal qualifications and schooling) during which people come
to internalise certain values.
The social requirements of a clan rely largely on the existence of shared
values and beliefs, and organisational commitment, which are also characteris-
tics of organisational cultures. The idea of organisational culture has increased
in popularity over the past decade. However, it is complex and problematic,
because of different understandings about the nature of the construct.

What is organisational culture?

Organisation culture has been addressed by many researchers (Pettigrew,

1979, 1985; Schein, 1984; Gagliardi, 1986; Meek, 1988) and it has been given a
182 Organisational culture and control

range of definitions. Meek (988), in a critique of organisational culture

research, described culture as encompassing multiple aspects: symbols
(including language, architecture and artefacts), myths, ideational systems
(including cognitive systems and ideology) and rituals. Morgan (986)
regarded 'shared meanings, shared understandings, and shared sense making'
as different ways of describing organisational culture. Organisational culture
was regarded as the 'process of reality construction that allows people to see
and understand particular events, actions, objects, utterances, or situations in
distinctive ways' (Morgan, 1986, 1.28).
Cultures have a stock of knowledge, including recipe knowledge and social
typifications (Sproull, 1981). Recipe knowledge includes the routine per-
formance programmes and standard operating procedures that guide actions
in particular situations. Social typifications are the shared understandings
which are acquired through socialisation and interaction of the people within
a group and are influenced by its distinct language, environment and history.
Culture is conveyed to organisational members through sentiment, beliefs and
attitudes and thus appeals to the emotional elements of individuals (Pfeffer,
1981). It is an elusive concept because of its intangible and amorphous nature.
While culture is manifested through behaviour, it should not be confused with
overt behaviour patterns.
Organisational cultures are perpetuated by myths or stories which act to
reinforce cultural values. These stories may be a series of memorable events
in the company's history that become part of the 'organisational folklore' .
For example, they may be concerned with the way the leader started the
organisation - his drive, his vision and the band of pioneers that surrounded
him (Clark, 1972). There is the element of myth surrounding some of these
organisational stories. In some situations organisational stories may be con-
sciously (or strategically) developed, but often emerge 'naturally' over time.
These stories may take on the character of myths and the content may subtly
change over time, especially when first-hand witnesses to the event become
fewer. Stories, or more specifically myths, may be used to present cultural
values to newcomers. These stories may present the social prescription of
'how things are done around here', the consequences of compliance or
deviance, and the social categories and statuses that are legitimate within
that organisation.
At Hewlett Packard, the stories surrounding Bill and Dave were stories that
emphasised and legitimated the management philosophy of the company
(Box 12.1). The symbolic nature of stories and myths makes it difficult to inter-
pret the underlying message or meaning, especially for people external to the
culture. However, these types of stories perform an important control function
as they transmit and reinforce the cultural values and shared perspectives of
organisational members. They provide a more effective method for co-
ordinating work and operations than formal bureaucratic control, as they can
be persuasive in their message and are easily remembered.
 Kim Langfield-Smith 183

Box 12.1 Hewlett-Packard

Hewlett-Packard was founded in the 1940s in the USA by Bill Hewlett

and Dave Packard and together they created a strong culture charac-
terised by innovation and strong team commitment. These founders
practised a hand-son management style that emphasised an enthusiasm
for work, and encouraged an atmosphere of problem sharing and open
exchange of information and ideas. These values were reinforced by
frequent meetings which provided opportunities for interactions and
sharing, and social interactions such as the ritual 'beer busts' and 'koffee
klatches'. Over time myths and stories were passed on to new members
which helped to sustain the cultural values created by the founders.
These include stories of the early days when Bill and Dave started the
company in Bill's garage and the family oven was used to make some of
the first products.
The overall message that was transmitted by the culture was that of
trusting and valuing employees. In the difficult times of the 1970s, the
strength of the culture and the commitment of the management and staff
was tested when instead of retrenching [sic] employees, staff took a 10
per cent decline in pay and worked a nine-day fortnight. Again the
emphasis on team spirit and sharing was possible, even in difficult times.

Source: Morgan (1986), p. 124, adapted.

Accounting and control systems can be an important component of the

organisational culture in that they can demonstrate the way that an organisa-
tion views the world, and help to reinforce a particular view of reality. An
example of the role of these systems in the changing culture of a railway
company is described in Box 12.2.

Organisational subcultures

The popular management literature (for example, Peters and Waterman's In

Search of Excellence) has encouraged the idea of strong unitary organisational
cultures. However it is more likely that organisations contain subcultures. A
variety of factors may contribute to the emergence of these subcultures within
an organisation: the introduction of new people from outside the organisation,
unique work roles and departmental perspectives, dense boundaries or
loose couplings of various organisational groups, the introduction of new
184 Organisational culture and control

Box 12.2 The symbolic role of accounting systems in a railway

company

In some organisations accounting and control systems have an import-

ant symbolic role, but in others they may have no particular significance.
Dent (1991) provided an interesting study where he illustrated the
significance of these systems in effecting a cultural change.
In a large European railway company the accounting activities and
financial reporting systems were incidental to the running of the busi-
ness, as the dominant culture of the company was focused on the railway.
For over a century the railway was viewed as a public service and the
purpose was to run trains. Profitability was secondary to railway engin-
eering and the logistics of running trains. The cultural knowledge was
that if the organisation provided the nation with a transport infrastruc-
ture then they would be financially supported by the government.
However, in an increasingly resource-constrained environment, busi-
ness managers were appointed and we were assigned responsibility for
developing strategies that would improve financial performance. The
pressure to adopt a business culture saw 'the train' as decreasing in its
symbolic significance and with the help of accounting activities there
was a shift of importance from the train to the customer. Rail transport
came to be regarded as a product or service and the railway was viewed
as a profit-seeking enterprise. Systems of accountability and organisa-
tional structures centring upon the business managers were created. The
assumptions underlying those systems - rationality, authority, organis-
ation and the time - gradually permeated the knowledge and the values
and beliefs of the organisation and became the dominant culture.

technologies, social and ethnic groupings, and professional affiliations. Some

of these subcultures (for example, those based on nationality or professional
affiliation) may have their sources outside the organisation's boundaries.
Subcultures within the one organisation may have some similarities, but
there may also exist disparate subcultures that actively compete for dom-
inance or persist in an uneasy symbiosis (Martin and Siehl, 1983). The dom-
inant culture is often associated with the senior management perspective,
which through access to power and resources is actively reinforced by man-
agerial decisions and actions. Some organisations successfully function with
discordant subcultures. These subcultures can maintain their distinct identities
by being 'loosely coupled' to other subcultures or subsystems. This can be
achieved by physical or geographical isolation of a department or, for
 Kim Langfield-Smith 185

example, by releasing a section of the organisation from the demands of the

budgeting and formal management control systems. For example, to foster
creativity the research and development division of a corporation may be
allowed to manage itself, free from the control systems imposed on the other
sections of the organisation.

National cultures and organisational culture

The impact of national cultures on organisational culture is a complex area for

research, but has implications for the design of control systems in multi-
national companies. Can the same structures, reporting systems and formal
controls that are used in the head office of a company be used effectively in
overseas branches?
It is tempting to associate Japanese cultures with Japanese companies, or
German cultures with German companies. However there may be distinct
differences in culture within the one nation. In these times of increasing
globalisation, and with the growth of transnational corporations, the isolation
of particular characteristics of national cultures becomes even more problem-
atic . It may be more useful to consider the influence of 'societal' cultures on
organisational culture. Societal cultures focus more on subcultures of a
nation, which may be based on ethnicity, region, class or profession. For
example, Birnberg and Snodgrass (1988) considered that in many Japanese-
style societies managers come to an organisation having already learned the
need for group consensus, the role of subset of actions and activities and the
nuances of language in conveying meaning. These societal values lead to a
control system that places less emphasis on bureaucratic rules and incentive
systems.
While national cultures may influence the style of organisational culture,
there are many examples of multinational companies that have more in
common with the dominant organisational culture of the parent company
than with their national culture. IBM has commonly been presented as a
company that has a strong global organisational culture that emphasises
respect for the individual, service orientation and the pursuit of excellence
(Hofstede, 1985). Internalisation of the dominant shared values in interna-
tional companies is often achieved by transferring staff between overseas
posts, conducting centralised training and conferences and imposing stand-
ardised management and control systems. Foreign subsidiaries may also
develop cultures that contain aspects of both their national and parent
company cultures. It is only recently that the differing national cultures within
multinational companies and their effect on control systems design have been
researched (see, for example, Chow et al., 1991).
186 Organisational cultureand control

What is the relevance of culture for control systems?

To understand the relationship between organisational culture and control,

we need first to consider different assumptions that can be held about the
nature of organisational culture (Smircich, 1983; Allaire and Firsirotu, 1984).
First, we can view an organisation as a sociocultural system. Under this
approach culture becomes a metaphor for the organisation. Culture is por-
trayed in terms of systems of cognition and beliefs, patterns of symbolic dis-
course or manifestations of unconscious processes (Smircich, 1983). Control
systems can be viewed as part of the sociocultural system, and will both reflect
and be a part of the organisational culture. To study organisational culture we
would consider organisational structures, functioning and evolutionary
processes, as organisational culture is assumed to be enmeshed within the
social structures of the organisation. Issues of dissonance and incongruity
between the culture and the other components of the organisation have no
relevance, as these are assumed to be part of an attuned whole.
The second approach is to regard organisational culture as a separate organ-
isational variable. That is, culture is a conceptually separate component of the
organisation, either being located in the minds of the members of the culture
or arising from shared meanings and symbols. The organisational culture is
revealed by the patterns of the attitudes and actions of the organisational
members. It is within this view that we can advance the idea that an organisa-
tional culture can adapt, and may be intentionally changed or designed. For
example, many popular management writers (Peters and Waterman, 1982;
Deal and Kennedy, 1982) view culture as a management control tool, which
can be changed to enhance organisational performance. Also, under this
approach, control systems can be viewed as an outcome of, as well as a pre-
condition to, organisational culture (Jelinek et al., 1983). Culture can be viewed
as providing a context for the design of control systems and may itself be a
source of control.

Culture as the context for control

Flamholtz et al. (1985) view organisational culture as an independent variable.

They present a model of organisational control that explicitly considers organ-
isational culture as a component of the control context (Figure 12.1). Culture
was described as facilitating control when the control system is consistent
with the social norms of the organisation - or as inhibiting control when it is at
variance with the shared norms, values, management philosophy and prac-
tices. Culture itself was recognised as a control mechanism in organisations
where knowledge of the transformation process is imperfect and where the
 Kim Langfield-Smith 187

ability to measure output is low. This is consistent with Ouchi's (1980) precon-
ditions for clan controls. Culture becomes the source of control through mes-
sages contained in rituals, stories and ceremonies which relay to
organisational members desirable behaviours. Thus culture is not only a
context for the design of control systems but, in certain organisations, may
itself be a mechanism of control.
Culture has also been described as a filter for perceiving the environment
that sets the action and decision premises of individuals within an organisa-
tion (Birnberg and Snodgrass, 1988). Culture influences the effectiveness of a
control system in two ways. First, culture can affect which aspects of a control
system are perceived by an individual, so that certain stimuli are sought and
others are ignored. Secondly, the culture can affect any value judgements
made about those perceived stimuli. Birnberg and Snodgrass focused on the

FIGURE 12.1 Framework for Organisational Control

r - - - - - - - - - - - - 7 External environment- - - - - - - - - - - - -
.... - - - - Control context ~ Organisational culture - - - - - - - - - - -
: - - - - - - - - - ~Organisationaistructure- - - - - - - - -
I
I
I 4.0Evaluation
I
Core control system •reward
I
I Extrinsic
I Intrinsic
I

I
I

1.0 Planning futrational Outcomes

Goals su system Performance
Standards
- attitude
Work behaviours

I
2.0
3.0Feedback Measurement

Information
process
I
------------------------------

Source: Flamholt, Das and Tsui (1985), p. 38

188 Organisational culture and control

degree of homogeneity of a culture and the differing values placed on co-

operation to explain the different types of control systems that may suit partic-
ular culture - co-operation mixes. A group is considered to have a
homogeneous culture when there are few differences between the values,
beliefs and behaviour of its members, and heterogeneous where differences
exist. Where there are heterogeneous cultures in the one organisation, those
groups will react differently to the same control processes - they will perceive
and interpret control mechanisms in different ways.
A model is presented in Figure 12.2 that explains the degree of formality of
the control system under four possible states. In cell 1 we have a homoge-
neous culture and high value is placed on co-operation by its members. In this
situation the members of the culture possess a common set of values, and
where these are consistent with the organisation's goals there is less need for
an implicit (formal) control system. One objective of formal controls is to com-
municate and reward certain behaviour; however this is not required in the

FIGURE 12.2 The Role of the Control System under Differing Culture-
Co-operation Mixes

Value placed on co-operation

High Low

1. Co-ordination I.Co-ordination
2.Avoid shirking
Homogeneous

Culture
1 2

3 4

Heterogeneous 1. Co-ordination 1. Co-ordinate

2.Motivate 2.Motivate
3.Avoid shirking

Source: Bimberg and Snodgrass (1988), p . 451

 Kim Langfield-Smith 189

situation in cell 1 because of the sharing of values. Therefore the role of the
control system becomes one of co-ordination, achieved by communicating
information for informed decision making. The cell 1 situation describes the
clan (Ouchi, 1979) and clan control. In cell 3 there is a high value placed on co-
operation; however, because of the heterogeneity of the culture, controls are
needed to both co-ordinate and motivate workers to achieve organisational
goals. In cells 2 and 4 the value placed on co-operation is low, so control
systems would include mechanisms such as participation, bureaucratic rules
and firm-wide incentive systems to facilitate goal congruence. According to
Birnberg and Snodgrass, the Japanese culture is an example of a cell 1 situ-
ation and US organisations are cell 4 (or perhaps cell 3). Compared to Japanese
companies, in many US organisations there are disparate values and individu-
als are more concerned with the way their actions affect themselves, in prefer-
ence to the group. Of course, these generalisations about the cohesiveness of
Japanese companies and the nature of the group ethos in both Japanese and
US companies have been questioned in recent years (see, for example,
Hammond and Preston, 1992).

Inferring the nature of the organisational culture

If we accept that the nature and homogeneity of an organisational culture

must be considered in the design of control systems, then we are assuming
that a culture can be separately identified and studied. However, if culture is
such a nebulous concept and if much of the culture resides in the unconscious
shared assumptions and values of its members, how can the nature of the
organisational culture be determined?
Schein (1985) suggests that an organisation's culture can be inferred and
understood by considering its visible artefacts which also act to reinforce the
culture. These include the organisational structure, the information and
control systems, formal goals and mission statements and myths, legends and
stories. However the inferences that we can make from observing formal
structures and responsibility systems are limited and their significance should
be interpreted with care. For example, a formal structure that is apparently
highly centralised with very little delegation of decision-making responsibility
could suggest that the senior management has low regard for and trust in the
abilities of the more junior managers, and believes that they cannot be trusted.
Alternatively this structure could be chosen because senior management
enjoys the day-to-day involvement in operational decisions. Another possibil-
ity is that the structure was imposed by the parent company or was an histor-
ical choice, and the formal structure may bear little resemblance to the informal
arrangements under which decisions are actually made. Management may not
bother changing the irrelevant formal structure, as it is not considered to be an
190 Organisational cultureand control

important issue. So, while the nature of formal structures might be important
in understanding a culture, there are many competing interpretations that can
be made.
An organisation's formal control system may reflect aspects of the organisa-
tion's culture - but only if the controls are actually attended to by managers.
Once again there may be little correspondence between the informal (actual)
control processes and the formal aspects. However even apparently unused
control processes may be regarded as rituals, and can perform a useful func-
tion within the organisation. Apparent contradictions between formal and
informal control systems may suggest conflict to an observer, but to insiders
may be important features of the culture. For example, in a company which is
dominated by engineering or marketing concerns, accounting controls may
form a ritualistic function. The accountants may be content to prepare cost-
based performance reports and all managers may take part in budgeting meet-
ings. However everybody may know that effective control and performance is
related to new product design and successful product launches. Alternatively
such apparent conflicts could be a cause of stress and disruption, creating an
even wider rift between functional cultures and engendering frustration
among accountants regarding their 'rightful place' in the organisation. An
example of the role of the finance function in the production culture of the
National Coal Board is discussed in Box 12.3. In this situation the finance func-
tion was allowed to exist although it conflicted with the dominant culture.

Box 12.3 Achieving control in the National Coal Board

In a study of management control in an area (region) of the National

Coal Board, Berry et al. (1985) provide an excellent account of the interac-
tion of control and organisational culture. Budgets and financial reports
were prepared for each colliery in the area and managers were held for-
mally accountable for results; however financial planning and control
systems were not the dominant mode of organisational control. The area
was dominated by a culture which emphasised production and technol-
ogy. This was heavily reinforced through socialisation processes which,
even for colliery or area managers, was reinforced by underground mine
work early in their careers, limited technical education and living in the
coal mining communities. A high value was attached to loyalty, local
knowledge, experience and skills which were reinforced through the
stories and myths of the coal mining industry. These myths were built
around the role of the battling miner who performed a dangerous and
difficult job in a hostile environment, but always with the support of his
colliery manager. Mine workers and area managers shared common
loyalties to 'our pit' and there was a strong belief in the protection of the
 Kim Langfield-Smith 191

miners' fair share of output targets and earnings. The maintenance of the
myth of the colliery as a 'self-contained, technical entity' also became an
effective way of controlling non-miners and non-engineers, who had not
experienced the same socialisation. Beside this strong culture, financial
considerations were marginal and tended to be the concern of head
office managers, or those few senior managers who were placed into the
area from the head office . Accounting reports and budgeting systems
were largely ignored by area managers and were
not allowed to intrude on the real production-related issues. The
researchers found the content and form of these accounting reports
curious, but suggested that the creation of more accurate economic state-
ments might not be welcomed as it could lead to a shift in power, and
thus affect the maintenance of the dominant culture.
The NCB was managing through vertical and horizontal decoupling
which allowed each part of the organisation to maintain its own distinct
identity and concerns. Strategic and planning issues and models, and
capital expenditure analyses were the domain of the head office and this
knowledge was generally not conveyed to areas . Any long-term plan-
ning at the area level was confined to geographical feasible capacities.
This lack of communication helped to maintain morale and stability in
the areas, in the face of a declining coal market and limited finance.
Horizontal decoupling occurred in the separation of finance and opera-
tions and while this partially reflected the differences in occupational
cultures it also satisfied the role of managing the inherent ambiguity and
uncertainty of the industry by again protecting the production core .
However the role of finance became one of legitimisation and ritual:
accounting reports satisfied external requirements and gave the appear-
ance that the area followed conventional accountability and efficiency
practices. The specific characteristics of the area, which depended on
and was reinforced by the maintenance of the strong mining culture,
provided a means for control in the face of change and uncertainty so
that any attempt to integrate systems and move towards a more finan -
cially-oriented business culture might be counterproductive.

The formal mission statement, or even the goals of an organisation, may not
derive from the shared beliefs and assumptions of the organisational members
but may reflect the senior management culture, or may indicate the new
future cultural direction for the company. However in designing control
systems we need to be aware of the way in which the values communicated in
the mission or goals differ from those of the dominant culture of the organisa-
tion. Organisational myths, legends and stories abound in any organisation
192 Organisational culture andcontrol

where there is a strong culture. Often these are used by senior management to
reinforce the new values of a culture. Sometimes they develop from lower
levels of the organisation. They may reflect the dominant culture or a subcul-
ture, and may even carry a subversive meaning. They are often some of the
earliest messages conveyed to new members of an organisation. In the earlier
example of Hewlett-Packard, stories and myths were very important in sus-
taining the strong culture.

Organisational culture as a management control tool

It is clear from the above discussions that there is a view that many aspects of
organisational culture provide control, and in some situations more effective
control than that based on explicit rules. Many popular management books
operate on the premise that culture is an important variable that can be
managed and manipulated in organisations, to achieve greater control and
effectiveness.

Creating a strong organisational culture

Organisational cultures have been described as 'weak' or 'strong'. Strong cul-

tures are those to which members are strongly committed, which is said to
provide control advantages. Strong and enduring organisational cultures are
often said to be initiated by a strong leader, for example the group's founder,
or an influential reformer (Clark, 1972; Pettigrew, 1979; Schein, 1985). There
are many theories about the way in which founders, or new managers, can
create a strong organisational culture. Gagliardi (1986) described the process
of 'idealisation' where beliefs are transformed into organisational core values
and assumptions. When organisations are created, leaders will have a vision
for that organisation and will direct operations using their specific sets of
beliefs and values. If the leader guides the organisation to success and the
members recognise this, over time this way of doing things will become
accepted by subordinates. This reduces the need for the leader to directly
control the members' activities. It is the shared experiences of success that help
to develop and strengthen organisational beliefs and values. The causes of the
successes become ideals or values that, over time, become important in their
own right. With successive generations of organisational members these
values are increasingly taken for granted, and they become the assumptions
that underlie the organisational culture. They influence the more tangible
aspects of the organisational culture - the work practices, the rituals and
symbols - and will protect and sustain the culture. The culture is perpetuated
and may grow in intensity as shared experiences increase and as the memories
 Kim Langfield-Smith 193

of success are enshrined as organisational sagas or myths. The Hewlett

Packard example, presented earlier in this chapter supports this theory,
involving a successful company with a strong culture led by strong founders.
It can be argued that the mere sharing of core values and assumptions pro-
vides a powerful source of control through the internalisation of corporate
objectives and goals, and reducing the problems of goal incongruence.
However Kotter and Heskett (1992) found that organisations with weak
organisational culture were likely to be just as successful as those with strong
organisational culture. One reason for this is that a strong culture may become
an obstacle to change and a strong but misdirected culture may 'lead intelli-
gent people to walk, in concert, off a cliff'. Also, strong cultures, or subcul-
tures/ within organisations may be at variance with the senior management
culture and conflict with the organisational goals .
A less harmonious view of cultural development is presented by Schein
(1985/ p . 222) who describes the situation of a new organisation where the
leader is surrounded by strong individuals who are unwilling to accept the
leader's vision, assumptions and beliefs . This may result in conflict, negotia-
tion/ compromise, resignations and either good or bad outcomes for the
organisation. Similarly, in an established organisation, the entry of a new
leader may be conflict-free - the process of embedding new assumptions and
values occurs through Gagliardi's processes of idealisation - or may be
fraught with resistance and conflict.
Strong organisational cultures have been held up as keys to success in many
newly industrialised nations, especially in Japanese companies. The example
of the Matsushita company is described in the Box 12.4. Many advanced man-
ufacturing philosophies based on Japanese models have been distilled and
promoted in Western companies. The success of these programmes in non-
Japanese companies depends on an ability to change the underlying manage-
ment philosophies and cultures of the workplace. However the historical
influences that contribute to Japanese culture are very different to the US or
UK experiences, and may cause Japanese practices to be less successful when
introduced into non-Japanese companies. For example, one theory is that
Japanese organisations combine the traditional cultural values of the rice field
with those of the samurai. This creates organisations that have an elitist man-
agerial hierarchy and feudal worker/master relationships. To Western
observers, this gives the appearance of a submissive and deferential work-
force . However, unlike the situation in many Western countries, where people
gain their self-respect through competing against a system to emphasise their
individuality, Japanese workers may achieve self-respect through service
within an organisation (Morgan, 1986/ p. 116). This is not to imply that
Japanese practices cannot be successfully introduced in non-Japanese coun-
tries/ or that practices from other countries than Japan cannot be adopted in
other countries. It merely highlights the influence that national cultures may
have on organisational cultures and the difficulties of implementing practices
within a different national context.
194 Organisational cultureand control

Box 12.4 The culture at Matsushita

In the early 1980s the Matsushita Electric Company was ranked as one of
the 50 largest corporations in the world. Its products were sold under the
brand names of National, Panasonic, Quasar and Technics. The founder,
Konosuke Matsushita, was very influential in shaping the future of his
company. He believed that the firm had an inescapable responsibility to
help employees' inner selves, which could best be realised by managers
being trainers and developers of character, not exploiters of human
resources. A technique of 'self-indoctrination' was used where every
employee was asked at least each second month to give a ten-minute
talk to his work group on the company's values and its relationship to
society. Matsushita was the first company in Japan to have a song and a
code of values. One executive said, 'It seems silly to Westerners, but
every morning at 8.00 a.m ., all across Japan, there are 87,000 people recit-
ing the code of values and singing together. It's like we are all a commu-
nity.' The basic principles, beliefs and values of the company were as
follows:

Basic business principles

To recognise our responsibilities as industrialists, to foster progress, to
promote the general welfare of society, and to devote ourselves to the
further development of world culture.

Employees' creed
Progress and development can be realised only through the combined
efforts and co-operation of each member of our Company. Each of us,
therefore, shall keep this idea constantly in mind as we devote ourselves
to the continous improvement of our Company.

Theseven spiritual values

1. National service through industry
2. Fairness
3. Harmony and co-operation
4. Struggle for betterment
5. Courtesy and humility
6. Adjustment and assimilation
7. Gratitude

Source: Pascale and Athos (1981) pp. 50-1.

 Kim Langfield-Smith 195

Cultural change

The term cultural change is used to describe the natural evolution of a culture,
or more commonly it is used to describe the process of introduced cultural
change. In the organisational theory literature, change is often described as
either first order or second order change. During first order change, the organisa-
tion searches for a solution to problems but is constrained by the limits of the
shared belief systems. Different behaviours from a range of possible behav-
iours are combined into different sequences, but these all lead to solutions that
lie within the boundaries of the existing cultural paradigm. A system, or an
organisation, 'cannot generate from within itself the conditions for its own
change; it cannot produce the rules for changing its own rules' (Watzlawick et
al., 1974, p.22). Thus, it is extremely difficult to break out of this self-directing
search, or even to conceive of alternatives that lie outside the existing cultural
assumptions. Janis (1972) described the concept of groupthink: when individu-
als in a group are dominated by collective belief structures, they will sample
the environment in an increasingly narrow manner, so that those beliefs
become reinforced. The inability to solve the organisation's problems may
result in confusion and in the culture losing its cohesion as the experiences of
failure intensify.
Second order change is said to take place when an organisation finds that it
is no longer effective in managing the problems within its environment. This
may involve a complete change of culture, a discontinuity or transformation of
the cultural paradigm, or incremental change where the organisation goes
through periods of transition. The new manager who might initiate a new
culture is not constrained by the existing cultural assumptions, so may be able
to question the current beliefs and values.
The success of any change will depend on a variety of factors, including the
extent to which new behaviours, and the new beliefs and values on which
they are based, conflict with the old cultural values and assumptions. The
degree to which the organisational members can collectively experience the
success of the new behaviour patterns is also an important factor . It must also
be remembered that, while there may be a change in the dominant culture of
the organisation, subcultures may remain unaffected. There is the view that
not all parts of an organisation need to change, and that a degree of cultural
fragmentation may even be desirable (Weick, 1976).
Advanced manufacturing philosophies such as total quality management
(TQM) and 'just-in-time' systems UIT) rely heavily on employee involvement
and empowerment. In particular, TQM practices require not only changes in
work practices, but changes in attitudes and values of all employees, which for
many companies implies a cultural change. Formal bureaucratic controls are
often regarded as inconsistent with these approaches, especially those that
concentrate on controlling the individual, rather than the group. The types of
196 Organisational cultureand control

controls that are cultivated in these systems are clan controls. Individuals are
encouraged to work in groups, to support their fellow workers in that group
and to be self-motivated though accepting responsibility and commitment to a
high-quality product or service.

Embedding and transmitting culture

Schein (1985, p. 224) described five mechanisms used by senior managers to

embed and transmit culture:

1. what leaders pay attention to, measure and control;

2. leader reactions to critical incidents and organisational crises;
3. deliberate role modelling, teaching and coaching by leaders;
4. criteria for allocation of rewards and status;
5. criteria for recruitment, selection, promotion, retirement and
excommunication.

These mechanisms may be used deliberately or may operate unintended. It is

interesting to note that, apart from the second item, all these mechanisms are
elements of the formal or informal management control processes. Thus the
processes of reinforcing or embedding a culture may act to create control.
The first mechanism is described by Schein as one of the best methods for
managers to communicate their orientation and the aspects that they believe
are important. This is achieved through designing formal control systems, or
simply by managers' remarks and questions that are consistently geared to
certain themes. Formal controls will include performance evaluation and
reward systems and routine reporting systems. An example of the informal
processes would be if the company's new orientation is to emphasise cus-
tomer service rather than market share, when the senior manager may give
important signals during a planning meeting by asking how certain product
changes or other activities will affect the level of customer service. Emotional
outbursts and other reactions may also be used by managers when they feel
that an important assumption is being violated. This will further reinforce the
priorities and values that they wish to transmit. Signals are also created by
what is not reacted to. However such messages will lose their effect if they are
not consistent with the managers' behaviour. The example of the quality drive
in a university department (Box 12.5) demonstrates the importance of provid-
ing a consistent message.
Simons (1990) explained the role of management control systems in
influencing the strategy formulation process and effecting change (Figure 12.3,
p. 198). He argued that senior managers choose to make certain control
systems interactive and to programme others. Programmed controls are dele-
gated to staff specialists, and senior managers are only involved when there
 Kim Langfield-Smith 197

Box 12.5 Promotional criteria in a university department

Traditionally universities have placed far greater emphasis on research

quality than on the quality of teaching when evaluating its academic
staff. However, with the increased pressures on universities for
increased accountability of funds, many institutions are placing greater
emphasis on evaluating the quality of teaching and explicitly consider-
ing teaching skills as an important promotional criterion. However
many senior academics who owe their high position to research exper-
tise, and not necessarily teaching skills, find it difficult to change their
priorities, especially when it comes to considering applications for pro-
motion of more junior staff. So, while research and teaching may be
given comparable status in formal documents of criteria for staff promo-
tion, in many cases these 'new values' are not carried through. For many
the shifting of priorities amounts to a change in values, and perhaps
even a cultural change.
Thus, while the new cultural message may be that quality research
and quality teaching are necessary for 'good academics', any inconsist-
encies over what is considered important by some senior university staff
when considering applications for staff promotions may weaken the new
initiatives.
Any control aspects that could be achieved through performance
evaluation and reward systems are also jeopardised by inconsistencies
between the espoused values and the enacted values.

are problems. When controls are used interactively, managers actively

monitor those controls and intervene in day-to-day decisions of subordinates.
This provides an active forum where important issues can be debated and
challenged, and managerial values and preferences are clearly signalled.
Interactive controls are those that focus on the strategic uncertainties in the
environment. Organisational learning takes place as managers become more
aware and respond to the opportunities and threats that the strategic uncer-
tainties present.
The way that an organisation, and a senior manager, deal with a crisis may
reveal important assumptions and closely held values. The leader's reactions
can provide a rapid learning experience for other managers and employees.
When the 5PC fruit canning company faced a crisis, the managers explained
the difficult situation to the shop-floor workers. The managers themselves
showed their commitment to the company by taking a decrease in pay and,
such was the level of trust between managers and workers, the workers also
agreed to accept lower pay levels. This is described in Box 12.6.
198 Organisational cultureand control

FIGURE 12.3 Process Model of the Relationship between Control Systems, Business
Strategy and Organisational Learning

Business strategy Strategic uncertainties

Organisationalleaming ~
"
Choice of interactive management
control systems by top management

Source: Simons (1990), p . 138.

Box 12.6 The SPC Fruit Canning Company

SPC Limited was established in 1917 in a fruit-growing valley of Australia

and by 1990 was the largest Australian canner and exporter of deciduous
fruit. However a series of expensive and poor decisions left the company
heavily in debt and the dissatisfaction with management led to a board
room coup. The shareholders, mainly local growers and employees of
SPC, sacked most of the directors and elected a new Chairman, who was
himself a grower. A local store owner, Jeff Tracy, was employed as the
General Manager and he admitted: 'I don't know anything about fruit and
I didn't know anything about production, but I do know about people'.
The previous culture was that of blaming the workers when things
went wrong, and there was resentment by workers. Tracy created a
climate for change. There was a flow of information to keep employees
informed of the companies activities and problems in an attempt to
improve the poor employee relations. An advisory committee of workers
met weekly with Board members and the employees were always
welcome to talk to Tracy in his office. However, the company was close
to bankruptcy and redundancies of 25% took place. The survival of the
town was dependent on the survival of SPC and the employees worked
with management to help solve the company's problems. In a secret
ballot 93% of employees voted to accept a lower wages package.
The new management demonstrated that they were serious about cost
cutting at all levels, and they were there on the job at all hours working
with employees. Reserved car parks were abolished, after-drinks ses-
sions were held with Tracy and the managers to get to know staff.
Board members were working on weekends on the hot and messy pro-
 Kim Langfield-Smith 199

duction lines. New ideas from employees were listened to and acted
upon and the Board and CEO were 'Jeff' and 'john'. The communica-
tions newsletter became very important and contained messages of unity
and commitment for all staff.
Profitability returned in 1991 and wages were restored, with the
promise to reimburse much of the wages lost, and a profit-sharing scheme
was negotiated. The Chairman stated that the new unity between employ-
ees and management was a major contribution to the turnaround.
The SPC is an example of a successful cultural change that took place
in a very short time. The survival of the local town and community was
heavily dependent on the survival of SPC. This factor, and the close per-
sonal identification of the employees, management and shareholders
with the company and their history of shared experiences were major
contributors to the success of the change.

Source: A. Sinclair and J. Baird, SPC: New Deal, by Melbourne Case Study
Services, University of Melbourne

Explicit role modelling, teaching and coaching are a part of training employ-
ees and communicating assumptions and values of the new culture. This is
where senior managers, by virtue of their access to resources and systems,
have distinct advantages over other competing cultures within the organisa-
tion. Strategic planning weekends, or retreats, are often used to present new
ways of thinking and to foster joint experiences in a setting removed from the
day-to-day company situation.
The criteria used for awarding rewards and granting status create a learning
experience for staff, especially if accompanied by formal appraisal sessions.
However the espoused rewards may differ from the actual reward system.
The formal rewards that accompany good performance may be emphasised in
performance appraisal sessions, but individuals may know that it is other
aspects that are actually rewarded. These types of inconsistencies may create
perceptions of inequity. This may cloud any 'cultural message' that manage-
ment is trying to reinforce. The example given of conflict in espoused values
and enacted values in a university illustrates this.
When people are assimilated into a culture, changes may need to be made
to their existing assumptions, values and beliefs before the values and beliefs
of the new culture can be accepted. By hiring people who already share some
of the cultural assumptions and values, a culture can be self-perpetuating and
the need for formal bureaucratic controls can be reduced. This is helped by
people self-selecting: choosing to work for organisations where they see the
company's values and vision as consistent with their own self-image.
200 Organisational cultureand control

Conclusion

In this chapter the notion of organisational culture has been considered and
various theories about the interaction of organisational culture and control
have been discussed. However, because of the intangible and ambiguous
nature of organisational culture, it is important when examining theories and
research findings about organisational culture and control that we consider
the assumptions made about the nature of organisational culture.
If organisational culture is the organisation, merely an idea that charac-
terises one organisation from another, then to talk about changing the culture
or control systems implies changing the very nature of the organisation. By
understanding the culture of an organisation, or the subcultures, we can gain
insight into the way control is achieved in an organisation; control is embed-
ded within the culture, within the organisation.
The control advantages that arise from a strong organisational culture are
an area that has not been addressed in conventional control systems theory.
However can it be assumed that strong cultures are convenient resources or
components of the organisation that can be managed to provide effective
control? From a managerial viewpoint, culture can be looked on as a tool of
control only if those shared values and beliefs are consistent with the dom-
inant management culture and the organisational goals. Whether organisa-
tional culture can be used or manipulated as a management tool is debatable,
and any notion of designing an effective culture rests on the idea that culture
is an identifiable artefact of an organisation. Theories of cultural change high-
light the importance of a leader with a strong vision being able to create a new
culture or change an existing ineffective culture. However empirical evidence
of the success of such ventures is sparse. If organisational culture is a
metaphor for the organisation, rather than a distinct organisational artefact,
then theories about effective or ineffective cultures, and prescriptions for
changing the culture, have little relevance!
 PART III

Practices of Control
 CHAPTER 13

Management control systems of

Japanese companies operating in
the United Kingdom
Istemi S. Demirag

Introduction

This chapter reports the results of a recently completed research project on

Japanese companies' operations in the UK. The project focuses on Japanese
companies' management control systems, corporate strategies and perform-
ance measures for their operations in the United Kingdom. The extent to
which these differ from those adopted by companies based wholly in the UK
is an issue to explore, given the current interest in Japanese approaches to
management.
The literature on management control systems of multinational companies
and their strategy indicated that the link between the two were not clearly
spelt out. This was mainly because the strategy literature used various typolo-
gies and definitions (Hambrick, 1984; Langfield-Smith, 1993). This resulted in
fragmented research findings, with often conflicting results. Similarly, differ-
ent definitions are given to management control systems. Moreover organisa-
tional effectiveness or performance are measured using different standards. It
is therefore hoped that the results of this study will help to fill this gap and
show how management accounting, when used as a control or performance
evaluation tool, can help to achieve strategic objectives (Dent, 1990a, 1990b;
Simons, 1987, 1990; Govindarajan, 1988). The overall intention of the project
was therefore to describe Japanese companies' strategy and their management
style and control systems for their UK operations (Daniel and Reitsperger,
1991; Demirag and Tylecote, 1992a, 1992b). Rather than impose a particular
strategy typology on companies, we have attempted to describe Japanese
203
204 Japanese companies in the UK

companies' strategies both at business and at operational level in Europe. The

study also highlights some of the problems these companies are experiencing
in the European markets and how they are dealing with these problems.
Where appropriate, significant differences found between this study and UK
company practice, as reported elsewhere, are highlighted in the chapter.
The first section provides an overall description of the research methods
used for the study. The second section describes the results of the study. It
highlights management control styles of Japanese companies' UK operations
and it describes how Japanese companies are using management control
systems to achieve continuous cost reductions and other strategic objectives
(see, for example, Goold and Campbell, 1987a). It also points out some of the
management control problems the companies are experiencing in the UK. For
the purpose of this study, organisational forms and responsibilities, corporate
strategies, capital investment techniques, and types of performance evaluation
and remuneration systems are used to describe management control styles of
Japanese companies. The last section provides an overall summary of the
results.

An overall description of the study

Case study approach

Given the complexity of the subject the research used a multilayered case
study approach with three Japanese multinational groups (companies) with
manufacturing subsidiaries located in the UK: Group A, Group B and Group
C. 'Face-to-Face' interviews with senior management located in the UK at the
regional and divisional subsidiary management levels were conducted for
each participating group. Participation in the study was obtained from the
Japanese parent company management by a research team based in Japan.
All interviews, which lasted on average four to five hours, were tape
recorded. I spent two years interviewing 31 managers, and participated in
several meetings. I also had access to some confidential management reports
and other documents. From the interviews, I developed and administered a
questionnaire among the senior managers in each of the companies in order
to cross- check my perceptions and interview notes. The study sought expla-
nation of the empirical situation. Throughout the study, for each manage-
ment level, I assured the confidentiality of the information, so that honest
opinions and perceptions were captured in the study. It is for this reason that
the results of the study are analysed in aggregate form only and there is also
no mention of the individual companies or persons who participated in the
study.
 Istemi S. Demirag 205

The companies

Two of the case study groups, Groups A and B, were in consumer electronics
and the third, Group C, was operating in the motor car industry. The compa-
nies were chosen on the basis of their willingness to participate in the inter-
views. For each of the selected three case study groups interviews were
carried out with the Japanese regional company management teams, cascad-
ing down to the lower level of management, within the same company. The
lower level management included both the UK divisional and business unit
management.

The results of the study

Organisation forms and management responsibilities

The study found that the Japanese companies at the parent company level had
strongly decentralised divisional profit responsibilities with highly
autonomous and powerful manufacturing plants (referred to as 'works')
focused on target results. This was supported by strong central management
committee functions and mobility of personnel throughout the organisation
within the units or between them. Regular monthly and quarterly financial
reviews were carried out, but these were not central to the planning process.
The UK operations can best be described as functional management. Each
functional management reported to general management in the UK and func-
tional and product management in Japan. In this complex matrix organisa-
tional structure, several levels of interaction in the decision-making process
were visible. This was particularly clear to us in the area of the design func-
tion. In general the companies' basic philosophy was that a design team was
responsible for profit. However, because the design function in the UK was
relatively new, that type of design department with profit responsibility was
not completely achieved in the UK. Instead the finance department supported
the design department and the other departments in producing profitable
products. This involved preparing profit forecasts for each product and then
discussing these forecasts with the design and other, related departments. The
finance department also assisted in the continuous costing of the models pro-
duced. In addition, in this matrix type of structure, purchasing departments
had the prime responsibility for materials used in the products. The compa-
nies also had, within their design function, a section known as 'Value
Engineering for the Customer' (VEC) responsible for reducing the cost of
products. Design teams were also knowledgeable about the alternative mater-
ials which could be introduced into current or future models.
206 Japanese companies in the UK

The Japanese companies were found to be rather single-minded, perhaps

rather bureaucratic, in many of the things they did. All the business units fed
the top management with information necessary for formulating and imple-
menting strategic plans. This single-minded commitment was a difference that
could be identified as being a Japanese strength relative to the Europeans.
There was much less flexibility in the strategic planning sense in the Japanese
companies. Once plans were established they were vigorously followed and
implemented. It is worth noting that the bureaucratic approach to business
plans is still maintained despite the growing internationalisation of Japanese
business. Underlying cultural attitudes in Japan still induce a conservative
approach to business overseas. The influence of parent company practices is
also apparent in terms of the power structure of the overseas divisions within
the organisation. Traditionally manufacturing and engineering functions have
higher status in Japan and this point is reinforced by one of the managers
interviewed:

Well, you are starting to touch on things like the work ethic. The Japanese system, as
you well know, is not one that encourages entrepreneurial flair. It comes up, it
doesn't come down in terms of the style of the management of an operation or a
company. Although in its environment the company was creating ideas, it was not
allowing the market momentum to filter upwards, which is what the Japanese have
done, I guess, in most segments they have approached . But remember that much that
we do here in the UK is a reflection of what is driven from the Tokyo side. I do not
think our company is a good example of a consumer-led company. I think that it is
more industrial manufacturingin its attitude. It may well be one of the Japanese com-
panies that will, during the course of the next 5 or 10years, not be in consumer prod-
ucts and who will find that its investment return is better suited in other industries.

The difficulty of co-ordinating products against the wishes of the powerful

manufacturing 'works' is to some extent alleviated by the way managers are
trained in Japanese companies. As managers often move from one manufac-
turing plant to another, from one division to another and, more recently, from
Japanese operations to overseas plants, the Japanese managers are expected to
understand the problems of communication and be able to co-ordinate prod-
ucts across d ivisions, manufacturing plants and overseas operations.
This type of management training contrasts very sharply with the training
of British managers, as is explained by one of the managers:

I know that many of the very high level people in our company, directors and others,
have an engineering background or technical background . A very important factor is
that people have worked in factories in manufacturing and have experience of
factory situations and how to create profit and run factories efficiently. Many of the
senior people are not based at Head Office, they are based at factories. To be General
Manager of a factoryis a very high position, very high. It is very common within our
company, and I think in many Japanese companies too, to move from headquarters
 Istemi S. Demirag 207

or from central office area to a factory and then to return to headquarters function.
There is no similarity to the Western style where people sometimesbecomeDirectors
having never worked in a manufacturing outlet, but have some academic or techni-
cal background of another type.

In addition to the unique Japanese style of management with bureaucratic

command structure and good bottom-up communication channels, actual
organisation structures of the companies also ensured that decisions made at
headquarters level were sensitive to all parts of the company and reflected
the interests of the organisation (see Figure 13.1). Even though most of the
businesses were divisionalised with profit responsibilities, in co-operating
and planning their activities managers at each level of the organisation put
the interest of the organisation first in the decisions they made. The fact that
each manager at one stage of his or her career may have worked in other
d ivisions seem to have influenced this strategic type of planning. Yet, in one
of the companies, I also came across problems relating to the co-ordination of
product policies within the European markets; as one of the sales managers
explained:

In our head officein Japan there is a division for consumerproducts which is respons-
ible for all consumer products . They should, of course, balance the interests of the
various product groups. The works, traditionally, in our company have had the
power and possibility to decide which product to develop, to manufacture, which
product to sell at which price. There was hardly any overall co-ordination and any
possibility to develop an overall strategy. The difficulty this division has on the pro-
duction side, I also have it in the SalesCompany here in Europe. I should be able to
say, 'Well, Germany perhaps is suffering a little bit, but Italy is making a profit and
overall in Europe we are still happy because we are still making a profit', but that is
not the case. When Germany is making a loss no-one cares about what we are doing
in Italy or Spain or other countries.

When I discussed this point further with managers, I found that the recently
set up European co-ordination centres were specifically designed to solve this
type of product co-ordination problem within Europe. In the next section, we
discuss these problems as part of the companies European strategies.

Corporate strategies and management control styles

All three of the companies had a mission statement which tended to be

expressed in terms of philosophy and ethics. However this was combined
with more specific strategy statements, pointing out the importance of, first,
market share and then profitability, at the local subsidiary levels. As one
manager commented:
FIGURE 13.1 N
o
00
HEADQUATERS
CORPORATE INTELLECTUAL PROPERTY
-STRATEGIC RIGHTS
PLANNING
HEADQUATERS
PUBLIC RELATIONS MARKETING
- DEPARTMENT
HEADQUATERS
ASSOCIATED ENGINEERING AND
COMPANIES DEPARTMENT MANUFACTURING

GENERAL AFFAIRS HEADQUATERS

DEPARTMENT RESEARCH AND
DEVELOPMENT
- LAW DEPARTMENT
ENERGY AND
PERSONAL AFFAIRS INDUSTRY SYSTEMS
CHAIRMAN - AND LABOUR RE LATIONS GROUP
PRESIDENT DEPARTMENT
EXECUTIVE AUTOMOTIVE PLANNING & QUALITY DISTRIBUTION EXTERNAL MARKET
BOARD VICE PRESIDENTS INFORMATION EQUIPMENT GROUP ADMIN DEPT ASSURRANCE PLANNING DEPT RELATIONS DEVELOPMENT
OF - - - SENIOR I SYSTEMS CONSUMER PRODUCTS
DIRECTORS MANAGING - DEPARTMENT
DIRECTORS GROUP
MANAGING _ CORPORATE ACCOUNTING HOME CONSUMER LIGHTING & AIR CONDITIONING
DIRECTORS DEPARTMENT INFORMATION AND APPLIANCES ELECTRONICS ARCHITECTURAL REFRIGERATION &
COMMUNICATION SYSTEMS PRODUCTS FIXTURES HOUSING SYSTEMS
- FINANCE DEPARTMENT GROUP DIVISION DIVISION MARKETING DIVISION
INTERNAL AUDITING ELECTRONIC PRODUCTS
- DEPARTMENT AND SYSTEMS GROUP

PURCHASING ELECTRONIC DEVICES GROUP

DEPARTMENT
g:,l~~~~J~~~~uP~---"I-----~I----rl----"I----rl----
HEADQUATERS
-STRATEGIC PRODUCTS ADMINISTRATION AFFILIATE OVERSEAS OVERSEAS ELECTRIC
CONTROL DEPT OPERATION MARKETING MARKETING EUROPE
PLANNING DEPT DIVISION - DIVISION - COORDINATION
ELECTRONIC INDUSTRIAL CENTRE
PRODUCTS AND
CONSUMER
PRODUCTS

ADMINSTRATION
TAIWAN
KOREA
HONG KONG
CHINA
EUROPE
SOVIET UNION
OCEANS
ASEAN & SOUTH EAST ASIA
INTERNATIONAL TRADE G RO U P
ASIA - MEC PROJECT SECTION
 Istemi S. Demirag 209

By my understanding the corporate mission of our company is to contribute to society .

That is our final target. To human beings, not only in Japan, all over the world. But we
need a profit to contribute to local society in Europe, the States or in Japan. So profit is
a means of achieving the final target. Of course, to increase the profit, to some extent
we need the market share or some position in the market, especially for the mass pro-
duction. We need some sort of penetration in the market. The most excellent compa-
nies in the world, not only Japanese companies, but European companies, need market
share in order to have profit and to contribute to their societies. So I think the market
share is a factor .

Japanese companies involved in consumer products in the project are now

posing the question, 'Should we stay in consumer goods?' In order to answer
the question, they are evaluating the return and the loss in terms of exposure
that the brand would enjoy, world wide and within the markets in which they
operate. The most valuable thing the companies appear to have is their brand
name. As one marketing manager explains:

Largely our company is not a marketing company, it is an engineering company. It

does not find the involvement in consumer products as interesting as its industrial
involvement or its chemical involvement or its nuclear involvement. It is not the
thrust of the business. But of course within Sony it is; Morita is driving a consumer
company and he is very unusual for a Japanese. He is perhaps the only entrepreneur
that I have ever met, he does drive the business, and the question will be of course
what happens when that entrepreneur eventually dies, or retires or moves on .
Because he is creating this momentum, there is a great difference between him and
the Japanese style. But he, Morita, is dependent on consumer products, we're not.
Sony is in consumer business; I don't know what the figures are, about 90% and the
other 10% is professional studio equipment; that might not represent 10% but that is
assuming that sort of ratio, 90% of their business is consumer products. Last year
13% of our business was in consumer products.

An understanding of the perception of UK subsidiary managers as to how

management control systems (particularly performance evaluation measures)
of their groups related to the overall corporate strategy was an important step
in this research. It was found that the Japanese companies in the study had
adopted a strategic planning type of management control style . Parent compa-
nies used divisionalisation but they did not fail to participate in and influence
the development of divisional strategies by establishing demanding planning
processes and making contributions of substance to strategic thinking. In
general companies placed less emphasis on accounting and management
control systems than on achieving smooth production or good quality prod-
ucts. Performance targets were set flexibly, and were reviewed within the
context of long-term strategic progress. The pay-off to strategic decisions was
sought in the long term, and it was accepted that there might be problems
along the route to building up the core businesses.
The companies appeared to have been organised by functional departments
supporting strong product groups in order to devolve strategic decision
210 Japanese companies in the UK

making. This matrix type of organisation required reviews at several levels:

functional, product group and, more recently, geographic levels with a variety
of inputs for the decision-making process. In these complex matrix organisa-
tions, strategic planning appeared to have sacrificed some clarity regarding
responsibilities, initiative and interaction in and among business units. This
was evident in the case of consumer electronic divisions and in particular in
Group A. However this approach proved useful and appropriate where the
existing portfolio of businesses had little overlap.
Relevant literature on UK companies indicated that many British compa-
nies are using a combination of management control styles, but I feel it might
be difficult to find a 'correct formula' for success. This is because all styles
must cope with basic conflicts or tensions that exist in the role of central man-
agement. Clear responsibilities are highly desirable as they enable tight
control on a business unit, which in turn leads to more motivation to perform.
Broad responsibility and autonomy at the business unit level leads to a greater
sense of responsibility and support efforts to meet long-term objectives, but at
the same time strong leadership produces purpose and direction.
Both strategic control and planning may be consistent with divisionalisa-
tion in organisations. However, if a profit centre is adopted with clear respons-
ibilities, as in many UK companies with the financial control style, then it is
difficult to have a matrix form of organisational structure designed to bring
people together over major decisions. In these organisations, development of
an organisational strategy might be carried out by managers closest to the
business, but at the same time imposing strong leadership from the centre
would not be consistent with the profit centre approach. Similarly tight con-
trols cannot be combined with softer response to performance evaluations. A
strict financial control regime might therefore be inappropriate with strategic
planning. This is probably why some UK companies strike a balance between
the extreme profit centre approach and strategic control which contrasts
sharply with the Japanese companies' strategic planning style.
There appeared to be a strong willingness among the Japanese parent
company managers to propose ideas for business units. Companies were gen-
erally concerned with market share, improved efficiency and cost reductions.
While all the companies were still largely dependent on parent company
functions and manufacturing, there appeared to be clear moves in all the com-
panies to have independent geographic centres set up. Although the basic
design in all the companies was carried out in Japan, companies were increas-
ing their design capabilities locally. All the companies had set up their
European co-ordination centres with the intention that these would become
independent units in the future.

European co-ordination centre: a means of compromise?

All the three companies have recently set up European co-ordination centres
in order to co-ordinate the sales and manufacturing efforts of their companies.
 Istemi S. Demirag 211

They all operated on the basis that the factory was a profit centre. One of the
roles of the European co-ordination centre was to redress the balance of power
between the powerful works based in Japan and individual manufacturing
and sales units scattered around Europe. Many Japanese companies started
production in the UK only after 1985 and have not yet been able to incorporate
British production into their previous sales network adequately. The setting
up of the European centres was also designed to improve the distribution of
the British production into other European sales companies. This reorganisa-
tion of the European control and distribution systems has been one of the
crucial and urgent problems for the Japanese companies.
The reasons behind setting up the European co-ordination centre have been
put by one of the managers in that new division as follows :

What we are trying to do is to set up an office in London with product specialists for
the European market. I would like to have two product specialists for each product
area, one Japanese product planner and one European product marketing manager.
The two, as a team, should understand the various European markets, Italy as well
as Scandinavia, Holland as well as Spain and be able to define on a European level
what product is necessary, a little bit more long term.

One of the managers indicated that the most important question in the last ten
months had been how to set up the European co-ordination centre:

In order to continue, to expand our bu siness here in Europe we have to know in

which direction Europe is going, to a single market perhaps. That is not so easy to
explain in a short time. The United States of America is also a very big market but
they are a single unit. In Europe, there are twelve different countries. They are going
to a single market with a lot of directives from the EC government, so conditions are
quite turbulent or changeable. In America when we proceed to business it is far
eas ier than in Europe. For example, now we are going to have digital TV. European
countries are going to have different systems, compared to the Japanese NHK
systems or the American systems. With such a trend we have to get information,
otherwise we cannot continue in the future.

One of the roles of the European co-ordination centre was to redress the
balance of power between the powerful works based in Japan and individual
manufacturing and sales units scattered around Europe. As one senior
manager from the European co-ordination centre explained:

The first step was to gather information on a European base. The second step was to
have a strong power base here to propose the requirements in Europe to Japan.
That's a very important role. So you see Europe as one market and you can speak
with one voice, rather than different people going to Spain and then Japan and
saying, we need this and that. The third step is to make decisions in Europe so that
we can conclude most of the management decisions here; how much or how to
invest the capital. Now this European Co-ordination Centre is the liaison office of
our head office. We have not got the authority to make decisions here yet. When we
212 Japanese companies in the UK

move into the third stage, we will have to re-organise for that part. This should be a
holding company or an independent organisation.

In order to accomplish the objectives of the European centre, one of the com-
panies established 'task forces' based at the European centre. This involved
establishing 18 different task forces ; nine product divisions/business units
and a further nine functional teams. The functional task forces included, for
example, distribution, tax or finance . Each task force had ten members. In all,
180 people were involved, but the same people could be in more than one task
force. The task force teams often met the consumer products task force team
from the parent factory, or the headquarters people. The objective was to get a
consensus or an understanding at the same level and to speed up communica-
tion. Each task force team had a leader. The leaders were Japanese people
working in the manufacturing side, and sometimes they came from the sales
companies in Europe.

Manufacturing versus sales corporation

It was a surprise to find that the Japanese companies separated their manufac-
turing and sales elements and allowed them to operate quite independently of
each other. Each had its own profit responsibility. The Japanese ethos seemed
to be that, when you keep the sales separate from the manufacturing, the
company as a whole is likely to benefit from the ensuing negotiations between
these units. Each side will drive the other to be more effective and efficient.
However I found that when there were small margins to be made it became
difficult to conclude a satisfactory negotiation. It was often the stronger side
which came out as the winner and traditionally it was the manufacturing
works which had the greatest power in these negotiations. The sales compa-
nies in the UK were responsible for the UK sales, but sales to other overseas
countries were handled from the manufacturing units. There were also links
with the consumer product group and group sales headquarters in Japan. The
problems of co-ordinating the activities of sales companies with manufactur-
ing plants were expressed by a senior UK sales manager:

Each factory, or each sales company, is trying to optimise its own result, of course
not looking at the total picture and the consequences that a certain decision may
have on other products. To give an example, we have a leading position in the
market for full-size camcorders. It is a relatively small segment, it is maybe 6 to 7 %
of the market in Europe, but it is a profitable segment and it is a very stable segment.
Now the factory has decided to drop one full-size camcorder, the top segment. This
influences sales and profitability of other products. We are most likely to lose our
market leader position in that small segment and thereby we become less interesting
for dealers. It is becoming much more difficult in some areas of Europe to try and
increase our television sales, because one factory decided that one product was not
profitable enough and forgot completely about the impact that decision would have
 lstemi S. Demirag 213

on other products. Although the sales companies were unhappy about it the factory
decided to stop it. Instances like this are constantly occurring .

The problems created by the separation of sales companies from the manufac-
turing plants also contributed to the failure of market information being fed
into the decision-making process of Japanese parent company managers from
relevant works. As one marketing manager commented:

People from the works and the factory, as from all Japanese companies, I think, have
alwa ys been very active in visiting countries and sales companies. One of the prob-
lems, I think, is that they are Japanese people, they generally don't speak too much
English and certainly they don't speak Italian and Spanish and French or German
even . So when they come into the country, they first of all visit our sales company -
most of the time they just spend in the office of the Japanese who is representing the
company in that country. When they go out they go out with the Japanese - they
visit a few stores, but they don't really talk to the salespeople, they certainly won't
talk to consumers and I think it is impossible for them to really get a good feeling
and understanding of the market.

Managers pointed out the problems of co-ordinating plans with the manufac-
turing plants in Japan. According to some managers interviewed, the negotia-
tions between the Japanese and the local manufacturing plants did not appear
to be independent. Furthermore certain preconditions such as providing a
profitable budget and meeting medium term plans had to be satisfied.

Capital investment decisions

In my discussions with the senior British subsidiary managers I found that the
senior parent company managers required long and complex formal proce-
dures for major investment decisions. They were also often involved in
detailed reviews aimed at key businesses. The centre appeared to be closely
involved in most areas and there was a definite 'top-down' approach. A
strong commitment to the corporate review of business strategies and plans
was apparent in all the companies. All the parent companies retained the right
to review and sanction capital investments but in most cases these were initi-
ated by the business units. This was normally a part of the strategy review
process. It was rather rare for the centre to reject proposals, and the process
was more like a check on 'homework'. Major new initiatives were funded and
managed centrally, with increasing willingness to invest massively behind a
new strategic direction. Managers appeared to have taken a company-wide
perspective which reduced the importance of the financial implications of a
particular project. Projects seemed to be evaluated in terms of their contribu-
tions to the whole company and the significance of the financial decisions on a
particular project was of secondary importance.
214 Japanese companies in the UK

Overall there appeared to be less emphasis put on financial control than

planning. As one manager indicated:

I think in our company we are very concerned with profit. Also we are concerned
with being seen to be a good employer and producing products and items which are
required for people to enjoy or to improve their lives. I think what we want is to
establish an operation which will exist comfortably in its environment; it will not be
subject to peaks and troughs of the employment cycle or severe growth/depression
periods. A company that will be able to operate over a long time and therefore will
give some security of employment to those people either directly employed or
involved through business relationships. To do that it has to be a company which is
growing in sales value or through profitability or both.

The companies appeared to have developed structures and processes which

allowed them to combine short-term financial focus with long-term strategic
vision and planning. While companies stated that they used discounted cash
flow and payback for the appraisal of new major projects, they did not seem to
provide the essential decision criteria. The pressures to meet short-term
financial targets were not allowed to detract from long-term progress. This
approach did not permit short-termism such as was found in some of the UK
companies studied by Demirag and Tylecote (l992a).
A senior manager in Group B also pointed out how strongly decentralised
divisional profit responsibilities focused on short-term results were balanced
by strategic vision and planning in his organisation.

Clearly you are not going to make a profit out of consumer products in terms of just
the hardware during the course of, say, the next 5 years. So it is very difficult for a
management board to actually make a decision to invest in something that they know
in the short term is not going to show them adequate returns. Because there is too
much capacity in the consumer electronics business. What you are potentially doing is
squeezing margins and some of the bigger ones will just drop off and say they don't
want to play this game. They can only keep pumping money in for so long and then
say, 'I give up, I don't want this any more: Clearly the industry is going through that
dip in terms of new development. We are now waiting for High Definition Television
(HDTV), being the longer-term goal for the consumer electronics industry, being a
profitable opportunity. Whether you get this depends on how much money you wish
to invest to get there, not just in the product but also in the brand. You can't just put it
on hold for four or five years and come back, by which time the brand will have been
desecrated. I think the fundamental question is: does a Japanese corporation believe
the consumer exposure is important to the rest of their business?

Performance measures

The findings showed that there was a considerable similarity in performance

measures used in all the three groups studied. Companies placed significant
 Istemi S. Demirag 215

emphasis on the overall corporate performance. Financial calculations

appeared to have been integrated into productive and market assessments.
Definition of performance was wider than that used in many British compa-
nies; I found that the companies studied placed much more emphasis on
design, production and marketing than on financial control measures. At the
same time I was told that increasingly profitability measures were becoming
important in performance evaluations. Given the recent emphasis on decen-
tralised operating policies for the foreign subsidiaries, this was not too sur-
prising. Profit budgets, sales budgets and cash flow potential from the
company to Japanese operations were also found to be important measures
used by the companies. However it is worth mentioning that return on invest-
ment (ROn was not ranked as a useful financial measure by any of the compa-
nies. I also found that companies often used several performance measures
relating to product costs; these included break-even point, fixed and variable
cost, personnel cost, sales per head, and profit and loss per head. Among the
comparative measures, comparisons with other similar manufacturing units
of the firm in different countries and earlier period balance sheet and income
statements were ranked as the most useful comparative measures.
Japanese companies are very conscious of the fact that, if it is not possible to
reduce costs or keep market share and prices high enough to earn a reasonable
margin, it may not be possible to stay in a particular business line . They use
market-driven product costing where prices are established on market condi-
tions. They appear to be more concerned with market position than short-term
profitability, as one manager argued:

In consumer products especially, you can only get more profitable, I think, through
increasing the volume of activity up to that level whereby you can be strong and
secure. What that level is in Europe, of course I can't tell you yet. The only way you
can make profit is through growth and cost reduction. As I said before, the level of
profitability in consumer products is not significantly high - we are not talking 10%
returns, we are talking about 3%.

This view is based on the principle that, if you can maintain or increase your
market share, you can reduce cost per unit produced. This in turn will enable
you to cut market prices and keep competitors out of business. In the longer
term this will lead to more profitable businesses. After establishing the market
price, companies determined a profit margin which was consistent with the
firms' long-term strategy. The difference between the market price and the
profit margin determined the cost of the products. Once the cost is deter-
mined, standard costs are established by the departments which contribute to
the production of the product, assuming no innovation. Between these two
estimates the organisation selects a target cost. When production starts, each
department tries to cut its cost down so that the target costs become closer to
the ideal. This is very different from the standard costing systems employed
in many UK companies, as the target costing employed by the Japanese
216 Japanese companies in the UK

companies is more a function of prices than of ideal standards, but the target
cost established for production managers is very close to an ideal standard.
The companies did not appear to be much concerned with the way over-
head allocation techniques measure costs. However much concern was
expressed as to how the allocation techniques encourage employees to reduce
costs. In one of the companies studied I found that overhead was allocated
according to the number of parts, and an overhead surcharge was applied
against the non-standard parts. This was a good example of a company which
was trying to cut the number of parts needed, especially the non-standard
parts, for producing a product.
There was some scepticism over the use of financial measures among the
senior management, who were more concerned with longer-term performance
indicators such as R&D activity and percentage of new products introduced.
While there was concern for profit-related financial measures, emphasis
appeared to be placed on market share growth and maintaining it, thus indi-
cating a longer-term perspective. It became apparent to me that companies
were paying more attention to their informal measures of performance than
formal ones. Companies indicated that their management viewed financial
measures only as an aid to decision making. As the companies did not care
much about financial performance I was not surprised to find that none of the
companies appeared to make the distinction between the performance of the
operating units and the managers' own performance.

Profit budgeting

A formal system of annual profit budgeting was common to all the case study
companies. UK management was held responsible by the parent company
management for meeting these budgets but in subsequent discussions with
the senior managers I found that parent company managers were understand-
ing and sympathetic towards the variances from the budgets.
The yearly budgets for the whole organisation were prepared in Japan after
each country independently reported their figures in local currencies. The
process of establishing budgets involved discussions and negotiations with
the parent company management after each manufacturing plant submitted a
proposal for its production plans to Japan, having consulted the European
co-ordination centre on market conditions in Europe. Once the production
plans were agreed, manufacturing units prepared their budgets for the
following year. In preparing the budgets, sales companies guaranteed
purchase of the next six months' projected sales. The Japanese manufacturing
plants 'works' were also involved in these production plans because approxi-
mately 30-40 per cent of materials and components were imported from
Japanese plants.
It was interesting to discover that Japanese parent company management
generally set financial targets. These were initially developed by the local
 Istemi S. Demirag 217

managers and they included profit, return on investment and return on assets .
The annual profit budgets were then prepared by the UK management after
consulting either the parent company management or, where applicable, the
European headquarters. In all cases these budgets were formally reviewed by
the Japanese parent company management who rarely changed the proposed
budgets. The overall philosophy appeared to aim at a consensus position,
through constructive dialogue and negotiations in which both top and lower
levels of management reached agreement on realistic levels of achievement.
The detail of the profit budgets depended on the situation but generally
they included income statement and balance sheet with appropriate support-
ing schedules which included cash flow and capital expenditure statements.
Profit objectives in budgets were determined mainly by improvements from
previous time periods and forecasts of economic conditions, and on the basis
of the company's global objectives. All the companies reported actual perform-
ance of profits against monthly budgets and profit budgets were revised every
six months, if necessary. I found that companies measured variances against
original and revised budgets. A detailed analysis of variance by causal factors
was carried out but the analysis of variance caused by exchange rate changes
was always ignored.

Currency choice of performance measures

All the three companies indicated that financial statements presented in terms
of sterling (local currency) provided them with better understanding of the
performance of their companies' operations and their management (see
Demirag, 1986, 1988). Statements prepared in local currency reflected the local
environmental conditions and were more useful in planning and control than
statements expressed in parent currencies after foreign currency translations.
From in-depth discussions with the senior managers it was found that compan-
ies generally preferred local currency statements for convenience reasons, as
sterling statements were more familiar to UK operation staff. However it was
surprising to find that their policy to evaluate foreign subsidiary performance
in yen or in local currency did not vary between subsidiaries. This might be
because foreign management control systems used by Japanese foreign manu-
facturing companies were not tailor-made for the requirements of each foreign
location and instead they were exported from their Japanese parent compa-
nies . None of the companies translated their profit budgets into yen for perfor-
mance evaluation purposes. Similarly companies did not adjust their profit
and loss account or balance sheet statements for price-level changes. I noticed
that none of the parent companies sent a copy of the translated yen state-
ments. This situation left foreign subsidiary managers unaware of their perfor-
mance in parent company currency terms. This may not matter much as
British managers are evaluated in local currencies. It is perhaps for this reason
that the budgets were sent to Japan in local currency and then they were trans-
218 Japanese companies in the UK

lated into yen by using some company fixed standard rate. When they evalu-
ated the actual performance relative to the budget they used the same corpo-
rate fixed rate used in setting the budgets. They also used the actual rate at the
end of the budget period to translate the actual results. This was generally for
financial reporting rather than internal evaluation reasons.
It would appear that the companies were trying to do two things: first, by
using the same budgeted fixed rate at the end of the budget period, they were
finding out the deviations in profit as a result of changes in local operating
environment; secondly, by using the actual exchange rate at the end of the
period, they were determining total deviations from the budgets, including
the full impact of exchange rate changes on managers' operating performance.
Companies in the survey rarely used hedging for foreign currency movements
and when they did this was not co-ordinated on a company basis.

Responsibilities for currency risk management

Foreign exchange risk management programme was either relatively decen-

tralised or completely decentralised. All the companies had a manager or
managers acting independently or as a group who were formally charged
with the task of managing the exchange risk for their companies.
The companies did not seem to be concerned with the policy impact of
accounting standards and were more interested in the underlying economic
activities. This is different from the findings of studies on UK and US compa-
nies where accounting policies and standards influenced companies' actual
operating decisions. The Japanese companies in this study and also in a recent
study by Bromwich and Inoue (1992) indicated their interest in areas like
target costing, achieving the desired 'yield ratio', reducing costs, and improv-
ing quality and distribution systems, rather than in financial controls. It is
perhaps for this reason that the Japanese companies in the study ignored
foreign currency translation gains and losses for the purpose of evaluating
managers' own performance in the UK. Similarly foreign currency transaction
gains and losses seemed to be managed according to the budgeted rates pro-
vided by the parent company management. While the gains and losses were
separately reported, they did not seem to be directly related to any particular
manager's performance.
It was interesting to notice that most of the manufacturing units in the UK
sold their products in local currency so they did not have any adverse affects
of exchange rate movements on their operating profits. This was probably
because all their output was sold to their sales companies in different coun-
tries and the individual sales companies abroad had the responsibility to
manage their own currency exposure. The companies' basic philosophy on
exchange rate was that there was no incentive to be in a gain or loss situation:
the objective was to be neutral, so you covered everything you could accord-
ing to the budget rate. The decision whether to cover the current situation or
 Istemi S. Demirag 219

not was determined by the subsidiary managers' own forecast of exchange

rates. If their forecast rate was more favourable than the budgeted rate then
they would have the incentive to take cover.

Performance rewards

All company members who were below the grade of middle manager had an
annual salary review by a 'company members' board' which included repre-
sentatives of a union, middle management, senior managers and an executive
manager. I saw their role as to try and come up with a consensus recommen-
dation on wage review. When they failed to reach an agreement, a negotiation
between the company and unions took place . Arbitration was used as the final
step if there was no agreement. It was also found that in addition to this
annual review there was an assessment procedure for every company member
twice a year. This was based on either nine or ten factors which had been
established for several years and which were subject to a review by a joint
working party under the auspices of ACAS, who came up with some amend-
ments to it, such as time keeping, absence and overtime working, and others
such as responsibility, creativity and reliability. Some other smaller schemes to
encourage participation and reward accordingly were also found . These
included methods of improving quality, stock control, efficiency and reducing
costs . Performance of senior and executive managers was generally rewarded
on informal evaluations, taking into account financial and non-financial
performance criteria as previously discussed under performance measures.

Conclusions

The strategic objectives of the Japanese companies operating in the UK were

on the whole directed towards sales and market share growth within the
European Community countries. In one of the companies separation of sales
companies from manufacturing seemed to have caused some problems in rela-
tion to product planning and design which were found to be generally
reserved for Japanese parent companies. All the companies have recently set
up their European headquarters partly to deal with these problems of central-
isation in these areas. Although investment decisions involved bureaucratic
procedures, strategic considerations dominated the decision process. In con-
trast to some of the divisionalised UK companies, Japanese companies operat-
ing in Britain were not yet subjected to short-term pressures. However it is
difficult to speculate as to what things will look like in a few years' time when
the vastly increased pressures for profits arising from the recent revaluation of
the yen and recession have worked through the system. It was found that in
220 Japanese companies in the UK

general companies paid less attention to financial control systems than to

achieving smooth production or good quality products. The Japanese compa-
nies in the study accorded market share the highest importance among the
performance measures used for internal evaluations. While profitability meas-
ures were given increasingly high importance, they were used with great care
so as to obtain consensus position through constructive dialogue and negotia-
tions in which realistic levels of objectives were attained. This approach to per-
formance evaluations was very different from those found in many of the UK
companies, where financial controls were strictly used in order to increase
short-term profits at the expense of longer-term benefits in areas such as
research and development and management training. Financial rewards were
awarded for reducing product costs and improving working environment, but
job security, opportunities for promotion and workers' welfare were also used
as other benefits available to the employees. Foreign currency risk manage-
ment focused on only large transactions as most sales were made in local cur-
rencies. The use of a company forward rate for operating budgets protected
local managers from any adverse impact of currency changes on their operat-
ing performance.

Notes

The author wishes to thank the research boards of the Chartered Institute of
Management Accountants, Matsushita Research Foundation and Sheffield University
Research Funds for providing financial support for the project.
The author is also indebted to many individuals for their encouragement, help and
advice, in particular to Kiyoshi Ogawa, Tomoaki Sakano and Yoshihide Iwabuchi, who
have also helped to set up interviews with senior Japanese managers in Britain; to Peter
Armstrong, Andrew Tylecote, Jane Broadbent, Olav [ull Sorensen, Richard Laughlin,
Tony Berry, and to Keith Harrison and others who commented on an earlier version of
the paper presented at the British Accounting Association Northern Accounting Group
in September 1993; and to the managers who co-operated and gave so much of their
precious time whose names will remain anonymous for confidentiality reasons.
A copy of the questionnaire is available from the author on request.
 CHAPTER 14

Management accounting: the

Western problematic against the
Japanese application
Karel Williams, Colin Haslam, John Williams, Makoto Abe,
Toshio Aida, Itsutomo Mitsui

Management accounting (MA) in Japan as practised by a group of

leading companies is often described and compared to the principles of
MA in the Anglo-American tradition as exemplified in leading texts.
Western textbook MA is shown to be a poor basis for productive inter-
vention because it rests on a defective implicit model of manufacturing
production which takes a single-product, single-process view of produc-
tion and neglects the gains in labour and capital efficiency which can be
realised by low stocks, small batches, faster set-up and improved layout.
Little attention is paid to new product planning and development,
whereas leading Japanese firms give these aspects of practice a privi-
leged place corresponding to investment project appraisal in Western
theory. The Japanese application of MA is shown to be of both intellec-
tual interest and practical importance: it enables the exploitation of
points of intervention which have been invisible (mostly) in the West,
allowing Japanese companies to generate continuous cost reduction.

The object of analysis in this chapter is the difference between Anglo-

American and Japanese management accounting (MA). We started to answer
this question by acquiring information on the Japanese application of MA in
manufacturing firms through an investigation of practices in some Japanese
car and electronics companies. During the spring of 1990, in Japan, we

221
222 Management accounting

interviewed senior management accountants in two large car companies

(Toyota and Nissan) and three electronics companies of varying size (NEC,
Sony TV division and Kenwood). These sources did not provide a straightfor-
ward answer: instead they revealed to us a methodological problem about
how to analyse similarity and difference.
Many of the elements of Western MA can be found in Japanese firms which
use calculative formulae like standard costing, budgeting or discounted cash
flow (DCF). But the frequency of the elements is different when, for example,
DCF is not universally used in investment appraisal. Furthermore the relation
between the elements is different: as Monden (1990, p . 30) observes, standard
costing is mainly used for financial reporting, not for MA. Increasingly it was
realised that these differences of relation amounted to much more than just
shifts in the nuances with which accounting formulae were interpreted and
used. When we reviewed our interview material, we were impressed by quite
radical differences in the articulation of accounting with (non-financial)
market and productive calculation, as well as with forms of intervention in
manufacturing practice.
What our research does show is that there is a substantial disjuncture
between the Anglo-American theory of MA and the Japanese application. We
argue that this disjuncture centres on a difference of representational presup-
positions where the practice of Japanese firms rejects the assumptions of
Western theory. This difference is of considerable importance to any Western
firm whose practice is consonant with Western theory and it may also have
significance for the way MA is generally taught in the West.
This said, it is necessary to enter a few cautionary comments and disclaimers.
This chapter concentrates attention on one particular aspect of manufacturing
activity, the part played by management accounting. We argue that the use
made of MA in Japan, together with the considerable constraints on its use, is
more beneficient and effective than the techniques highlighted in the leading
Western MA texts. From this perspective the Japanese companies are rightly
credited with intellectual foresight and productive virtue. If Japanese manufac-
turing was being viewed more broadly then other, rather darker, aspects of
Japanese practice would demand more attention, as well as a more critical
approach to the exaggerated claims often made by Western writers about the
extent of Japanese productive superiority. We have attempted to confront each
of these broader issues elsewhere, in work which questions the exaggerated and
uncritical Japanolatry of much Western social science (Williams et al., 1991, 1992).

The problematic of Western MA

Our account of this problematic is based on a reading of a number of American

and British MA textbooks . In them it is clear that the standard techniques not
only limit the field of the visible but also promote an implicit model of produc-
 Karel Williams et al. 223

tion which is largely contradicted by the current engineering knowledge of

manufacturing production (see, for example, Schonberger, 1986). This engineer-
ing knowledge discloses many distinctive features of manufacturing activity
which are mostly invisible in the MA textbooks. Point by point, the assump-
tions and assertions made in the textbook exposition of calculative formulae
can be contrasted with the knowledge of production in mainstream production
texts such as Hayes et at. (1988) and in articles such as our own on specific
aspects of manufacturing (Williams, 1989a, 1986, 1991). In the six points below,
the theses of production engineering are set against the antithesis of MA.
Thesis 1: modern manufacturing is a complex multi-process activity where
many of the problems and opportunities arise from the reduction of waiting
time and queues between process points.
Against this, in conceptualising production, Anglo-American MA focuses
on the brief moments when value is added through conversion rather than the
long intervals of waiting when costs are incurred. The standard techniques
promote an analytic perspective focusing on the single product, process
department or investment project. Synthesis through documents like the
master budget is profoundly unilluminating because estimates of sales and
production have to be broken down so that they connect with the analytic
techniques and categories used in lower-level responsibility centres.
Thesis 2: stock reduction is an imperative which realises large benefits
because 'stocks hide problems' like line imbalance and machine unreliability.
The cost of stocks is high mainly because stock movement requires (unneces-
sary) indirects.
Against this stocks are virtually invisible in MA texts which still sometimes
treat inventory turnover only as a liquidity ratio which measures debt-paying
ability (for example, Ricketts and Gray, 1988, pp. 601, 607-8). MA texts never
realistically compute the cost of stocks: the general assumption is that inven-
tory costs equal the opportunity cost of the money tied up in stocks
(Horngren, 1981, pp. 494-7) . This is true even in texts which seem to cover
other possibilities. One such recent text gives a list of costs of holding stocks: it
is headed by the opportunity cost of the capital which is then followed by
'Insurance and Property taxes'. Not only is 'handling' low on the list, it is
unexplained and appear not at all in ensuing discussions which feature only
the capital and ordering costs (Wolk et al., 1988, pp. 307-11) .
The point about high stocks and more indirects is never taken on board
because overhead is a vague residual category which includes everything
except materials and direct labour. Overhead is allocated (rather than decom-
posed), so overhead items can be listed without examining their relative
importance (Ricketts and Gray, 1988, pp. 28,559; Wolk et al., 1988, pp. 165,
173, 181; Horngren and Sundem, 1991, chs 10 and 16; Horngren and Foster,
1991, pp. 35-7, 40-1). The presentation of overhead as 'joint costs which
benefit several products' (for example, Ricketts and Gray, 1988, p. 94) encour-
ages the notion that all overhead is necessary. Discussion does not focus on
how overhead can be reduced because 'the most important thing to note about
224 Manag~ment accounting

manufacturing overhead is the fact that ... it is not easily traceable to specific
units of output' (Wolk, et al., 1988, p. 19).
Thesis 3: with fixed capital equipment in manufacturing, the most important
issue is 'not what you have but how you use it' . Capital and labour productiv-
ity can usually be steadily raised by modifying machines for specific tasks and
designing short-travel shop layouts.
Against this, MA texts assume that capacity and efficiency are fixed at the
level set by 'currently available standards based on efficient operating condi-
tions' (Ricketts and Gray, 1988, pp. 341-2, 357). Machine modification and
better layout are ignored presumably because low-cost improvement is too
paradoxical for the discourse (Horngren and Sundem, 1991; Allen and
Myddleton, 1987; Anthony, et al., 1984; Horngren and Foster, 1991). And the
implicit technical determinism of the upper limit is reflected as language slips
towards a narrow concept of control as policing the negative variances
(Ricketts and Gray, 1988, pp. 321,350,354,356) or implying that, if there are no
exceptions or deviations from planned performance, management need not
spend precious time where it is not needed (Wolk et al., 1988, p. 302).
Thesis 4: faster set-up and changeover is essential if high rates of machine
utilisation are to be maintained as batch size is reduced. Small batches are
essential in a low-stocks, mixed-model factory.
Against this, set-up and changeover do not figure at all in Ricketts and
Gray's main text where factories run without incurring these costs (Ricketts
and Gray, 1988, chas 2-5) . In other MA texts, set-up is discussed in the sections
on EOQ which treat set-up as a fixed parameter in relation to which batch size
should be adjusted as a dependent variable (Drury, 1988; Horngren 1981) or,
having said set-up time can be substituted in Economic Order Quality (EOQ)
formulae for order costs if management is interested in optimum batch size,
evade any need to evaluate this alternative by asserting that 'the switch in
application is straightforward' (Wolk et al., 1988, p. 311).
Thesis 5: modern manufacturing remains a remarkably labour-intensive
activity where internally controlled labour typically accounts for 30 per cent of
costs and 70 per cent of value added.
Against this, the Anglo-American MA texts rarely have anything sensible to
say about labour's share in costs or value added because labour, apart from
direct labour, is bundled into such general categories as overheads as part of
the process of allocating costs. Standard costing focuses attention on direct
labour which is generally supposed to account for less than 10 per cent of
manufacturing costs. This diverts attention from the more important issues
about indirect labour and the embodied labour in materials.
Thesis 6: materials and components usually account for 50 per cent or more
of costs and these items can be controlled through design to cost in the devel-
opment stage and through constructive relations with suppliers.
Against this, MA texts are nearly 100 per cent preoccupied with production.
They say little about the design and development phase and nothing about
the scope for cost reduction in that phase, while the standard category of
 Karel Williams et al. 225

'research and development' either does not appear (Wolk et al., 1988) or hardly
appears (Horngren and Foster, 1991, p. 411) or suggests that development is a
necessary expense (Ricketts and Gray, 1988, pp. 180,495). In discussions of
make or buy and other issues (Wolk et al., 1988, pp . 416-19; Ricketts and Gray,
1988, p . 270) it is throughout assumed that assembler-manufacturers have an
arms'-length financial relation with suppliers.
The problem is not simply that MA restricts the field of the visible; it also
comprehensively misspecifies the points of intervention and causal relations
in the activity of manufacturing. For these reasons we cannot accept MA's
own assurance that it delivers neutral relevant information for decision
making. In manufacturing, this claim is an unsustainable pretension which
covers up a tendentious account of the interests of the firm and a completely
inadequate guide to the problems and possibilities of manufacturing activity.
Any subject (academic or manager) who attempted to use the discourse to
define the field of the visible would literally not know what to do in a manu-
facturing firm . When Ricketts and Gray (1988, pp. 179-80), for example,
discuss how managers might reduce costs they immediately focus on mater-
ials and direct labour; the two options presented are to use less expensive
materials or buy capital equipment to take out (direct) labour. Their field of
the visible does not include the two more interesting and effective options of
improving process efficiency through layout and set-up improvements or
reducing overhead through stock reduction which saves indirect workers.
The broad characterisation of Anglo-American MA which has been pre-
sented is substantially confirmed in the findings of a recent study based on a
large-scale examination of the textbooks (Scapens, 1991). The 'core material
has not changed substantially during recent decades' (ibid., p. 12); 'the terms
cost accounting and management accounting now tend to be used synony-
mously in textbook titles' (p. 10); 'current management accounting textbooks
give particular attention to the classification of costs as fixed or variable'
(p, 15); 'the measurement of performance in terms of variance ... is the primary
method of managerial control' (p, 17); economics was the dominating
influence and profit maximisation 'accruing to the owners of the business' the
basic motivation (p . 23); whilst practitioners found it difficult to implement
the techniques being taught (p. 33).
'Textbooks, in particular, and the accounting education process, more gen-
erally, are still firmly grounded in the traditions of the past' (Hopwood, in
Clark et al., 1985, p. 230) and the major changes in 'manufacturing operations
are hardly anywhere reflected in contemporary cost accounting courses and
materials' largely because it was exceptional 'for an accounting major to take a
course in production, manufacturing processes or technology (Kaplan, 1985,
pp. 223-4). The new directions which were flagged related to accounting's
involvement in 'organizational power structures' and especially the 'rise to
power of the corporate accountant and financial manager' as a 'phenomenon
of the Atlantic fringe countries' (Hopwood, 1985, pp. 232-3). If these issues
were important in themselves, they only related obliquely to our concern with
226 Management accounting

MA's relevance to the productive aspects of manufacturing; though it is fair to

acknowledge Mackey's (1991) chapter in Ashton et al. (1991) which directly
addresses production.
Much the same message is conveyed by recent books bringing together
essays from a number of academic researchers in America and Britain (Bruns
and Kaplan, 1987;Ashton et al., 1991). These register the dissatisfaction felt by
'leading edge' scholars at the gap between accounting research and practice. It
is also recognised (Ashton et al., 1991, p. 4) that what has been embodied in
the textbooks 'entailed the application of neo-classical economic theory to the
problems of decision-making and control'. The models and techniques were
attempts to programme the decision-making and control processes. It was
assumed that the objective of management decision making is to maximise
the wealth of shareholders. Such professional self-criticism is to be welcomed
but it does not make redundant our own strictures because the 'issues in
management accounting' are mostly concerned with the techniques of, and
technical developments in, Western MA . In this sense the approach
(Anglo-American) is insular. The key issue is to demote purely economics-
based techniques from their overriding dominance and to make more room
for social, legal and behavioural considerations. The allocation of costs
(Ahmed and Scapens, 1991) and transfer pricing (Ezzamel, 1992) are
reworked on this basis.
Such concern might appear to be more directly confronted in other exam-
ples of the 'worrying about MA' literature, exemplified by Johnson and
Kaplan's (1987) recent well-received text. It is part of our contention that this
literature proposes what might indelicately be termed 'largely irrelevant'
reforms of MA because it remains within the original problematic which
defines largely imaginary problems. To illustrate this point we will take a brief
closer look at the text of Relevance Lost.
Johnson and Kaplan's title embodies their conservative problem definition.
These authors do not query the object of profit; indeed they insist quite explic-
itly that the role of MA inside hierarchies is to generate substitutes for the
price signals which the market would otherwise provide. Johnson and
Kaplan's problem is simply that twentieth-century American management
accounting has used relatively crude techniques which generate incorrect and
irrelevant signals for internal use . They particularly criticise the general adop-
tion, after 1918, of simple systems which absorb overhead by applying a mul-
tiplicand to the direct labour used in products and processes. In their view
this simplifying procedure reflects the malign influence of external financial
reporting requirements and the high costs of pre-electronic, manual data pro-
cessing. The multiplication exercise is adequate for financial reporting pur-
poses as long as all the overhead is absorbed, but the exercise allocates
overhead to individual products on an arbitrary basis and thus does not
generate adequate and accurate MA information on the cost of individual
product lines. Procedures such as standard costing may have adequately rep-
 Karel Williams et al. 227

resented manufacturing in an earlier era, but various developments, such as

the growth of overhead and the substitution of capital for labour, have ren-
dered the old formulae obsolete.
Johnson and Kaplan's solution is unitary with their problem definition.
After accepting the validity of the neo-Kantian correspondence in an earlier
historical era, their reform proposal must take the form of a neo-Kantian
project. Their solution is to refine the techniques of MA and attach costs more
precisely to products and processes so as to restore the lost correspondence
between accounting reason and productive being. In product costing, the aim
should be to determine long-run costs of production by subdividing overhead
costs and attaching them to individual products. In process control the aim
should be to understand short-run variation in costs by identifying the hetero-
geneous cost drivers specific to the activity. This second project is pressed by
the current advocates of activity-based costing who take up the internal
reform project of restoring a lost correspondence. Through its insistence on
identifying relevant cost drivers, activity-based costing (ABC) does make new
concessions to the diversity of different businesses, but that still allows the old
technical objective of attaching costs to products to be secured and the privi-
lege of the technique to be reasserted.
Critical MA does not succeed as a defensive move within the problematic.
Many of the technical problems of cost accounting are practically insoluble
when for example, it is not always possible to allocate selling expenses accu-
rately to individual products. The general MA concern with predicting the
future creates further problems. In DCF, for example, the financial values
depend on productive and market estimates of operating efficiency and
market share. More fundamentally, even if all the technical estimation prob-
lems could be solved, the reform proposals do not confront the problem that
having better financial numbers within the MA problematic does not mean
knowing what to do .
This issue is nicely brought into focus by one of Johnson and Kaplan's
examples where they discuss the case of a manufacturing firm which, after
MA reform, discovers that 23 per cent of a plant's products account for 85 per
cent of total dollar sales and 400 per cent of the plant's profits (Johnson and
Kaplan, 1987, pp. 240-41) . The activity of throwing light on dark corners tends
to be treated as an end in itself in critical MA. But, of course, the practical
question is, what should the firm do next? The costing information provides
no basis for deciding whether the company should accept low profits or losses
on some of its lines, whether the unprofitable lines should be deleted, or
whether prices should be raised. There can hardly be a simple financial
answer to the question of whether the firm should carryon as a fullter) line
producer. Furthermore the costing information provides no guidance as to
whether and how margins on the unprofitable lines might be increased by
expedients such as shorter, cheaper set up which would improve margins on
short-run, low-volume products.
228 Management accounting

In response to such criticisms, the classic Anglo-American MA defence is

that it is 'better to know the financial numbers'. As Johnson and Kaplan (1987,
p . 241) observe in their discussion of the example we have cited, 'there are
many actions we can contemplate or take once we understand the problem'.
That defence may have common sense on its side, but it is hardly convincing
after the inadequacies of the MA problematic have been exposed. A manufac-
turing firm which knew the numbers in a MA problematic would be disabled
rather than empowered by that knowledge. It would be altogether more sensi-
ble to reject the problematic of MA and, in our view, this is what Japanese
manufacturers have done to their considerable advantage.

The Japanese application of MA in a production and marketing-centred

system

In practice it is seldom possible to make a complete break. Thus the Japanese

do not eschew or proscribe all MA techniques, but they apply these techniques
in a different way in a production and marketing-centred manufacturing
system where the techniques are used by firms that break with the problem-
atic of MA. In Japanese firms financial calculations are integrated into produc-
tive and market calculations; the result is a three-dimensional view which
denies the universal representational privilege of financial numbers.
Furthermore the integration of different kinds of calculation broadens out the
definition of performance and identifies new points of intervention in a way
which undermines the privilege of financial guidance techniques; in Japanese
firms the main practical emphasis is on productive and market intervention
rather than orthodox financial control. The absence of financial control can
partly be explained by the economic context: the interviews were conducted in
the spring of 1990 before the onset of the recession. But this particular chrono-
logical context does not explain the preoccupation with productive and
market interventions. These points can be illustrated and exemplified if we
begin by considering the difference of Japanese practice in the appraisal of
divisional or affiliate performance and in the appraisal of investment projects.
There is less emphasis on the appraisal of divisional or affiliate performance
in Japan because of institutional differences; Japanese manufacturing favours
a loose affiliate structure rather than tight divisional organisation and, at every
level, there is less stockholder pressure for the distribution of dividends. In
appraising affiliates, Japanese companies do not use ROJ and RI measures
exclusively, as in the theory of MA or much current American practice. As
Monden emphasises, Japanese companies use a variety of heterogeneous mea-
sures; these include productivity, sales growth and market share as well as
ROI and RI. This is a significant difference because the Japanese measures
 Karel Williams et al. 229

incorporate a broader multidimensional view of market and productive as

well as financial success.
Investment project analysis is again much less important in Japan for insti-
tutional and practical reasons. Most large Japanese manufacturing companies
have been cash-rich and rationing funds between competing investment pro-
jects has not been a real problem; for much of the 1980s, firms like Toyota
diverted funds into various kinds of financial activity because their manufac-
turing businesses could not absorb the operating surplus which they gener-
ated. When interviewed in 1990, senior Japanese management accountants
who had operated in this context had difficulty in answering questions about
the appraisal of abstracted, individual projects . They insisted that it was neces-
sary to take a company-wide perspective which made the financial numbers
on the particular project a secondary consideration; at Toyota we were told :

if the figures do not appear to be favourable, we try to understand the merit from the
point of all Toyota. And if we judge it worthwhile, we do it ... Because the project is
evaluated in [terms of its contribution tol the whole company, the importance is cal-
culated not only in figures .

Again, this is a significant difference because the company's interest is not

defined as the sum of an iterative series of financial decisions on individual
projects.
The DCF calculations which feature so prominently in Western textbooks
and in American practice have had a relatively minor place in Japanese com-
panies. It is possible to find large, well-run companies like NEC which never
make DCF calculations, and other companies like Toyota which make DCF
calculations, but quite explicitly do not use them as a decision principle. The
Japanese approach to investment is to think in terms of cash flow and
payback: in all three electronics companies, the simple rule of thumb was
payback 'within three years' or 'in two to three years'. For Westerners, it
seems paradoxical that firms which take a relaxed whole company view of
investment should at the same time insist that they expect, and generally
achieve, a short-term payback. For Japanese managers, the two positions are
not paradoxical or incompatible. Rapid payback is regarded as unproblematic
because they put great emphasis on practical intervention for cost reduction,
which is very different from the Western textbook stereotype of financial cal-
culation for cost maintenance. Cost reduction is achieved in two phases: first,
in the product planning and development phase and, second, through manag-
ing operations.
The Japanese break with the Western problematic of MA because they
attach great importance to the phase of development before production
begins, which is almost completely ignored in the Western MA textbooks. The
irrefutable argument for this is that most of a product's cost is determined in
the design stage before it goes into production. Japanese management account-
230 Management accounting

ants also insist that model changeover and the introduction of new models is
the point at which the firm naturally makes its major investment choices.
Generally, in Japanese practice, product planning and development occupies
the place corresponding to investment project appraisal in the Western prob-
lematic of MA. In that problematic it is assumed that firms can be guided to
financial success by the allocation of capital to more profitable projects. In
Japanese practice, the emphasis is on the attainment of target levels of cost
which will not only safeguard profit but also maintain or increase market
share and sales revenue. As we will argue later, capital expenditure is con-
trolled indirectly through the procedures which limit and reduce costs in
model development.
The general Japanese approach to new product development could be
described as design for cost reduction and the role of financial calculation in
this process is to serve as the relay between productive and market calcula-
tion. Design for cost reduction is implemented in different ways because the
problems and opportunities vary according to the nature of the product. Cost
reduction is relatively straightfoward in quasi-commodity product lines. In
television, for example, Sony told us that the main consideration was simply
to reduce the number of components in the new chassis which are introduced
at three-yearly intervals. Kenwood's approach was similar, but rather more
abstract and financial, because the aim here was to reduce the ratio of material
costs to sales or cost of sales. But in the development of complex mechanical
engineering products, like motor cars, Japanese firms use elaborate systems of
cost planning on spreadsheets which are designed to dramatise the internal
trade-offs at the same time as they illustrate the connection to the market.
In Nissan, the 'product manager', who is responsible for the development of
one new model, is given a 'contribution target' for profit by top management.
The unorthodox definition of contribution (as revenue - materials + direct
labour + depreciation + R&D) had been introduced because it suited the task
of new model cost reduction in this particular business; the manager who had
designed the spreadsheet knew, and did not care, that it did not define contri-
bution in the correct Western way. At Nissan, the basic rule is that the product
manager has to make the contribution which is set so that overall costs have to
be reduced. But the product manager is free to spend more on one item such
as capital equipment if he can make savings elsewhere, for example in direct
labour. The connection to the market is established by the provision for
detailed projections of units sold and prices realised on cars produced and
sold at home and overseas; this allows the product manager to consider the
consequences of different kinds of pricing strategy in various markets. The
spreadsheet also allows the product manager to consider feedback linkages;
for example, more expenditure on development may generate product fea-
tures which increase sales revenue.
Some Japanese companies put most of their effort into cost reduction in the
product planning and development phase. Nissan, whose factories have never
 Karel Williams et al. 231

operated at anything like the Toyota level of integration, claimed: 'We changed
our approach from the old approach of minimising cost in the production
phase to the new approach of taking account of cost reduction in the planning
phase.' However the other four companies we interviewed saw cost reduction
in the production phase as a necessary follow-up to cost reduction in planning.
A firm like Toyota maintains its traditional obsession with increasing opera-
tional efficiency and continuously improving labour utilisation.
Generally Japanese management accountants (like the line managers we
interviewed in an earlier research trip) are convinced that the aim of produc-
tion management should be cost reduction through 'Kaizen' steps to improve-
ment, rather than cost maintenance through the normalisation of negative
variance. As Toyota's senior management accountant observed: 'We always
seek Kaizen, not only maintenance of present cost.' And there can be no doubt
that reductions in production cost are regularly achieved in firms which privi-
lege this objective. Indeed in some firms a Kaizen assumption is built into the
firm's internal operating procedures; at Sony TV, for example, the division
sells its products to the parent organisation at a transfer price which is
reduced by 10 per cent each year.
Japanese practice demotes and marginalises cost accounting to show which
product is more profitable: Nissan told us cheerfully 'we threw it away' and
some companies like Toyota appear never to have used cost accounting for
internal MA purposes. In the Japanese system profitability should be guaran-
teed by design intervention before the product goes into production. And,
once the product is in production, product line or batch costing is of little
interest to volume manufacturing companies whose factories operate mixed
model lines and feature rapid low-cost changeover. The more relevant consid-
eration is costs incurred or resources consumed at the factory and shop level
where companies track and enforce progress through budget and target
systems. If budgeting is used extensively in Japanese manufacturing firms, its
articulation to production is quite different. In the Western problematic of MA
(negative) financial variance against a (constant) standard is used as an indica-
tor of the need for intervention. In the Japanese practice of factory and shop
Kaizen (tightening), financial standards are used to enforce constant produc-
tive improvement which is indicated by physical measures.
The philosophy of Japanese manufacturing companies in MA, as in organ-
isational design, is to strip out unnecessary intermediate layers. In the pro-
duction phase, the aim is to move directly from financial targets, which are set
by top management for each factory, to direct physical measures of process
efficiency at plant or shop level. The overall picture in our five companies cor-
responds to what we sketched in an earlier article on car press shops
(Williams, 1991). As a result, in Japanese practice, the balance between
financial and physical calculation is tilted very strongly towards the physical.
The incorporation of the physical breaks the universalism which is such a
characteristic feature of the Western MA problematic because the physical
232 Management accounting

measures vary according to the nature of the plant and process. In a NEC
semiconductor plant the measures are yield and machine utilisation, whereas
in a car company press shop they would typically be power-strokes per man-
hour and die changes per month.
As far as one can judge, there is no general retreat from labour-centred meas-
ures. Many Japanese firms remain preoccupied with the reduction of (direct
and indirect) labour hours. Toyota told us that labour supplied 50 per cent of
the Kaizen cost reduction inside the company, although labour only accounted
for 30 per cent of total costs. And, although each TV set now contains only one
and a half hours of labour, Sony claimed the reduction of labour hours was
still a priority because all the company's competitors were capable of
redesigning chassis to incorporate fewer components. The focus on labour in
production arises partly because materials and component savings are usually
obtained in the design and pre-production phase when investment is also
fixed; thus the responsibility of the factory managers in companies like Toyota
and Sony is simply to ensure 'labour and machine utilisation'. Whatever the
operational targets, the factory level imperative is the same: production man-
agers must intervene continuously and positively to improve the physical
indicator so that the budget target becomes attainable.
If we sum up the argument so far about Japanese practice, our account of
enterprise calculation shows how MA is broken into pieces and used instru-
mentally in Japan. The recombination of elements realises productive and
market objectives which have the same status as financial results; it also finds
new points of intervention and new outcomes such as continuous improve-
ment. These valuable results, which arise from the integration of productive,
market and financial calculation are safeguarded by a number of heuristic
principles which are respected in most Japanese manufacturing firms. Four of
these principles are briefly described below.

Organisational integration of the MA function as a staff contribution to a

team effort

It is well known that the Japanese economy operates with a tiny number of
externally qualified accounting specialists; in 1986 the total number of profes-
sionally qualified accountants in Japan was 7837 (Choi and Hirarnatsu, 1987,
p. 186). From our point of view, we would argue that an external pedagogy of
accounting and the mass production of certificated accounting specialists is
unnecessary in a national manufacturing sector which depends on company-
specific articulations of financial, productive and market calculation; the com-
panies we researched met their own needs for MA specialists who could apply
MA in their individual house style and would thus know what to do .
The detail of departmental organisation varies. In some firms one head
office department is responsible for both management accounting and
 Karel Williams et al. 233

financial accounting; in other firms the two functions are carried out by separ-
ate departments. Whatever the organisational form, the number of head office
specialists engaged in MA is very small. The larger of our two car companies,
Toyota, employed more than 90 000 worldwide; its central accounting depart-
ment employed just 80 executives in management accounting (excluding
female clerical workers). The largest of our electronic companies, NEC,
employed 40 000 worldwide; its central accounting department employed 180
executives in management accounting (excluding female clerical workers).
The number of executives employed in staff management accounting posts
at the plant level is not very much larger in companies such as Toyota or NEC.
Accountants are often to be found on detachment in product or project-
centred teams. Much of the important work on new product development is
carried out in teams which include an accountant. Nissan and NEC both
operate a system of product managers who run mixed function teams; while
Toyota explained, 'We make a team consisting of product planner, marketing,
manufacturing and accounting.' Intractable problems of shop floor Kaizen are
attacked in the same spirit by a project team.

The dissolution of accounting knowledge into other forms of knowledge and

practices of intervention

This is the discursive principle which is the corollary of organisational integra-

tion. The dominance of this principle is given away by a variety of indicators,
most important of which is the attention which Japanese companies give to
non-accounting problems and priorities. A good example of this order of pri-
orities is the attention to low stocks and small batch production which is such
a striking feature of Japanese manufacturing practice. The issue of low stocks
is not problematised because Japanese management accountants have an alto-
gether more sophisticated accounting knowledge of the cost of stocks than
Western MA. The priority arises because Japanese management accountants
are part of a team which is prepared to accept production engineering argu-
ments to the effect that 'stock hides problems' because the team gives produc-
tion engineering knowledge an equal or higher status than accounting
knowledge. Thus, at Toyota, the senior management accountant argued that
the 'visible benefits' of stock reduction in Profit and Loss terms from reduced
inventory investment were outweighed by the 'more important ... invisible
benefits' of production benefits, especially the ability to run mixed model lines
in a small batch factory.
When the invisible benefits are accepted as real , our conclusion is that
Toyota and other Japanese firms are operating outside the problematic of
Western MA because accounting knowledge does not delimit the field of the
visible or define the priorities of the firm. The practical assertion of different
priorities then depends on the integration of different forms of calculation
234 Management accounting

plus the support of appropriate practices of intervention. For example, cost

reduction targets in development are only achieved because they are sup-
ported by the techniques of value analysis and value engineering which are
generally used to take out materials and component cost in the design stage.
The ruthless financial short-termism of Japanese manufacturing does not lead
to productive retreat (as in so many Western companies) because it is tied to
constructive productive intervention.

Intelligibility of accounting information so that (cost reduction) tasks, which

are inherently difficult, do not appear unnecessarily complicated

The emphasis on intelligibility is clearest at the plant and shop level where
the preference for simple physical measures is always justified on the grounds
that such measures are intelligible at the shop-floor level. And in the Japanese
scheme of things every worker must be his own accountant, just as he is his
own engineer. Thus, in Sony TV, the view was that even labour productivity
measures were too complicated, abstract and indirect for shop-floor use; the
division's factory managers were expected to translate a request for 10 per
cent labour productivity improvement into 'remove one worker in ten within
so many months'. From this point of view product costing was disparaged
because it was always unintelligible to the uninitiated. Nissan claimed to have
abandoned standard costing partly because 'the workforce didn't .. .
understand it' and costing failed to provide 'incentives or motivation'.
The principle of intelligibility is not confined to the shop-floor; it is hon-
oured at all levels of the companies studied, even in their top echelons where
capacity for abstract, conceptual thought is not a problem. Considerable effort
is put into the development of company-specific analytic spreadsheets which
list pertinent variables (rather than standard accounting categories) and thus
show at a glance what has gone right or wrong. Table 14.1 shows the spread-
sheet summary of divisional performance which is used at main board level in
Nissan. In this income statement, used for both marketing and manufacturing
divisions, the director can at a glance take in the trend of production and sales,
the importance of exchange rate fluctuation, the progress of in-house Kaizen
and cost reductions on newly developed models as well as the prices of
bought-in components.

Adjustment of accounting practice to productive and market circumstances

We have already described the adjustment of accounting categories to vari-

able productive circumstances, so it may be worthwhile to add something
here about the incorporation of the market into enterprise calculation. A
pattern of differences and similarities emerged from our interviews with the
 Karel Williams et ai. 235

TABLE 14.1 Divisional income statement

No . units sold
Total revenue
Net sales
Marketing costs
Operating profit (A)
Expected no . of units
Expected profit (B)
Difference
(A) - (B)
Sales increase/decrease
Production change
Sales rate changes
fluctuation in exchange rate
Shipping costs
Total
Changes in invoice price
Discounts
Change in marketing costs
Profit/loss a/ c of fixed costs
Cost reduction of in-house manufacturing
Fluctuations in purchasing prices
Cost reduction of newly developed models
R&D

three electronic companies. Only large, successful Japanese companies have

the internal financial resources to make long-term market calculations. But all
three of our companies make the same short-term payback calculation which
is determined by an external variable, the speed at which the product market
moves.
Companies like NEC and Sony can afford to carry unprofitable product lines
on national markets if they believe that presence in these areas is important for
their productive capability or global marketing plans. NEC is the world's
largest producer of semiconductors but its managers admitted that investment
in semiconductors, as in computers, did not meet their payback criterion. Sony
admitted much the same about its presence in the American TV market, which
others had abandoned, but then argued: 'it is very important for us, the US
market is so huge and only a company looking for long-term business and the
236 Management accounting

company which has enough money to invest and survive, can buy the market.'
The small Kenwood electronics company was not in this category of being able
to 'buy the market'. Although Kenwood was using its established hi-fi busi-
ness as a cash cow, the company was so short of funds that it could only afford
research into the next generation of production.
Despite the difference in financial resources, the same basic 'within three
years' payback rule was adopted in all the companies because they faced the
same market and product-life time horizons. Two to three years (with cos-
metic facelifts) is the typical chassis life of consumer electronics products and
all companies aim to recover their investment (including model-specific R &
D) in the production run so that investment can be rolled over into the next
generation of product. The companies also believe that it is impossible to fore-
cast market volumes and prices for specific products over a period longer than
three years . Thus a Kenwood manager argued that all the consumer electron-
ics firms made the same payback calculations because 'even in bigger compa-
nies, the market situation is the same ... [and Japanese] companies do not
chase dreams'. Significantly he then added the point that a five-year payback
would be financially acceptable to his company but that was irrelevant
because his company could not forecast the market that far ahead.
If our view of heuristic principles and forms of calculation has any novelty,
it is because Japanese accounting practice is so often discussed as though it
represented an incremental development of Western MA or can at least be
added to Western MA as a supplement. Thus Bromwich and Bhimani com-
fortably conclude that what the West needs is evolutionary not revolutionary
change in Western MA:

If firms in the West place greater emphasis on meeting cost objectives from the
Japanese, this should not be taken to mean that Western firms are at fault and ought
to adopt the Japanese view automatically. A combination of the best of the two
approaches seems a sensible aim. (Bromwich and Bhimani, 1989, p. 43)

But that conclusion can only be justified by an analytic reading which breaks
down Japanese practice into a series of disconnected elements which can be
added to any other kind of MA practice. We would dispute the conclusion
and deny the premise. As we have argued, the reading of Japanese practice
should be focused instead on the relation between the accounting elements,
their relation to non-financial calculation and forms of intervention as well as
the supporting heuristic principles. It is the combination of internal system
and external connection which generates the possibility of continuous cost
reduction for a manufacturing sector which is not operating in the textbook
problematic of Western MA. Against this background, the project of borrow-
ing and adding discursive elements to Western MA would be intellectually
incoherent and practically ineffective.
 Karel Williams et al. 237

Conclusion

As we explained in the introductory section, our aim was to contrast the

Western (Anglo-American) theory of MA and the Japanese application. We
have not yet investigated the application of management accounting in
American or British firms, where the range of variation in practice is probably
larger than in Japan. The large company British spreadsheets which we have
seen are within the textbook problematic and take no account of Japanese
practice. But it is not possible to come to any conclusion on American or
British practice without much more knowledge. It is only very recently and
slowly that 'management accounting research is starting to confront manage-
ment accounting practice' (Hopwood, 1985, p. 227).
The verdict on the Western theory of MA, as represented in the common
texts, is however clear enough and depressingly unfavourable. Any student
who had acquired the principles of management accounting from a Western
textbook would need re-education before he or she could work usefully in an
accounting role inside a Japanese manufacturing company. If the aim is to
encourage Western manufacturing companies which wish to close the perform-
ance gap with Japan, as we have argued elsewhere, then that is a difficult task.
But Britain and America might usefully begin by discouraging the profession-
alisation of management accounting through external certification and re-
directing the university departments and schools which do little except teach
orthodox accounting techniques.

Note

The British authors have worked together as a team for the past ten years. Team publi-
cations include many articles and books such as Why are the British bad at
Manufacturing? (1983) to 1992: The Struggle for Europe (1989). Their first article with I.
Mitsui was published in Critical Perspectives (1991) .
 CHAPTER 15

Management control in an airline

C. Wilkinson

Introduction

This chapter describes and critically evaluates some of the management

control processes in a large international airline, specifically those controls
which concern the production and usage of management accounting informa-
tion . The major activities to be examined include those of planning and bud-
geting, cost management and the evaluation of performance of the airline as a
whole, its operations and the managers who are responsible for them.
The airline industry is highly cyclical in nature as demand for its product is
heavily influenced by world economic trends, and as a result the profitability
of airlines is extremely volatile by most standards. The industry is undergoing
considerable structural change through a process of deregulation which is
intended to remove many practices which have been perceived as anti-
competitive by governments and international agencies such as the European
Community. The combination of these factors has resulted in an unprece-
dented increase in competition between airlines on a global scale as markets
which were traditionally protected become subject to 'open-skies' policies in
the 1990s. Thus the need for efficiency in operation and effectiveness in
resource allocation decisions have become major concerns for airline managers
in recent years. Management accounting practices which are intended to serve
these ends by providing useful and relevant financial information and perform-
ance measures clearly have a vital role to play in the overall management
control process of any airline.
However it can be argued that many of the standard techniques of manage-
ment accounting are not well suited to an airline environment. An airline can
be seen as a network of interdependent operations in which many costs are
common to the whole network rather than directly attributable to any specific
product or service. Responsibility for operations follows functional roles
238
 C. Wilkinson 239

rather than products because of the highly specialised nature of those func-
tions, so costs are accumulated and controlled within functional areas, such as
engineering or catering. Revenues, on the other hand, arise and can only be
meaningfully accumulated by specific route products, so any attempt at
matching costs to revenues for the purpose of evaluating operational
efficiency must be heavily dependent upon arbitrary processes of allocation
covering almost all significant cost items except fuel.
Airlines are a service industry in which product is consumed as it is pro-
duced. The product cannot be stored, so ideally production and consumption
should match. In practice production levels and their associated costs are
determined well in advance of consumption, when the flying schedule is
planned and published. Most operational costs then become fixed, so in the
short term it is demand management rather than output decisions which
determine the financial efficiency and profitability of the airline. The manage-
ment accounting system must address the needs of managers who are respons-
ible for short-term demand management and longer-term capacity planning as
well as the needs of those primarily concerned with operational efficiency,
cost management and adherence to financial plans or budgets. This chapter
attempts to assess how effectively the management accounting and related
control systems in one large airline help managers to deal with these issues.

The airline industry and its environment

Fuelled by an ever-increasing demand for business, personal and leisure trans-

port, the world airline industry has grown steadily throughout the latter half of
the twentieth century. In some periods the rate of growth has been spectacular
and at other times more modest, broadly in step with changing world macro-
economic trends, but industry pundits confidently expect demand to continue
to grow until well into the twenty-first century. Against this background of
expected long-term rising demand, airline managers have from time to time to
cope with significant year-to-year fluctuations in passenger numbers.
Another key characteristic of the industry is the trend towards deregulation
in what has traditionally been a highly regulated environment. The structure
of the industry is such that most nations have one national airline, often state-
owned, which dominates its domestic market for internal and international
flights. The dominance of the national carriers in their respective markets is no
accident, since market entry has been rigidly regulated by governments since
the early days of the industry. International routes in particular have tradi-
tionally been subject to a high degree of regulation, not only in safety and
security matters, but also in terms of which airlines can operate between
countries, at what capacity and the level of fares which can be charged for a
standard class journey. Capacity constraints are usually the result of inter-
240 Management control in an airline

government agreements, where governments are keen to protect the market

share of their own airline, and standard fares are regulated by the
International Air Transport Association OATA). In effect no airline is allowed
to fly just wherever it wishes, or to determine unilaterally the prices charged
to its customers on many of its routes. Some routes are subject to revenue-
sharing agreements, under which total route revenue is shared between carri-
ers in some predetermined ratio, irrespective of the share of route passengers
actually achieved by each carrier.
By such forms of regulation and collaborative actions, many airlines have
been shielded from the normal competitive pressures of commercial life. Once
an international route licence had been granted, the airlines involved (typi-
cally two, one from each country concerned) enjoyed a duopoly within which
price competition was forbidden. This situation still continues today in many
cases, but now an increasing number of routes are being 'deregulated', that is,
opened to new airlines and freed of restrictive price controls.
Deregulation has already happened within the USA on internal routes,
though these routes are still closed to foreign airlines by international agreement.
The other major route network which is in the process of deregulation is the
international network between member countries of the European Community.
Some routes within the EC, such as London-Amsterdam, are now officially open
to any EC-based carrier, without price control. The inevitable consequence of
such changes is that increasing competition will threaten the traditional market
dominance of established carriers on routes, at the same time providing oppor-
tunities for airlines to compete in new markets from which they have previously
been excluded. Although the process of deregulation is still just beginning, it is
clear that its effects upon the world's major airlines will be profound, as they
evolve to meet the new threats and opportunities presented to them.
For the large national airlines such as Lufthansa, Air France and British
Airways it is unlikely that their dominance of their respective domestic
markets will be seriously challenged in the short term. However the threat of
increas ing competition is of considerable concern to their managers. Profitabil-
ity levels can vary quite widely between routes, depending upon levels of
demand, price and cost structures. It is the more profitable parts of an airline's
route network which will attract the attentions of competitors and which
therefore need to be strenuously defended.
Given free market access by competitors and the possibility of real price
competition, the defence of market share on a profitable route is both vital and
problematic. Established carriers would prefer to avoid price competition,
with its immediate effects on profit margins, and so attempt to shift the focus
of competition to quality of service factors such as punctuality and comfort.
New entrants tend to focus on price in order to gain a foothold in the market,
but then also shift to service quality to improve their margins, as evidenced by
Virgin Airlines on the North Atlantic routes. Competition can also be expected
to lead to increased costs in marketing and advertising. Before deregulation,
 C. Wilkinson 241

passengers had less choice of carrier and so, in times of healthy demand, mar-
keting expenses could be relatively light. All of these factors indicate increas-
ing pressures on costs as well as price levels across the industry. The overall
effects of deregulation have thus been to increase consumer choice while
placing many airlines under severe financial pressure. The fates of Pan Am
and TWA are testimony to the size of the challenge faced by airline managers
as they adapt to the new order with limited degrees of freedom to act.
Cost structures are largely fixed in the short to medium term, the major cost
items being those associated with owning, maintaining and flying a fleet of
aircraft and the cost of employing staff at the home base and at many locations
around the world. In the short term, those costs which vary in relation to pas-
sengers flown are little more than a small proportion of fuel and catering costs,
so total costs are fairly insensitive to fluctuations in output and revenues. In
times of declining demand or market share, airlines are generally reluctant to
reduce capacity, and in any case their ability to do so is constrained by practi-
cal cons iderations. For example, in the short term, there would be little point in
using a smaller aircraft on a particular route unless another route could benefit
from using a larger one, given a stable aircraft fleet. An alternative tactic
would be to reduce the frequency of flights on a particular route, but this has
its own drawbacks. Quite apart from the difficulties of coping with a
rearranged schedule in mid-season, failure to utilise previously allocated
landing and take-off slots could jeopardise an airline's right to those slots in
the future. If a slot is not used, it can be reallocated to another airline which
then takes over the right to continue using it. Therefore any decision to cut
capacity by withdrawing a flight at a busy airport could have long-term impli-
cations for competitiveness on the route concerned. For these reasons, airlines
have little scope to change their total costs in response to short-term variations
in demand. They must therefore ensure that they run their operations as
efficiently as possible at all times and maintain capacity utilisation rates in
order to minimise unit cost per passenger.
The follow ing sections describe the management control systems in 'Inter
Continental Airlines' (lCA), a major international airline, to illustrate and
comment upon the way these control objectives are pursued in practice. The
controls which are discussed are those of a financial or management account-
ing nature. This focus is not intended to imply that other types of controls are
not important, but merely reflects the particular concerns which are upper-
most in senior managers' minds at the time of writing.

Management structure and responsibility accounting

ICA is a large and complex organisation, and as such any attempt to describe
it in detail would be both lengthy and quickly outdated because of the
242 Management control in an airline

frequency of change in the organisation structure. However there are certain

characteristics of the organisation which seem to endure, and which are
described below.
The board of directors of ICA consists of the chairman, the chief executive,
the finance director and several non-executive directors. Although the board
fulfils the statutory responsibilities of a board of directors, day-to-day man-
agement responsibility lies with the executive board, consisting of the senior
managers who are responsible for the various functional areas of the
company, including sales and marketing, flight crew, cabin crew, engineering,
legal, logistics, human resources management, information management and
so on. The size of each functional area in terms of costs or people employed
varies from the relatively small, such as legal, to the very large, such as engin-
eering. Within each functional area there are a number of smaller units and
subunits, each headed by a senior manager.
In responsibility accounting terms, the functional areas can be regarded as
being either revenue centres or cost centres. Where profit centres exist, these
tend to be in peripheral areas of activity, such as running a travel agency
chain, or selling various services to other airlines. Within the major business
area the marketing function, which includes all sales activity, is the aggregate
revenue centre. All other functions, either operational or support areas, are
cost centres. Within the marketing function, there are product managers who
have responsibility for the major products or brands, which are the various
classes of cabin accommodation - first class, business class, and economy. In
this context, a product is a brand, not a route. These managers are responsible
for decisions concerning the advertising and promotion of the products,
pricing and levels of service provided, but are not directly accountable for all
of the cost implications of their decisions which make demands upon the
operations area. For example, if a product manager decided that the
maximum acceptable length of a check-in queue for business class is three
people, this would have staffing and hence cost implications for ground oper-
ations. Though such a decision may be justified in terms of maintaining a
quality of service differential over executive class services offered by competi-
tor airlines, the cost of its implementation would not be borne by the product
manager's main budget, which is concerned with revenue items only.
The major responsibility for revenue generation lies with the sales function,
which is divided into geographical regions and then into individual countries
or groups of countries. Thus the initial aggregation of revenues is on the basis
of where the sale of tickets occurs, rather than by individual route. In many
cases the apportionment of the value of a ticket to particular routes is
inevitably arbitrary. For example, if a customer in Mexico buys a ticket for a
journey from Mexico to London, on to Cairo then back to Mexico direct, the
last leg of the journey would have to be travelled with another carrier, as ICA
does not fly Cairo to Mexico. There are inter-airline rules in operation which
apportion revenue between ICA and the second carrier, but for the K'A legs of
 C. Wilkinson 243

the journey there is no objective, non-arbitrary method of allocating the

remaining revenue between Mexico-London and London-Cairo. (The fare
would be different to the sum of three single tickets, purchased separately, for
the separate legs of the journey.)
Within the operational areas cost responsibility follows functions, not
routes. The majority of costs are incurred by activities which are common to
many or all routes, rather than by routes directly. For example, the cost of
baggage handling is a common cost of all routes which start or finish at a par-
ticular airport. The cost of maintaining an aircraft is a cost common to all
routes on which that aircraft flies, of which there may be several.
The purpose of including the above examples at this stage is twofold.
Firstly, in relation to the issue of organisation structure and responsibility
accounting, they illustrate the difficulty of assigning profit responsibility for
individual routes, as only a small proportion of costs can be influenced or
managed at the individual route level. Secondly, they illustrate that the assess-
ment of route profitability which is done (see later) has an inevitable degree of
arbitrary cost and revenue apportionment contained within it, and therefore
its results need to be interpreted with considerable care.
Given the degree of separation between responsibility for costs and rev-
enues described above, the question arises of who is ultimately responsible
for the profitability of the airline as a whole. The answer to this can only be
that it is the chief executive, to whom the various cost and revenue centre
managers report. Below this level, the managers are not in positions from
which they can make well-informed trade-offs between cost decisions and
their likely revenue implications for individual products, be they brands or
routes. This is a rather unusual state of affairs. In typical manufacturing
organisations it is a fairly simple task to analyse revenue by product line and
the majority of costs will be either direct or apportionable on some reasonable
basis between a limited number of products. In many service organisations
profitability can be computed at subunit level as both costs and revenues are
generally attributable to the same organisational subunit at least, if not specific
products. However the problem is not unique to airlines. Any organisation
which operates a large network-based service, be it a railway company, a tele-
phone company or a mail delivery service, faces a similar situation in which
costs are incurred by the activity of maintaining the network as a whole, and
are therefore largely fixed, whereas revenues are earned from customers who,
individually, use only a small part of the total network operation.

The budgeting process

Budgets and the operating plans upon which they are based are prepared for
the two seasons which make up the ICA year, summer (April to October) and
244 Management control in an airline

winter (November to March). The seasons are budgeted for separately because
the scheduled flying plan and the scale of operations differ significantly
between them. The budgeting process is the same in each case.
The first stage in the process is the specification of broad targets for revenue,
expenses and manpower levels by the chief executive, based upon the desired
level of improvement over the previous corresponding period. This sets expec-
tations lower down the organisation as to what levels of performance will be
acceptable. Then follows the issue of assumptions concerning macroeconomic
trends and market forecasts, which are used by marketing to determine the
flying plan for the season. The plan concerns routes to be flown, their frequency
and capacity. The plan is then developed into a detailed schedule of operations
by the logistics department, a process which, though technically complex, is
handled smoothly by a team of specialists who are experts in the area.
In practice, changes to the schedule are relatively few from year to year,
partly because of lATA's system of allocating landing rights at airports. Any
airport has a maximum capacity in terms of the numbers of aircraft move-
ments and passengers which can be handled in a given period of time. All
major airports already operate at or near their capacity, so landing rights are a
scarce resource. From season to season, landing rights are allocated on the
basis of previous use of a specific landing 'slot'. For example, if K'A had a
landing slot at 4.00 p.m. on Tuesdays at New York JFK last year, they have
first claim on the same slot this year. Given the scarcity of slots at major air-
ports, airlines are reluctant to give them up because of the difficulty of getting
them back later, and because keeping slots is a very effective way of prevent-
ing competitors from establishing new routes.
Once finalised, the operating schedule provides the basis upon which many
operating costs can be forecast, just as a production plan precedes a produc-
tion budget in a manufacturing organisation. It is also the basis from which
the revenue budget can be estimated, given the assumption of price levels and
load factors for passengers and cargo. The revenue budget is submitted to the
chief executive for approval at this stage, and is either accepted or returned for
revision, which may involve subsequent changes to the schedule.
The next stage in the budgeting process is the issue of business plan packs
to cost centre budget holders, which contain pro formas showing the cost
information to be provided. The pro formas contain details of actual expendi-
tures for the corresponding season in the previous year. Each budget holder is
required to compile the budget for the coming period and to provide com-
mentaries on significant variance from the previous the year's plan. Individual
cost centre budgets are consolidated into departmental budgets by the finance
function for review by department heads. The budgets are then submitted to
the chief executive and finance director for review, and any amendments
required by them will be communicated back down the chain.
The following quotation from a budget holder provides a view of this
process from a line manager's perspective.
 C. Wilkinson 245

In practice, what tends to happen is that the budgets are constructed in some detail
by line managers and reviewed upwards through the hierarchical chain with adjust-
ments at most stages, usually reducing the costs. It seems that these cost budgets are
always reduced even further by the chief executive - the cuts were particularly severe
in the Summer 1989 plan - on a top-down basis . The line manager is then faced with
an arbitrary cut to the plan despite having formulated it in significant detail. This
then creates demotivation and low morale amongst the management team .

This quotation illustrates the difficulty faced by functional managers who

do not have a global view of cost and revenue trends across the airline. Whilst
they may strive to be more efficient than before, and may have a good under-
standing of what is feasible in terms of local performance, they cannot, by the
nature of their work, have as clear an understanding of necessary performance
levels across the whole organisation as the chief executive who has to balance
the forecast revenue, cost and required profit equation.
The reality of the budgeting process is that, once the schedule has been
fixed, the level of activity in many parts of the airline is also fixed. For
example, the flying schedule effectively determines the engineering main-
tenance workload. Therefore most cost savings have to be the result either of
efficiency gains or of cuts in discretionary activities such as staff training,
refurbishment of ground facilities or purchasing new equipment. Whilst
senior management would generally prefer saving to result from efficiency
gains, which generally means reductions in staff numbers, cuts in discre-
tionary activities are generally less painful to implement and are likely to be
preferred by many cost centre managers.

Cost reporting systems and cost management practice

At the level of the operating unit or departments within the main functional
areas, there is no single company-wide cost reporting system. Each functional
area has its own cost-reporting system which draws data from the common
general ledger system.
Responsibility for the provision of cost data lies with the accountants within
each function. Although there is a central management accounting depart-
ment, its role is concerned with overall company performance and route per-
formance monitoring, not with cost monitoring in functional areas. Each
function has its own team of accountants, whose role is to provide a service to
that area. This seems to be sensible in principle, as there is little commonality
between the accounting information needs of, say, engineering, catering and
marketing. According to the accountants who provide the service, managers
are able to receive cost reports to their own specification. The general ledger
system is capable of producing reports with a high level of detail, with
246 Management control in an airline

comparisons between actual spending and budget the previous year, or

whatever is required by local management.
So, in principle, each functional area and every individual cost centre
manager within it have the means of rece iving cost reports to their own
specification of content, layout and frequency. However in practice the system
does have its critics amongst the managers who receive the reports. To work
effectively, the system depends upon managers being able to specify their
requirements to their accountants, and to ensure that the data from which
reports are compiled are accurate. Some managers appear to lack sufficient
technical understanding of accounting and or motivation to review and
improve the cost information they receive. Some report that entries frequently
appear under incorrect cost codes, causing managers to regard the system
as unreliable and contributing to some degree of ambivalence about the
budgetary control process.
Both of the above points reveal that in some cases cost control is not always
regarded as a high priority, even amongst managers who, nominally at least,
hold budget responsibility. This tendency is not evenly spread across
the whole airline, however. Those functional areas which are essentially
production-oriented, such as engineering and catering, appear to be more
budget-conscious than others. These are areas in which cost standards can be
most easily set using industrial engineering principles, as opposed to areas in
which the 'right' cost level is largely a matter of judgement based upon
changing notions of appropriate levels of customer service.
ICA has undergone a marked shift in senior management priorities during
the past decade, from a preoccupation with increasing customer perceptions
of quality throughout the mid-1980s, which was achieved with great success,
to a focus on cost control as the recession of the early 1990s began to take effect
in the form of a decline in demand across the whole industry. It does appear
that, in those parts of the organisation where the drive to improve quality of
service levels was most apparent, little attention was paid to cost management
during the period of rising passenger numbers and high profit levels. Lack of
attention to budgetary control in these areas allowed cost reporting systems to
deteriorate through lack of necessary updating as business needs and operat-
ing practices continued to change. When the market downturn carne and man-
agement needed to focus on cost control to achieve profit targets, some parts
of the organisation found that they no longer had the reporting systems neces-
sary to manage costs effectively.
This problem was recognised by ICA's senior management, who in 1990
instigated a wide-ranging review of overhead spending. This activity value
analysis exercise was akin to a form of zero-based budgeting, in which all
activities were evaluated and questioned, with a view to cutting out those
activities which were not considered essential, and thus saving their associated
costs. The first phase of the exercise considered 11 'overhead' departments,
which employed approximately 4000 (that is, about 40 per cent of the non-
 C. Wilkinson 247

operational staff). The analysis and evaluation took about four months, using a
team of 30 managers who were seconded to the project from their usual jobs.
The exercise identified £40 million of potential cost saving, or 16 per cent of the
cost of all activities evaluated in this phase, which could be achieved within a
two-year period. The majority of these savings were payroll costs.
A common concern amongst managers involved in the project was that, even
where savings were achieved as a result of this exercise, there were insufficient
ongoing budgetary or cost control measures to ensure that overhead costs did
not steadily creep back up to their former levels. The majority of managers
recognised that better cost control was essential for the airline to remain
profitable, since costs had been rising faster than revenues for some time . On
the other hand, many doubted that they had the control systems required to
manage costs on an ongoing basis and thought that they would take some time
to develop. Ad hoc attempts to reduce the fixed cost base such as that
described above may be necessary from time to time when costs are perceived
to be 'out of control', but in themselves are not substitutes for adequate
controls in the hands of those who bear responsibility for cost management.

Monthly management accounts

Management accounts are prepared each month by the central accounting

function for the executive board. These accounts consist of a series of reports
which cover several aspects of the financial performance of the company and
its operations. They include monthly financial accounts (income statement and
balance sheet), cash flow analysis, detailed cost and revenue centre reports,
schedules of capital expenditure and an analysis of route performance. As is
usual in most companies the purpose of these accounts is to enable senior
management to monitor performance and progress towards the anticipated
year end outturn and to identify functions (or routes) which are underper-
forming in some respect. Heads of functions are expected to account for
significant operating variances by providing explanations as to the causes and
what actions they propose or are taking to remedy the problems.
The production and purposes of these accounts is fairly unremarkable in
that those activities would be regarded as standard practice in almost any
organisation. They provide a mans of 'keeping the score', a comforting ritual
which confirms that all is well or, if not, at least the impression that things are
being controlled, that attention is being directed at problems which need to be
addressed. The one slightly unusual aspect of this management accounts
system is the lack of integration between these company-wide summary
accounts and the cost-reporting systems specified by and tailored to the needs
of lower-level managers. In most organisations the management accounts
would serve as the major cost-reporting vehicle. This is not the case in K'A .
248 Management control in an airline

Whether or not this hinders control to any significant extent is difficult to

assess, but it does have the potential to create translation difficulties as the
level of analysis moves from the company to function to department.

The evaluation of route profitability

leA evaluates the financial performance of each of its routes by allocating

operating costs to them and comparing these costs with route revenues. The
costs allocated in this process are all costs which are directly attributable to the
activity of flying aeroplanes, such as aircraft depreciation and maintenance,
fuel, cabin crew and flight crew costs, catering, insurance, landing fees and so
on. All non-flight related costs such as marketing, selling and the costs of
service departments are excluded. By means of a rather complicated process of
allocation, all operational costs are assigned to routes using a variety of alloca-
tion bases. For example, crew costs are allocated on the basis of duty time,
fuel costs on the basis of distance flown and according to type of aircraft,
catering costs on the basis of passenger numbers and so on . The extent of cost
allocation means that route costs cannot be precise, but the system seems to be
a reasonable compromise between accuracy and expediency.
Within these limits route costing serves a number of useful purposes in
addition to allowing the airline to meet certain obligations to provide data to
industry authorities. It provides a basis for evaluating fare levels, and it serves
as a cost model which is complementary to that which is based on functional
costs. Most importantly, it provides an indication of the relative profitability of
different routes when the cost data are combined with route revenue. Like
route cost data, route revenue data can in some cases be difficult to assess with
precision, for reasons already explained, but, within the limitations of the
exercise, the degree of accuracy is considered to be acceptable. The difference
between revenues and costs is normally referred to as route profitability, but
because marketing and other central costs are not allocated to routes, the
'route profit' is actually a measure of route contribution to central overheads
and profit. Nevertheless as such it still provides a reasonable guide to the rela -
tive profitability of routes and an indication of the economically strong and
weak parts of the network. Route performance is evaluated regularly, using
three key measures which affect profitability:

1. yield - the average revenue earned per passenger carried on the route;
2. unit cost - the average cost of providing each seat on the route;
3. load factor - the proportion of route capacity sold.

Relative route profitability is expressed in the form of a 'profit index', which is

defined as the ratio of total route revenue to total route costs . Thus a break-
 C. Wilkinson 249

even route will have an index value of one . Using the three measures defined
above, the profit index can be computed using the formula:

profit index = yield x load factor/unit cost

Analysing route results in this way facilitates understanding of route econ-

omics and can provide pointers to ways in which the performance of a route
may be improved. For example, improvement on a route which already has a
high load factor must come from either increased yield or reduced unit costs,
whereas a route which has a low load factor may be improved by increasing
the seat occupancy rate, perhaps at the expense of reducing yield.
The relationship between yield, unit cost and load factor is critically
important in airlines and in many other service industries whose costs are
largely determined by their planned level of operations which, once decided,
cannot be changed easily in the short term. If a route's load factor falls, it is
unlikely that a smaller, cheaper aeroplane could be substituted unless
another route required a larger plane at the same time. Therefore the capacity
and costs associated with any route are effectively fixed for the season at
least, and typically a year in advance owing to planning constraints. This
means that, in the short term, route performance can only be improved by
increasing capacity utilisation, or yield, or both together. Achieving both may
be difficult, since this would mean selling more seats while raising the
average fare paid.
In the longer term, the relationship between a route's load factor and
profitability provides a useful analytical device for aiding decision making
about future resource allocation. For example, a route with a high load factor
but low profitability may indicate that capacity should be reduced, whereas
high profits associated with a low load factor would suggest the need for
further marketing expenditure to stimulate demand. So long as the limitations
of the cost allocation and route profitability computations are recognised and
taken account of in the decision-making process, this type of analysis appears
to serve a useful role in the long-term planning of the route network.

Managerial performance evaluation and reward system

Throughout K'A every manager's performance is formally assessed each year

by reference to a set of criteria known as 'key result areas' (KRAs) which are
individually agreed between managers and their direct superiors. A typical
manager would have between five and seven KRAs against which their per-
formance is evaluated. This method of evaluation was introduced on the
advice of external consultants in the mid-1980s as part of a performance-
related reward system. The general idea behind the system is that the most
250 Management control in an airline

senior managers first agree their KRAs with their director, and then agree
KRAs with and for their subordinates so that, providing all subordinates actu-
ally achieve their KRAs, the senior managers will also achieve theirs. By this
process, a senior manager's KRAs cascade down the organisational hierarchy,
and in doing so become translated into more specific targets for each reporting
subordinate manager. The type of KRAs that any manager has is largely
influenced by the specific responsibilities associated with their job, although
some may relate to the personal developments needs of the manager con-
cerned. A typical set of KRAs might include projects to be completed by a
fixed date, levels of operational efficiency to be achieved, customer service
standards to be met, staff development objectives and so on. Within the
ground rules of the system every manager's KRAs should include at least one
'financial' objective, which for many managers is simply to manage their
department within their budgetary target. The weighting of each KRA in the
overall assessment of the manager is a matter for discussion at the time when
KRAs are set.
The formal assessment takes place in interviews between managers and
their superiors and results in an overall grading of performance, which can
range from less than satisfactory to excellent. This grading is used to deter-
mine the level of performance bonus earned, but in a rather indirect manner.
The bonus earned is actually determined by reference to the average per-
formance rating of a manager's peer group, so that a 'very good' performer in
a 'very good' peer group could receive less than a 'good' performer in a
'satisfactory' peer group.
The total bonus pool available for distribution under the scheme is deter-
mined by the accounting results of the airline. The details of precisely how
individual bonus payments are determined are not generally well understood
by managers, who tend to see the bonus calculation exercise as a rather myste-
rious process which is performed by a computer programme. Thus the link
between the degree of achievement of KRA targets and the amount of bonus
received seems somewhat tenuous to many managers. A number of managers
expressed the opinion that, whilst they did their best to meet their KRA
targets, the money itself was not much of a motivating factor because the
amount of bonus earned was difficult to reconcile with the quality of their
achievement.
Apart from this aspect of the system, the use of KRAs was generally seen by
managers as a useful control device. The system is flexible enough to incorpor-
ate a wide range of targets, and by careful choice of KRAs and their relative
weightings, performance targets can be individually tailored to the respons-
ibilities of any managerial position. For many managers, the inclusion of one
or more financial KRAs was the only means by which they considered them-
selves to be held accountable for budget achievement. However, given that
the performance review is a once-a-year exercise, that the penalties for not
meeting a financial KRA are extremely difficult to quantify and that the accu-
 C. Wilkinson 251

racy of the cost-reporting system is open to question, it is unlikely that the

KRA system in itself can provide an effective incentive to motivate budget
achievement. This was not a major objective at the time when the KRA system
was introduced.
The performance evaluation system appears to play an important role in
integrating and communicating business objectives throughout ICA, and the
review process provides some degree of accountability for managers' overall
performance. The extent to which the prospect of financial rewards help to
motivate performance is difficult to assess, but there seem to be reasons to
doubt that they provide a major source of motivation, given the general lack of
understanding of the way the system works. The inclusion of financial KRAs
provides some degree of budget accountability, but does not compensate for
the lack of a comprehensive budgetary control system in an organisation
which is seeking to manage its costs more effectively.

Evaluation of the management accounting system

The earlier sections of this chapter have described the major components of
ICA's management accounting system (MAS). The processes of planning and
budgeting, cost reporting, cost management, and evaluation of product
profitability and managerial performance contain nothing which is particu-
larly unusual. They are standard control processes which will be familiar to
almost any middle or senior manager in any reasonably large organisation.
They are rooted in the concept of responsibility accounting wherein an organ-
isation is divided into subunits which are accountable for the achievement of
local operational and financial objectives.
ICA is a 'unitary' (as opposed to divisionalised) organisation; all its major
activities are concerned with the maintenance of a complex network of opera-
tions which transports passengers and freight between hundreds of locations
around the world. It is a classic functional bureaucracy (see Chapter 4) with a
high degree of task specialisation within functions and a requirement for con-
siderable integration between functions at the operational, control and deci-
sion-making levels. Most of the elements of ICA's MAS are operationalised
within functions; for example, the planning and budgeting, cost-reporting and
managerial evaluation components are all conducted within functional
departments. These accounting practices and the accounting model on which
they are based provide the technical rationality of the organisation within
which management control is exercised. This particular rationality tends
to promote and reinforce the primacy of functions to those who work
within them, at the expense of the vision of the organisation as a system of
interdependent parts.
252 Management control in an airline

Some components of the MAS do cut across functional divisions, as for

example, with the budget review process which is conducted at executive
board level and the evaluation of route performance. However these are quite
remote from the vast majority of managers and the processes involved are not
well understood, because they are simply not made visible to the majority. On
one level everyone can appreciate the product as an amalgam of functional
inputs which fuse together at the point of service delivery and that product
quality is totally dependent upon successful integration between functions at
this point. However the controls which have the most immediate impact upon
the majority of managers are concerned with their own area of functional
responsibility only, which creates a certain tension between what is perceived
to be desirable from the organisation's viewpoint and what is required to
satisfy more parochial functional objectives.
The existence of such tensions raises the fundamental question of how an
organisation like ICA identifies appropriate units of accountability upon
which to base its MAS. In many organisations the choice is fairly straightfor-
ward: accountability units may be divisions, operating sites or product
groups, based upon the assumption of a reasonable degree of independence
between units. This independence allows for autonomy in local decision
making and optimisation at unit level. Such assumptions are not tenable with
regard to the functional constituent parts of an airline which together form a
complex operational system. The sheer complexity of the system and the
bounded rationality of decision makers preclude the possibility of systemic
optimisation coinciding with optimisation at the subunit level. There is no
easy answer to this problem, but recognising its existence can at least sensitise
managers to the limitations of accounting models as vehicles for optimisation
across systems whose operational and economic logic is not well represented
by the MAS. In ICA, it can be argued that the MAS tends to obscure the func-
tional interdependencies which exist at the operational level, and tends to
promote suboptimising behaviours by its concentrated focus upon within-
function operations.
The evaluation of route performance can also be argued to be essentially
reductionist in approach. Focusing on the performance of individual routes
rather than functions is just another way of cutting the cake from an overall
systems perspective, notwithstanding its usefulness as an aid to short-term
demand management and seat capacity planning on existing services. The
accounting method used is heavily dependent upon arbitrary cost and
revenue allocation processes. It is therefore unlikely to provide a suitable basis
for major decisions concerning the overall size and shape of the route network
or the capacity of individual operational functions for overall system optimis-
ation purposes. But it is exactly these issues which the management of ICA
need to address when contemplating their response to the environmental chal-
lenges now facing them and their competitors.
 C. Wilkinson 253

Implications for management control systems design

In Chapter I, management control was defined as 'the process of guiding

organisations into viable patterns of activity in a changing environment' . This
short definition encapsulates the fundamental control issue within ICA, that of
how to encourage managers to adopt behaviours which are appropriate to the
degree of environmental change to which the organisation is exposed.
Environmental changes necessitate changes in priority amongst various
dimensions of organisational performance, as well as the need to reconfigure
the organisation to take advantage of new competitive opportunities that arise.
The picture which emerges from the description of its control procedures
suggests that ICA now has a problem in finding a style of control which is
appropriate to the demands of its present situation. The key issue for the
company concerns how simultaneously to ma intain its high quality image,
increase its operational efficiency and compete effectively in new markets as
they become open to it. This final section is an attempt to relate the issues of
finding an appropriate style of control to the nature of the business and its
environment.
In the classification of business types described by Goold and Campbell
(1987a, 1987b), ICA is clearly a 'core business' . Its major activities are all con-
cerned with the provision of air transport and there is no significant
diversification outside this primary activity. Goold and Campbell suggest that
under these circumstances, a strategic planning style of management control
works best. By this they mean that top central management should be closely
involved in setting the strategic direction of the company and creating an
appropriate organisational culture of shared common values to support the
strategy. Top management in ICA provided just such a style of leadership
throughout the quality improvement initiative of the 1980s.
Goold and Campbell further note that there are potential weaknesses inher-
ent in any core business. Firstly, lack of diversity across industries exposes a
company to greater variability in earnings than would be expected of a
diversified company, as industry fortunes wax and wane. International air
transport is notoriously vulnerable to world economic conditions, so consider-
able variability in earnings is to be expected. Secondly, changes in the business
environment can make a particular style of management less appropriate,
eventually creating a need for difficult changes in style. Both of these points are
clearly relevant in the ICA case. The recipe for success throughout the 1980s
which was based on quality first and foremost was appropriate in a buoyant
market, but required radical amendment to include cost and efficiency consid-
erations as market conditions deteriorated. The process of introducing further
changes in management style and control processes in response to the
demands of a more competitive environment is unlikely to prove any easier.
254 Management control in an airline

McCosh (1990) draws the distinction between what he refers to as 'positive'

controls and financial controls. Positive controls are those control system compo-
nents which are aimed positively at the fulfilment of corporate goals, particu-
larly goals concerning quality of output and market share. Financial controls, he
argues, are aimed at avoiding corporate catastrophe and tend to be based on his-
toric accounting measures. The nature of competition within the international
airline industry indicates that ICA will in the future have to compete in terms of
both quality of service and price. Quality controls are already well established,
but need to be maintained and further developed as consumers become more
demanding and competitors improve their own quality standards.
Increasing price competition indicates that ICA will have to become more
efficient in order to satisfy the return requirement of investors. It has been
argued that existing financial controls at operational levels in the company are
distrusted by managers, are underutilised and have not been developed to
meet the requirements of those managers who are in a position to control costs
effectively. More fundamentally, as has been argued in the previous section,
the accounting model of the organisation is more likely to encourage subopti-
misation of operations than to promote informed debate about future resource
allocation decisions.
This case illustrates the tension which exists in any business between the
pressures for short-term financial results and the need to build long-term, sus-
tainable competitive advantage. In the present recession this tension seems to
be particularly acute in ICA because of their high fixed-cost base, which can be
argued to be largely a consequence of their long-term competitive strategy.
The precise form of ICA's strategy for obtaining sustainable competitive
advantage has yet to emerge. Several possible strategic alliances with other
airlines have been considered and rejected. Such alliances could lead to more
efficient utilisation of the route network if a suitable partner with a comple-
mentary network could be found. Even if a suitable alliance was to emerge,
the problem of long-term system optimisation would remain, and in a yet
more complex organisational context.
ICA needs effective controls at the operational and strategic levels. This case
study has identified some of the problems of control system design in a
network organisation, where interdependence between constituent parts of
the network needs to be recognised and co-ordination facilitated through the
control system at both of these levels. The case should stimulate debate about
the extent to which existing management accounting models and techniques
are capable of satisfying such needs, or whether new approaches are necessary
within the MAS and other parts of the management control system.
 CHAPTER 16

Control and the National Health

Service: some psychology of
managing health care with budgets
and cash limits
Derek Purdy

Introduction

As the opening chapters indicated, the issue of control is one which can be
approached from a variety of perspectives. The approach taken in this chapter
is through a psychological perspective. It is a psychological perspective because
the focus is upon the way in which individual managers have responded to
change in relation to the accounting system in their hospital, and the individual
choices which these managers have made in relation to the financial manage-
ment information with which they have been provided. In coming to this work,
the author accepts the personal construct psychology of George Kelly which
notes that it is possible for individuals to construe and respond to the same
event in different ways, amongst other things (Kelly, 1955). The notions used to
model this work are simplified because of the limited space available for this
contribution, the limited time available to work with individuals, and the fact
that it is not possible to detail all psychological issues.
The Otley and Berry (1980) model of control has been selected because the
author considers this a simple but sound explanation of the basic psychologi-
cal approach of individuals to control. The model suggests:

1. The need for an objective.

2. The fact that there is an objective simultaneously requires (an individual
to produce) a predictive model about the objective.
255
256 Control and the National Health Service

3. In order to ascertain if the objective has been attained requires the object-
ive to be measurable.
4. Finally, there must be an opportunity for any appropriate intervening
action, in order to harmonise the objective and the actions being undertaken.

This last phase is important, because the mechanistic approach to control gen-
erally portrays the objective as something which is preset and unalterable, and
consequently the actions concerned with control are directed towards produc-
ing conformity to the unalterable objective. In the Otley and Berry concept, it
may be that the object remains unaltered and conformity is sought through
amending actions; however the more dynamic aspect of the notion is that it
also allows for the alteration of the objective to bring about conformity, so that
it is possible (for an individual) to say that things are under control.
Examining the effect of something like budget information upon a manager
is a potentially very complex undertaking. It will involve the background of
the manager and his or her current state of knowledge as well
as the organisation in which the manager works. To simplify the analysis of
data from conversations with health service managers, eight psychological!
organisational issues were formulated (Purdy, 1993a, 1993b) and these are
explained later.
The author wanted to conduct research work which would aid the under-
standing of the way individual health service managers deal with financial
management accounting data, and explored the possibilities with some 13
hospitals and health authorities before gaining access to a teaching hospital. In
the hospital in which the author eventually worked, a senior member of staff
was appointed his liaison person. She ensured that the author did not conflict
with any ethical issues, through discussing what he wanted to do, then
arranging for a cross-section of managers with different approaches to
handling accounting data, finally reading and commenting upon drafts of
reports, both those given to staff and those for external publication. Although
she had an overall control, nothing the author wanted to do, or wanted to
write, was altered.
One of the reasons that the author wanted to examine accounting practices
and effects within the NHS was because the NHS was changing. Previously it
was considered that one of the most suitable ways of studying accounting prac-
tices and effects was at times of change (Wildavsky, 1975; Hopwood, 1983). The
next section briefly outlines the context of these changes to the NHS .

A brief context

For a long time UK governments have been concerned about the increasing
costs of NHS health care, so, to help control expenditure, the Labour govern-
 Derek Purdy 257

ment introduced cash limits in 1976. Simultaneously the report of the

Resource Allocation Working Party (RAWP), which introduced the idea of a
mechanistic formula for funding health districts, was implemented (Perrin,
1988). Since 1979, Conservative governments have reduced expenditure on the
NHS and altered its structure so that central government has direct adminis-
trative control over resource allocation to Regional Health Authorities
(McGuire et al., 1991).
Until the early 1980s, the NHS had many functions devoted to the curing
and care of patients, with hard-working and well-meaning doctors with
ambiguous responsibilities. When new nursing management was introduced
in 1974, a vast bureaucracy arose, with a wide range of responsibilities which
people were not trained to assume. Some nurse managers lacked the technical
grasp to handle complex nursing problems, and the budgetary grasp to
accompany the financial responsibilities of nursing, training and deployment
of staff (Strong and Robinson, 1988).
A study of the senior managers in a regional health authority in the mid-
1980s found that they varied in their ability to understand financial manage-
ment accounting data and this affected their ability to use such data. The most
conspicuous example was the regional nurse manager, who initially had a
limited ability to construe financial management accounting data, but
improved this ability through natural learning while working with the
accountants (Purdy, 1991).
In the early 1980s the government wanted information about both what had
caused expenditure increases and what would be the costs of the anticipated
patterns of health care. In general this information could not be provided
because the NHS lacked the necessary management structure, control mecha-
nisms and accounting systems. To promote change the government imple-
mented the findings of its enquiry into NHS management which
recommended general management at all levels . The report wanted responsi-
bility pushed to the point of delivery of care, where effective action could be
taken, and anticipated the development of budgeting at unit level (National
Health Service, 1983). The concept of a 'unit' could vary from a hospital to a
group of wards within a hospital. These changes were recommended to
provide direction and personal responsibility for developing management
plans, so that subsequent output could be measured against objectives, plans
and budgets. Apart from each hospital being headed by a general manager,
the organizational form was open to individual interpretation. This appeared
to be a simplification of the control mechanism compared with the previous
team management, operated since 1974 (Perrin, 1987),by the most senior man-
agers from administration, estates, finance, medicine and nursing.
The government wanted to reduce total health expenditure, as well as
increase the efficiency of the NHS . It wanted more accountability and more
information for this and for planning. When the existing team management did
not move to produce these changes, it instituted management changes which
258 Control and the National Health Service

would. It replaced one hierarchical system with another and brought in better-
paid general managers to install general management systems. These systems
would encompass more objectives, more rules and more standardisation to
provide more information which would demonstrate the accountability of the
service. The government's gradual imposition of absolute cash limits can be
considered to be a form of rule-based control. The Conservative government
planned to move the hospital service to a form of market-influenced control
with the notion of a split in the service between providers of health care, such
as hospitals, and a new orientation of purchasers, for example general practi-
tioners. One part of the new forms of information was to facilitate this market.
The way in which the government has viewed the issues in the NHS is in an
autocratic top-down manner. This would appear to be no different to the rest
of the public sector. It made policy to set the cash rules for general managers
from the top. In turn successive levels of general managers were constrained
with a cash limit by their superiors, throughout the hierarchy. In this way, the
top-down objective for managers, at the lower levels, was to provide direct
patient treatment and care within the cash constraint.

A framework for the study

Having considered the more overall and generalised context of the NHS, the
focus now moves to the individual managers of ward nurses. This was the
level of management which was reported to be of poor quality and weak with
financial matters between 1974 and 1984 (Strong and Robinson, 1988).
The enquiry conducted with the ward unit managers (WUMs) was a longitu-
dinal study. It comprised two conversations held with each individual manager
during December 1988 and again in May 1990. The conversations dealt with
financial management accounting data, the associated systems, how these were
handled and any perceived changes. This was an approach based upon the
individual WUM, so that differences between the WUMs were expected, even
though it might be anticipated that the types of issue raised by them might be
common. For example, the way in which an individual expressed their interest
in budgets, and the ways in which the individual described them and used
them, was considered to be directly related to the individual's personality and
pattern of leadership (Argyris, 1952). Consequently there is a simple listing of
eight psychological!organisational issues which were anticipated:

1. The budget is imposed without the manager's influence.

2. This budget is more related to cash limits than to the unit's work.
3. The financial management accounting data received are from a source
hierarchically above the manager in the hospital, and are based upon a
custodial accounting system for financial accounting.
 Derek Purdy 259

4. The accounting data are inadequate and untimely for the manager. The
accounting system does little to alter this situation and it is evolving to
deal with cash-restricted budgets.
5. The manager has little appreciation of budgets and associated matters,
and has not been trained in these areas.
6. The manager does not have financial awareness and does not under-
stand the data, and consequently he or she has only a limited ability to
use it.
7. The ability of a manager to control the expenditures from the cash-
restricted budget will depend upon the extent to which he or she can
control the activities of the area, and the extent to which the expenditures
on these activities coincide with budget.
8. The manager considers that specific knowledge about the budget should
not go any further, because, for example, the manager is responsible for
the budget.

It is not desirable to classify these eight issues into other groupings, because
of the manner in which some of the organisational and the psychological
issues combine. For example, the notion that 'the budget is imposed with-
out the manager's influence' is at the same time a notion about the
organisational structure and about the nature of the manager within that
structure.
In this context, a budget can be considered as a financial quantification of
a plan, that is, a plan explained in monetary terms. A budget can also act as
a control, where the financial components of the budget are an objective
which can be measured and acted upon. The budget objective can form
an unalterable quantity that has to be attained, or it can represent some-
thing quite flexible which moves with events. These issues are re-examined
later.
Of course these are not the only issues which could arise, but, from the
author's experience with others and from a position outside the NHS, these
seemed to be the ones which might occur with 'naive' individuals. This frame-
work was based upon the notion that, assuming that most of the managers
were likely to corne from within the NHS, the WUMs would be 'naive' in rela-
tion to financial management accounting data in general, and any new data in
particular. Furthermore, it is impossible to know the situation in an organisa-
tion like the Hospital in advance of any study.
From this perspective, it was anticipated that a naive recipient of financial
management accounting data would have difficulties in handling the data,
and as a possible consequence of that naivety could choose to ignore the data.
It was also possible that a naive recipient might improve his or her awareness
of that data, and might learn through using the data, or through explanations
about the data, or through formal training, or any combination of these
actions.
260 Control and the National Health Service

The study: the ward unit managers

The ward unit managers worked in a teaching hospital which claimed to be

one of the earliest to introduce

1. clinical budgets (budgeting based around the idea of a doctor as a centre

of activity);
2. resource management (a concept of planning the deployment of funds in a
hospital, rather than allowing them to go into the same areas that they
had in perhaps previous years); and
3. unit budgets. The unit budget was the provision of funds for the structure
which the hospital designated to be a unit. (During the course of the
study, one of the original ward unit managers left and was replaced.)

In 1985/6 the hospital took the then rare decision to form clinical directorates.
With this structure, doctors became responsible as clinical directors for the
care and the costs of a unit of wards. Ward unit managers were appointed to
help administer each unit and occupy the position of budget holder for the
unit. They also took over some parts of the roles of the existing nursing
officers. The majority of WUMs in this hospital had been nursing officers. In
those instances where the WUM was not a nurse there was a senior clinical
nurse to advise.
In the four units studied, the WUM was responsible for a group of between
four to nine wards. They had the management responsibility for all of the staff
on the wards from the registrar downwards. The WUMs were responsible for
administering the unit's budget for all ward activities. The exceptions to this
were the student nurses, and central services such as meals and cleaning. In
the previous three years each WUM had received a cash-restricted budget for
the whole unit. When the study started, the central accounting system pro-
vided each WUM with a monthly computer printout for the whole unit, but
nothing about the individual wards.
The printout contained a summary of staff, an expenditure summary outlin-
ing staff, drugs, medical supplies and surgical equipment (ward supplies) and
the amount of the annual budget. The expenditure summary also included the
overall monthly total of cash paid out by the finance department for the unit,
and the cumulative total of cash paid out during the financial year.
The conversations with the WUMs were approached in an open manner.
They were told of the author's interest in financial management accounting
data and asked to talk quite freely about their work in relation to any aspects
of these. Notes of each conversation were taken, then written up and given to
each manager for amendment or comment. (See Purdy, 1993a, for further
details of the methodology and notes). These notes form the basis for the
 Derek Purdy 261

remainder of this section, which is a summary of the WUMs' position in rela-

tion to the framework of eight issues formulated earlier.

Imposition of the budget

The WUMs observed that the unit's budget had been imposed upon them.
This was after all of them had in some way questioned, with the hospital's
general manager, the way in which the budget had been set. Consequently the
managers felt that they had no influence upon the budget and the amount of
funds with which to operate the unit. The only way in which a budget might
be increased was if a manager could identify and make a successful applica-
tion for funds in an area which the government considered a priority.

The budget is a cash limit

The budget was imposed in the form of a cash limit, and it had been derived
using the assumptions of the government and NHS systems, as well as the
previous practices of the hospital towards allocating funds. The WUMs con-
sidered that there were problems with such a derivation: it was based upon
outdated precedents which understated each unit's needs for staff; the govern-
ment's cash limits calculated staff costs using the mid-point of the salary scale,
but usually staff were above this point. Consequently WUMs recognised that
the budget was faulty and would be immediately exceeded as soon as the
financial year started. To deal with this, WUMs had reconsidered staffing in
the context of maintaining safe nursing levels . Such was the pressure to reduce
staff that one WUM produced a new nursing establishment to do so. In order
to work sensibly, each WUM produced their own independent budget related
to the unit, up to the cash limit.

The source of financial management accounting data

Each WUM received the budget from the general manager and the monthly
printout of other data from the finance department. The WUMs felt these
printouts were of little use, being expenditure summaries.

The adequacy of the data

This monthly printout was a problem for the WUMs, since it invariably con-
tained inaccuracies, was always late and did not structure the data adequately.
262 Control and the National Health Service

By way of improvement some WUMS wanted less aggregation and more

analysis, and some wanted separate ward data.

The appreciation of budgets and training

Initially, although the budgets did not make sense, the WUMs thought they
had an appreciation of budgets. Over time they produced their own indepen-
dent budgets yet each operated their own budget in a different manner. Two
WUMs had been promoted from within the hospital and had received in-post
training concerned with budgeting; the other WUMs had pre-hospital training
which included budgeting, yet they still could not understand the unit budget.
Each of the WUMs had fostered budget understanding on their own initia-
tive.

Financial awareness and data use

The WUMs' ability to use the data was hampered initially by their lack of
appreciation of the derivation of the budget. Those from outside the hospital
were additionally hampered by a lack of understanding about the working of
the unit. In contrast to the others, one manager had discussed the monthly
unit printout with the ward sisters. This WUM went on to obtain similar
monthly details for each ward which were passed on to the ward sisters, so
that they could see the financial effects of their actions. This manager's idea
was that the responsibility for certain expenditures fell to the ward sister, and
she would be in a better position to control expenditures if she could construct
a budget for her ward.

Expenditure control

Initially the WUMs found it difficult to control expenditures because they bore
no relationship to the unit's activity. By way of reaction they produced their
own independent budgets. The way in which these were controlled varied
with the manager. One manager was keen to pursue budgeting with the ward
sisters so that they became more involved with financial affairs and expendi-
ture control. Another produced very precise budgets which were closely moni-
tored against actual costs, and closed beds at times to remain within the
budget. A third had prepared a budget, but the subsequent control was
centred upon ensuring that any expense was currently necessary, irrespective
of its inclusion in the budget. The fourth had controlled expenditure by chang-
ing the nursing system, budgeting on this basis, then monitoring events
closely.
 Derek Purdy 263

Financial data and responsibility

The WUM was responsible for the administration of the unit's budget, and
was the recipient of the financial data about the unit, but the actual responsi-
bility was with other people for the expenditures in some areas such as the
ward sister in relation to direct patient care. It is in the areas of responsibility
that the greatest diversity of practice occurs amongst the WUMs. This diver-
sity occurs with the control of activity on the wards, the control of the budget
for these activities, the manager's perception of the manager's responsibility
for these activities, the manager's perception of the responsibility of ward
sisters and the ward sisters' influence on the activities on the wards and the
resulting costs.
Like many of the other seven issues, discussion about financial data and
responsibility cannot be taken in isolation. These elements are also entwined
with other issues. One WUM had discussed the unit's monthly financial data
with the ward sisters in order to further the ward sisters' responsibility and
involvement with the budget so that eventually they could create their own
ward budgets. These would have defined the activities of the ward sisters and
perhaps made these activities more visible to the ward sister, as well as pro-
viding her with the basis of a control mechanism. Ward budgets would also
enable the manager to have some greater insights into ward activities and like-
wise provide the opportunity for the manager to exercise more specific
control.
Initially a second WUM had involved the ward sisters with the financial
management accounting data about ward supplies, and medical staff with
drugs, because these were considered to be their areas of responsibility where
they had the opportunities to keep expenditures within the budget. At the
second conversation this involvement was considered to be greater than ever,
and the ward sisters had profiles of staff levels for their own ward and were
trying to minimise expenditures on ward supplies and drugs.
The third WUM, initially, was providing the ward sisters with monthly
budget data about staff, and discussing these at monthly meetings, together
with the costs of ward supplies and drugs, in order to get them to contain
expenditures. The formal budgets were not discussed as they were considered
to be irrelevant to the unit. The WUM did not want to pressurise the ward
sisters into cash restriction, but rather to encourage them to minimise expendi-
ture and keep within the budget.
The fourth WUM did not intend to pressurise the ward sisters as the WUM
had been subject to enough pressure. Initially the WUM intended to obtain
ward data and delegate the budget for ward supplies. Although the ward data
were received, they were for the sole use of the WUM, and the delegation to
the ward sisters did not occur. The WUM had tried to involve the doctors in
minimising expenditure, but essentially the WUM was the person who had
the responsibility to control the unit's expenditure.
264 Control and the National Health Service

Control at an individual level: cash limits, plans and budgets

There are a number of entwined notions relating to the individual and control
which come out of these find ings. It can be seen that the WUMs have been
provided with a 'budget'. Previously a budget was defined as the financial
quantification of a plan, but as far as the WUMs were concerned there was no
plan relating to the unit. There was a plan by the general manager, since each
WUM had been provided with a cash limit which had been imposed on each
unit. The general manager's plan was to ensure these cash limits were not
breached by any WUM. Although it took them several years, the WUMs even-
tually produced their own independent budget, which, from their perspective,
was a financial quantification of a plan concerning the unit. Thus the cash limit
was not really a budget for the WUMs until they each produced their own
independent budget for potential expenditures. It would appear that the total
cash limit of the individual units was a budget for the general manager,
because it was the financial quantification of the planned allocation of the hos-
pital's total cash limit . It would further appear that, in order for there to be
control of the control mechanism of a budget, there needs to be some personal
recognition of or identification with that mechanism. If this personal link does
not exist, then the managers are left to manage a cash limit and not a budget.
The Otley and Berry (1980) control model has been cited as a simple but
sound surrogate for an individual's psychological processes relating to
control. It has also been observed that there are likely to be other factors, such
as learning and training, which are associated with their control model, and
which make it more representative of an individual's control processes. The
ways in which these factors combine could make control either more or less
effective at any point. The factors concerned will vary from the context of the
organisation to the individuals involved.
The hospital's general manager had been given a cash limit, which was both
a plan and a budget, and which had been imposed by the regional health
authority. In terms of the Otley and Berry model, this was the objective which
the general manager had to control. The operational objective for the general
manager was to allocate the hospital's cash limit. This appeared to be rela-
tively straightforward since it followed the ways in which the hospital's cash
limit had been apportioned in previous years. Both the objective received by
the general manager and the objective delegated by the general manager were
cash limits. Both could be considered as plans and budgets, since they were
identical in kind and they were identical in total. There were predicted object-
ives which could be measured and altered according to the amounts of money
calculated to be spent and the amounts actually spent by the hospital in total.
When the general manager imposed the plan and the synonymous cash
limit upon a WUM, a WUM only perceived the cash limit part, which was an
objective received. The operational objective for a WUM was not the same as
 Derek Purdy 265

for the general manager, because the WUM had to allocate cash to the actual
matters of patient treatment and care, and not merely produce another cash
limit. Such a situation meant that, if a WUM ignored the planning perspective,
there would be no objective, no prediction, no measurement and no alteration
of these matters. At any point there would only be spent cash, and then
unspent cash up to the extent of the cash limit. The underlying activities
would not be specified or specifiable.
This meant that a WUM needed an objective, or even a series of objectives,
that the WUM could relate to, such as a plan which indicated and represented
the actual work of patient treatment and care. When there were predicted
treatment and care operational objectives, it was possible to specify the under-
lying activities, then to attach financial quantities to form a budget, and then
examine both the activities and the financial quantities. Another outcome of
setting these operational objectives was that it was then possible for the WUM
to investigate potential levels of treatment and care, the potential costs of
these, and to vary these as the WUM wanted.
Some of these issues fall within the scope of the Otley and Berry model,
whilst other issues are an elaboration of their model. The areas of elaboration
include the ability of a WUM to learn or receive training about operational
objectives, in order to predict these and the levels of finance attached to them.

Control at an individual level: influence, responsibility and accounting data

In order for an individual to be able to handle the financial management

accounting data received and then to use the data for control, the individual
needs to be in a position to influence the affairs within the context of the
control issue. The ability to use both data and influence is associated with the
individual's competence to understand the data and the context of the data.
The conversations with the WUMs showed that they were not in a position
to influence the cash limit imposed upon their unit. They could manage the
cash limit by producing a plan and a budget which they could then influence
and control in various ways. The manner in which they chose to control the
budget was bound up with the way they related to those concerned with the
delivery of care, and where each manager considered the boundary of respons-
ibility was positioned. These notions about responsibility revealed themselves
in different ways.
All of the WUMs were concerned about the control of expenditure in their
unit and to keep within the cash limit imposed. One of the largest areas of
expenditure for a unit was for the provision of nursing care. All of the WUMs
had examined this and made alterations. The most changed situation was in
the unit where the WUM had considered providing the ward sisters with data
about ward activities, w ith a view to getting the ward sisters to control their
266 Control and the National Health Service

ward expenditures. This was not carried out because the WUM did not want
to subject the ward sisters to the pressures felt by the WUM. Also the WUM
felt strongly that the issue of controlling expenditure to keep within the cash
limit was the WUM's responsibility. Having considered alternative objectives,
in a manner consistent with the Otley and Berry control model as modified by
the author, the WUM decided that the only way in which the unit could keep
within the imposed cash limit was to change the work of nursing so that the
system fitted the cash limit. In this way the WUM planned the system, set a
precise budget and then personally monitored all nursing changes and all
expenditures.
The other three WUMs had made alterations, but to the existing staff
pattern, still in a manner consistent with the modified Otley and Berry model.
These three WUMs also mentioned that they wanted the ward sisters to take a
more active part in controlling expenditure, but in these units it was carried
out. Each WUM had an individual approach to the issue of expenditure
control and the methods through which the ward sisters could exercise their
influence and control.
One WUM had provided the ward sisters with some data about the unit at
the monthly meeting of the unit. At these meetings, the WUM had discussed
these data with the ward sisters and the issue of cash restraints imposed upon
the unit, and then urged the ward sisters to restrain activities to curtail expen-
ditures. The second WUM had initially provided the ward sisters with the
unit's financial management accounting data. These were discussed at staff
meetings and at other times with the ward sisters, when they were asked to
keep all spending to a minimum. In order to keep within the imposed cash
limit, the WUM closed a ward for a period. At a later time, after experiencing
difficulties with the financial management accounting data from the finance
department, the WUM had produced plans for the unit and kept detailed cost-
ings and precise budgets of activities. The staffing profiles for each ward were
agreed with each ward sister, then provided to each ward sister who was
urged to minimise the ward expenditures. The WUM then personally moni-
tored all budgets with actual activities and expenditures. Both of these WUMs
acted in a manner consistent with the modified Otley and Berry model.
The third WUM wanted the ward sisters to acquire some understanding
about how the decisions and the actions that they took on the ward became an
expenditure for the ward and then the unit. The WUM wanted the finance
department to provide the ward sisters with data about their wards, and antici-
pated that in time the ward sisters would be in a position to formulate plans
and budgets for their ward. This WUM was seeking to establish a further tier
of management and responsibility with budgets. The procedure would have
been for each ward sister to have their own control mechanism, through
which they established plans and budgets for their ward, with suitable assist-
ance and agreement of the WUM, and which operated according to the Otley
and Berry model, as modified earlier. When established, these plans and
 Derek Purdy 267

budgets would have formed a part of the WUM's Unit plans and budgets. It
seems likely that such procedures would have provided the WUM with finer
explications of control than currently existed. At the same time it might have
led to closer financial control, and perhaps other controls, at the point of
nursing care .
The exercise of control through influence and the use of data in context are
also associated with the individual's ability to understand the data. Even if an
individual can exercise influ ence and has received data, it may not be possible
for the individual to act, and hence control, because the data are not under-
stood. In this case the cash limits, in the form of budgets, were not understood
by the WUMs or seen to be related to their own unit. Additionally, even
though they had all received training about budgets, this apparently had done
little to prevent their confusion and perhaps not helped them to exercise the
most effective control. It seems as though control was not most effectively
exercised until the WUMs had passed through a process of natural learning
about budgeting, which also led them to produce their own budgets. Thus the
Otley and Berry control model requires to be associated with learning to
enable, for example, objectives to be established.
Also related to this issue is the timeliness of data provision, and the content
of the finance department's data sheets. The WUMs considered that the
unsuitably structured historic data sheets were of little use, and always too
late . Sensible control can only be exercised when timely data are available and
when the data are material which the users can understand and utilise.

Concluding observations

The earlier parts of this chapter have sought to demonstrate some of the issues
faced by managers as they deliver health care and control cash budgets. The
emphasis has been on a psychological perspective concerned with the way
managers have responded to change with the provision of financial manage-
ment accounting data and the imposition of cash limits . The chapter has been
concerned to outline a general framework of psychological! organisational
issues in which such managers can be individually located, and to summarise
their common positions in relation to this . It has then moved on to detail and
examine each individual's approach to the control of the workplace issues
which each individual has perceived. This has been based upon conversations
with the managers of four ward units.
For simplicity the issue of control has been explored using the four phases
of the Otley and Berry control model, as a surrogate model of individual
control. This model has been modified or elaborated to include a consideration
of influence, information and training, but more particularly the learning asso-
ciated with control. It has been found that the explanations of their actions by
268 Control and the National Health Service

the ward unit managers can be encompassed by the elaborated model, even
though this meant that the actual forms of control varied with each individual.
The model also encompasses the actions of the general manager, although the
evidence is based upon the WUMs' view of the general manager. These
findings and analysis provide a method for explaining the sequential mechan-
ism of the imposition of cash limits in the NHS by the government, which
appears to be unproblematic until the delivery of treatment and care is
required.
The government sets the overall cash limit for the NHS which is then allo-
cated through the regional health authority to the general manager of a hospi-
tal unit. It would appear that all of the individuals who are concerned with
these sub-allocations, up to and including the general manager, can act in a
similar manner. Their immediate boss imposes a cash limit upon their area of
responsibility and it is the individual's job to accept and to allocate the
imposed cash limit. This seems to work in a straightforward manner because
both the objective which is received and imposed and the objective which has
to be operationalised are in the form of a cash limit. Also, in general, these
allocations are simply based upon historical precedent and the previous year.
After the hospital general manager has allocated the cash limit to the WUMs it
is presented to them as a cash budget with which to finance the activities
under their responsibility.
It was impossible for the WUMs to manage their cash budget by simple
reallocation, as there was no management structure to allocate it to. This was
because they were responsible for the health care activities in their units and
needed to acquire some operational understanding of their unit's activities.
The accounting systems were not set up to provide them with the types of
data which they needed for both planning and control. The existing account-
ing systems concentrated upon stewardship control. In general the WUMs
understood neither how the cash limit/cash budget related to their unit, nor
what was the meaning or usefulness of the reporting data provided by the
finance department.
In order to make the cash budget/cash limit meaningful, the WUMs had to
create their own understanding of a budget and put it into their own perspec-
tives, and so the WUMs created their own budgets. They did so in different
ways which reflected the individual approach of the WUMs and local circum-
stances in the unit. Faced with a limit on the cash they could spend, the
WUMs had to examine the alternative operational activities and prepare both
an operational plan and a budget which did not exceed the cash limit. It was at
this point that a plethora of issues confronted the WUMs. They had to plan for
treatment and care which would match the cash limit, rather than plan
wholely in a way which was consonant with the needs of patients, as per-
ceived by medical and nursing staffs. This was because none of the WUMs
considered that their cash limit would finance their unit's current level of
activities. They had to find ways of immediately saving cash, or producing
 Derek Purdy 269

longer-term cash savings, or more effective treatment and care regimes which
would allow an increase in the number of patients seen .
Over several years, the WUMs found ways of dealing with the cash limit, so
that one WUM changed the nursing system, another closed beds, a third
closed a ward and a fourth urged the ward sisters to restrain their expendi-
tures. The project specifically sought neither complete details of the processes
involved nor the views of the general manager about the performance of the
WUMs. The WUMs indicated that their actions were accepted by the general
manager and, in some cases, the gains which they had brought to the system
were acknowledged. Confirmation of this would appear to be the continuation
of the WUMs in their jobs.
The introduction of a general management structure into the NHS, along
with the allocation of cash limits throughout the system, has had a variety of
effects, and a few are considered here. There have been different types of con-
trols introduced, in particular the control of a rigid unalterable cash limit at
unit level. These in turn, and amongst other things, seem to have fostered the
discussion of crucial issues, some of which perhaps went undiscussed previ-
ously/ such as bed closures, or were not even considered, such as a changed
nursing system. Although it is unlikely that the WUMs took any far-reaching
decisions without discussions with others, the individual ways in which the
problems were resolved may enable somewhat arbitrary and very localised
decisions to be taken about treatment and care, which may resolve an issue in
one location but have undesirable effects elsewhere.
There is a greater awareness of activities undertaken, a greater consciousness
of the costs and expenditures of these, and a shift in the way in which financial
matters are handled in the NHS. At the forefront of these reforms has been the
introduction of cash limits and cash budgets at the operational level of the
WUMs, together with their plans. One matter which seems to have caused a
great deal of confusion, misunderstanding and misdirected anger at accountants
is the fact that the notion of a budget was introduced into this operational level
of management in the form of a cash limit. The general manager and the finance
department provided the WUMs with an imposed cash limit, which all of the
WUMs referred to as a budget. A budget was defined earlier as the financial
quantification of a plan. This would seem to be a reasonable working definition
because all of the WUMs initially thought that the imposed cash limit, called a
cash budget, had a plan relating to their unit underpinning it. Of course this
was not the situation: the general manager had a plan for allocating cash, and
this did not resemble the operation of the unit. It was the limitation of the cash
which had caused and was causing problems for the WUMs, not the notion of
budgeting. Since the cash limit was presented to them as a cash budget, the
notion of a budget took the blame for the notion of a limit. The notion of a
budget was not introduced in a positive context; the budget was synonymous
with the cash limit and, because of this synonymity, it seems likely to have pre-
vented the useful aspects of budgeting from becoming apparent a lot sooner.
270 Control and the National Health Service

The positive aspects to the planning and budgeting practised by the WUMs
are that now there are plans at the operational level of the patient which make
available patterns of treatment and care; these plans can indicate where treat-
ment and care are considered to be ideal and where this is not the situation; all
of this can then be set into a financial context. A system such as this would
facilitate planning from the level of the patient upwards, through the NHS to
government. This seems to represent a more sensible approach than the uncer-
tain results from cash limits . The government has sought to keep the amount
of cash spent to its prescribed limits. This means that the basic issue which is
being controlled is the cash limit. The overall cash limit for the NHS is cen-
trally determined and this amount is apportioned throughout the NHS until it
arrives at a manager, such as a WUM, who is to finance care and spend the
cash.
It would appear that the method of allocating the cash does control the cash,
but it does not take into consideration any consequences for care. It is left to
those managers responsible for directly spending cash to arrange that this care
is kept within the cash limit, as opposed to ensuring that the perceived care is
available. It is not clear that this is the best way of managing health care, as
opposed to managing cash limits.
 CHAPTER 17

Management control in schools

Jane Broadbent

In this chapter we shall look at management control in schools, not only as an

example of control in a specific situation but also as a vehicle by which to
reflect upon the framework of control suggested by Ouchi (1979, 1980) which
we referred to in Chapter 3. We shall use Ouchi's framework as a heuristic
device giving 'id eal types' of different approaches to control and we show
how these ideals of hierarchical (or bureaucratic), market and clan control can
be used to understand the approaches to management and control in schools.
The chapter will also illustrate the effect of organisational participants with
deeply held values on political and legislative attempts to change control
mechanisms in particular organisations. Before the introduction of legislation
in 1988, schools in the UK had been involved with management control only
in an educational sense. They are now having to grapple with broader man-
agement issues, following the introduction of local management of schools
(LMS), the details of which we shall examine later in the chapter.

Introduction and historical background

The management structures of schools have been radically altered in recent

years by a series of legislative acts which have been designed to change the
accountabilities which exist in the educational sphere, as well as the location
of decision making. Following the 1944 Education Act, which set up the
postwar system of state education, a three-tier bureaucratic or hierarchical
structure was developed. In this structure the levels of strategic, manage-
ment and operational control described by Anthony and discussed in
Chapter 2 may be identified. The control system was therefore hierarchical
in nature. At the top of the hierarchy was the Department for Education! (the

271
272 Management control in schools

central government level), next was the local education authority (at local
government level) and finally came the schools themselves. Strategic control
in the guise of policy decisions was located at the Department of Education
on a national basis, but the local education authority (LEA) as a subsection of
the elected local authority (LA) had a great deal of power to set local policies
and to influence the operation of schools in their control. The LAs were the
legal owners of the schools and the equipment within them; they also were
the direct employers of the staff. The LEA was the main focus of manage-
ment control, controlling, for example, funding for the education services
and deciding how to allocate the budget which had been set by the local
authority and which was raised from local taxation and made available as
block grants from central government. The LEA also had a role in the inspec-
tion of the educational standards of the school it controlled and could set
educational policies. Advisory services acted to develop the educational
aspects of the service and there were specialised departments looking, for
example, to the maintenance of schools and the employment of teachers.
Other departments of the local authority provided services such as cleaning,
grounds maintenance and school meals. Whilst policy issues were decided
both at central and local government levels by elected representatives, the
headteacher dealt with the day-to-day conduct of the school, the operational
control, and was responsible for the educational provision therein. The
incumbent of that role had limited responsibility to commit resources, acting
to channel requests to the LEA who could then determine priorities for the
area as a whole. A board of governors existed for each school, and they had
some input into discussions of local matters pertaining to the school and
how the LEA policy was implemented at school level. There was no respon-
sibility on this body to manage resources and the membership was limited,
often biased towards LEA representatives. Control of teaching and learning
processes within the classroom cannot easily be directly supervised and
professional standards and values, which we might see as clan control, can
be argued to have a large part to play in the operational control of the activ-
ity of teaching.
New control structures have now been introduced which can be argued to
have changed the location of strategic and management controls and thus the
relative influence of the different parties (particularly the LEA): some attempts
to influence the professional control of the teaching process have also been
made.

The new philosophy of control

Before turning to examine the nature of the new controls which have been
imposed on schools it is useful to make quite explicit the philosophy which
 Jane Broadbent 273

has underpinned many of the recent changes in control structures within the
public sector in the UK (and in many other parts of the world: Hood, 1991;
Broadbent and Guthrie, 1992). This changing philosophy has been referred to
earlier (Chapter 3) and is characterised by a claim to move away from a
system of hierarchical control to a system which lays claim to a reliance on a
market-based one, but which still exists in the context of a hierarchy. Whether
it achieves this market approach is debatable, but this does not undermine the
extent to which the claims for the legitimacy of the approach rest in the
context of 'the market' and reject the hierarchical or bureaucratic.
The roots of change in the public sector can be found in the Financial
Management Initiative (FMI) the basic elements of which were laid out in
Cmnd 8616, 'Efficiency and Effectiveness in the Civil Service'. The PMI implic-
itly uses a notion of private sector management to inform the approach to
managing the public sector. It seeks to provide a framework within which
individuals are made accountable and the use of financial resources is a
central element of accountability. Hood (1991) lists seven characteristics of
what he terms 'the new public management model':

1. The importance of financial devolution to service units.

2. Explicit standards and measures of performance for those units.
3. A clear relationship between inputs, outputs and performance measures.
4. Increased accountability requirements on the units.
5. A stress on private sector management styles.
6. A stress on competition and contracting between units.
7. A stress on efficiency and parsimony in resource usage.

As can be seen from the list, one element of FMI is the need to link input
resources to outputs. This requires the ability to measure outputs, which is
also a requirement of the market approach to control. The philosophy of
market-based control is thus mixed into 'the new public management'
model. It is perhaps clear that what is desired is control of the outputs of the
system, but there is no real clarity as to whether this should be achieved
through the direct (and perhaps hierarchical or bureaucratic) control of
linking inputs and outputs (resources being given on the basis of the outputs
actually achieved in a more bureaucratic way) or whether a market should
be used as a means to mediate this relationship. The use of the market
control has the advantage of the possibility of delegating blame for poor out-
comes away from the controller (in this case the government) as any
outcome is the responsibility of 'the market' or of the service to which
the work has been delegated. We will demonstrate how a market-type
logic has been used in designing the control systems for schools, with-
out abandoning the possibility of the use of output controls through a
hierarchical approach.
274 Management control in schools

Changing processes of control

In summary we can argue that the imposition of FMI logic and a market
approach in schools has led to operational and management control now
being located at school level. The LEA's role in management control of the
service is much reduced. Clan control of the activity of teaching has been put
under some pressure by the requirements of the national curriculum and
testing regimes (see below) . Strategies are still controlled in a bureaucratic or
hierarchical fashion by central government.
The Education Act 1986 started the process by which control of schools was
to be taken away from the LEA and redistributed between the central govern-
ment on the one hand and the schools and parents on the other." This act set
up the principle of parent representation on school governing bodies, a situ-
ation which some LEAs had already voluntarily adopted. This change sig-
nalled the perceived importance of parents in the process of educational
management. However the most significant changes were started with the
Education Reform Act 1988 (ERA). This Act introduced educational changes,
the national curriculum with a programme of national testing and reporting of
test and examination results, alongside LMS. It can be argued that the educa-
tional issues of the national curriculum and national testing form the basis for
a challenge to the clan control of the teaching process. The central concern of
this chapter will be with LMS and its implications for management control,
rather than the educational changes, but it will be seen that LMS is closely
related to education issues through the funding mechanisms.

LMS and formula funding

The practical implications of LMS are extensive. While responsibility for

capital expenditure still lies with the LEA (who have a 'landlord' role) a school
is now a responsibility centre with the task of preparing annual budgets and
controlling the day to day expenditure of the organisation.
The issue of resourcing the school is rather more complex and to under-
stand this we need to consider the manner in which resources are distributed
to schools. The LEA still has some influence in this respect as it has to decide
the overall amount which will be spent by all schools under its control. A
small number of central services are still retained by them (home to school
transport, for example) but 85 per cent of the overall educational budget
must be distributed to schools (this proportion will increase to 90 per cent in
April 1995). The LEA has the power to decide the distribution of the balance
of the educational budget to the individual schools, but has to do this on the
basis of a publicly stated formula (the aim being to give an objective basis on
which schools can predict their budget share for planning purposes) . An
 Jane Broadbent 275

important element of the formula is that 75 per cent of the budget to be

distributed to schools must be distributed on an 'age-weighted pupil
numbers' basis.
The outcome of the implementation of formula funding is that the basis of
provision of resources to schools is measured largely by pupil numbers; thus
the ability of LEAs to give priority to other bases of need (for example, social
deprivation) has been considerably reduced. The outcome has been that some
schools have received relatively more, whilst others have received relatively
less than they might otherwise have expected. In a time of tight control over
the whole of local government spending the ability to protect the schools who
were relative losers has been highly constrained and some painful readjust-
ments have been inevitable. Given that the schools spend about 85 per cent of
their budget on salaries, the outcome has been loss of jobs in some schools.
Thus, not only have headteachers and governing bodies had to control their
own budgets for the first time, they have often had to do so in the most
difficult of situations, that of a declining resource base.

Local management of schools, delegated authority and standardisation

Local management of schools has also changed the location of decision

making. In line with the philosophy of the new public management, the
authority to run the school was delegated to the school level, to the governing
body. Thus management control is delegated to schools and LEAs, as a conse-
quence, have lost much of their role in this respect. The constitution of the
governing body was defined to ensure representation from parents, local busi-
ness people and teachers as well as the locally elected political parties. This
was meant to increase the extent of accountability to different constituencies
and to bring in experience from the private sector. Governing bodies have the
right to hire and fire staff and to spend the resources allocated to them in
whatever way they wish, provided the requirements of the national curricu-
lum are met. The duties and responsibilities of the governing body are com-
prehensive but, de facto, the day-to-day operation of the school is now
effectively in the control of the headteacher. Research has shown that the
implementation of the LMS initiative has been soaked up by a small group of
the school community seeking to 'protect' the rest of the school from these
management changes, which they see as peripheral to the main educational
aims of the school (Broadbent et al., 1993; Laughlin et al., 1994). One aim of the
legislation was to try to ensure that a wider set of views were represented in
the decision-making processes and it is unclear whether this has de facto been
achieved. Empirical research also suggests that LMS has generated much
stress for teachers who find themselves having to deal with financial and man-
agement issues for which they have had little training (Laughlin et al., 1993;
Broadbent, forthcoming).
276 Management control in schools

By delegating responsibility to schools decision making was intended to be

given to those who 'know best' - those who work at the operational level. The
delegation of responsibility also provides the opportunity for the delegation of
blame (Armstrong, 1989) for cuts to the local level and away from central gov-
ernment. This is a significant factor, given not only a downward pressure on
LA spending from central government, but also a declining school population,
leading to overprovision of school places and a need to close some schools.
However it must be emphasised that the delegation downward to schools was
accompanied by the centralised national curriculum. This ensured that,
although schools can choose how best to use their resources to deliver the
national curriculum, they have little choice of what to deliver. The delegation
to local level is therefore countered by a strong centralising tendency which
also emphasised a high level of standardisation. The process of standardisa-
tion is a familiar tactic for control and is illustrated in the scientific manage-
ment approach of F. W. Taylor. It enables the comparison of actual
performance against standard, any deviation can be highlighted and steps can
be taken to 'control' that variation. Thus standards of achievement for pupils,
albeit minimum standards, are set which will be checked through national
testing regimes. In the case of schools the philosophy of LMS and the stand-
ardisation inherent in it further seems to imply that very similar resources can
achieve the same educational outputs from different children. Finally we
should note that the process of standardisation can also be seen as a strong
challenge to the professional or clan control of the teaching process.

LMS and the philosophy of market-based control

Whilst the setting of standards provides a control mechanism through the

comparison of actual performance against the standard set, it also provides the
information which is deemed to be required in the attempt to implement a
control strategy based on the 'market' . The attempt to introduce a 'market' has
been facilitated by linking together financial and educational issues. This has
been made possible because of the standardisation processes introduced as a
consequence of the national curriculum which allows some measurement of
'output'. Added to this is the introduction of formula funding, which means
that resources are closely linked to pupil numbers and a policy of open enrol-
ment (see below). The market-based mode of control is premised on the exist-
ence of competition between different suppliers, and the possibility for those
who require the service in question to have the ability to choose a supplier. It is
assumed that schools will now compete to provide a service to pupils and that
parents will choose more actively where to send their child to school. Popular
schools will expand and unpopular ones will be forced to close (or improve).
Under the previous LEA bureaucratic system, schools received the children
in their catchment area and the ability of parents to send their children to
 Jane Broadbent 277

other schools was constrained. Now, under open enrolment, there is, de jure,
much more opportunity for parents to exercise choice. De facto, this may not
be the case; the physical capacity of schools still imposes some constraint on
pupil numbers and geographical location or the socioeconomic status of the
area may also playa part. In both urban and rural districts, travelling time to
school will be a factor in constraining choice; for some parents with work com-
mitments or with limited choices of transport there may be no choice at all.
The fact that much of the funding which a school receives is based on age-
weighted school numbers means that a school has to be attractive enough to
the pupils and their parents to enrol sufficient pupils to ensure enough funds
to run a viable school. A viable school must be able to provide the national
curriculum for its pupils and this means it must have sufficient numbers of
staff to cover all aspects of the syllabi. It is assumed that schools will, there-
fore, compete for pupils to ensure they generate resources. It is also assumed
that parents will exercise choice. In order to make a choice about the school
which a child should attend it is assumed that information about the perform-
ance of the school will be required. The information which is assumed to be
needed is not just the 'grapevine' information that exists in any community
about the schools within it, but more 'objective' and public performance-
related data. The data which are deemed to give an indication of the output of
schools are statistics of examination performance - the national GCSE and A-
level examinations are already reported in league table form and the testing
which is planned to accompany the national curriculum is meant to supple-
ment this - along with details of absenteeism. It is assumed that parents will
make their choice of school on this basis and because schools need to retain
sufficient pupil numbers they must take these indicators seriously.
This new approach to control in schools is a powerful one which uses ideas
of delegation of responsibility, a philosophy of market-based competition at
the operational level and a strong tendency to centralise and standardise. It
gives an indication that in practice the dichotomy between hierarchical and
market-based control is not clear-cut (see the critique offered in Chapter 16)
and that control processes are often complex rather than tidy. The next ques-
tion to be asked is the extent to which these controls produce the control
required, to use Drucker's terminology (Drucker, 1964). Put another way , we
may ask whether the controls produce the desired outcomes.

Approaches to the evaluation of controls in schools

The possibility that schools will 'raise standards' by placing greater emphasis
on examination results based on a centrally controlled curriculum is, arguably,
a focus which underlies the whole initiative. This interpretation is based on
the assumption that government policy is geared towards a belief that better
278 Management control in schools

examination results will mean 'better' education. Accepting this argument

allows us to evaluate LMS from two different perspectives. First is the issue of
whether examination results do give an indication of the quality of the educa-
tional experience provided by a particular school. This leads us to consider
whether a system based on this output control is appropriate for education.
Second is the issue of whether the controls which are implemented achieve
the control which is desired. The first issue might be argued to be a 'political'
question of choice rather than a central control issue, but we will consider it
here because it does impinge on the way in which the controls have an effect
in a practical situation and because the need to define an output measure is
central to the FMI approach to control. The second issue is important as it
illustrates the dysfunctions associated with the practical working of the FMI as
a control system.

Is there a linkage between examination results and school performance?

Perhaps one of the most contentious issues in the whole of the implementation
of LMS and other elements of ERA is the issue of the testing of children's
progress and the reporting of the test results. Results of both public examina-
tions and the periodic tests associated with the national curriculum (standard
attainment tests or SATS as they are called) are to be reported, and not just to
individual pupils and their parents: but the overall results are to be collated
and national league tables are to be drawn up. These results can, it is argued,
provide the output measures which are necessary for a mode of control which
is based on the philosophy of FMI and which wishes to see managers respon-
sible for the outputs of their operational unit. The main issue of contention is
whether the output measures selected do reasonably represent that which
education sets out to achieve. It can be argued that the examination results are
not a good proxy for the outputs of the educational process and that it is ques-
tionable whether the outputs measured bear any relation to the transformation
of the inputs which occurs in the processes of teaching within the organisa-
tion . Teachers would argue that while examination results are important they
are not the only outcome of education and may be more important for some
children than others. The examination results tables only measure a limited
range of a child's competence as a human being. Thus the 'value added' to a
child by the school is not measured and the visibility attached to examination
results emphasises one particular set of achievements (perhaps 'examination
competences added') over the many other achievements a child may have.
In a critique of 'economic reason' Andre Gorz (1989) suggests that there are
areas which should not be driven by economic reason because they do not
create commodities. The caring professions such as teaching do not produce
commodities because (using Gorz's definition) they are not activities geared
 Jane Broadbent 279

towards providing items in a measurable amount of time at as high a level of

productivity as possible. He argues that/to follow this pattern might in fact be
detrimental as the quality of care may well not be quantifiable in relation to
output measures. He argues that service 'depends on a person to person rela-
tionship, not on the basis of quantifiable actions' (Gorz, 1989, p.143). Research
(Broadbent et al., 1993; Broadbent, 1992) has revealed much resistance to the
idea that the outputs of the schools can be measured. This belief is justified by
appeals to the value set of education. This value set is claimed by teachers to
be rooted in relationships and based on a desire to help individual pupils
achieve their own unique best.
It is important to highlight the fact that there is a lack of clarity in the link-
ages between inputs and outputs which are inherent in the way LMS is cur-
rently conceived. If we consider the way in which LMS and the market based
system work, the implied relationship is between input resources (the amount
of money which a school is allocated by the age-weighted formula) and output
educational results (such as exam results or truancy rates). This relationship is
mediated, as explained earlier, through the choice of the parent, the reported
results providing the basis of a more informed choice. Thus a market-based
approach is being applied. However the reported results establish no linkage
between the status of the child which is put into the system to be educated and
the eventual educational achievement of that child. This is seen as unfair to
teachers, as the output results may well be the result of the status of a given
intake of children rather than a 'better' educational process within the school.
Put crudely, the ability, say, to recognise colours may be an incredible achieve-
ment for one child, but a matter of little consequence to another. Measuring
both by the same scale does little justice to the achievements of either child -
or the educators of either child.
This means that the information supplied to the 'market-place' can be irrel-
evant, misleading or biased in a technical sense. It should therefore lead to
questions about whether the 'market' can really work in this area. However, if
the 'market' fails to produce the results which the legislation requires, that is
an increase in standards as measured by the outputs reported, there is the poss-
ibility of adopting a more hierarchical PMI approach based on the linking of
input resources to outputs. This would result in schools being resourced, not
on the numbers of pupils attending, but on the basis of the reported output
measures. Because of the perceived crudity of the output measures, it is this
possibility which schools most fear.

The control implications of the new system

In Chapter 6 we discussed research into the dysfunctions of budgeting and

the unintended consequences of applying controls. Dysfunctions may arise for
280 Management control in schools

many reasons, but it is often noted that a successful control system will need
to align individual objectives with those of the organisation (as in Hopwood's
idea of self-control, discussed in Chapter 2). In commercial organisations some
element of self-control is often seen as being best achieved through promoting
the self-interest of managers who are offered salary bonuses for achieving
objectives such as profit or sales targets. This type of linkage is not yet being
applied in education, although the granting of some broad flexibility to gov-
erning bodies to set the salary point of individual teachers (subject to a
national scale and some broad guidance) may be seeking to change this .
Even if these linkages were made, it is unclear whether teachers would
respond in the desired way. A fundamental issue, perhaps, is whether the
objectives of individual teachers, informed by a particular value set, can ever
be seen to be in alignment with objectives imposed by the Education Reform
Act in general and LMS in particular, for the professional work of teachers
seems to be based on a very different value set (Broadbent, 1992). It has
always to be remembered that a pathway of change is not one which can be
unproblematically chosen. Laughlin (1987) emphasises the importance of
changes in the life world (which includes the values) of the organisation in
analysing the extent of organisational change. He notes that, without a funda-
mental shift in the life world, only first order change can be said to have taken
place . By first order change we mean that an imposed change is either rejected
outright or absorbed so that the organisation reorients itself to the change in a
way which does not change the fundamental life world. The latter situation
seems to describe schools at the moment, where the existence of a small
'absorbing group' of individuals who soak up the changes protects the rest of
the school from their impingement (Laughlin et al., 1994).
As well as leading to an absorption of change rather than fundamental
changes in values or life worlds, the imposition of particular control systems
can lead, not to control, but to manipulation of a system. Thus, it can be
argued, as in any budgeting system, 'games' will take place in schools which
will not particularly ensure that the system works as it is intended by those
designing it, but will secure the required results for those operating it. For
example, good examination results can be achieved by being selective about
the intake of pupils rather than by giving higher standards of education.
Further the possibility that the emphasis and corresponding visibility given to
particular elements such as examination results will in turn make other ele-
ments invisible and less valued heightens the concern of teachers about the
implementation of these new controls. Already there is public debate about
whether more children with behavioural problems such as truancy are now
being excluded from schools because of the likelihood that they will adversely
affect the schools' scores on the league tables of measured outputs such as
truancy rates. Thus selectivity of admissions and examination entries may
ensue, as well as manipulation of results.
 Jane Broadbent 281

In summary we might conclude that an evaluation of LMS as a control

system is problematic. Whilst it promotes a focus on certain outcomes, the
desirability of those outcomes to teachers and pupils and what those outcomes
actually 'measure' are open to question, because of both educational issues
and the possibility of manipulating results.

Summary

This chapter illustrates a particular example of an attempt to change the

control systems of an organisation. It illustrates how the previous systems of
bureaucratic and clan controls have been challenged and how there has been
an attempt to impose on schools a type of control based on markets in the
context of a hierarchy. Thus the extent to which control systems in any organ-
isation are often an amalgam of approaches is demonstrated. The issue of
output controls in the context of schools is discussed. Problems associated with
their use in this type of organisational context (in which agreement about the
nature of outputs is not easily achieved and in which outputs are not easily
quantified) have been highlighted. The example of LMS gives a useful illustra-
tion of the way in which the values of organisational members (or the organ-
isationallife world) can affect the extent of any change in that organisation.
It could be argued that LMS gives an example of an ideological approach to
control and that the legislation which has imposed LMS has not taken account
of the values of the teaching profession. Instead it has sought to impose the
'new public management' as a consequence of the ideological stance of the
government. Thus we have a situation in which one set of deeply held values
is in conflict with another. While government has the power to legislate we
can also see that there is always a possibility of mobilising the resistance of
those within the organisation which is the focus of the legislation.
We can also see the problems of control in a human service organisation
where the essential inputs, children, are also the outputs. The task of learning
has to be substantially achieved by the children themselves, under the guid-
ance of teachers and the possible encouragement of parents. The fact that this
transformation process is imprecise and that there is immense difficulty in
measuring inputs and outputs makes the available control models very
difficult to use. Because there is a great desire to achieve particular outcomes
and there is a recognition of the limited extent to which control of the transfor-
mation process can be achieved, attempts to try to quantify and measure the
desired outcomes have been made. The question with which we are left is
whether we are attempting to measure the unmeasurable in an attempt to
apply a control model which is inappropriate in the circumstances.
282 Management control in schools

Notes

1. This is the current title of the department; it has had several different names
throughout the period.
2. The desire to provide PMI and market-based control was perhaps one element
which gave rise to changes. Another element might be argued to be a desire to
eliminate the control of the LEAs over education. This might be seen as a response
designed to neuter the power of the left-wing local authorities. Whilst this aspect of
the changes is recogni sed it will not be d iscussed at length as the concern of the
chapter is with the changing structures of control rather than the reasons for the
changes.
 CHAPTER 18

Management control in the

financial services sector
Kim Soin

Introduction

The UK financial services industry is one of the .largest sectors in the UK,
employing around 400 000 people (more than any other industry except for
the civil service and the health service). The productivity revolution that
swept through the manufacturing industry a decade ago has now reached
financial services. Because the industry is coming to restructuring late, the
changes are being pushed through fairly fast. At Lloyds, the chief executive
has predicted that up to 100 000 jobs will be cut from the industry this decade.
The general consensus among the major UK banks is that there is enormous
scope for rationalisation and a need to keep costs at an acceptable level in rela-
tion to income.
The development of management control systems in the UK banking indus-
try has been inextricably linked to the changes that have taken place in the
economic, political and regulatory environment over the last decade. This
chapter outlines these changes and the way in which banks have responded,
and explains how management control systems have changed and developed
in relation to the new ethos in clearing banking. This new ethos is embodied in
profitability, cost awareness and shareholder value. This is a marked shift
from the past, when a banks' success was measured by the size of its assets.
The marketing philosophy has moved from the mass marketing of the 1980s to
market positioning and segmentation. There is now more interest in a
profitable product line than in a full product line .
The detail of this chapter is based upon cost control systems, in particular
activity-based costing (ABC). Traditional product-costing systems trace direct
costs to products and allocate or apportion the remaining (indirect) costs to
283
284 The financial services sector

products, normally through a two-stage process. The indirect costs are first
allocated into pools, and then allocated to products by methods such as direct
labour costs or direct labour hours which are based on production volume.
These traditional costing systems are being criticised for producing mislead-
ing product costs (Johnson and Kaplan, 1987; Cooper, 1987) which no longer
reflect the resources consumed to produce them. ABC is the rejection of the
conventional treatment of overhead in favour of identifying specific services
(resources) being put into the process (activities) being costed. The concept
underlying ABC systems rests on the premise that products utilise activities
and activities consume resources. Emphasis is placed on the important role
that activities play in 'causing' costs to be incurred. As in a conventional
system, ABC is based on a two-stage procedure (Innes and Mitchell, 1991).
The first stage is charging overhead cost to activity-based cost pools. The
second stage is deriving and using a series of cost driver-based rates to attach
the pooled costs to product lines. The design and operation of ABC is depend-
ent upon three key factors: the choice of cost pools; the selection of means of
distributing overhead cost to the cost pools; and the choice of cost driver for
each cost pool.
This chapter will first examine the general nature of management control in
the UK clearing banks from a managers' and consultants' point of view. Using
contingency theory, the following section will identify the major classes of
contingent factor which affect the organisations' control system. The chapter
then moves to a consideration of the contingent factors influencing the UK
banking environment, in particular the effect of the new capital adequacy
requirements and the effect of deregulation on the financial services industry.
A product of the environmental changes, activity-based costing is discussed in
the fourth section, which is followed by case study evidence on the role of
management control systems, and in particular ABC, in one of the UK clearing
banks. The final section provides a critical evaluation of the role of manage-
ment control systems in the UK clearing banks.

Management control

Banks are clearly facing considerable challenges and there is, and has been, a
need for an 'appropriate response'. This response is manifested, in part, in the
management control systems that are being introduced into the banks.
Interviews with managers across the four UK clearing banks suggest that the
principal types of control systems found in banks are budget control systems,
cost control systems, risk control systems, security and staff control systems,
premises control systems, personnel control systems and product control
systems.
 Kim Soin 285

In practice the control systems listed above are operated separately but they
also have an impact on each other. An example of these interrelationships is
provided by the case of budgetary control. Product control affects the branch
manager's budget; standard products are offered across the UK, and the inter-
est rates and margins linked to those products are centrally controlled, with
little scope for local deviation. Centralised personnel and premises functions
mean that a large proportion of annual expenditure may be outside a branch
manager's control. Risk control can restrict what a manager sees as potentially
profitable business being taken on if those higher up the discretionary ladder
disagree.
Management control systems, in banks as elsewhere, are complex. In one
study Smith (1987) distinguishes between management controls and pro-
cedural and accounting controls. Management controls, he suggests, are
designed to promote operational efficiency within the organisation and ensure
adherence to management policies and minimise business risk. Management
controls are 'essential in defining the environment in which the more detailed,
procedural and accounting controls can operate' . They include:

1. formal allocation of responsibility and lines of reporting,

2. clearly defined and properly understood operating procedures,
3. approved authority levels,
4. permitted exposure limits,
5. budgetary control systems,
6. regular management reporting and
7. systems of internal check and internal audit.

Central to this understanding of control systems is the idea of a control envir-

onment. A control environment encompasses the attitudes, abilities, percep-
tions and actions of the institution's personnel, in particular management. A
'favourable' control environment requires the support and leadership of
senior management and an acceptance of the need to ensure that the laid
down procedures operate in practice.
The management and informational requirements for effective control and
management in banks (Smith, 1987) should be:

1. accurate, consistent and reconcilable, where appropriate, with accounting

records;
2. timely and sufficiently frequent for their purpose;
3. relevant to the recipient's decision-making requirements;
4. flexible and readily adaptable to changing needs;
5. inclusive of all group operation, risks and expoiures;
6. forward-looking. as well as concerned with the past; and
7. capable of comparison to the main competitors.
286 Thefinancial services sector

Smith (1987) then goes on to develop a 'matrix of management information

and control' which relates to planning, performance measurement and opera-
tional control. Appropriate key performance criteria are determined as a
means of measuring the progress of the institution. Key variables here include
inflation, interest rates and exchange rates. By reporting key finance indica-
tors, management will not be swamped with information. Trend analysis is
cited by Smith as a powerful control mechanism, as are control systems that
monitor, identify and develop practical techniques to manage risk. This frame-
work suggests the external environment will be of importance in looking at
management control systems.

Contingency theory

As we have seen in earlier chapters of this book, management control can be

looked at through many different lenses. Building on the suggestion in Smith's
work that external issues are important, the definition of a management
control system used in this chapter is that management control seeks to align
the decision-making behaviour of individuals with the organisation's goals
and strategies, which are in turn influenced by the internal and external envir-
onment (Middaugh, 1988). This is essentially a contingency approach to man-
agement control. As an organisation's structure and environment changes, so
must its control systems.
The contingency approach to management accounting is based on the
premise that there is no universal accounting system applicable to all organ-
isations in all circumstances, (Emmanuel et al., 1990). General all-purpose
systems are unlikely to be uniformly successful: the management control
system needs to fit the specific circumstances of the organisation for which it is
intended. The major classes of contingent factor which have an impact on the
organisation's control system have been identified as the environment, organ-
isation structure, technology, strategy and culture.
The external environment will affect the nature of the control system.
Accounting researchers have identified various environmental factors which
will affect the type of management control system used. These include the
degree of competition faced in the market-place (Khandawalla, 1972), the
number of different products markets and the type of environment; that is, a
'tough' or 'liberal' environment (Otley, 1978). Gordon and Miller (1976)
identify three main environmental characteristics hypothesised to affect
control systems: dynamism, heterogeneity and hostility. A high rate of change
will require frequent control reports. The number of different product markets
served will lead to a decentralised control system with quasi-independent
responsibility centres. In times of severe competition or market hostility, a
more sophisticated control system is required. Other factors which may shape
 Kim Soin 287

an organisation's environment include government, customers and sharehold-

ers. Furthermore the existence of powerful groups in the organisation's
environment may increase the level of uncertainty it faces.
Technology as a contingent variable is characterised by its complexity and
degree of predictability in producing desired results. The technology the
organisation utilises is particularly important. Organisation structure is essen-
tially concerned with matching task interdependence with the structure
imposed on it. Strategy is important for defining appropriate goals . Simons
(1987a) found that firms following different strategies employ (accounting)
control systems in different ways. Dent (1990) argues that, rather than taking
the objectives of the organisation as determined, the control systems are
influencing the formation of objectives. This suggests there may be a more
complex relationship between organisation and environment than contin-
gency theory allows. However the importance of the environment is still
emphasised.
Contingency theory, therefore, allows for some explanation of the way par-
ticular control systems have been developed. The control systems in the UK
clearing banks have evolved in the context of the changes that have taken
place in the external environment. These changes will be considered by focus-
ing on the environmental changes that have taken place over the last decade
and then illustrated in a UK clearing bank case study.

The UK banking environment

Banks playa central role in the UK economy by providing financial intermedi-

ation services. The big four commercial or clearing banks are Barclays, Lloyds,
Midland and the National Westminster Bank. These banks are service organ-
isations which provide acceptance of deposits, a payments mechanism and
provision of credit. Their role has , however, changed over the last decade and
into the 1990s is continuing to change: the effects of deregulation, increased
competition and the introduction of new technology have introduced a
number of new entrants to the UK banking market: for example, building soci-
eties, major retailers and insurance firms . Banks, too, have been moving into
new lines of business. The focus of banking has moved towards sales and mar-
keting (particularly in the branches) and towards a new cost awareness.
For a long time clearing banks in general paid almost no attention to the
development of pricing strategies or to the measurement and control of costs.
This is evidenced by conversations with bank managers who state that at the
beginning of the 1980s cost control had not even entered the decision maker's
conception. There was no emphasis on cost or cost awareness. The simple
reason was that they did not have to bother: competition as well as the banks'
possibilities of doing business were very restricted. This was because of the
288 The financial services sector

oligopoly arising from the clearing banks' control of payment systems. In the
1980s however, the UK banking industry found itself facing problems associ-
ated with increasing competition from other sectors of the capital market.'
Many activities previously conducted by banks could now be undertaken by
others and it is not evident that banks have any permanent competitive
advantage in these areas.
The 1980s was an important decade for the UK clearing banks: it was a time
of overexpansion and neglect of prudence for the banking industry.
Deregulation and 'Big Bang', which removed barriers to competition in the
UK financial services industry, encouraged banks to purchase stockbroking
firms, stock-jobbing firms and estate agencies. Barclays Bank, for example,
combined its merchant banking arm, the brokers deZoete and Bevan, and the
jobbers Wedd Durlacher Mordaunt. Lloyds Bank entered the estate agency
market by acquiring a number of estate agencies and, at one stage, became the
biggest estate agency in the country. Inevitably marketing was given a higher
organisational priority in the 1980s, reflecting the intense competition in all
segments of the market. Segmentation and branding of products were recog-
nised as being applicable to banking and banks had felt it necessary to carry a
whole range of products.
However the 1990s have added a new dimension to the UK clearing banks'
strategy: they have lost some of their traditional informational, structural and
reputational advantages. Now they are dediversifying because they have dis-
covered that their diversity did not bring strength. It brought lack of control
and severe losses. Diversification created overheads and placed an additional
strain on the central management. Banks have pruned product lines and have
become conscious of product profitability. Many of the banks' main products
were, by 1989, in the mature or declining phases of their life-cycles (Morrison,
1989). In response to severe losses, the dominant financial ethos in UK clearing
banking today is profitability, cost awareness and shareholder value.
The profound changes that have been taking place in the banking industry
have been produced by a complex, interrelated series of components, includ-
ing capital adequacy, the issue of costs and competition, risk and reward and
intangibles such as quality of service.

Capital adequacy

The Basle Agreement (1988) on capital adequacy was implemented in the UK

by means of the Banking Act 1987. This is essentially a form of control by gov-
ernment rules. Capital adequacy rules relate the amount of a bank's capital or
'own funds' to its risk exposure. The imposition of the capital adequacy
requirements has had major implications for the required profitability of
banking. This issue has proved to be of great importance because it is a
restraint on banks' balance sheet growth and put pressure on banks to
 Kim Soin 289

perform in line with stock market expectations in order to ease the raising of
capital if it is necessary. Llewellyn (1991) emphasised that, as 'both internal
and external sources of equity ultimately depend on the banks' profitability,
the imposition of equity capital requirements has major implications for the
required profitability of banking, the ability of banks to compete, and the type
of business they are likely to conduct'. The effect of the capital adequacy rules
was to increase the banks' cost of capital just at the time when the banks' had
and still do have, an increased need for capital.
Lloyds Bank, for example, was determined to earn an adequate rate of
return on capital that is invested in order to enhance shareholder value.
Attracting capital at an adequate cost means that the bank has to create share-
holder value by producing a rate of return on equity in excess of their cost of
equity. Lloyds is also committed to the idea of creating shareholder value.
This means increasing dividends from earning a return on capital in excess of
others in the market, managing risk and trying to gain market price growth.
The capital adequacy requirements have been a significant factor in the change
to a performance-based culture in the UK clearing banks because of this new
emphasis on market value.

Deregulation

Significant deregulation of the UK financial services industry which took place

in the 1980s came through the Financial Services Act 1986 - 'Big Bang' and the
Building Societies Act 1986. The Financial Services Act removed barriers to
competition in the UK financial services industry and significantly increased
the level of competition. Banks were, for the first time, allowed to participate
on the UK securities market. The Building Societies Act gave UK building
societies the opportunity to move away from their traditional mortgage
lending business and into the personal finance arena, competing head-on with
banks.
The effect of these two Acts was to increase competition for the personal
customer account and hence reduce the profit margins that could be achieved
on such businesses. A former Director of Retail Services in Barclays Bank
believed that the most powerful catalyst for the changes in the banking indus-
try was the deregulation of the building societies, the effect being, for the first
time, to expose some of the clearers' core activities - such as across-the-
counter personal services - to competitive pressure from new entrants.
Deregulation and the 1990-93 recession has led to structural overcapacity in
the financial services industry. This has led banks into programmes of ratio-
nalisation, for example, the pursuit of efficiency gains.
Clearing banks have faced more competition from non-traditional suppliers
of financial services, such as Marks and Spencers Financial Services. Non-
finance companies have a greater capacity to diversify into banking services
290 Thefinancial services sector

than banks have to diversify out of finance. The effect of increased competi-
tion and declining profits meant that cost measurement and cost reduction
have taken on great importance. The deregulated financial markets and the
ensuing increase in competition has put pressure on cost structures which
evolved during decades when competitive pressures were less intense.
Additionally cost reduction is one component of attempts to raise operating
profits to compensate for lending losses in the 1980s.
These changes in the banking environment have introduced an awareness of
rationalisation, efficiency and cost control which in the past were not so impor-
tant. The resurgence and recognition of control systems is largely a response to
the need to embrace these concepts and ABC in particular is one control system
which has gained great prominence in financial services. Awareness of ABC in
financial institutions came from the financial/management accounting press
and from management consultants. Additionally these developments in
product costing, performance measurement and cost management systems in
the retail financial services sector have lagged many years behind the
development of similar techniques in manufacturing.

Activity-based costing

ABC as an approach to cost information was thought to provide accurate and

relevant cost information as a guide for making profitable decisions about
products and services, revenues and costs. It provided a way to organise the
collection, processing and reporting of cost information that supports decision
making and strategy formulation. Furthermore it was thought to yield accu-
rate and relevant information for managers who may not be cost accountants,
but who must utilise cost information to improve the competitive position of
their financial institution. It is important to remember that costs are not just a
threat to profitability; they can also be used as a competitive weapon by pro-
viding a barrier to entry (Bromwich, 1990).
Fitzgerald et al. (1991), from their empirical survey of service businesses,
argue that there are five factors which occur in many, if not all, services:

1. the usual presence of the customer in the service delivery process;

2. the intangibility of many aspects of the service package;
3. the heterogeneity of service staff performance and customers' service
expectations;
4. the usual simultaneity of service production and consumption, which
means that services cannot be counted, measured, inspected, tested or
verified in advance of sale;
5. the perishability of most services, which obviates the use of inventory as a
buffer between peaks and troughs in demand
 Kim Soin 291

When they occur together, these five factors cause particularly difficult prob-
lems for service managers in scheduling operations, controlling quality,
measuring performance and tracing and controlling costs .
The ABC approach is not, however, a panacea which will solve all manage-
ment cost information needs per se. Its value is situationally dependent. ABC
produces historic cost information which only has an indirect relevance to
managerial decisions. Its role in decision making requires careful specification.
Furthermore it does not overcome all of the procedural problems of conven-
tional costing. Sephton and Ward (1990) believe that the introduction of ABC
will provide the potential for retail financial services to be at the 'leading edge'
of management accounting development and gain for themselves a competi-
tive advantage. With its concentration on the relationship of overhead costs to
products and customers, ABC is suitable for retail financial services with its
complex product-to-process relationship and high fixed cost base. Sephton
and Ward highlighted three areas in particular where retail financial services
could gain considerable benefit from using ABC: (a) as part of the strategic
management process, understanding cost behaviour and analysing profitabil-
ity; (b) in the calculation of meaningful product costs; (c) in budgeting, fore-
casting and performance measurement in overhead departments.
There are two types of ABC systems, one with a strong strategic focus and
one with a strong behavioural focus (Spicer, 1990). Strategic systems tend to be
the most complex, involving many more activity-based cost pools and cost
drivers than do behavioural systems. Accurate product costs seem to be of
primary importance. Documented cases of firms experimenting with these
systems reveal that all these firms are under severe competitive pressure, with
diverse product mixes and facing a variety of pressing strategic decisions
about the rationality of their product mix, pricing, make-or-buy and/ or the
disposition of their market resources (Spicer, 1990). Behavioural ABC systems
tend to drive or reinforce behaviour which is consistent with the achievement
of an existing, clearly defined strategy.
The next section discusses the reasons for the implementation of ABC in a
UK clearing bank. The case outlines the changes that have taken place in the
organisation and the changes ABC has brought to the organisation.

Management control: ABC in the context of a UK bank

The bank is a UK clearing bank which, since 1992, has undergone significant
changes in its Payment Services Croup.' These changes relate to reorganisa-
tion, restructuring and reworking. The person responsible for managing the
change is the general manager, whose responsibility is to meet the strategic
objectives agreed by the board. The strategy was to increase profitability in a
competitive market and provide what is perceived by the bank to be an
292 The financial services sector

adequate return to shareholders: that means raising the share price and
increasing dividends.
The tactical elements employed to build up to meeting the overall strategy
included the development of costing control systems, personnel control
systems, communication control systems and internal audit control systems.
Control and measurement of cost were the key strategic elements. At present
ABC and the resource engineering systems have been developed and are in
use. The control system which is the focus of this case study is ABC, which
was being used as a catalyst to induce fundamental cultural change in the
organisation: that is, instilling a new cost awareness in the organisation.
Organisation structures changed before the introduction of the new control
systems and a number of new people had been brought into the organisation
at very senior levels.
ABC was implemented across all parts of the bank. The bank perceived that
in the current competitive environment there is a need for more information
on costs, particularly when the bank is tendering for clearing contracts. When
the project was initiated the bank was not in a position to break down these
costs. This cost management approach has generated a great deal of interest
from all the UK clearing banks. The aim of the ABC project in the bank was to
be able to identify what exactly the costs and services in the clearing depart-
ment were and to identify the cost drivers of the activities in the value chain.
The foci of the ABC system are as follows:

- establishing what people did in terms of activities and analysing all jobs,
departments and activities;
- establishing why the costing systems that were in place were very poor;
- establishing which activities were expensive and why were they being used;
- establishing what effect increased volumes being put through the system
had on cost.

The practical reasons cited for the introduction of ABC included:

- understanding how costs increased and decreased with increased volumes

of cheques being put through the clearing system;
- appreciating how resources were consumed and used in order to generate
cost savings;
- establishing what is the process behind the cheque clearing process and
how they changed;
- providing more accurate costing measures for the clearing process. Further
advantages of using an ABC system were perceived to lie in its links to
competitive advantage, market share and positioning.

Other reasons for the introduction of ABC are probably related to the per-
sonal ambitions of those promoting its introduction and seeing it as a means
 Kim Soin 293

of developing a power base. Initiating new ideas is looked upon very

favourably, particularly when they involve cost-saving identification devices.
The ABC system implemented in bank operations was both a strategic and a
behavioural system. From the behavioural perspective it enabled the bank to
make better use of costs and resources. The strategic perspective allowed for
reorganisation, for example with respect to marketing. This in turn has led to a
greater awareness of costs and cost control which has influenced the pricing
and marketing decisions of services.
The management control systems were all interlinked and to a large extent
dependent on each other. The ABC systems were related to other management
control systems. Between the resources engineering control system and the
ABC system there was an information overlap. The communications systems
enriched the control systems, and this synergy is related closely to technol-
ogy. Note that technology has been an important element of the new control
systems, for it has enhanced communication and made the flow of information
more efficient. Technological change has meant that changes in procedures
can be dealt with more efficiently. Information technology has provided the
necessary software to make ABC a feasible control system.
ABC has, therefore, been introduced as a part of the management control
systems in this particular bank. Its introduction was part of the attempt to
change the culture of the bank (promoting cost awareness) which has arisen as
the result of changes in the competitive environment and is enabled by tech-
nological changes. Whether there will be success in achieving the bank's aims
remains to be seen. The system remains in its infancy at the time of writing.

Critical reflections on management control systems in banks

The contingency theory of management accounting suggests that the environ-

ment facing an organisation has an impact upon the organisation and its
control systems. Environmental changes may therefore have implications for
the control system in the case of banking. Financial services environmental
changes have taken place. These are largely manifested in the deregulation
and capital adequacy rules that have been invoked. The contingency approach
to management accounting provides a useful framework for the case study for
suggesting why these changes arise, but there are other critical issues which
need to be addressed in relation to the nature of the MCS in this case study.
As Dent (1990) suggests, in addition to being affected by factors in the
organisation's environment, management control is an interactive process.
Contingency theory fails to take account of the two-way interaction between
organisation and environment, seeing the pressure as only flowing from the
environment to the organisation. It is a determinist model which ignores
organisational choice. Contingency theory has also failed to consider the
294 The financial services sector

nature of the process of organisational control. Preston (1991)/ for example,

argues that organisations are able to playa part in the creation of their envir-
onment/ rather than simply responding to it. He proposes that environments
may be contingent upon the actions of the organisation and, in turn, an organ-
isation's structure and process may be contingent upon events in the environ-
ment. Where an environment begins and an organisation ends is not clear-cut
and the relationship between organisations and their environments is more
complex and interrelated than contingency theory would suggest. In the
context of this complexity, strategy, culture and organisational change are
important elements which relate to management control.

Strategy

Simons (1990) emphasises the potential power of management control in the

process of strategy formulation as opposed to just strategy implementation.
Control and strategy are linked in a dynamic process through language and
discourse: strategy forms control systems and control systems form strategy.
The link between strategy and control systems is not a simple one-way process
but a highly interactive one (Dent, 1990; Simons, 1990). Dent suggests: 'New
planning and control systems provide a means for re-orchestrating respons-
ibilities and linkages into the environment, facilitating organisational change,
not just in a passive way, but proactively' (1990/ p . 20). Rather than taking the
objectives of the organisation as determined, the management control system
can actually influence the formation of objectives. Strategic change is por-
trayed by Dent (1990) as essentially an interpretive phenomenon involving the
uncoupling of structures, systems and strategies from already existing para-
digms and their recoupling to new ones. Uncoupling and recoupling is cast as
an emergent process.
In the case of the bank discussed, ABC has been introduced as an element of
overall strategy as a reaction to environmental change. Changes that have
taken place in the environment are largely manifested in the deregulation and
capital adequacy rules that have been invoked. Both have had an effect on
strategy: in the 1980s, expansion and diversification; in the 1990s, contraction,
with an emphasis on cost/ efficiency and rationalisation. The change in strat-
egy placed a new emphasis on the role of the management control systems.
ABC was used as both a strategic and competitive weapon and as a means of
instilling a fundamental cultural change in the organisation.

Culture

A new culture can be an important source of power and influence. Beliefs

about being able to control costs/ to rationalise and to create efficiency can
 Kim Soin 295

often bestow privilege and status upon those who can achieve this through
cultural change. The notions of efficiency, rationalisation and cost (which are
embodied in ABC) are 'infiltrating the organisation settings, leading to the cre-
ation of particular agendas' (Dent, 1991, p. 707). ABC was being used to instil
a fundamental cultural change in the organisation, one where the emphasis
was on cost control.
The change in culture in the bank can be conceptualised as the 'uncoupling of
organisational action from one culture and its recoupling to another' (Dent,
1991). The bank has gone from virtually no measurement and control of costs to
a strict programme of cost control and rationalisation. It is an emergent process
brought about by a change in the accounting system, one that is perceived to be
a cost-saving identification device, namely ABC. The control system has there-
fore introduced a new concept of costs and resources. It has sought to change
the whole banking culture from a process culture (a culture where only
sufficient effort was devoted to getting through the work) to a business culture.
In the bank, a change in leadership, promoted by the demands placed on
the organisation by its external environment, has been the impetus for the
change in control systems and the ensuing cultural change. These particular
MCS are 'facilitating organisational change, not just in a passive way, but
proactively' (Dent, 1990). The following section will examine the extent of the
change that has taken place in the bank, using a theoretical model developed
by Laughlin (1991).

Organisational change

Laughlin (1991) in his model of organisational change puts forward the thesis
that the processual dynamics of change in an organisation are conceptualised
in relation to an environmental disturbance or 'jolt' . This disturbance triggers
transitions and transformations along different tracks or pathways. The key
assumption here is that organisations are naturally change-resistant, with a
strong tendency to 'inertia', and will only change when 'forced' or 'kicked'
into doing something. There is no single result for any disturbance, the degree
of transformation will differ over time and across different organisations;
there are a number of possibilities for the end result.
When viewed in terms of a theory of inertia (Starbuck and Hedberg ,1977;
Jonsson and Lundin, 1977; Mintzberg, 1978; Miller and Friesen, 1984),
attempts to introduce organisational and cultural change has been precipi-
tated by a crisis: government regulation (capital adequacy) and government
deregulation (Financial Services Act 1986 and Building Societies Act 1986).
The real threats were only perceived when the financial services industry was
deregulated and the capital adequacy rules were imposed.
The banking environment has itself become more unpredictable. The com-
bined effects of deregulation, the capital adequacy requirements and the
296 The financial services sector

harmonisation of the EC have made the banking environment very competi-

tive and this in tum has led to uncertainty. Bank-wide there seems to be a con-
certed effort to work with these uncertainties. Many decisions are
'unprogrammed' and rely on the judgement of managers. ABC is being imple-
mented bank-wide and resources engineering has been 'piloted' in other areas
of the organisation. This provides the 'kick' promoted by the environmental
change.
Central to Laughlin's framework is the way organisations are conceptu-
alised: as being an amalgam of 'interpretive schemes', 'design archetypes' and
'sub-systems'." The interpretive schemes, design archetypes and subsystems
are potentially, at some point in time, in some dynamic balance. This means
that at some level there will be certain characteristics that bind the organisa-
tion together and make it a coherent whole. It is only an environmental disturb-
ance which will require the organisational participants, however reluctantly,
to shift the inert characteristics of organisational life.
Laughlin develops four models of organisational change: rebuttal, reorienta-
tion , colonisation and evolution. The first two are categorised as first order
change; things are made to look different while remaining basically as they
always have been. The interpretive schemes will not be affected. Second order
change will penetrate so deeply into the 'genetic code' that all future genera-
tions will acquire and reflect these changes. Second order change will affect
the interpretive schemes, design archetype and subsystems.
In the case of rebuttal, the disturbance is deflected, so that the organisation
is maintained exactly as it was before the disturbance. Reorientation changes
are assumed to affect not only the design archetype but the subsystems as
well. Change of a colonisation nature is not chosen but forced on the organisa-
tion and is change of a second order type. In the case of evolution, change is
chosen and accepted by all organisation participants freely and without coer-
cion. These categories are not meant to describe all possibilities, but to provide
a framework upon which to develop understandings of the actual situation.
The nature of the change that is taking place in the bank is not as well
defined as the 'ideal types' of the model. Interviews with managers and senior
managers suggest that changes of different types have been taking place at
different levels of the organisation and it is too early to gauge whether cultural
change of a second order nature has actually taken place. A senior manager
interviewed sees the changes taking place at two levels: while senior man-
agers, policy makers and the board see these changes as fundamental cultural
changes, at the lower levels of the organisation change remains of a first order
nature. Underneath the mantle of changes the old organisation still exists. In
some cases (where activities have not been changed drastically) people are
working in the same way as before and 'rebutting' the changes. In other situ-
ations the disturbance is accepted into the workings, but the heart of the
organisations remains unchanged. This senior manager believed that changes
that are decided at senior management level do not necessarily lead to change
 Kim Soin 297

of a second order nature. Different types of change will occur at different

levels. Individuals at lower levels of the organisation will often refuse to
accept management changes.
The second manager interviewed sees the change process happening in a
way similar to that described in Laughlin's model: employees initially fought
against the changes (rebuttal) and tried to ignore the benefits because a more
competitive internal environment was being promoted. Eventually the
changes were accepted into the organisation, but the heart of the organisation
remained unchanged (reorientation). With the wave of redundancies and the
level of uncertainty, it was a no choice situation' . Though the changes were
forced upon the organisation and its participants, it seemed to be that the
change had permeated into the very heart of the organisation (colonisation).
The systems are now being used by organisational participants because they
are experiencing the benefits for themselves.
There has been a definite attempt to change the culture of the organisation
(by the new general manager). To enhance this cultural change, new managers
have been imported at very senior levels and in placed key positions in the divi-
sion, to ensure and support the changes that were taking place. The control
systems were embedded into the wider organisational system, and were
playing a key role in the organisation's development, formation of goals and
strategic objectives and their subsequent achievement. Despite these attempts to
colonise the organisation, it is clear that second order change is not universal.
Thus the formal control systems of ABC were supported by informal control
mechanisms. Much of the behaviour observed in the bank seems to be shaped
by emerging social standards and interaction, rather than by the formal
control system (Hopper and Berry, 1983).

Conclusion

This chapter has provided a descriptive overview of change in management

control in the financial services sector and has provided some case study evi-
dence. The banking environment of the 1990s has served as a catalyst for the
emergence of new control systems. These control systems are a direct result of
the environmental changes and the mistakes made by banks in the 1980s.
Clearly management control in the financial services sector is being carried
out in a wide variety of ways and it is evident that the process of manage-
ment control cannot be separated from the process of organisational, opera-
tional or strategic control, and is therefore a complex and highly interactive
process. The relationships between control systems and strategy, culture,
organisational change and emerging social standards are highlighted as being
particularly important when assessing the nature of management control
298 The financial services sector

systems in a bank. Management control in banking is a dynamic process

where control systems are influencing the formation of objectives.

Notes

1. Institutions, other than banks, that canalize the supply and demand for long-term
capital and claims on capital. For example, the Stock Exchange, insurance compa-
nies and other suppliers of core banking services.
2. The focus of this study was on the cleaning department in which an ABC system
was being implemented.
3. Where the interpretive schemes are the underlying set of values and belief; design
archetypes are the organisation structures and management systems; and the sub-
systems are the tangible organisational elements (buildings, people, machines,
finance).
 CHAPTER 19

Manufacturing accountability
T. Colwyn Jones and David Dugdale

This chapter is concerned with roles of accounting in management control,

especially in large-scale UK manufacturing enterprises. We focus on discus-
sion of 'the new manufacturing environment' which many people have
identified as emerging in the 1980s and claim constitutes a major reorganisa-
tion of production. Some argue this is a transformation as profound as the cre-
ation of 'the factory system' at the end of the eighteenth century, or the rise of
'the modem corporation' at the end of the nineteenth. If so, we would expect
this change to have a strong impact on accounting and management control in
the 1990s. We explore this through discussion of 'regimes of accountability' -
sets of theories and practices aimed at ensuring that employees of the enter-
prise are held accountable for their activities and/or achievements.
Accounting, as a distinctive discipline or occupation is not the only body of
theory and practice relevant in this. However its contributions to the determ-
ination of ends (as objectives, targets or criteria) and the monitoring of means
(through measurements, analyses and reports) have led to its gaining a central
position in management control.
The title of the chapter was prompted by Burawoy's (1979) 'Manufacturing
Consent', although our emphasis will be rather different. Burawoy was con-
cerned with demonstrating how workers, by playing the game of 'making-
out' within management control systems, came to accept their rules. This he
referred to as 'consent' . He did not explore how managers had come to con-
struct these systems or whether they had deliberately manufactured them to
achieve consent. In our view, his workers might more accurately be seen as
demonstrating lack of dissent rather than actively consenting to management
regimes. Our focus is upon accountability and how managers have attempted
to create it through management control systems in which there are different,
and changing, balances between consent and dissent, in differing social con-
texts. We will examine ways in which accounting operates not just as hard
accountability 'via financial results and numerically rendered information' but
299
300 Manufacturing accountability

also .as soft accountability through its role in shaping 'contextual, subjective
evaluation stressing human values and superordinate goals' (Hoskin and
Maeve, 1988, p. 68). We will argue that accounting is important not only in
formal control systems as an external check on people's performances and
achievements, but also in attempts to construct employees who are 'respons-
ible' - or self-accountable. Both, to varying degrees, are present in regimes
concerned with manufacturing accountability.

The nature of management control

Management control is inseparable from its organisational context. Ouchi

(1980) classified economic organisation in terms of three mechanisms for
mediating transactions between individuals associated through bureaucracies,
markets and clans . We use a similar three-dimensional model but with funda-
mental differences in interpretation and explanation.

Hierarchies and rules

Large-scale manufacturing enterprises are frequently vertically and horizon-

tally integrated units of production administered by management hierarchies.
Discussion of such organisations has been heavily influenced by Weber (1922)
who characterised modern society as one where non-rational modes of
thought (traditional and affective) were increasingly being replaced by rational
forms based on calculation. These involved the conscious thinking through of
cause and effect; of means and ends. The clearest demonstration of this was
the capitalist enterprise where accounting was its purest expression.
Enterprises had developed a specific form of rational administration -
'bureaucracy' - in which a hierarchy of offices, consisting of official duties,
were regulated by a system of rules (see Box 19.1). These rules might take legal
and/ or technical forms and constituted the legitimate basis for authority to
the extent that they were ,\ccepted as rational by members of the organisation.
Officials acted, not on the basis of customary or habitual responses, or
emotion, but by impersonally calculating the means and ends of action by
applying the abstract body of rules to particular cases.
Weber predicted that this form of administration would become dominant
in twentieth-century organisations and his concern was to analyse its structure
and processes. Others were more actively involved in constructing such forms
through their advocacy of management 'principles'. Taylor (1911) in the USA,
Fayol (1949) in France and Urwick (1947) in the UK prescribed forms of
administration which had much in common with Weber's model of bureau-
cracy. Together with other writers and management consultants such as
 T. Colwyn Jones and David Dugdale 301

Box 19.1 Weber's model of bureaucracy

Fixed jurisdictional areas where there are detailed official duties, whose
conduct is limited by rules and which are methodically provided for
each organisational activity.
A firmly ordered hierarchy of superior and subordinate offices.
Offices are both a centre of written documentation of activities and a post
(or position) whose public resources are clearly separated from the
personal funds of the office-holder.
A prescribed course of expert training (through formal education and/ or
experience) which prepares the individual to hold office.
The holding of office is the primary activity of the office-holder who is
rewarded by salary.
The conduct of each office, and relations between them, are covered by
general rules (which may also state how rules are to be changed).

Follett (924), Gulick and Urwick (937), and Barnard (938), they produced
what is now termed 'classical management theory'. Whilst their prescriptions
differed in detail, they collectively advocated functionally separate depart-
ments, with specific duties for managers, to replace the wide-ranging activities
in earlier forms of organisation. This, they claimed, would generate efficiency
within enterprises.
There has been considerable controversy over bureaucratic organisation.
Psychologists in the human relations traditions (for example, Mayo, 1933,
1945; McGregor, 1960; Argyris, 1964) objected to the mechanistic and imper-
sonal nature of bureaucracy and saw it as stifling people's initiative, imagina-
tion and creativity, thus leading to inefficient use of human resources.
Sociologists argued that bureaucratic forms were only efficient under certain
conditions, such as large-batch assembly and mass production (Woodward,
1958a, 1958b) or in stable market and technological environments (Burns and
Stalker, 1961) or where routine production resulted from the use of uniform
raw materials transformed by well-understood processes (Perrow, 1970).
Elsewhere they were inefficient. However, for economists adopting the 'trans-
action costs' approach, the rise of bureaucratic hierarchies in general had to be
seen in terms of their economic advantages (Chandler, 1962, 1977;Williamson,
1975). If the 'invisible hand' of the market had been replaced by the 'visible
hand' of management (Chandler, 1977) then this must be due to higher
efficiency. In modern production raw materials are transformed into semi-
finished goods and then into final products within multi-function enterprises
without passing through markets. The reason must be that, for complex econ-
omic activities, the costs of market transactions rise to a point where they are
302 Manufactu ring accountability

unbearable and bureaucracies (despite their own inefficiencies) are preferred

(Ouchi, 1980).
Whatever the explanation for the rise of bureaucratic hierarchies, the impli-
cation is that management control is to be achieved by specifying the actions
required of enterprise members. Rules specify official duties, how they are to
be carried out, the authority and resources available to managers and to
whom (and for whom) they are responsible. Here the essential meaning of
'performance' is the conduct of duties in accordance with predetermined pro-
cedures; it is this which is to be monitored and reported. This emphasises
formal rationality which is 'the extent of quantitative calculation or accounting
which is technically possible and which is actually applied' (Weber, 1922,
p. 85) and 'subordinate managers are ... evaluated on the basis of their con-
formity to organizational rules' (Hopwood, 1974, p . 19).

Markets and contracts

In recent years there has been increasing concern that large-scale enterprises
have reached the limits of efficiency achievable under regimes of hierarchies
and rules. One set of prescriptions for this involves the creation of market con-
ditions within the enterprise. Organisations should be segmented into divi-
sions and then departments which operate as mini-businesses 'buying' from
internal suppliers and 'selling' to internal customers. This construction of sub-
units of firms may be seen as an attempt to drive 'the discipline of the market'
deep within enterprises, encouraging managers to see their interactions with
each other as 'trading' relationships, and thus treating them as entrepreneurs
whose objectives and activities are micro-level equivalents of traditional entre-
preneurs. The problem, of course, is that departments are not independent
businesses, and managers do not own them as personal property. Thus
attempts to create market relations within enterprises can only be simulations
- pseudo-markets with pseudo-entrepreneurial activity.
One influential approach to constructing management control which aims at
constructing such pseudo-market conditions within the enterprise is agency
theory (Hogler & Hunt, 1993). This begins with the assumption that remote
owners (principals) have delegated decisions to managers (agents) whose
interests are different and who must therefore be controlled. The organisation
is seen as a network of contracts between individuals. Principals lack the time,
expertise and detailed knowledge to specify exactly what their agents should
do (Shapiro, 1987). Instead principals are concerned to set targets and monitor
the outcomes of action. The theory assumes that agents are utility maximisers
and, without control systems, will be inclined to be lazy and/or divert
resources and efforts toward non-owner goals. Agency theorists have been
concerned with payment systems as incentives for agents to pursue the goals
of owners, and information systems which enable principals to monitor their
 T. Colwyn Jones and David Dugdale 303

achievements. The creation of such systems entails a cost to principals but is

economically advantageous if this is less than the benefits of contractual
control. However a central problem with this approach is that monitors are
also agents and they in turn must be monitored (Armstrong, 1991). Thus the
logical outcome of agency theory is an endless chain of regulatory relation-
ships which eventually must become prohibitively expensive.
Emphasis on control through contracts and monitoring of outcomes focuses
on Weber's concept of substantive rationality: 'the degree to which the provi-
sioning of given groups of persons with goods is shaped by economically
social action under some criterion of ultimate values ... [which are] bases from
which to judge the outcomes of economic action' 0922, pp . 85-6). Unlike
formal rationality, which is concerned with the means adopted by managers,
substantive rationality focuses upon the relationship of means to ends.
Unfortunately for those who wish to explain enterprises in terms of the
pursuit of efficiency, these two forms of rationality can conflict. By following
the letter of the law, decisions may go against the spirit of the law. Reliance on
information about the outcome of managerial action may lead to managers
attending only to those items which are reported, or diverting their efforts into
constructing information which generates the preferred message.

Trust and values

There are limits to the extent to which enterprises can be controlled through
formal application of either rules or contracts. Managerial jobs may be
described as 'high discretion roles' (Fox, 1974) where 'if some crucial organiza-
tional roles are essentially complex, containing discretionary elements, and
requiring skills and judgement, then the commitment and 'trustworthiness' of
these members is most important' (Salaman, 1979, p. 98). Ouchi refers to organ-
isations which exemplify such trust relationships as 'clans', where 'a variety of
social mechanisms reduces differences between individual and organizational
goals and produces a strong sense of community' (1980, p. 136). Central in this
are shared values within the organisation supported by systems of shared
meanings, language, symbols, myths and stories - the 'cultu re' of the organi-
sation (Langfield-Smith, this volume). This might appear to circumvent the
need for management control since organisation members share views on
means and ends. Culture may be seen as 'a pervasive way of life, or set of
norms' (Handy, 1985, p . 186) which preconditions relationships between organ-
isational members. However culture is also viewed as an outcome of existing
enterprise structures and processes (thus reflecting its rules and contracts) and
as something which senior executives can design and implement in order to
change organisations (Handy, 1985). Thus culture may be 'manipulated or cul-
tivated to achieve better management control' (Langfield-Smith, this volume)
if organisational members can be socialised with shared values.
304 Manufacturing accountability

If formal control can never be complete then an issue for management

control is 'deciding who - not whether - to trust' (Armstrong, 1991, p . 13).
What social mechanisms are available to identify or create trustworthy
members? Salaman (1979) pointed to five features of the employment of man-
agers which facilitate trust. First, managers often come from family back-
grounds with strong business connections. Second, many have been educated
on university courses, steeped in business ideology, and aimed at producing
potential managers. Third, considerable care is taken in the recruitment of
managerial personnel, selecting those who express empathy with the values of
the organisation. Fourth, managers experience socialisation in corporate
values through training courses and mission briefings. Fifth, managers are
rewarded, not only for their adherence to rules and achievement of targets,
but also for their commitment to 'appropriate' ways of thinking and acting.
These rewards include their salaries, but also encompass the work itself (more
intrinsically satisfying and less controlled than that of lower-level members)
and their work conditions (which offer better surroundings and facilities,
security, status, career prospects, pension rights, and so on). Overall, manage-
ment control is directed at generating trust through the selection and socialisa-
tion of managers with a commitment to the values of the enterprise. As they
explain and justify their decisions and actions - to superiors, subordinates and
to themselves - they articulate these values which become internalised fea-
tures of self-control. In this way management control is concerned with the
social construction of trusted persons whose means and ends require less close
direction and monitoring.
This is not confined to managerial levels. In the creation of Fordism - a
dominant manufacturing paradigm of the twentieth century - Ford intro-
duced not only a 'profit-sharing plan' (the $5 day of 1914) but also a sociology
department to monitor the social life of employees and an English school to
train immigrant workers in appropriate attitudes and habits. As one of the
Ford educators put it, 'as we adapt the machinery in the shop to turning out
the kind of automobile we have in mind, so we have constructed our educa-
tion system with a view to producing the human product in mind' (Marquis,
1916, quoted in Meyer, 1980, p. 74). Thus, as Ford created a mass production
system, he also attempted to construct the kind of worker to be employed
within it. This is not to suggest that such attempts can be fully successful. Ford
had 'labour problems' from the beginning and although Beynon's (1973) study
of Fords in Liverpool some 50 years later found personnel managers still
trying to recruit a 'responsible' workforce, production could hardly be
described as demonstrating 'consent'.
More recently there has been considerable interest in the 'empowerment' of
workers where instead of formal control 'everyone's work is guided by and
aligned to a common vision that company leaders shape and project by their
own example' (Johnson, quoted in Jayson, 1992, p . 31). Again the aim is to
shape workers to enterprise interests and values (the common vision) as a pre-
 T. Colwyn Jones and David Dugdale 305

condition of self-control. We may be sceptical of the possibilities of completing

this project - fully aligning worker interests with owner interests in a
'common vision' - but as Burawoy noted the everyday life of enterprises is
characterised more by co-operation than by conflict.

Contradictions in management. In identifying three aspects of management

control - rules, contracts and values - we are not proposing a model of three
types of organisation (such as bureaucracies, markets and clans). Instead we
see manufacturing enterprises as characterised by various forms of hierarchi-
cal, market and trust relationships which co-exist (see Box 19.2). Bureaucratic
organisation rests upon the legitimacy of rational rules (Weber, 1922) and
hence is concerned with values. Market relationships exist within hierarchies
of position and power. Trust alone cannot be relied upon since socialisation is
never complete (Wrong, 1961) and because different hierarchical positions
generate differing interests. Thus we would expect enterprises to display a
balance of factors rather than a single 'style'. Nor do we argue that 'an
efficiency criterion would make it possible to predict the form organisation
will take under certain conditions' (Ouchi, 1980, p. 129). In order to explain
how management control systems develop we need to explore issues which
go beyond 'efficiency' and 'values' in enterprises and encompass the politics
of management.
Marx (1868) distinguished two functions of management. Given any kind of
division of labour - the separation of tasks and their allocation to different
individuals or groups - then some form of organisation is required to relate
and unify these discrete elements. Marx referred to this as co-ordination and
unity. In capitalist enterprises there are also differences of interest between
individuals and groups because they occupy different economic positions, and
there are struggles over how resources are to be used and for whose ends.
Here management is also concerned with control and surveillance. These
twin functions of management are contradictory since enterprises are simulta-
neously collective, social organisations (in terms of production) aimed at

Box 19.2 The nature of management control

Social Control Focus of

relations mechanism control
hierarchical rules activities
market contracts resources and outcomes
trust values persons
306 Manufacturing accountability

individual, private interests (increasing shareholder wealth). This creates ten-

sions in any management control system since 'employers and managers are
faced with the inescapable problem of achieving co-operative activity by
antagonistic means' (Armstrong, 1991, p. 6). Owners must entrust their
resources to people who do not share their interests and who therefore cannot
be completely trusted. Attempts to resolve this by tighter control through
rules or contracts may result in increased resistance from employees who thus
become less trustworthy and require even greater control, thus increasing
costs. Attempts to escape this by creating more 'high discretion roles' create
more scope for employees to pursue their own interests. In this view enter-
prises are shaped not by some universal, abstract pursuit of efficiency but by
the twin political processes of managers' control of workers, and owners'
control of managers (Hopper et al., 1987; Jones, forthcoming). In struggles over
this two-sided management control, tensions and conflict imply that any
balance between rules, contracts and trust will be unstable and changing.

Accounting and manufacturing enterprises

In UK (and US) manufacturing enterprises accounting has become an import-

ant, possibly the dominant, form of management control. Since this is not true
of enterprises in Germany and Japan (Armstrong, 1985, 1987; Jones et al., 1993)
this cannot simply be seen as the logical and inevitable outcome of the need
for financial information and usefulness of accounting. Hence it requires
explanation.

The rise of accounting in management control. Armstrong 0985, 1987, 1993)

traced the rise of accounting in UK management, locating developments in
historical social context. By the early years of this century the dispersal of
shares among many owners, and active capital markets, had led to auditing
becoming an important activity. From this base the accountancy profession
extended its intervention in enterprises. Financial accountants gained impor-
tance in reporting the impact of economic slumps in the 1920s and the depres-
sion of the 1930s. They interpreted corporate crises in terms of problems of
profitability and liquidity, and recommended tight financial controls.
Typically this was through the adoption of 'holding company' forms of organ-
isation where the information used to monitor subsidiaries paralleled that
already familiar to accountants through annual reporting and auditing. Over
the same period there was a development in costing. UK government policies
during the First World War, aimed at restraining profiteering on cost-plus
wartime contracts, led to the development of costing techniques in ministries
and supplying enterprises. In 1919 the Institute of Cost and Works
 T. Colwyn Jones and David Dugdale 307

Accountants (ICWA) was set up, recruiting members active in enterprises and
stimulating the growth of budgeting and cost accounting systems. As financial
and cost accounting became more important in enterprises career opportuni-
ties were opened up for accountants within management and on boards of
directors. Accountancy was establishing an important position in UK manu-
facturing enterprises.
In the period of industrial reconstruction, following the Second World War,
UK enterprises sent 'productivity teams ' to the USA to study production organ-
isation and methods. Accountants on these teams returned with a new set of
accounting theories and techniques, and a new concept: 'management
accounting' . They advocated 'divisional' rather than holding company organ-
isational structures, and resource allocation and performance - reward
payment systems to regulate activities. The ICWA became the Institute of Cost
and Management Accountants (1972) and then the Chartered Institute of
Management Accounting (1986). Increasingly it promoted its members as pro-
fessionals concerned with 'executive' and 'strategic' issues. Management
accountancy's role was defined far more broadly than cost accounting and its
members became established at higher managerial levels, gaining ascendency
over engineering, marketing and personnel occupational specialisms.
Thus Armstrong explained the rise of accounting in management control as
the outcome of successful strategies by the accountancy profession to promote
the services of its members to enterprises. Beginning with issues of resource
allocation - regulating both external investment (financial accounting) and
divisional investments (management accounting) - the role of accounting
spread more generally in management control.

The fall of management accounting? Management accounting in UK enterprises

has been considerably influenced by ideas and techniques pioneered in the
USA. However in recent years there has been a major debate over the role of
this accounting in US enterprises. Johnson and Kaplan (1987) argued that as
enterprises developed in the nineteenth century many stages of the process of
converting materials into finished products were integrated within organisa-
tions. With the absence of market exchange between these stages there was a
demand for measures to determine the price of 'output' from internal opera-
tions. It was from this that management accounting developed. Drawing on
transaction cost theory, they explained the rise of large hierarchical organisa-
tions in terms of their efficiency advantages over markets. However the full
benefits of organisation would not have been gained without an increase in the
quantity and quality of management accounting information. In responding to
this need, especially in companies such as Du Pont and General Motors,
accountants created effective information systems so that: 'By 1925 virtually all
management accounting practices used today had been developed: cost
accounts for labor, material and overhead; budgets for cash, income and capital;
308 Manufacturing accountability

flexible budgets, sales forecasts, standard costs, variance analysis, transfer

prices, and divisional performance measures' (Johnson and Kaplan, 1987, p. 12).
At this point innovation in management accounting appeared to stop. After
1925 financial accounting dominated, so that cost accounting became designed
to meet the requirements of annual reporting and auditing rather than cost
management. Cost information was increasingly less useful to managers since
it ignored aspects of business (such as research and development, marketing,
distribution) which were growing in importance, and because the attribution
of overhead to products depended on bases such as labour hours or labour
costs which were becoming inappropriate. The result was that 'today's man-
agement accounting is too late, too aggregated, and too distorted to be rel-
evant for managers' planning and control decisions' (ibid., p. 1). Having
explained the rise of management accounting in terms of its contribution to
efficiency, Johnson and Kaplan now abandoned this theoretical framework
and attributed its fall to the dominance of financial accounting and to out-
dated academic theories. The latter continued to assume one product/one
process production, whereas manufacturing enterprises were now typically
multidivisional, multi-product organisations. Management accounting had
not responded to change after 1925 and had thus lost its relevance.

The new manufacturing environment

This critique by Johnson and Kaplan was given greater urgency by the view
that the pace of change was quickening and precipitating a crisis in manufac-
turing enterprises. Changes in manufacturing in the 1980s were widely
identified as a fundamental shift in the nature of modern production.

Flexible specialisation. In an influential overview of this transformation, Piore

and Sabel (1984) argued that changes in consumer taste were leading to the
break-up of mass markets. The demand for products was becoming increas-
ingly fragmented (with many market 'niches') and volatile (with more rapid
'fashion' changes) . This produced strain in mass production systems designed
to produce high-volume, long-run, low-cost, standardised products. In the
new 'post-Fordist' environment there was a need to develop production
systems capable of making a broad mix of products, moving between them as
market conditions dictated, and changing rapidly to new ones . The need to
respond to the new environment was made urgent by the increasing intensity
of international competition - especially from Japan. Fortunately new technol-
ogy (especially computer-based innovations) provided an opportunity to
develop new production systems based on 'soft automation' (reprogramma-
 T. Colwyn Jones and David Dugdale 309

ble, flexible} to replace the 'hard automation' (dedicated, inflexible) of mass

production. This could be linked to more flexible forms of work organisation
using multiskilled workers, and de centralised management organisation
which enabled rapid, local decision making responsive to particular markets.
Together these would facilitate 'flexible specialisation', a new manufacturing
strategy to replace mass production.
This view of a fundamental shift in manufacturing has been challenged both
as to the scale and the nature of the changes identified, and the explanations
advanced. Smith (1989) for example, noted that, while Sir Adrian Cadbury
(Chairman of the Cadbury-Schweppes Group) stated in 1982 the need to adapt
to more varied consumer tastes, his company actually reduced the number of its
products by half in the 1980s. Shaiken et al. (1986) found that, as well as its
potential for flexibility, computer-controlled technology offered new opportuni-
ties for integrated production so that batch production systems which had once
been fragmented were becoming more like mass production. Hopper and
Armstrong (1991) argued that the driving force behind change was not con-
sumer taste. In the period after the Second World War the strength of organised
labour had forced US giant corporations to come to terms with their workforces
in a 'management - labour accord', the costs of which they were able to pass on
to customers (in quasi-monopolistic US markets) and to smaller firms (with
non-unionised labour). Increasing international competition in the 1980s put
pressure on these enterprises to change their production systems, whilst the
weaker position of workers (and their unions) in the new economic/political
climate gave them the opportunity to change working practices.
These criticisms suggest that change in manufacturing has not been as wide-
spread, unilinear or fundamental as much discussion of flexible specialisation
implies. Nevertheless there is sufficient evidence of the development of new
production practices for us to see the 1980s as period of significant change and
to suggest that this has important implications for management control in the
manufacturing of the 1990s.

Changes in manufacturing. Innovation in production technology - usually

labelled advanced manufacturing technology (AMT) - has attracted consider-
able attention. Although developments are patchy, many enterprises have
seen the arrival of computer numerical control machine tools (CNC), robots
and computer-aided design - computer-aided manufacturing (CAD/CAM)
systems. A few have adopted flexible manufacturing cells (FMC) or flexible
assembly cells (FAC) and have aspired to flexible manufacturing systems
(FMS) which link to management information systems (MIS). These technolo-
gies incorporate computer-based information and control systems and offer
the long-term prospect of computer-integrated manufacturing (CIM).
Alongside these technological innovations there has been the adoption of
new production techniques, particularly those associated with Japanese 'just-in-
310 Manufacturing accountability

time' manufacturing (JIT). This, in Japan, is seen as a 'philosophy' aimed at

reducing manufacturing costs by delivery of materials to production areas
when they are required. It relies on materials being 'pulled' through the plant
under the regulation of 'kanban' plates or cards which prompt workers to
produce batches of the required materials, rather than the creation of stock
which is then 'pushed' into the next stage of production. The system depends
on smooth operation requiring 'zero defect' materials (since there are no just-
in-case alternative stocks) . In its most spectacular practice - 'jidoka' - produc-
tion halts when faults are discovered and attention is focused on eliminating
the error before production recommences. The 'kaizen' approach of continual
improvement is aimed at achieving (or overachieving) levels of quality
required to construct and maintain the JIT system, and quality circles of super-
visors and workers monitor and improve the process. Although it is unlikely
that many UK enterprises have implemented the full range of JIT activities
(Jones et al., 1993), techniques such as total quality control (TQC), total preven-
tative maintenance (TPM) and statistical process control (SPC) have been
introduced.
There have also been changes in job design and work organisation which may
be seen as moves away from Taylorist and Fordist production. Kelly (1982)
identified three widespread changes: first, reorganisation of flowlines where
long chains of work roles are broken up either into several shorter chains or
are replaced by individual workstations; second, the creation of flexible work-
groups, where a number of work roles are amalgamated and distributed to
groups where workers are allocated (or allocate themselves) to tasks as and
when required; third, vertical role integration where a number of work roles
once carried out by different workers are amalgamated in one job.
Employment relations have also changed. Atkinson (1984) identified three
forms of flexibility pursued by UK enterprises in the 1980s: functional (involv-
ing the movement of 'multiskilled' workers between tasks); numerical (aimed
at rapid increase and decrease in employment numbers depending on product
market conditions); and financial (tying pay levels more closely to product
output and labour market conditions). In his view, this represented a new
employment model - 'the flexible firm' - where a core group of full-time, per-
manent employees was separated from a more marginal group of temporary,
self-employed and subcontract workers, beyond which workers in other firms
were increasingly being used for outsourcing of materials. His view that this
amounted to a new managerial strategy has been criticised by Pollert (1988) on
the grounds that it merely represents a redistribution of traditional practices,
reflecting the growth in female employment and the decline in traditional male
occupations. In terms of both work organisation and employment relations,
the highly publicised activities of Edwardes (British Leyland), McGregor
(British Steel and the National Coal Board), Shah (Today newspaper) and
Murdoch (Times newspapers) may have been extreme cases not representative
of management initiatives elsewhere. Nevertheless, under the encouragement
 T. Colwyn Jones and David Dugdale 311

of the Thatcher government, there does seem to have been some kind of
'employer offensive' (Hyman and Elger, 1981) in the 1980s where the weaker
position of workers in labour markets and trade unions enabled management
to make significant changes to working and employment practices.
There have also been signs of managerial reorganisation popularly labelled
'delayering'. This involves the stripping away of what are now seen as un-
necessary layers of middle management and placing more responsibility in
supervisory leaders of workgroups. Hopper and Armstrong argued that this is
a consequence of the declining difficulty of controlling labour in the harsh econ-
omic climate of the 1980s and 1990s and thus 'the bureaucratic and costly appa-
ratus of control in large core conglomerates, which had emerged in more
benign economic conditions of the last fifteen years or so, is increasingly being
questioned' 0991, p. 434). The achievement of this delayering is often accom-
panied by discussion of 'em powerment' , where responsibility is delegated to
lower-level employees - but only if they can be relied upon to pursue the
goals set by senior management.
Overall there is evidence of significant change in production technologies
and techniques, work and employment patterns, and managerial organisation
in manufacturing enterprises in the UK. All of this implies a new environment
for management control.

Accounting and formal management control

Most discussion of the contribution of accounting to management control

focuses on its formal procedures - the way its measurements, techniques and
criteria produce accounting information which contributes to setting targets,
allocating resources, monitoring performance and rewarding achievement.
Here accounting both reflects and shapes the rules and contracts of enterprise
in the context of hierarchical organisation and attempts to create pseudo-
market relations within the enterprise.

Controlling activities. Traditional management accounting practices are built

upon standard costing systems (originally developed in Taylorist and Fordist
production) which employ a formula of standard labour cost plus standard
material cost plus standard overhead to identify product costs . Control is then
achieved through the monitoring of variances between these standards and
actual costs. In the new manufacturing environment these standard costing
systems have been criticised as outdated, misleading and irrelevant.
With increasing use of AMT direct labour is seen as a diminishing
proportion of total costs. Yet it remains of central importance in standard
costing both in itself and as a base for the recovery of overheads. In flexible
312 Manufacturing accountability

low-volume production systems machine set-up and changeover times take

on increased importance, but conventional accounting does not highlight
these and they may become lost in an aggregate of indirect labour. TQM aims
at zero-defect, but material standards may imply an acceptable level of scrap.
JIT strives towards minimal stock based on small batches. However emphasis
on material price variances may encourage bulk buying and measures of
labour efficiency can lead to long production runs, both of which increase
stock. In these and other specific ways traditional costing may be seen as out
of step with new production systems, and overall the very notion of 'standard'
has been seen as inappropriate in companies attempting to achieve flexible
specialisation.
Alongside these criticisms of conventional accounting there was a stream of
advice from academics and management consultants on more appropriate
new approaches. The most influential of these has been activity-based costing
(ABC). Shifting the emphasis from direct to indirect labour, Cooper and
Kaplan argued that, 'Virtually all of a company's activities exist to support the
production and delivery of today's goods and services. They should therefore
be considered product costs' (1988, p. 96). The central aim of ABC is to trace
costs from resources to activities to specific products by identifying cost
drivers and measuring the costs which ensue. For example, the decision to
produce a new part drawing triggers not only drafting, but also inspection,
data processing, quality control, stock-keeping and parts-ordering activities
(Johnson, 1992b). When the costs of all such activities are measured managers
will be in a position to 'measure costs right: make the right decisions' (Cooper
and Kaplan, 1988, title). The outcome of an ABC analysis typically shows that
products made in small batches may consume a small number of direct labour
hours but require almost as much indirect labour as large batches. Thus stan-
dard costing underrepresents their costs - they are not as profitable as tradi-
tional information suggests. This should not lead managers automatically to
drop low-volume lines; instead the new information may highlight the need to
develop flexibility and encourage investment in AMT which is capable of
improving this.
As ABC developed, producing spin-offs of activity-based budgeting and
activity-based management, it appeared to provide a powerful new means of
management control more suited to the new manufacturing environment.
However its critics argued that its 'solution' to the problems of standard
costing was misleading or irrelevant. Its promise was that, by discovering the
'right' costs of products, managers would be able to produce more profitabil-
ity . For some this was illogical, since 'Products are not profitable or
unprofitable, businesses are' (Waldron, 1988, p. 1). Attention should be
focused not on costs but on throughput.
The theoretical base for this view was supplied by Goldratt (1984) in devel-
oping the theory of constraints (TOC). In his view profit was simply revenue
less costs. These costs have two elements: materials and operating expense.
 T. Colwyn Jones and David Dugdale 313

Investment shows itself in the company in the form of inventory. Thus the
measure for rate of return on investment (ROD is:

ROI = (revenue - materials) - operating expense

investment/inventory

A company may increase its ROI by reducing operating expense, or invest-

ment/inventory, but both are inherently limited since some minimal levels
must be maintained. The alternative approach is to increase revenue-less-
materials by achieving greater throughput from existing facilities. (Or to put
this another way, revenue-less-materials is essentially a measure of the
amount of throughput.) Since, in principle, throughput can be increased
without limit, managerial attention should be focused on this, and not on
minor issues of costing. In particular, managers should identify any factors
which hinder the increase of throughput which are the system 'constraints'. In
Goldratt's early work, these were identified as bottlenecks - often (but not
necessarily) focal machines. The stripping of stocks (under JIT systems) from
non-constraint locations in production would be beneficial since it should
encourage reduction in lead times and increased customer responsiveness,
thus promoting throughput. The bottlenecks, however, must be protected by a
'buffer' of stock to ensure that production never pauses. Management of con-
straints was thus the key issue and here there was a fundamental disparity
between views from 'the cost world' and from 'the throughput world' . The
emphasis in management control should shift from reducing costs to increas-
ing throughput.
In a break-away move, Galloway and Waldron (l988a, 1988b; 1989a, 1989b)
advocated throughput accounting (TA) which would reform costing in line
with an emphasis on bottlenecks and throughput. This return to a more tradi-
tional emphasis was rejected by Goldratt and now there is considerable hos-
tility not only between supporters of TOC and ABC, but between TOC and
TA. So far, ABC seems to have made the largest impact in UK manufacturing,
but even here many companies seem content to take the first step of
analysing activities without moving to implement new costing systems
(Lyne, 1993).

Controlling resources. Standard costing has also been the basis for traditional
budgeting techniques. In large hierarchically organised enterprises, divisions
and departments were defined as 'cost centres' and managers were held
responsible for costs incurred within their boundaries. Budgeting processes
can begin with complex and prolonged negotiations between higher and
lower management concerning the distribution of resources. Once deter-
mined, this is typically controlled through variance analysis, which has
314 Manufacturing accountability

FIGURE 19.1 Detailed Variance Analysis

Manufacturing CostVariance

Materials Variance Production Variance

n
Price Usage
Labour
Overhead

Mix Yield
Rate Efficiency
Variable Fixed

Expenditure Efficiency Spend Volume

Capacity Efficiency

developed into a highly elaborate examination of divergencies of 'expected'

(standard) from 'actual' costs (see Figure 19.1).
Whilst the logic of this approach may appear sound, its implementation can
be problematic. Where it is applied rigorously in a 'budget constrained style'
(Hopwood, 1974) management control outcomes may be viewed as dysfunc-
tional. These problems are widely noted in textbooks. Effective negotiators
may acquire a disproportionate share of resources; budget padding may be
employed to provide slack; buck-passing may be rife. In order to stay within
current budgets managers may delay or veto expenditure on items with
longer-term benefit - for example, by failing to carry out regular machine
maintenance.
In order to combat this, some companies have set up 'profit' or 'investment'
centres instead. Here divisions or departments are monitored using profit or
ROI figures in attempts to align subunits' contributions more closely with the
overall profitability of the enterprise. Again many textbooks note the distor-
tions which can creep in. Under ROI managers may retain assets which have
little 'book value' and investment in new equipment may be delayed or
avoided. Decisions may be directed at changes which improve local ROI but
make little contribution to the overall goals of the enterprise. In some cases the
companies have attempted to resolve this by relying instead on residual
income (RI). However, as Emmanuel and Otley note, 'there are technical
 T. Colwyn Jones and David Dugdale 315

defects in any accounting measure, and to the extent that these are incorpor-
ated in controllable RI, the manager's commitment to it will be diminished'
(1985, p . 175).
The development, first, of cost centres and then of profit and investment
centres can be seen as attempts to create pseudo-market relations as divisions
and departments are constructed as though they were mini-businesses within
the enterprise. However these units are not autonomously operating in
markets but exist within hierarchical relations. This generates contradictions
and each solution generates new problems which have been addressed by
refining ever more sophisticated formal accounting techniques. These contra-
dictions become more intense when control of divisions and departments is
closely linked to appraisal of their managers.

Controlling performance. If enterprises are viewed as pseudo-markets with

networks of contracts, crucial issues in management control are the setting of
targets and the monitoring of their achievement by individual managers.
Advocates of agency theory in accountancy (Jensen and Meckling, 1976; Fama,
1980; Watts and Zimmerman, 1986) argue that accounting offers principals
(owners and executives) the potential for regulating their relationship with
agents (managers) at minimum cost/maximum benefit. One key element here
is the tying of managerial reward to performance - as measured by financial
indicators. The payment of performance bonuses (averaging about a quarter
of annual salary) to profit centre managers seems to have been a widespread
practice in US enterprises for some time, and in the majority of cases these
bonuses are decided solely or partially by a financial measurement formula
(Vancil, 1979). This appears to have been far less common in the UK, but
Ezzamel and Hilton (1980a, 1980b) found that profit measures were seen as a
major factor in the promotion of managers in over half the companies they
surveyed. Thus accounting frequently provides information which is con-
cerned both with co-ordinating the activities of units and at the same time
encouraging competition between managers (for pay and promotion).
The tension created by this contradiction is recognised in textbooks:
'Accountants often face a dilemma because they are supposed to fulfil two
conflicting roles simultaneously. Firstly, they are seen as watchdogs for top
management. Secondly, they are seen as helpers for all managers' (Horngren,
1977, p. 11). Practitioners recognise the problem this causes through the
popular slogan, 'you get what you measure'. This implies either that managers
concentrate their efforts on improving only those items which are highlighted
in performance appraisal or that they manipulate information supplied to
senior management to give the desired impression. As one financial controller
told us, 'the minute you say that 'that' is what you are measuring, if you
haven't got a balance and a check in the system 'it' will get better!' Providing
such balances and checks leads to more elaborate accounting systems and
316 Manufacturing accountability

increases costs, so that (eventually) proliferation of controls becomes prohibi-

tivelyexpensive.

Control and formal controls. Since the Second World War management
accounting has undoubtedly been important in management control in UK
manufacturing. It offered a means of measuring and reporting adherence to
rules and fulfilment of contractual obligation through techniques of standard
costing, budgeting, variance analysis and performance appraisal. In recent
years, however, there has been increasing disquiet about accounting's role.
This may be summarised as concern over the specific techniques employed,
the focus of their attention and the way increasingly elaborate accounting, if
rigorously applied, can produce ritualistic and/ or antagonistic responses from
managers. The last issue raises questions about whether the proliferation of
formal 'controls' may generate greater difficulty in securing overall 'control'
(Drucker, 1964) and ultimately become self-defeating (see Box 19.3).
Hopwood (1974) noted that some enterprises did not apply accounting in
the rigorous manner of the 'budget-constrained' style. In some cases a looser,
more flexible 'profit-conscious style' was adopted; in others a 'non-accounting
style' was preferred. Assuming a rigorous application of accounting would be
appropriate, Kaplan and others have sought to build new accounting systems
in tune with the new manufacturing environment. An alternative prescription
is that the emphasis in the 1990s should be to decrease reliance on any account-
ing system. The second view has been advanced by Johnson (a one-time close
colleague of Kaplan) who not only attacked the development of activity-based

Box 19.3 Control and formal controls

It is increasingly .,. being realised that the distinction between controls

and control is a crucial one . In many cases, for example, managers have
actually achieved less overall control .. . as a result of using the ever-
increasing number of individual controls. Subordinatets) ... have used
their ingenuity to find ways of satisfying or even beating the controls,
resulting in both intended and unintended consequences. The paradox
often goes further. After recognising the reduction in real control many
managers have quite naturally become even more concerned with
regaining it. In order to do so they have called upon yet further controls
with the result that the desired degree of control has become even more
distant.

Source: (Hopwood, 1974), p. 18.

 T. Colwyn Jones and David Dugdale 317

management but also questioned the value of all accounting information. He

identified 'a new competitive environment - call it the global economy - in
which accounting information is not capable of guiding companies toward
competitiveness and long-term profitability' (Johnson, 1992b, p. 31). Responses
in enterprises to this new environment were seen as constituting a broad trend
from 'top-down control to bottom-up empowerment' (Johnson, 1992a, sub-
title) in which accounting would have less importance.
Whilst we agree that there are limits to the effectiveness of accounting as
formal controls, its importance is not confined to these . We argue that it is also
important in other, less formal aspects of control. To explore this we must
move away from an emphasis on rules and contracts, and become more con-
cerned with accounting as an articulation of values contributing to the con-
struction of trust relations.

Accounting and informal management control

Management accounting is more than a collection of specific measurements

and techniques applied in particular ways. It constitutes a distinctive way of
looking at, and talking about, enterprises. This has implications for manage-
ment which go far beyond the formal control issues dealt with so far.

Manufacturing the worker. To begin with standard costing, Miller and O'Leary
(1987) note its roots in scientific management (Taylorism) which they argue
was itself part of a broader movement of viewing, measuring and controlling
people which was gaining momentum at the beginning of the twentieth
century. Its origins may go back to the development of educational practices
in the eighteenth century, when students became increasingly subject to
regimes of writing, examination and grading (Hoskin, 1993). What these stu-
dents learned was not only particular subjects - arts, sciences and social sci-
ences - but ways of learning which they later reapplied in industry when they
developed systems of measuring, reporting and rewarding workers and man-
agers (Hoskin and Maeve, 1988). Scientific management offered a means by
which workers could not only be compared against each other and customary
work effort, but also judged against a standard of abstract human effort. When
the standard times for work were converted into standard labour costs man-
agers created a new way to 'know' workers - in terms of their detailed contri-
bution to production. Miller and O'Leary argue that, as workers came to be
seen as objects measured against abstract standards, they themselves were
altered in the process so that they became 'governable persons'. They were
not totally docile, obedient employees but they were controllable through
formal accounting systems. Even if standard costing systems were completely
318 Manufacturing accountability

swept away in the 1990s (a highly unlikely prospect) the legacy of seeing
people as aggregates of measurable activities and achievements will probably
endure for many years.
An early indication of change in accounting systems resulting from a break
with the scientific management tradition comes from Malmberg (980). In a
study of one Swedish company he found standard costing systems in place
but their use had changed. The company had developed work groups which
accountants treated as 'profit-centres' and workers were encouraged to
improve on material and labour standards by the award of a symbolic 'profit'.
This was reported to the group supervisors eight times a year and then dis-
cussed with workers. Malmberg, using the fashionable management language
of the time, described this in terms of 'motivation' and claimed it increased
workers' psychological satisfaction. The supervisors he quoted, however, saw
it in terms of workers gaining a better understanding of 'financial realities'
and the outcome was that 'many more workers become conscious of, and
involved in, the ways in which their own work influenced the financial results'
(Malmberg, 1980, p. 81).
Although there has been little research on such accounting change in the
UK, we know of a number of companies which are currently developing
accounting systems based on work groups, usually labelled 'cells'. Sometimes
cells are merely a convenient accounting device for identifying and attributing
costs, but often result from a physical reorganisation of workers and
machines. In either case these cells are seen as receiving inputs from 'internal
suppliers' and delivering output to 'internal customers'. Each cell 'owns' its
costs, is supplied with financial information in order to 'understand' these and
encouraged to work collectively to reduce them (as well as more generally
improving production). In different economic and social conditions and with a
differing managerial emphasis this is now not described in terms of 'motiva-
tion and satisfaction' but as 'responsibility and accountability'.
Here we detect a subtle, but perhaps significant, shift in management
control. At the centre of this is a reconstruction of workers from producers of
products to producers of profit. Under scientific management the amount of
labour and materials required for production was predetermined and man-
agement control was directed at matching 'actual' to 'expected' performance.
It was (in this sense) a fixed contract for a specified level of production. In cell
accounting, standard costs are notional and workers are exhorted to continu-
ally improve levels of production, and control may be characterised as moving
to 'responsible autonomy' (Friedman, 1977). Accountants are seen as advisors
(helping cells to understand how to improve their profitability) and reporters
(recording the financial outcomes of group efforts). This is accounting's
version of 'empowerment' - the allocation of responsibility to workers and
their supervisors to pursue corporate objectives defined in financial terms.
Managers talk of the need for a change of 'mind-set' for this to be successful.
Broadly this involves the construction of self-conscious workgroups, and per-
 T. Colwyn Jones and David Dugdale 319

ceptions of managers as advisors rather than controllers. There are symbolic

aspects to this: one company colour coded its cells as blue team, red team, and
so on; another renamed its supervisors 'coaches' . Another element is changed
practices. When presented with financial information on their performance
and comparisons with other groups, cells can be called upon to explain
current results and suggest improvements. In order to do this they need to
investigate their performances through a financial perspective. In doing so
they may come to reconceptualise their workplace as a world of profit rather
than a world of production.
All of this may be exaggerated of course. In a general sense workers have
always been aware that their employment is aimed at profit. The novelty we
are suggesting is that managers may be deliberately constructing forms of
accounting control in which workers and supervisors are encouraged to per-
ceive their detailed activities in financial terms and are aware of themselves as
financial actors. The empirical evidence for such change in management
control is, at best, sketchy (but see Munro & Hatherly, 1993, for similar obser-
vations). The extent to which workers actually remodel themselves in line
with this new conception of them is unclear. Some research suggests that this
may be the case. Hopper et al. argued that UK coalminers had internalised 'the
logic that profit and loss were absolute measures of performance - the bottom
line' (1986, p. 126) and this had considerable impact on their campaign, and its
public presentation, in the 1984-5 strike. Similarly Knights and Collinson
(1987), in a study of a UK heavy vehicle manufacturer, identified accounting as
a major element in workers' response to the closure of a division. Faced with a
'financial audit' which portrayed the division as loss-making, the workforce
did not resist and accepted redundancy. In both cases the authors argued that
alternative accounts could have been created, showing a different 'reality', but
that workers (and their unions) did not challenge the managerial numbers.
These studies are suggestive of the potential of accounting to influence
workers' views of their workplaces.

Manufacturing the manager. Accounting may be more powerful in shaping

managerial perspectives. Here we may see it as a language of business shared
by managers from many different occupational specialisms. More than this, it
is embedded in managerial practices. The setting up of profit centres and
investment centres is not only a technical means of creating pseudo-market
relations through mini-enterprises within the enterprise. It is simultaneously
an attempt to construct managers as mini-entrepreneurs (or 'intrapreneurs').
In one company a series of three-day finance schools was run for these man-
agers. The course dealt with ABC, TA, target costing, marginal costing, and
many other techniques. However the financial controller responsible for the
course places more emphasis on the first morning where, in a series of case
study exercises, managers are required 'to think like shareholders' . Only once
320 Manufacturing accountability

they have adopted the appropriate orientation to finance (the perspective of

owners) would financial knowledge be relevant. In his view, accounting tech-
niques in themselves are of little importance; it is their purpose which crucial.
Such rather formal, one-off, approaches to constructing managers as
financial actors are unlikely to be effective unless they are embedded in actual
practices. In a detailed study of another company (Dugdale and Jones, 1993,
1994) we explored the way one such practice - investment decision making-
incorporated accounting in the social construction of managers. Here there
were no specific requirements for either the measurements or the techniques to
be used for investment appraisal. Instead engineering staff proposing invest-
ment were expected to decide for themselves what financial information was
appropriate. Nor was it believed that accounting numbers provided the 'right'
answer to questions of costs and benefits of investment. Instead accounting
information was used as the basis of probing proposals through the interroga-
tion of proposers. The process was described as 'confidence building', gaining
'credence and credibility' and making sure people have 'thought about it' .
Finance was not the only aspect of this interrogation. There would be ques-
tions on product and production technology, and markets. There would also
be a general concern with corporate strategy. However accounting was a nec-
essary element of the interrogation. The aim, as one manager put it, is that as
proposer you must 'convince me that I should trust you'.
What is the basis of this trust? In our study it was concerned with shared
perception and values within management. Proposers should see things from
a business point of view. Managers are suspicious that engineers propose
investments because they 'love to engineer things', 'build empires', 'do an
interesting project', and so propose investment in 'singing, dancing, wonder-
ful' machines. These are not 'business' reasons and so are not acceptable. What
they must do is propose 'cost-effective, reliable solutions' . In this, the appro-
priate use of financial information is crucial: not because accounting produces
right answers, but because it shows the proposer is thinking in appropriate
ways. Here then we see accounting, not as an assembly of calculative tech-
niques for formal control, but as a mode of thought, reflecting and expressing
managerial perspectives and values, implicated in the construction of trust
relations. Where these are established there need be less reliance on formal
controls. But this does not imply their abandonment. Even trusted managers
are encircled by financial controls. Indeed it is through the practices associated
with these controls that trust is constructed.

Accounting and accountability

In this chapter we have discussed accounting's contribution to management

control in terms of formal and informal practices. Accounting has been por-
 T. Colwyn Jones and David Dugdale 321

trayed both as a collection of measurements and techniques, and as a mode of

thought. We have related this portrayal to three mechanisms of control- rules,
contracts and values - which are embedded in three patterns of social interac-
tion - hierarchical, market and trust relationships. Such neat categories, whilst
useful for the exposition of thoughts, carry the danger of becoming rigid
frameworks for understanding enterprises. We do not intend them in this
way. First, we are not suggesting that there are three 'types' of organisation.
Instead we argue that enterprises can be analysed in terms of different
patterns of social relations which co-exist and interact with each other. Second,
we do not suggest that there are 'better' (or even 'best') forms of control, either
from a moral viewpoint or in terms of the maximisation of some abstract
achievement of efficiency. Rather we suggest that different patterns of control
and surveillance and co-ordination and unity are the outcomes of particular
social actions in specific social contexts. Third, we do not see these as univer-
salistic rational responses to stimuli from objective external realities . While not
wishing to dismiss notions of rationality and reality (see Jones, 1992) we recog-
nise the importance of differing meanings in social action and how social con-
texts are differently interpreted in these meanings. Here ways of seeing and
talking (such as accounting) both reflect and shape actions and contexts.
Fourth, we do not regard the patterns we have described as clear-cut, self-
contained or fixed . On the contrary, the contradictions of management control
imply that social relations are ambiguous, interactive and unstable. Overall
we view categories as a useful means of analysing enterprises, but not to be
confused with representations of enterprises themselves.

Accounting and the new manufacturing environment. To the extent that enter-
prises have been experiencing a period of transformation, we would expect
this to be represented in changing regimes of accountability. We would not
wish to draw direct causal connections between specific changes in the manu-
facturing environment and particular developments in accounting control.
Instead we see the last few years as a period in which managers have become
more intensely and urgently concerned with both the need and opportunity
for change. This general concern has, in its various specific forms, been the
product of their awareness of new technologies and production techniques,
different forms of job design and work organisation, and changing managerial
structures. These may be linked to wider identification of changing, and
increasingly competitive, market conditions. Together they have produced a
widespread feeling that existing organisational patterns are inappropriate to
'the new manufacturing environment'.
We would characterise this as a period of organisational 'unfreezing' - a time
when existing patterns are no longer seen as capable of improvement by incre-
mental change but must be transformed. Although traditional formal controls
may remain in place they are seen as less relevant to current needs. However
322 Manufacturing accountability

the outcome of this unfreezing cannot be predicted. On the one hand, the lack
of certainty may hinder the construction of rules and contracts since managers
areunable to specify exactly what is tobe achieved and how this is to be done.
Thus they may be forced to rely more heavily on trust relations, relying on
people to do the 'right thing' - as suggested by the rise of interest in 'empower-
ment' and the need to develop 'a common vision'. Current changes may lead to
less reliance on formal accounting control in the long term (Johnson, 1992a). On
the other hand, uncertainty also offers a market-place for new forms of formal
control claiming to be relevant to the new conditions, Hence the proliferation
of new accounting systems (based on ABC, TOC, TA and many more). Thus
management control may, after a short transition, refreeze around new
accounting systems embodying formal controls.
However this is unlikely to be a stark alternative between increased reliance
on informal control or the adoption of new formal controls. In such a period of
change and uncertainty many enterprises appear to be considering, and exper-
imenting with, several forms of change simultaneously. Thus companies claim
to be 'using' many new accounting techniques (which their supporters con-
sider incompatible) all at the same time. This is not so that managers can dis-
cover which one provides the 'right answer' to accounting issues, but because
they offer a range of 'tools' to enable managers and supervisors to 'under-
stand' their activities and achievements in financial terms - as part of an
attempt to 'empower' them as financial actors pursuing an (as yet) ill-defined
'common vision'. If new patterns of formal and informal control are emerging
from the unfreezing of manufacturing enterprises we anticipate that these will
also prove contradictory and thus create new tensions, though in ways which
cannot be predicted today.

Accountability. At the centre of our discussion is the question of accountabil-

ity (and its relation to accounting) as management control. Our analysis has
suggested a number of ways in which accounting may operate as manage-
ment control. Here the concept of 'accounting' has different meanings. At one
level accounting is about counting- the conversion of activities and outcomes
into numbers. At a deeper level these activities and outcomes are accounted for
- they are reported upon (usually to superiors). At the same time the reporters
may be required to make an account of their activities. This has the dual
meaning of providing a narrative (how things happened) and an explanation
(why they happened) . Deeper again is the notion of being held accountable
where the person who is counted, accounted for and makes the account is
deemed responsible not just for the accounting but for the events and circum-
stances which have led to that accounting. Taken together these different
facets of accounting may be seen as a regime of accountability. This hints at
the deepest level of all- the construction of the accountable person. This would
be someone who is so deeply enmeshed in practices of counting, accounting
 T. Colwyn Jones and David Dugdale 323

for, making accounts and being accountable that their view of the world, and
of their own self, has become deeply permeated with accountability as a mode
of thought - someone who, in a very precise sense, is self-accountable: who
sees the world, interprets it, acts in it and evaluates his or her action in ways
which mirror their external regimes of accountability.
But - and this is a very big 'but' - this is merely to explore the concept of
accountability. It is most certainly not a description, let alone an explanation,
of what is happening in manufacturing enterprises. There is no simplistic pro-
gression from counting to the accountable person, from hard accountability to
soft accountability or even from formal controls to informal control.
Nevertheless the delineation of the term 'accountability' may provide insights
which prompt a different view of accounting from that provided by its equally
absurd, but more usually discussed, counterpart. That absurdity is, of course,
the view that accounting is an objective, neutral, factual statement of reality,
entirely untouched by, and not touching, those persons who count and those
who are counted.
To move away from absurdity, views of accounting as a technical practice
and as a discursive discipline (when not extremist and doctrinaire positions)
are both relevant to our understanding of management control. Accounting
has more usually been recognised for its contribution to formal control
through its array of techniques. We have pointed to some of its more
neglected aspects in terms of constructing a vocabulary of action or language
of business which carries a distinctive view of people as financial actors in
financial contexts. Burawoy (1979) argued that when workers play the game of
'making out' in production they come to accept the rules of that game and
that this implies that consent is manufactured-it is a product of the process.
We would not be so bold. In 'playing the game' of formal accounting control,
managers and workers construct practices which influence both the game and
the players. This may be termed 'manufacturing accountability' . But the
players are never fully constructed by the game, the game changes, and
accountability is always an elusive and fragile phenomenon. It is never made,
but always in the making - though no less real for that.
 CHAPTER 20

Endpiece
Anthony J. Berry, Jane Broadbent and David Otley

In this chapter we will provide a brief overview of the three sections and
explore some of the questions and issues that have arisen. In the language of
Fayol, control appears as one of the universal activities of organisations and of
managing. It had, for him, a central place in the list of planning, leading,
organising, controlling and motivating. In this sense management control
exists as an aspect of all domains of practical managing: marketing, produc-
tion operations, personnel, purchasing, selling and so on. For we can discuss
the control of the production process, control of personnel practices, control of
the budgeting process and so on. Figure 20.1 illustrates this richness. The
plane created by the axis of activities and the axis of domains is the plane of
practice. From an inspection of this figure you can see how it is that Fayol and
others could conceive of universal activities and of course in a straightforward
practical sense they are right to do so.
In Figure 20.2 we add the third axis, that of academic disciplines. In so doing
we create a plane which represents the multiple disciplines from which stu-
dents may seek to understand the activities and their interactions; for example,
we may examine planning through mathematical modelling. On closer inspec-
tion this plane might be found very abstract. We also create a plane which rep-
resents the likewise multiple disciplines from which students may understand
the domains and their interrelationships; for example, we may explore market-
ing from the discipline of linguistics. The combination of the three axes of
activities, domains and disciplines adds up to a three-dimensional space which
shows, for example, that we can study control of production using social psy-
chology or control of personnel using psychoanalytic theory.
So it is clear that management control as a plane of practice may be studied
from the standpoint of any number of disciplines. To illustrate this point we
note how economic theory led Williamson to explore control from the stand-
point of markets as a counterpoint to hierarchies. The discourse of accounting

324
 Anthony J. Berry, Jane Broadbent and David Otley 325

FIGUR E 20.1 Activities and Domains: A plane of action

FIGURE 20.2 Activities, Domains and Disciplines

Disciplines

Psychology

Domains

Here, for example, we can see that the activity of

Leading in the Domain of finance may be studied
using the discipline of Psychology.

has provided concepts such as cost absorption in activities to enable the flow
of resources to be traced through to products and shown how the budget
model may be used as a basis for understanding the relationship of inputs to
outputs of a system. In addition anthropology, the study of human societies,
has provided the concepts of culture from which Ouchi extended
Williamson's framework and aided Langfield-Smith in her chapter on culture
and control. And of course sociology and psychology have underpinned much
326 Endpiece

of the behavioural analysis of control processes. We would also note that these
disciplines are not free from the sometimes vigorous epistemological debates
taking place among scholars. While this is not primarily a research text, it is
important to note that the research literature of these disciplines has informed
our understanding of management control and shaped some of the contribu-
tions in this volume. For a wider discussion the reader is referred to Chua,
Lowe and Puxty, Critical Perspectives in Management Control.
Parts I & II of this book were designed to explore control in a variety of
ways which illustrate some of these different possibilities. In Part III, we pre-
sented examples of control in the context of various kinds of organisations. In
this review of the volume it is not our intention to provide an integrating
framework, but we do suggest that this simple framework of managerial activ-
ities, managerial domains and disciplines provides a very rich map of the
field. It also provides a rationale for the inclusion of the types of issues
covered which may seem, at first sight, to be somewhat diverse.
In Part I we reviewed how the subject of management control has been
approached by different authors and theorists. In Chapter 1 we were con-
cerned to establish our task as being to do with control in purposive organisa-
tions. From that standpoint we explored management control through the
idea of domains in which control is encountered. This was followed by a
discussion of the way control has been described and understood in some of
the literature of organisation theory. Structures and procedures for control
were examined prior to a discussion of the context for control.
We do not claim to have provided an overarching theory but we have
sought to provide a reasonably coherent picture of some basic approaches to
management control. The aim was to broaden the view of management
control from that in the standard texts. The material is complex, mainly
because the practical world of organisations is complex and we cannot give a
normative answer to the question as to how we should control. Further we
have endeavoured to indicate the limits of rational approaches as the prob-
lems of uncertainty, ambiguity and multiple and differentiated values are
addressed. These are the reasons for both the multiple theoretical approaches
put forward and for our unwillingness to offer clear prescriptive stances. We
have sought, in Part I, to present our views in a manner that is complementary
to those put forward in standard accounting and control texts. This text can be
seen as an attempt to broaden the theoretical perspectives traditionally used to
analyse the design and operation of management control systems.
In Part II, where the focus is on the examination of management control
systems, we were concerned to present the variety of ways in which control
has been understood in relation to particular issues. We have examined man-
agerial issues of accounting, divisional management, strategy, operations
management and performance measurement. While the varied styles and
stances of the authors demonstrate some differences of approach, there is in
these chapters an exploration in depth of some of the familiar problems of
 Anthony J. Berry, Jane Broadbent and David Otley 327

organisations. These managerial problems were complemented by discus-

sions of control from the standpoint of the discipline of economics and the
influence of culture. The latter was illustrated by studies of management
accounting practices in Japanese companies in Japan and in the UK as a
vehicle for exploring the differences in enacted culture and control between
East and West.
In accordance with this approach, Part III contains some case studies of
control in contemporary organisations, not as examples of perfect practice but to
provide a practical complement to the more theoretical material of Part I and II.
The examples were drawn from private and public, manufacturing and service
organisations to provide some rich material for teaching and exploration.

Some final comments to students and managers

To students of management we would like to say that, in this book, we have

been concerned to provide a counterpoint to the prescriptive and technical
material found in the standard texts of management control which tend to
equate control with management accounting. We do not, however, seek to
dismiss these technical approaches for it is clear from the case material and
from some of the chapters that of the ubiquity of accounting indicates its
central importance. It is so because of its great virtue in giving a transorganisa-
tionaI language for describing some aspects of organisational activity.
Moreover accounting provides a description in terms of some common
notions of value expressed in monetary terms. It is a vehicle for control which
can be argued to increase in importance as its logic is applied in more and
more situations. What is clear from the considerations in this volume is that
the social and cultural limits to the accounting discourse mean that great care
must be taken in transferring the meanings attributed to accounting concepts
when they are applied to organisations which are not predicated upon liberal
market economics. In brief the notions of value which might exist in a wide
range of organisations are not readily comparable and to assume they are
because it is possible to record some interactions or transactions in accounting
terms is a serious error. In any case accounting reduces the complexity in
activity systems into a unidimensional depiction and, as such, oversimplifies
many details essential to effective management control.
This notion of accounting as quantification of activity presupposes that all
organisational matters of interest can be quantified and that that quantification
is adequately meaningful. For example, it is possible to produce quantified
management accounts for the process of control in schools, churches, profes-
sions and hospitals. However to assume that the activities and outputs of such
organisations and their social significance are measured by such statements is
to be misled as to the limitations of management accounting. Hence to base
328 Endpiece

decisions about such organisations upon such management accounts would

be a grave mistake.
The variety of the approaches to understanding control presented in this
volume provides the possibility of reflecting upon them in the context of dif-
ferent organisational settings. Should you, the reader, generally see organisa-
tional analysis as a task of understanding the technical systems of acquisition
of resources, production and distribution then it is likely that you will be out
of sympathy with much of the material in this book, even though you might
be able to make some progress with your problems. Should you see organisa-
tions as primarily social structures, you will also, surprising as it may seem, be
out of sympathy with much of the material herein. For what we are setting out
to encourage you to do is to make connections across the technical and social
processes and to explore the significance of the issues raised, in order that you
may develop a rich rather than an impoverished concept of managerial control
in your organisation. So it is important that you explore the cases in Part III to
broaden understanding through interpretation rather than analysis. They are
not meant to be seen as essays in the simple application of ideas.
If you are currently managing in an organisation, we hope that you can
build upon your own experience with the aid of the material in this book.
Again we make no apologies for the discursive rather than prescriptive
approach of the editors and authors. The literature of managing democratic
organisations, empowered organisations, creative organisations and efficient
and resourceful organisations has led scholars away from rather technical pre-
scriptions. We set out with the premise that the underlying approach which
essentially rejects simple technical prescriptions is necessary to effective man-
agement. The puzzle, as we see it, is that people in organisations which do
have rather simple technical accounting systems for control have to locate
them in the context of the markets and internal social structures and processes
of their organisation, whether they recognise this or not. Hence we are encour-
aging the understanding of the significance and utility of processual
approaches to management control.
This might seem to be a simple point but in practice it seems that managers
do tend to a reductionist approach to information as the common use of
targets and budgets as benchmarks for plans and action demonstrate (see, for
example, Chapter 11). It seems that the perceived financial goals, because they
are readily quantifiable, even if there is much slippage between what is known
and what is represented, are commonly used as the ends which dictate present
behaviour. It is unusual to find commercial organisations in the West which
use process as the focus of management control, with goals as outcomes of the
uncertainty.
Further there is in the privatisation movement a temptation to apply the
technical process of management accounting to organisations which can in no
sense be comprehended by such a curiously limited calculus. Yet the assump-
tions of the privateers is that such a calculus is not only universal but that it is
 Anthony J. Berry, Jane Broadbent and David Otley 329

somehow morally and politicoeconomically more effective and efficient than

the more subtle, insightful processes which it replaces. Indeed there are
always political processes in organisations and between organisations and
their environments. Choosing not to recognise them cannot mean that they
will be removed. The ideas in this volume provide at least a beginning for
understanding from a widely drawn framework how management control
may be developed in very varied organisations.
Of course you will have noticed that we have avoided much of the issue of
power in organisations. Mostly we are concerned in this book with the issue of
authority and its legitimacy, its explication in varied social and cultural con-
texts. Power presents some particular problems for the academic theorists and
text writers. Central to these is the problem that control reduces to doing what
one is ordered to do. There is no problem, other than the incompetence of the
power holders. Of course such closed systems of power regress to warfare as a
mode of control.
Conflict is, however, part of our concern. So a further problem with techni-
cal management accounting as a basis for management control is the assump-
tion that all conflict is dealt with prior to the recording and measuring of
transactions. This is an unlikely scenario. There is plenty of evidence and
experience to show that the processes of management control carry within
organisations some of the necessary conflict suppression and quasi-resolution
or actual resolution of conflict.
In the material as we have presented it there is of course the problem of
how anyone individual relates to the ideas . Almost all accounting procedures
have within them the risk of objectifying the experiencing self, as though the
truth of human experience can be set aside by some such calculation. There is
the risk then of reducing the very human concerns for our work, for each
other, for our organisations and for the society in which we live to abstractions
and then pretending that such abstractions are the stuff of intelligent manage-
ment. This is to mistake the models for the substance, a common enough exer-
cise when one group is seeking dominance over another.
Bibliography

Ackoff, R. (1974) Redesigning the Future (New York: John Wiley).

Ahmed, M.N. and Scapens, P.W. 'Cost Allocation Theory and Practice: The Continuing
Debate', in Ashton, D., Happer, T. and Scapens, R.W. (eds), Issues in Management
Accounting (Hemel Hempstead: Prentice-Hall pp. 39-60 .)
Alchian, A. and H. Demsetz (1972) 'Production, Information Costs and Economic
Organisation', American Economic Review, Vol. 62, pp. 777-95.
Aldrich, H.E. (1979) Organisations and Environments (London: Prentice-Hall).
Allaire, Y. and M.E. Firsirotu (1984) 'Theories of Organizational Culture', Organization
Studies, Vol. 5, No.3, pp. 193-226.
Allen, M. and D. Myddelton (1987) Essential Management Accounting (Hemel
Hempstead: Prentice-Hall).
Alvesson, M. and L. Lindkvist (1993) 'Transaction Costs, Clans and Corporate Culture',
Journal of Management Studies, Vol. 30, No.3, pp. 427-52 .
Arney, L.R. (1969) 'Divisional Performance Measurement and interest on Capital',
Journal of Business Finance, Vol. 1, Spring pp. 1-7.
Arney, L.R. (1979) Budget Planningand Control Systems (London: Pitman).
Anderson, C. and F. Paine (1975) 'Managerial Perceptions and Strategic Behaviour',
Academy of Management Journal, Vol. 18, pp. 811-23.
Andrews, K. (1971) The Concept of Corporate Strategy (Homewood, Ill.: Dow
Jones/Irwin).
Ansoff, H.I. (1965) Corporate Strategy (New York: McGraw-Hill).
Ansoff, H.I. (1979) Strategic Management (London: Macmillan).
Anthony, R.N. (1965) Planning and Control Systems: A Framework for Analysis
(Cambridge, Mass.: Harvard University Press).
Anthony, R.N. and D.W. Young (1984) Management control in nonprofit organizations, 3rd
edn (Homewood, Ill.: Richard D. Irwin).
Anthony, R.N., J. Dearden and N.M. Bedford (1984) Management Control Systems, 5th
Edn (Homewood, Ill.: Richard D. Irwin) .
Anthony, R.N., J. Dearden and N. Bedford (1989) Management Control Systems, 6th edn
(Homewood, Ill.: Richard D. Irwin) .
Anthony, R.N., J. Dearden and. Govinjarajan (1992) Management Control Systems, 7th
edn (Homewood, Ill.: Richard D. Irwin) .
Aoki, M. (1984) The Co-operative Game Theory of the Firm (Oxford: Clarendon Press) .

330
 Bibliography 331

Aoki, M. (1986) 'Horizontal vs Vertical Information Structure of the Firm', American

Economic Review, December, Vol. 76, No .5, pp. 971-83.
Aoki, M. (1990a) 'The Participatory Generation of Information Rents and the Theory of
the Firm', in M. Aoki, B. Gustaffson and O.E. Williamson (eds), TheFirm asa Nexus of
Treaties (London: Sage) .
Aoki, M. (1990b) 'Towards a Theory of the Japanese Firm', Journal of Economic
Literature, Vol. xxvii (March) pp. 1-27.
Argyris, C. (1952) The Impact of Budgets on People (Ithaca, New York: The Controllership
Foundation).
Argyris, C. (1964) Integrating the Individual and the Organisation (New York: John Wiley).
Argyris, C. (1985) Strategy,Change and Defensive Routines (Cambridge, Mass.: Ballinger).
Argyris, C. (1990) Overcoming Organisational Defences: Facilitating Organisational
Learning (Boston : Allyn & Bacon/Prentice-Hall).
Argyris, C. and D. Schon (1978) Organisational Learning: A Theory of Action Perspective
(Reading, Mass.: Addison-Wesley).
Armstrong, P. (1985) 'Changing Managerial Control Strategies: the Role of Competition
between Accounting and other Organisational Professions', Accounting, Organizations
and Society, Vol. 10,No.2, pp. 129-48.
Armstrong, P. (1987) 'The Rise of Accounting Controls in British Capitalist Enterprises',
Accounting, Organizations and Society, Vol. 12, No.5.
Armstrong, P. (1989) 'Variance Reporting and the Delegation of Blame: A Case Study',
Accounting, Auditing and Accountability Journal, Vol. 2, No.2, pp. 29-46.
Armstrong, P. (1991) 'Contradiction and Social Dynamics in the Capitalist Agency
Relationships', Accounting, Organizations and Society, Vol. 16, No .1, pp. 1-25.
Armstrong, P. (1993) 'Professional Knowledge and Social Mobility: Postwar Changes in
the Knowledge-base of Management Accounting', Work, Employment and Society,
Vol. 7, No.1, pp. 1-21.
Arnold, J. and T. Hope (1990) Accounting for Management Decisions, 2nd edn, (Hernel
Hempstead: Prentice-Hall).
Ashby, E. (1956) An Introduction to Cybernetics (London: Chapman and Hall).
Ashley Smith, D. and P. Smith (1987) 'Local authority annual reports', Public Money,
Vol. 7, No .3, pp. 52-6.
Ashton, D., T. Hopper and R. Seapens (eds) (1991) Issues in Management Accounting
(Hemel Hempstead: Prentice-Hall).
Atkinson, J. (1984) 'Manpower Strategies for Flexible Organisations', Personnel
Management, August, pp. 28-31 .
Audit Commission (1992) Citizen's Charter indicators: charting a course (London: HMSO).
Baiman, S. (1982) 'Agency Research in Management Accounting: a survey', Journal of
Accounting Literature, Spring, pp. 154-213.
Baiman, S. (1990) 'Agency research in managerial accounting: a second look',
Accounting, Organizations and Society, Vol. 15, No .4, pp. 341-371.
Baldamus, W. (1961) Efficiency and Effort: An Analysis of Industrial Administration
(London: Tavistock).
Barnard, C. (1938) The Functionsof the Executive (Cambridge, Mass.: Harvard University
Press) .
Bator, F.M. (1957) 'General Equilibrium, Welfare and Allocation', American Economic
Review (March) pp. 22-59.
Beer, S. (1966) Decision and Control (New York: John Wiley).
332 Bibliography

Beer,S. (1979) TheHeart of the Enterprise (New York: John Wiley).

Beer,S. (1972) Brain of the Firm (Harmondsworth: Allen Lane).
Benbassat, I. and A.S. Dexter (1979) 'Value and Events Approaches to Accounting: An
Experimental Evaluation', The AccountingReview, Vol. LIV, No.4, pp. 735-49.
BenefitsAgency (1993) Business plan, 1993/1994 (London : Department of Social Security).
Berry, AJ. (1983) 'Management Control and Methodology', in E.A Lowe, and J.L.
Machin (eds), New Perspectives in Management Control (London: Macmillan).
Berry, A.J. and D.T. Otley (1986) 'The Aggregation of Estimates in Hierarchical
Organisations', Journal of Management Studies, Vol. 12, pp. 175-93.
Beyer, J.M. (1981) 'Ideologies, Values and Decision Making in Organisations', in P.C
Nystrom and W.H. Starbuck (eds), Handbook of Organisational Design, Vol. 2, (Oxford :
Oxford University Press).
Beynon, H., (1973) Working for Ford (London: Penguin), reprint 1975 (Wakefield: EP
Publishing).
Birnberg, J.G. and C Snodgrass (1988) 'Culture and Control: A Field Study',
Accounting, Organizations and Society, Vol. 13, No.5, pp. 447-64.
Blau, P.M. (1955) The Dynamics of Bureaucracy (Chicago: University of Chicago Press).
Boulding, KE. (1956) 'General Systems Theory - the skeleton of a science', Management
Science, Vol. 2, pp. 197-208.
Bourn, M. and M. Ezzamel (1986) 'Organisational Culture in Hospitals in the NHS',
Financial Accountability and Management, Vol. 2, No.3, pp. 203-25.
Bower,J.L. (1986) Managing the Resource Allocation Process: A Study of Corporate Planning
and Investment (Boston: Harvard Business School Press) .
Boyatzis, R (1982) The Competent Manager: A Model for Effective Managers (New York:
John Wiley).
Braverman, H. (1974) Labor and Monopoly Capital (New York: Monthly Review Press).
Braybrooke, D. and CE. Lindblom (1963) A Strategyfor Decision (London: Macmillan).
Broadbent, J. 'The Values of Accounting and Education: Some Implications of the
Creation of Visibilities and Invisibilities in Schools', Sheffield University
Management School Discussion Paper, 92/41; forthcoming in Advances in Public
Interest Accounting, Vol. 6, No.4.
Broadbent, J. and J. Guthrie (1992) 'Changes in the Public Sector: A Review of Some
Recent"Alternative" accounting Research', Accounting, Auditing and Accountability
Journal, Vol. 5, No.2. pp. 3-31.
Broadbent, J., RC Laughlin, D. Shearn and N. Dandy (1992) 'It's a long way from
teaching Susan to read', in G. Wallace, (ed.) Local Management of Schools: Research and
Experience, Bera Dialogues, (Clevedon: Multilingual Matters).
Broadbent, J., RC Laughlin, D. Shearn and N. Dandy (1993) 'Implementing Local
Management of Schools: A theoretical and empirical analysis', Research Papers in
Education, Vol. 8, No.2, pp . 149-76.
Bromwich, M. (1990) 'The Case for Strategic Management Accounting: The Role of
Accounting Information for Strategy in Competitive Markets', Accounting,
Organizations and Society, Vol. 15, No.1 /2, pp.27-46.
Bromwich, M. and A Bhimini (1989) Management Accounting: Evolution not Revolution
(London: CIMA).
Bromwich, M. and S. Inoue (1992) 'Management Practices and Cost Management
Problems in Japanese-Affiliated Companies in The United Kingdom', Working
Paper, London School of Economics.
 Bibliography 333

Brownell, P. (1981) 'Participation in Bugdgeting, Locus of Control and Organisational

Effectiveness', Accounting Review, Vol. 56, pp. 844-860.
Bruns, W.J. and R.S. Kaplan (eds) (1987) Accounting and Management - Field Study
Perspectives (Boston: Harvard Business School Press) .
Bruns, W.J. and J.H. Waterhouse (1975) 'Budgetary Control and Organizational
Structure', Journal of Accounting Research, Vol. 13, pp. 177-203.
Brunsson, N. (1985) The Irrational Organisation (Chichester: John Wiley).
Buckley, W. (1968) Modern Systems Research for the Behavioural Scientist (Chicago:
Aldine),
Burawoy, M. (1979) Manufacturing Consent: Changes in the Labor Process Under Monoploy
Capitalism (Chicago: Chicago University Press).
Burns, T. and G. Stalker (1961) The Managementof Innovation (London: Tavistock) .
Burrell, G. and G. Morgan (1979) Sociological Paradigms and Organisational Analysis
(London: Heinemann).
Capps, T., T. Hopper, J. Mouritsen, D. Cooper and T. Lowe (1989) 'Accounting in the
Production and Reproduction of Culture', in W.F. Chua, T. Lowe, and AG. Puxty,
(eds), Critical Perspectives on Management Control (London: Macmillan).
Carter, N. and P. Greer (1993) 'Evaluating agencies: Next steps and performance
indicators', Public Administration, Vol. 71, pp. 407-16.
Chandler, AD. (1962) Strategy and Structure: Chapters in the History of the American
Industrial Enterprise (Cambridge, Mass .: MIT Press).
Chandler, AD. (1977) The Visible Hand: The Managerial Revolution in American Business
(Cambridge, Mass. : Harvard University Press).
Chandler, R. and P. Cook (1986) 'Compliance with disclosure standards in published
reports and accounts of local authorities', Financial Accountability and Management,
Vol. 2, pp. 75-8 .
Chartered Institute of Public Finance and Accountancy (1982) Local government compar-
ative statistics 1982 (London: CIPFA).
Checkland, P. (1981) Systems Thinking, Systems Practice (Chichester: John Wiley).
Choi, F. and K. Hiramatsu (1987) Accounting and Financial Reporting in Japan
(Wokingham: Van Nostrand Reinhold).
Chow, C.W., M.D . Shields and Y.K. Chan (1991) 'The Effects of Management
Controls and National Culture on Manufacturing Performance: An Experi-
mental Investigation', Accounting, Organizations and Society, Vol. 16, No .3,
pp.209-26.
Christenson, C. (1983) The Methodology of Positive Accounting', Accounting Review,
58 (January) pp. 1-22.
Chua, W.F., T. Lowe and T. Puxty (1989) Critical Perspectives in Management Control
(London: Macmillan).
Clark, B.R. (1972) 'The Organisational Saga in Higher Education', Administrative Science
Quarterly, Vol. 17, pp. 178-183
Clark, K., R. Hayes and C. Lorenz (eds) (1985). The Uneasy Alliance (Boston: Harvard
Business School) .
Coase, R. (1937) 'The Nature of the Firm', Economica, Vol. 4, pp. 386-405.
Cohen, M., J. March and J. Olsen (1972) 'A garbage can model of Organisational
Choice', Administrative Science Quarterly, Vol. 17, pp. 1-25.
Committee of Vice Chancellors and Principals and University Grants Committee (1987)
University management statistics and performance indicators (London: CVCP).
334 Bibliography

Cooper, D. (1981) 'A Social and Organisational View of Management Accounting'. In

M. Bromwich and A Hopwood (eds), Essays In British Accounting Research (London:
Pitman).
Cooper, D. (1983) 'Tidiness, Muddle and Things: Commonalties and Divergencies in
Two Approaches to Management', Accounting Organizations and Society, pp. 269-86.
Cooper, R (1987) 'The Two Stage Procedure in Cost Accounting, Part 1', Journal of Cost
Management , Summer, pp. 43-51.
Cooper, Rand R.S. Kaplan (1988) 'Measure Costs Right: Make the Right Decisions',
Harvard Business Review, September-october.
Covalesk, M.A, Dirsmith, M.W. and White, C.E. (1987) 'Economic Consequences: The
Relationship between Financial Reporting and Strategic Planning, Management and
Operating Control Decision', Contemporary Accounting Research, Vol. 3, No .2,
pp.408-29.
Cyert, R.M. and ].G. March (1963) A Behavioural Theory of the Firm (Hemel Hempstead:
Prentice-Hall).
Cyert, RM., J.G. March and W.H. Starbuck (1961) 'Two Experiments on Risk and
Conflict in Organizational Estimation', Management Science, Vol. 7, pp. 254-64 .
Dalton , M. (1959) Men Who Manage (New York: John Wiley) .
Daniel, S.J. and W.O. Reitsperger (1991) 'Linking Quality Strategy with Management
Control Systems: Empirical Evidence From Japanese Industry', Accounting,
Organizations and Society, Vol. 16, No.7, pp. 601-18.
Deal, T.E. and AA Kennedy (1982) Corporate Cultures: The Rites and Rituals of Corporate
Life (Reading, Mass.: Addison-Wesley).
Dearden, J. (1962) 'Limits on Decentralized Profit Responsibility', Harvard Business
Review, July / August, pp. 89-90.
Dearden, J. (1968) 'Time span in management control', reprinted in Readings in Cost
Accounting, Budgeting and Control, 4th edn Thomas, W. E. Cincinati: South Western
Publishing, pp. 340-54.
Demirag, I.S. (1986) 'The treatment of exchange rates in internal performance evalu-
ation', Accounting and Business Research, Spring, Vol. 16, pp. 157-64.
Demirag, IS. (1988) 'Assessing foreign subsidiary performance: The currency choice of
UK MNCs', Journal ofInternational Business Studies, Vol. 19, No.2, Summer, pp. 257-275.
Demirag, I.S. and A Tylecote (1992a) 'The Effects of Organisational Culture, Structure
and Market Expectations on Technological Innovation: A Hypothesis', British Journal
Of Management, Vol. 3, No .1, pp . 7-20.
Demirag, IS. and A Tylecote (1992b) 'Short-termism: culture and structures as factors
in technological innovation', in R Coombs, V. Walsh and P. Saviotti (eds),
Technological Change and Company Strategies (New York: Academic Press).
Dent, j.F, (1987) 'Tension in the Design of Formal Control Systems: A Field Study in a
Computer Company', in W.J. Bruns and RS. Kaplan (eds), Accountingand Management
- FieldStudy Perspectives (Boston: Harvard Business School Press), pp. 119-45.
Dent, J.F. (1990) 'Process model of relationship between business strategy and man-
agement control systems: Strategy, Organisation and Control: Some possibilities for
accounting research', Accounting, Organizations and Society, Vol. 15, No. 1/2, pp. 3-25 .
Dent, J.F. (1991) 'Accou nting and Organisational Cultures: A Field Study of the
Emergence of a New Organisational Reality', Accounting, Organizations and Society,
Vol. 16, No .8, pp. 705-32.
 Bibliography 335

Department of Education and Science (1991) The parent's charter (London: DES).
Department of the Environment (1981) Local authorityannual reports (London : HMSO).
Department of Health (1991) The patient's charter (London: HMSO).
Department of Health and Social Security (1983) Performance indicators: national
summary for 1981 (London: DHSS).
Department of Health and Social Security (1985) Performance indicators for the National
Health Service:guidancefor users (London: DHSS).
Department of Health and Social Security (1987) The performance indicator analyst,
version 1.0 (London: DHSS).
Department of Health and Social Security (1988) Comparing health authorities (London:
DHSS).
Dermer, J. (1977) Management Planning and Control Systems (Homewood, Ill.: Richard
D. Irwin).
Dermer, J. (1988) 'Control and Organisational Order', Accounting Organizations and
Society, Vol. 13, No.1, pp. 25-36 .
Dermer, J. (1990) 'The Strategic Agenda: Accounting for Issues and Support',
Accounting, Organizations and Society, Vol. 15, No . 1/2, pp . 67-76.
Dermer, J.D. and R.G. Lucas (1986) 'The Illusion of Managerial Control', Accounting,
Organizations and Society, Vol. 11, No.6, pp . 471-82.
Dess, G.G. and A Miller (1993) Strategic Management (New York: McGraw-Hill).
Dow, G. (1987) 'The Function of Authority in Transaction Cost Economics', Journal of
Economic Behavior and Organization, Vol. 8, pp. 13-38.
Drucker, P. (1964) 'Control, Controls and Management', In CP. Bonini, R.K. [aedicke
and H .M. Wagner (eds) Management Control: New Directions in Basic Research (New
York: John Wiley) pp. 286-96 .
Drury, C (1985) Management and Cost Accounting (London:.Van Nostrand Reinhold).
Drury, J. (1988). Management and Cost Accounting, 2nd edn (London: VNR
International).
Drury, C (1992) Management and Cost Accounting, 3rd Edn (London: Chapman and Hall).
Dugdale, D. and T.C Jones (1993) 'Investment Decisions and the Social Construction of
Trust', Interdisciplinary Approaches to Accounting Workshop, January Manchester
Conference Centre.
Dugdale, D. and T.C Jones (1994) 'Finance, Strategy and Trust in Investment
Appraisal', Management Accounting, April, pp. 52-4, 56.
Emmanuel, CR. and D.T. Otley (1985) Accounting for Management Control (London:
Chapman and Hall) .
Emmanuel, CR., D.T. Otley and K. Merchant (1990) Accounting for Management Control,
2nd edn (London: Chapman and Hall).
Enthoven, AC (1985) Reflections on the Management of the NHS: an American looks at
incentives to efficiency in health services management in the UK (London: Nuffield
Provincial Hospitals Trust).
Etzioni, A (1961) A Comparative Analysis of Complex Organizations (New York: Free
Press) .
European Commission (1991) Directive on the adoption of standards for satellite broadcast-
ing of television signals, COM (91) 242, doc. C3-290/9 1, SYN3 50 (Brussels).
Ezzarnel, M. (1992) Business Unit and Divisional Performance Measurement (London:
Academic Press in association with CIMA).
336 Bibliography

Ezzamel, M. and H. Hart (1987) Advanced Management Accounting: An Organisational

Emphasis (London : Cassell).
Ezzamel, M. and K. Hilton (1980a) 'Divisionalization in British Industry: A Preliminary
Study', Accountingand Business Research, Spring, pp. 197-214.
Ezzamel, M. and K. Hilton (1980b) 'Can divisional discretion be measured?', Journal of
Business Finance and Accounting, Summer, pp . 311-29 .
Ezzamel, M., K. Hoskins and R. MacVe (1990) 'Managing it all by numbers: A review
of Kaplan and Johnson 's Relevance Lost', Accounting and Business Research, Vol. 20,
No.7, pp. 153-66.
Fama, E. (1980) 'Agency Problems and the Theory of the Firm', Journal of Political
Economy, Vol. 88, No.2, pp . 288-307.
Fayol, H. (1949) General and Industrial Management (London: Pitman).
Fiske, S.T. and S.E. Taylor (1991) Social Cognition (New York: McGraw-Hill).
Fitzgerald, L., R. Johnston, S. Brignall, R. Silvestro and C. Voss (1991) Performance
Measurement In Service Businesses (London: CIMA).
Flamholtz, E.G., T.K. Das and AS. Tsui (1985) 'Toward an Integrative Framework of
Organisational Control', Accounting, Organizations and Society, Vol. 10, No .1,
pp.35-50.
Follet, M.P. (1924) Creative Experience (London: Longman and Green).
Forester, J. (1961) Industrial Dynamics (Cambridge, Mass.: MIT Press) .
Fox, A (1974) Beyond Contract: Work, Power and Trust Relations (London: Faber and
Faber).
Friedman, A (1977) Industry and Labour: Class Struggleat Workand Monopoly Capitalism
(London: Macmillan).
Gagliardi, P. (1986) 'The Creation and Change of Organisational Cultures:
A Conceptual Framework', Organization Studies, Vol. 7, No.2, pp. 117-34
Galloway, D. and D. Waldron (1988a) 'Throughput Accounting - 1: The need for a new
language for manufacturing', Management Accounting, November, pp. 34-5.
Galloway, D. and D. Waldron (1988b) 'Throughput Accounting - 2: Ranking products
profitably', Management Accounting, December, pp. 34-5.
Galloway, D. and D. Waldron (1989a) 'Throughput Accounting - 3: A better way to
control labour costs', Management Accounting, January, pp. 32-3 .
Galloway, D. and Waldron, D. (1989b) 'Throughput Accounting - 4: Moving on to
complex products', Management Accounting, February pp.40-1.
Ganley, J.A and J.S. Cubbin (1992) Public Sector Efficiency Measurement: Applications of
Data Envelopment Analysis (Amsterdam: North Holland).
Geertz, C. (1973) The Interpretationof Cultures (New York: Basic Books).
Glaister, K., and D. Thwaites (1993) 'Managerial Perception and Organisational
Strategy', Journal ofGeneral Management, Vol. 18, No.4, Summer, pp. 15-33 .
Goffman , E. (1959) The Presentation of Self in Everyday Life (Garden City, New York:
Doubleday Anchor) .
Goldratt, E. (1984) The Goal (London: Gower).
Goold, M. (1991) 'Strategic Control in the Decentralised Firm ', Sloan Management
Review, Winter, pp . 69-81.
Goold, M. and A. Campbell (1987a) 'Managing diversity: Strategy and control in
diversified British companies', Long RangePlanning, Vol. 20, No .5, pp . 42-52 .
Goold, M. and A Campbell (1987b) Strategies and Styles: The Role of the Centre in
Managing Diversified Corporations (Oxford : Basil Blackwell).
 Bibliography 337

Goold, M. and J.J. Quinn (1990a) Strategic Control: Milestones for Long Term Performance
(London: Hutchinson).
Goold, M. and J.J. Quinn (1990b) 'The Paradox of Strategic Controls', Strategic
Management Journal, Vol. II, No. I , pp. 43-57.
Goold, M., A, Campbell and K. Luchs (1993) 'Strategies and Styles Revisited: Strategic
Planning and Financial Control', LongRange Planning, Vol. 26, No .5, pp. 49-60.
Gordon, L.A. and D. Miller (1976) 'A Contingency Framework For The Design of
Accounting Information Systems', Accounting, Organisations and Society Vol. I, No . I,
pp.59-70.
Gorz, A. (1989) A Critique of Economic Reason (London: Verso).
Govindarajan, V. (1984) 'App rop riateness of Accounting Data in Performance
Evaluation: An Empirical Examination of Environmental Uncertainty as an
Intervening Variable', Accounting, Organizations and Society, Vol. 9, No .2, pp. 125-35.
Govindarajan, V. (1988) , A Contingency approach to strategy implementation at the
business-unit level : integrating administrative mechanisms with strategy', Academy
of Management Journal, Vol. 31, pp. 828-853.
Govindarajan, V. and Gupta A.K. (1985) 'Linking Control Systems to Business Unit
Strategy: Impact on Performance', Accounting, Organizations and Society, Vol. 10
No . I, pp. 51-66.
Grabner, G. (1993) The Embedded Firm (London: Routledge).
Gray, J., D. Jesson and N . Sime (1990) 'Estimating differences in the exam performance
of secondary schools in 6 LEAs - a multilevel approach to school effectiveness',
Oxford Reviewof Education , Vol. 16, pp. 137-58.
Grinyer, J. (1986) 'An alternative to maximisation of shareholders' wealth in Capital
Budgeting', Accounting and Business Research, Autumn, pp . 319-26.
Grinyer, P.H. and J.C Spender (1979) 'Recipes, Crises and Adaptation in Mature
Businesses', International Studies of Management and Organisation, Vol. 9, pp. 113-23.
Gulick, L.H. and L. Urwick (eds) (1937) Papers on the Science of Administration
(Columbia: Columbia University Press) .
Haka, S., L. Gordon and G. Pinches (1985) 'Sophisticated Capital Budgeting Selection
Techniques and Firm Performance', The Accounting Review, Vol. lx, No .4, pp. 651-69.
Hall, W.K. (1978) 'SBU's : Hot new topic in the Management of Diversification
Business Horizons', February, reprinted in D.J. McCarthy, R.J. Minichiello,
J.R Curranteds), Business Policy Strategy, rev . edn, (1979) (Homewood Ill.: Richard
D. Irwin).
Hambrick, D.C (1984) 'Taxonomic Approaches to Studying Strategy: Some Conceptual
and Methodological Issues', Journal of Management, Vol. 10, No . I, pp. 241-78.
Hammond, T. and A. Preston (1992) 'Culture, Gender and Corporate Control: Japan as
"Other" ',Accounting, Organizationsand Society, Vol. 17, No.8, pp. 795-808.
Handy, C (1985) Understanding Organizations (Harmondswoth: Penguin).
Hannan, M.T. and J.H. Freeman (1977) 'The Population Ecology of Organisations',
American Journal of Sociology, Vol. 82, pp. 929-64.
Harrison, R (1972) 'Understanding your organization's character', Harvard Business
Review, May/June, pp. 119-28.
Hayes, R, S. Wheelwright and K. Clark (1988) Dynamic Manufacturing: creating the
learning organisation (New York: The Free Press) .
Hertz, D.B. (1964) 'Risk Analysis in Capital Investment', Harvard Business Review,
Jan./Feb., pp. 175-86.
338 Bibliography

Hertzberg, F., B. Mausner and B.B. Snyderman (1959) The Motivation to Work (New
York: John Wiley).
Hirst, M.K. (1981) 'Accounting Information and the Evaluation of Subordinate
Performance', The AccountingReview, Vol. 56, pp. 771-84.
Hofer, CW. and D. Schendel (1978) Strategy Formulation: Analytical concepts (S1. Paul,
Minn : West Publishing Co).
Hofstede, G. (1967) TheGame of Budget Control (London: Tavistock).
Hofstede, G. (1985) 'The Interaction between National and Organizational Value
Systems', Journal of Management Studies, Vol. 22, No .4, pp. 347-57.
Hood, C (1991) 'A Public Management for All Seasons?', PublicAdministration, Vol. 69,
pp.3-19.
Hopper, T. (1988) 'Social Transformation and Management Accounting', in Proceedings
of the Second Interdisciplinary Perspective in Accounting Conference, Manchester.
Hopper, T. and P. Armstrong (1991) 'Cost Accounting, Controlling Labour and the Rise
of the Conglomerates', Accounting, Organizations and Society, Vol. 16, No. 5/6,
pp.405-38.
Hopper, T.M. and AJ. Berry (1983) 'Organisational Design and Management Control'.
In E.A Lowe and J.L.J. Machin (eds), New Perspectives In Management Control
(London: Macmillan) .
Hopper, T., J. Storey and H.C Willmott (1987) 'Accounting for Accounting: Towards
the Development of a Dialectical View', Accounting, Organizations and Society,
Vol. 12, No.5, pp . 437-56.
Hopper, T., D. Cooper, A Lowe, T. Capps and J. Mouritsen (1986) 'Management
Control and Worker Resistance in the National Coal Board: Financial Controls in the
Labour Process '. in D. Knights and H.C Willmott (eds) (1986) Managing the Labour
Process (London: Macmillan).
Hopwood, A (1972) 'An Empirical Study of the Role of Accounting Data in
Performance Evaluation', Empirical Research in Accounting, Supplement to Journal of
AccountingResearch, Vol. 10, pp . 156-82.
Hopwood, AG. (1974) Accounting and Human Behaviour (Englewood Cliffs, NJ:
Prentice-Hall).
Hopwood, AG. (1983) 'On Trying to Study Accounting in the Contexts in which it
Operates', Accounting, Organizations and Society, Vol. 8, No . 2/3, pp. 287-305.
Hopwood, A G. (1985) 'The growth of "worrying about management accounting:". in
K. Clark, R. Hayes and C Lorenz (eds), The Uneasy Alliance (Boston: Harvard
Business School) ,
Hopwood, A.G. (1987) 'The Archaeology of Accounting Systems', Accounting,
Organisations and Society, Vol. 12, No.3, pp. 207-34 .
Hopwood, AG. (1990) 'Accounting and Organisations Change', Accounting, Auditing
and Accountability, Vol. 3, No .1, pp. 7-17.
Horngren, CT. (1962) CostAccounting: A Managerial Emphasis (London: Prentice Hall) .
Horngren, CT. (1977) Cost Accounting: A Managerial Emphasis, new edn (London:
Prentice-Hall).
Horngren, CT. (1981) Introduction to Management Accounting, 5th edn (London:
Prentice-Hall). .
Horngren, CT. (1984) Introduction to Management Accounting, 6th edn (Englewood
Cliffs: Prentice-Hall).
 Bibliography 339

Horngren, C. T. and G. Foster (1991) Cost Accounting: a managerial emphasis, 7th edn
(Englewood Cliffs: Prentice-Hall)
Horngren, C.T. and G. Sundem (1991) Introduction to Management Accounting, 8th edn
(Englewood Cliffs: Prentice-Hall).
Horovitz, J.H. (1979) 'Strategic Control: A New Task for Top Management', Long Range
Planning, Vol. 12, pp. 2-7.
Hoskin, K.W. (1993) 'Education and the Genesis of Disciplinarity: The Unex-
pected Reversal', In E. Messer-Davidow, D.R. Shumway and D.J. Sylvan (eds),
Knowledges: Historicaland Critical Studies in Disciplinarity (London: University Press
of Virginia).
Hoskin, KW. and RH. Maeve (1988) 'The Genesis of Accountability: The West Point
Connection', Accounting, Organizations and Society, Vol. 13, No.1, pp. 37-73.
Hrebiniak, L.G. and W.F. Joyce (1986) 'The strategic importance of managing myopia',
Sloan Management Review, Fall, pp. 5-14 .
Hurst, E.G. (1982) 'Contro lling strategic plans', in P. Lorange (ed.), Implementation of
Strategy Planning (Englewood Cliffs, NJ: Prentice-Hall) .
Hyman, Rand R Anderson (1982) 'Solving Problems', in D.A. Kolb, I.M. Rubin and
J.M. Mcintyre (eds), Organisational Psychology (Englewood Cliffs, NJ: Prentice-Hall).
Hyman, Rand T. Elger (1981) 'Job Controls, the Employer Offensive and Alternative
Strategies', Capital and Class, Vol. 15, Autumn, pp. 115-45.
Imai, K and H . Itami (1984) 'Interpenetration of Organization and Market',
International Journal of Industrial Organisation , Vol. 2, pp. 285-310.
Innes, J. and F. Mitchell (1991) ABC: A Review With Case Studies (London: CIMA).
Ivanevitch, J. (1976) 'Effects of Goal Setting on Performance and Job Satisfaction',
Journal of AppliedPsychology, Vol. 61, pp. 605-12.
Jackson, P. and B. Palmer (1989) First steps in measuring performance in the public sector
(London: Public Finance Foundation).
[alland, RM. (1989) 'Plan-making in multi-divisional Companies' unpublished PhD,
thesis, University of Manchester.
Janis, I.L. (1972) Victims of Groupthink (Boston: Houghton Mifflin).
Janis, I.L. (1982) Groupthink: Psychological Studies of Policy Decisions and Fiascos (Boston:
Houghton Mifflin) .
Janis, I.L. and L. Mann (1977) Decision-Making (New York: Free Press) .
Jaworski, RJ. and S.M. Young (1992) 'Dysfunctional behavior and management control:
an empirical study of marketing managers', Accounting, Organizations and Society,
Vol. 17, No.1, pp. 17-35.
Jayson, S. (1922) 'Focus on People - Not Costs', Management Accounting (US),
September, pp. 28-33 .
Jelinek, M.L. Smircich and P. Hirsch (1983) 'Introduction: A Code of Many Colors',
Administrative Science Quarterly, Vol. 28, No.2, pp. 331-8
Jensen, M. (1983) 'Organization Theory and Methodology', AccountingReview, Vol. 50,
April, pp. 319-39 .
Jensen, M. and W. Meckling (1976) 'Theory of the Firm : Managerial Behavior, Agency
Costs and Ownership Structure'. Journal of Financial Economics, Vol. 3, pp. 30!H)0.
Jensen, M. and W. Meckling (1979) 'Rights and Production Functions: an Application to
the Labor Managed Firm', Journal of Business, Vol. 52, pp . 469-506.
Johnson, G. (1987) Strategic Change and theManagement Process (Oxford: Basil Blackwell).
340 Bibliography

Johnson, G. and K. Scholes (1993) Exploring Corporate Strategy (Hemel Hempstead:

Prentice-Hall).
Johnson, H.T. (l992a) Relevance Regained: From Top-down Control to Bottom-up
Empowerment (New York: Free Press).
Johnson, H.T. (1992b) 'It's Time to Stop Overselling Activity-Based Concepts: Start
Focusing on Customer Satisfaction Instead', Management Accounting (US),
September, pp. 26-35.
Johnson, H.T. and RS. Kaplan (1987) Relevance Lost: The Rise and Fall of Management
Accounting (Cambridge, Mass.: Harvard Business School Press).
Jones, T.e. (1992) 'Understanding Management Accountants: The Rationality of Social
Action', Critical Perspectives on Accounting, Vol. 3, No .3, pp. 225-257.
Jones, T.e. (forthcoming) Accounting and the Enterprise: A Social Analysis (London:
Routledge).
Jones, T.e., W.L. Currie and D. Dugdale (1993) 'Accounting and Technology in Britain
and Japan: Learning From Field Research', Management Accounting Research, Vol. 4,
No.2, pp . 109-37.
Jonsson, S. and RA. Lundin (1977) 'Myths and Wishful Thinking As Management
Tools', in P.e. Nystrom and W.H. Starbuck (eds), Prescriptive Models of Organisation
(Amsterdam: North Holland) pp . 157-70.
Kaplan, RS. (1984) 'Yesterday's Accounting Undermines Production', Harvard Business
Review, July / August, pp . 95-101.
Kaplan, R.S. (1985) 'The Obsolescence of Cost Accounting System' in K. Clark,
R Hayes, and e. Lorenz, (eds), The Uneasy Alliance (Boston: Harvard Business
School).
Kaplan, R S. and D.P. Norton (1992) 'The balanced scorecard - measures that drive
performance', Harvard Business Review, Vol. 70, No .1, pp. 71-9 .
Kaplan, RS. and D.P. Norton (1993) 'Putting the Balanced Scorecard to work', Harvard
Business Review, September-October, pp. 134-47.
Katz, D. and RL. Khan (1978) The Social Psychology of Organisations (New York:
John Wiley).
Kelly, G.A. (1955) The Psychology of Personal Constructs (New York: W.W. Norton).
Kelly, J. (1982) Scientific Management, Job Design and Work Performance (London:
Academic Press).
Kenis, 1. (1979) 'Effects of Budgetary Goal Characteristics on Managerial Attitudes on
Performance', The AccountingReview, Vol.liv, No .4, pp. 707-21.
Kennedy, A. and R. Mills (1992) 'Post-completion auditing: a source of strategic direc-
tion?', Management Accounting, Vol. 70, No.5, May, pp. 26-8 .
Kerr, S. (1975) 'On the folly of rewarding A while hoping for B', Academyof Management
Journal, pp. 769-83.
Khandawalla, P.N. (1972) 'The Effects of Different types of Competition on the Use of
Management Control', Journal of Accounting Research, Vol. 10, pp. 275-85.
King, P. (1975) 'Is the Emphasis of Capital Budgeting Theory Misplaced?', Journal of
Business Finance and Accounting, Spring, pp. 69-82 .
Knights, D. and D. Collinson (1987) 'Disciplining the Shopfloor: A Comparison of the
Disciplinary Effects of Managerial Psychology and Financial Accounting',
Accounting, Organizationsand Society, Vol. 12, No.5, pp. 457-77.
Kornai, J. (1992) The Socialist System: The Political Economy of Communism (Oxford:
Clarendon Press).
 Bibliography 341

Kotter , J.P. and J.L. Heskett (1992) Corporate Culture and Performance (New York: The
Free Press) .
Kuhn, T.S. (1970) The Structure of Scientific Revolutions (Chicago: University of Chicago
Press) .
Langfield-Smith, K. (1993) 'Linking Management Control Systems And Strategy: An
Examination of The Strategy Construct and Critical Review' / paper presented at the
European Accounting Association Conference, Turku, Finland, 28-30 April.
Langlois, A. (1986) Economics as a Process (Cambridge: Cambridge University Press).
Laughlin, RC (1987) 'Accounting Systems in Organisational Contexts: A Case
for Critical Theory', Accounting , Organizations and Society, Vol. 12/ No . 5/
pp. 479-502.
Laughlin, RC (1991) 'Environmental Disturbances and Organisational Transitions and
Transformations: Some Alternative Models', Organisation Studies, Vol. 12/ No.2.
Laughlin, R. and J. Broadbent (1993) 'Accounting and Law: Partners in the
Juridification of the Public Sector in the UK?', Critical Perspectives on Accounting.
Vol. 4/ No. 4/ pp. 337-68.
Laughlin, Rand R Gray (1988) Financial Accounting:Method andMeaning (London, VNR).
Laughlin, R , J. Broadbent, D. Shearn and H. Willig-Atherton (1994) 'Absorbing LMS:
The Coping Mechanism of a Small Group', Accounting, Auditing and Accountability
Journal, Vol. 7/ No . 1/ pp. 59-85 .
Lawler, E.E. (1973) Motivation in Work Organisations (Monterey (Caly): Brroks Cole
Publishing Co.).
Leijonhufvud, A (1968) On Keynesian Economics and the Economics of Keynes (Oxford:
Oxford University Press) .
Levi-Strauss, C (1967) The Scope of Anthropology (New York: Jonathan Cape) .
Lilienfeld, R (1978) The Riseof Systems Theory (New York, Chichester: John Wiley)
Lillis, AM. (1992) 'Sources of influence on capital expenditure decisions: a contextual
study of accounting measurement' / Management Accounting Research, No. 3/
pp.213-27.
Lindblom, CE. (1977) Politics and Markets (New York: Basic Books).
Llewellyn, D. (1991) 'Is There A Credit Crunch?', Banking World, May, pp . 23-6.
Locke, E.A (1968) 'Towards a Theory of Risk Motivations and Incentives',
Organizational Behaviour and Human Performance, Vol. 3/ pp . 157-89.
Loft, A (1991) 'The History of Management Accounting: Relevance Found', in
D. Ashton, T. Hopper and RW. Scapens (eds), Issues in Management Accounting
(Englewood Cliffs, NJ: Prentice Hall) pp . 17-38.
Lorange, P. (ed) (1980) Implementation of Strategic Planning (Englewood Cliffs,
NJ: Prentice-Hall).
Lorange, P. (1988) 'Monitoring strategic progress and ad hoc strategic modification', in
J. Grant (ed), Strategic Management Horizons (Greenwich, Conn.: JAI Press).
Lorange, P. and D. Murphy (1984) 'Considerations in Implementing Strategic Control',
Journal of Business Strategy, Vol. 5, pp. 27-35.
Lorange, P./ M.F. Scott Morton and S. Goshal (1986) Strategic Control (St Paul, Minn .:
West Publishing Co).
Lorsch , J.W. and S.A Allen (1973) Managing Diversity and Interdependence Division of
Research, Graduate School of Business Administration, Harvard University.
Lowe, E.A. (1971) 'Budgetary Control: An Evaluation in a Wider Managerial
Perspective', Accountancy, November, pp. 764-9.
342 Bibliography

Lowe, E.A. and W.P. Chua (1983) 'Organisational Effectiveness and Management
Control', in E.A. Lowe and J.L.P. Machin (eds), New Perspectives in Management
Control (London: Macmillan).
Lowe, E.A. and J.L.P. Machin (1983) New Perspectives in Management Control (London:
Macmillan).
Lowe, E.A. and RW. Shaw (1968) 'An Analysis of Managerial Biasing : evidence from
company's budgeting process', Journal of Management Studies, Vol. 5, pp. 304-15.
Lowe, E.A., A.G. Puxty and RC. Laughlin (1983) 'Simple Theories for Complex
Processes: Accounting Policy and the Market for Myopia', Journal of Accounting and
Public Policy, Vol. 2, pp. 19-42.
Lowe, T. and T. Puxty (1989) 'The Problems of a Paradigm: A Critique of the Prevailing
Orthodoxy in Management Control', in W.P. Chua, T. Lowe and T. Puxty (eds),
Critical Perspectives in Management Control (London: MacMillan).
Lyne, S. (1993) 'An Evaluation of the Roles of Activity-Based Accounting Information',
Management Accounting Research Group Conference, September, University of
Aston .
Machin, J.L. (1983) 'Management Control Systems: Whence and Whither?', in
E.A. Lowe and J.L.P. Machin (eds), New Perspectives in Management Control (London:
MacMillan) pp. 22-42.
Maciariello, J.A. (1984) Management Control Systems (Englewood Cliffs, NJ, London:
Prentice Hall) .
Macintosh, N.B. (1985) The Social Software of Accounting and Information Systems
(Chichester: John Wiley).
Mackey, J. (1991) 'MRP, JIT and Automated Manufacturing and the Role of Accounting
in Production Management', in D. Ashton, T. Hopper and R Scapens (eds), Issues in
Management Accounting (Hemel Hempstead: Prentice-Hall) .
Malinowski, B. (1922) Argonauts of the Western Pacific (London: Routledge and
Kegan Paul) .
Malmberg, A. (1980) 'The Impact of Job Reform on Accounting Systems' . in
G. Kanawaty (ed.), Managing and Developing New forms of Work Organisation, Geneva:
International labour organization, 2nd edn, pp. 79-94 .
March, J. and J. Olsen (1976) Ambiguity and Choice in Organisations (Bergen:
Universitetsforlaget).
March, J. and H . Simon (1958) Organisations (New York: John Wiley).
Marglin, S.A. (1974) 'What do Bosses do? The origins and functions of Hierarchy in
Capitalist production', Review of Radical Political Economics, Vol. 6, pp. 33-60 .
Martin, J. and C. Siehl (1983) 'Organisational Culture and Counterculture: An Uneasy
Symbiosis', Organisational Dynamics, Vol. 12, pp. 52-64.
Marx, K. (1972) Capital (London: Lawrence and Wishart).
Marx, K. (1868, reprinted 1976) Capital: Volume One (Harmondsworth: Penguin).
Maslow, A.H . (1954) Towards a Psychology of Being (New York, London: Van Nostrand
Rheinhold).
Mayo, E. (1933) The Human Problems of an Industrial Civilization (Cambridge, Mass.:
Harvard University Press).
Mayo, E. (1945) The Social Problems of an Industrial Civilization (New York: Macmillan).
McCosh, A. (1990) 'Positive Control', Economia Aziendale, Vol. ix, No .3, pp. 405-427 .
McGregor, D. (1960) TheHuman Sideof Enteprises (New York: McGraw-Hill).
McGuire, A. et al. (1991) 'The Economics of Health Care', in A. McGuire, P. Penn and
K. Mayhew (eds), ProvidingHealth Care (Oxford: Oxford University Press).
 Bibliography 343

Mead, M. (1928) Comingof Age in Samoa (New York: Morrow).

Meek, V.L. (1988) 'Organisational Culture: Origins and Weaknesses', Organization
Studies, Vol. 9, No .4, pp. 454-73
Merchant, K. (1981) 'The Design of the Corporate Budgeting System: Influences on
Managerial Behaviour and Performance', The Accounting Review, Vol. lvi, No.4,
pp.813-29.
Merchant, K. (1985) Control in Business Organizations (London: Pitman).
Meyer, S. (1980) 'Adapting the Immigrant to the Line: Americanization in the Ford
Factory, 1914-1921', Journal of Social History, Vol. 14, No. I, pp. 67-81.
Michael, D. (1973) On Learning to Plan and Planning to Learn (San Francisco: [ossey-Bass).
Middaugh, J.K. (1988) 'Management Control In The Financial Services Industry',
Business Horizons, May-June.
Milani, K. (1975) 'The Relationship of Participation in Budget Setting to Industrial
Supervisor Performance and Attitudes: A Field Study', The Accounting Review, Vol. I,
No.2, pp. 274-83.
Miles, RE. and c.c. Snow (1978) Organisational Strategy, Structure and Process (New
York: McGraw-Hill).
Miller, D. (1990) The Icarus Paradox: How Excellent Organisations Can Bring About Their
Own Downfall (New York: Harper Business) .
Miller, D. and P.H. Friesen (1980) 'Momentum and revolution in organisational adapta-
tion', Academyof Management Journal, Vol. 23, pp. 591-614.
Miller, D. and P.H . Friesen (1984) Organisations: A Quantum View (Englewood Cliffs,
NJ: Prentice-Hall).
Miller, E.J. (1976) 'Task, Territory and Technology', in E.J. Miller Task and Organisation
(London: John Wiley) .
Miller, P. and T. O'Leary (1987) 'Accounting and the Construction of the Governable
Person', Accounting, Organizations and Society, Vol. 12, No .3, pp. 235-65.
Mintzberg, H . (1973) The Nature of Managerial Work. Harper and Row.
Mintzberg, H . (1978) 'Patterns in Strategy Formation', Management Science, May,
pp.934-48.
Mintzberg, H. (1979) The Structuringof Organisations (Englewood Cliffs, NJ: Prentice-Hall).
Mintzberg, H . (1983) Power In and Around Organisations (Englewood Cliffs, NJ:
Prentice-Hall).
Mintzberg, H. and J. Waters (1985) 'On Strategies Deliberate and Emergent', Strategic
Management Journal, pp. 25-37.
Monden, Y. and M. Sakurai (eds) (1990) Japanese Management Accounting (New York:
Productivity Press).
Morden, T. (1993) Business Strategyand Planning (London: McGraw-Hill).
Morgan, G. (1986) Images of Organisation (London: Sage, Beverly Hills).
Morgan, G. (1979) 'Internal Audit Role Conflict: A Pluralist View' Managerial Finance,
Vol. 5, pp. 160-70.
Morrison, I. (1989) 'The Cultural Revolution In Banking: A 20-Year Perspective',
Chartered Institute of Bankers' Cambridge Seminar.
Munro, RJ.B. and D.J. Halterly (1993) 'Accountability and the New Commercial
Agenda', Critical Perspectives on Accounting, Vol. 4, No.4, pp. 369-95.
National Health Service (1983) Management Inquiry Report (The Griffiths Report)
(London: Department of Health and Social Security).
Nelson, Rand S. Winter (1982) An Evolutionary Theory of Economic Change (Cambridge,
Mass .: Harvard University Press).
344 Bibliography

Newman, W.H. and J.P. Logan (1971) Strategy, Policy and Central Management
(Cincinnati: South-Western Publishing).
Northcott, D. (1991) 'Rationality and Decision Making in Capital BUdgeting ', British
AccountingReview, Vol. 23, No.3, pp. 219-234.
Northcott, D. (1992) Captial Investment Decision Making (London: CIMA/Academic
Press).
Otley, D. (1978) 'Budget Use and Managerial Performance', Journal of Accounting
Research Vol. 1, No.6, pp. 122-49.
Otley, D. (1990) 'Issues in Accountability and Control: some observations from a study
of colliery accountability in the British coal corporation', Management Accounting
Research, Vol. 1, pp. 101-23 .
Otley, D. and AJ. Berry (1980) 'Control, Organization and Accounting', Accounting,
Organizationsand Society, Vol. 5, No.2, pp. 231-46.
Ouchi, W.G. (1977) 'The Relationship between Organizational Structure and
Organizational Control', Administrative Science Quarterly, Vol. 22, pp. 95-112.
Ouchi, W.G. (1979) 'A Conceptual Framework for the Design of Organisational Control
Mechanisms', Management Science, Vol. 25, No .9, pp. 833-49
Ouchi, W.G. (1980) 'Markets, Bureaucracies and Clans', AdministrativeScience Quarterly,
Vol. 25, No.1, pp . 129-41.
Ouchi, W.G. (1981) Theory Z (Reading, Mass.: Addison-Wesley) .
Parker, L.D. (1979) 'Divisional Performance Measurement: Beyond an Exclusive Profit
Test', Accounting, and Business Research, Autumn, pp. 309-19.
Parsons, T. (1964) Social Systems (New York: The Free Press).
Pascale, RT. (1990) Managing on the Edge: How Successful Companies Use Conflict to Stay
Ahead (London: Viking Penguin).
Pascale, RT. and AG. Athos (1981) The Art of Japanese Management (New York: Simon
and Schuster).
Pask, G. (1961) An Approach to Cybernetics (London: Hutchinson).
Perrin, J. (1987) 'Resource Management and Clinical Budgeting', Journal of Management
in Medicine, Vol.2, No.2, pp. 99-106.
Perrin, J. (1988) Resource Management in the N .H.S. (Woringham: Van Nostrand
Reinhold (UK), in Association with Health Services Management Centre).
Perrow, C. (1967) 'A Framework for the Comparative Analysis of Organisations',
American Sociological Review, April, pp. 194-208.
Perrow, C. (1970) Organisational Analysis: A Sociological View (London: Tavistock).
Perrow, C. (1986) Complex Organisations, 3rd Edn (New York: Random House).
Peters, T.J. and RH. Waterman (1982) In Search of Excellence (New York: Harper and
Rowe, New York).
Pettigrew, AM. (1979) 'On Studying Organisational Cultures', Administrative Science
Quarterly, Vol. 24, pp. 570-81
Pettigrew, AM. (1985) The Awakening Giant: Continuity and Change in Imperial Chemical
Industries (Oxford: Basil Blackwell).
Pettigrew, AM., L. McKee and E. Ferlie (1989) 'Hints on How to Ring the Changes',
Health ServiceJournal, 16 February.
Pfeffer, J. (1981) 'Management as Symbolic Action: The Creation and Maintenance of
Organisational Paradigms', in L.L. Cummings and B.M. Staw (eds), Research in
Organizational Behavior (Greenwich, Conn: JAI Press) pp. 1-52 .
 Bibliography 345

Pfeffer, J. and c.R. Salanick (1977) The External Control of Organisations: A Resource
Dependent Perspective (New York/London: Harper and Row).
Pike, R. (1983) 'The Capital Budgeting Behaviour and Corporate Characteristics of
Capital-Constrained Firms', Journal of Business Finance and Accounting, Vol. 10,
pp.663-72.
Pike, R. (1988) 'An Empirical Study of the Adoption of Sophisticated Capital Budgeting
Practices and Decision-Making Effectiveness', Accounting and Business Research,
Vol. 18, pp. 341-51.
Piore, M.J. and C.F. Sabel (1984) The Second Industrial Divide: Prospects For Prosperity
(New York: Basic Books).
Pollert, A. (1988) 'The "Flexible Firm": Fixation or Fact?', Work, Employment and Society,
Vol. 2, No .3, pp. 281-316.
Porter, M.E. (1985) Competitive Advantage, Creating and Sustaining Superior Performance
(New York: The Free Press).
Preston, A.M. (1991) 'Budgeting, Creativity and Culture', in D. Ashton, T. Hopper and
R.W. Scapens (eds), Issues In Management Accounting (Hemel Hempstead: Prentice Hall).
Purdy, D.E., (1991) 'Natural Learning in Committee B', Management Education and
Development, Vol. 22, Part 1, pp. 60-70 .
Purdy, D.E. (1993a) 'Accounting, Controls, Change and the Perceptions of Managers: A
Longitudinal Study of Ward Units in a Teaching Hospital', Financial Accountability
and Management, Vol. 9, No .1 , pp. 45-66
Purdy, D.E. (1993b) 'Ward Sisters and Financial Management Accounting', Financial
Accountabilityand Management, Vol. 9, No .4, pp. 279-96.
Putterman, L. (1986) The Economic Nature of the Firm: a Reader (Cambridge: Cambridge
University Press) .
Putterman, L. (1988) 'The Firm as Association versus the Firm as Commodity',
Economics and Philosophy, Vol. 4, pp. 24~6 .
Quinn, J.B. (1980) Strategies for change: Logical Incrementalism (Homewood, Ill.: Richard
D. Irwin).
Radcliffe-Brown, A.R. (1952) Structure and Function in Primitive Society (Oxford : Oxford
University Press) .
Reece, J.5. and W.R. Cool (1978) 'Measuring investment centre performance', Harvard
Business Review, May-June, pp. 29-49 .
Ricketts , D. and J. Gray (1988) Managerial Accounting (Boston: Houghton Mifflin).
Roethlisberger, F.J. and W.J. Dickson (1964) Management and the Worker (New York:
John Wiley).
Ronen, J. and J.L. Livingstone (1975) 'An Expectancy Theory Approach to the
Motivational Impact of Budgets', The Accounting Review, Vol. 50, pp. 893-903.
Ronen, J. and S. Sad an (1981) Smoothing Income Numbers (Reading: Addison-Wesley».
Rose, M. (1975) Industrial Behaviour: Theoretical Development SinceTaylor (London: Allen
Lane) .
Roy, D.F. (1960) "'Banana Time" Job Satisfaction and Informal Interaction', Human
Organization, pp. 158-68.
Roy, D.F. (1973) 'Banana Time, Job Satisfaction and Informal Interaction', in G. Salaman
and K. Thompson (eds), Control and Ideology in Organisations (Milton Keynes: Open
University Press) .
Salaman, G. (1979) Work Organization: Resistance and Control (London; Longman).
346 Bibliography

Seapens, RW. (1991) Management Accounting: a review of contemporary developments, 2nd

edn (London: Macmillan).
Scapens, RW. and J. Roberts (1993) 'Accounting and Control: a case study of resistance
to accounting change', Management Accounting Research, No .1, pp. 1-32.
Schein, E.H. (1984) 'Coming to a New Awareness of Organisational Culture', Sloan
Management Review, Vol. 25, Winter, pp. 3-16
Schein, E.H. (1985) Organisational Culture and Leadership (San Francisco: [ossey-Bass).
Schiff, M. and A.Y. Lewin (1970) 'The Impact of People on Budgets', The Accounting
Review, Vol. 45, pp . 259-268.
Schoderbeck, P.P., CG. Schoderbeck and A.G. Kefalas (1975) Management Systems:
Conceptual Considerations (Dallas: Business Publishers).
Schon, D.A. (1983) The Reflective Practitioner (New York: Basic Books).
Schonberger, R (1986) World Class Manufacturing (New York: The Free Press).
Schreyogg, G. and Steinmann, H. (1987) 'Strategic Control: A New Perspective',
Academy of Management Review, Vol. 12, No .1, pp. 91-103 .
Schwenk, C (1984) 'Cognitive simplification processes in strategic decision-making',
Strategic Management Journal, Vol. 5, pp. 111-28.
Schwenk, C. (1988) The Essenceof Strategic Decision Making (Lexington, Mass .: Lexington
Books).
Scott, B.R (1971) 'Stages Four of Corporate Development - Part 1', Harvard Business
School Case Services.
Seal, W.B. (1993) Accounting, Management Control and Business Organisation: an institu-
tionalist perspective (Aldershot: Avebury).
Senge, P. (1993) Fifth Discipline: Art and Practice of the LearningOrganisation (New Canon
CT: Business Books).
Sephton, M. and Ward, T. (1990) 'ABC in Retrail Financial Services' Management
Accountant, Vol. 13, pp. 29-33 .
Shaiken, H., S. Herzenberg and S. Kuhn (1986) 'The Work Process Under More Flexible
Production', Industrial Relations (US), Vol. 23, No .2, pp. 167-83.
Shapiro, S.P. (1987) 'The Social Control of Impersonal Trust', American Journal of
Sociology, Vol. 93, No.3, pp. 623-58.
Sheldon, A. (1980) 'Organisational Paradigms: A Theory of Organisational Change',
Organisational Dynamics, Vol. 8, No .3, pp. 61-7l.
Shillinglaw, G.D. (1982) Cost Accounting Analysis and Control (Homewood, Ill.: Richard
D. Irwin) .
Simmonds, K. (1981) 'Strategic Management Accounting', Management Accounting,
April, pp. 26-9 .
Simon, H. (1957) AdministrativeBehavior (New York: Free Press) .
Simon, Herbert, Guetzkow, H ., Kozmetsky, G. and Tyndall, G. (1957) Centralization
versus Decentralization in the Controller's Department (New York: Controllship
Foundation).
Simons, R (1987a) 'Accounting Control Systems and Business Strategy: An Empirical
Analysis', Accounting, Organisations and Society, Vol. 12, No.4, pp. 357-74 .
Simons, R (1987b) 'Planning, Control and Uncertainty: A Process View', in W.J. Bruns
and RS . Kaplan (eds), Accountingand Management - Field Study Perspectives (Boston,
Mass .: Harvard Business School Press) pp. 339-62.
Simons , R (1990) 'The Role of Management Control Systems in Creating Competitive
Advantage: New Perspectives', Accounting, Organizations and Society, Vol. 15,
 Bibliography 347

~o . 1/2,pp. 127-43
Sims, H . and D. Gioia (1986) The Thinking Organisation (San Francisco: [ossey-Bass).
Skinner, RC. (1990) 'The role of profitability in divisional decision making and perform-
ance evaluation', Accounting and Business Research, Spring, Vol. 20, pp . 135-41
Smircich, L. (1983) 'Concepts of Culture and Organizational Analysis', Administrative
Science Quarterly, Vol. 28, ~o . 2, pp. 328-58.
Smith, P. (1987) 'Control In Banks', Banking World , Feb., pp. 2l.
Smith, C. (1989) 'Flexible Specialization, Automation and Mass Production', Work,
Employmentand Society, Vol. 3, ~o . 2, pp. 203-20.
Smith, P. (1990) 'The use of performance indicators in the public sector', Journal of the
Royal Statistical Society, Series A, ~o . 153, pp. 53-72.
Smith, P. (1993) 'Outcome-related performance indicators and organizational control in
the public sector', British Journal of Management,Vol. 4, No.3, pp. 135-52.
Smyth, p.s. and P.B. Checkland (1976) 'Using a Systems Approach: The Structure of
Root Definitions', Journal of Applied Systems Analysis, Vol. 5, No . l.
Solomons, D. (1965) Divisional Performances: Measurement and Control (Homewood, Ill.:
Richard D. Irwin).
Spicer, B. (1990) 'N ew Directions In Management Accounting Practice and Research',
Management Accounting Research, Vol. 1, pp. 139-54.
Spicer, B.H. and V. Ballew (1983) 'Management Accounting Systems and the Economics
of Internal Organization', Accounting, Organizations and Society, Vol. 8, pp. 73-96.
Sproull, L.S. (1981) 'Beliefs in Organizations' in P.c. Nystrom and W.H. Starbuck (eds),
Handbook of Organisational Design, Vol. 2 (Oxford : Oxford University Press) .
Stacey, RD. (1993) Strategic Management and Organisational Dynamics (London: Pitman
Publishing).
Starbuck, W.H. and B. Hedberg (1977) 'Saving an Organisation from a stagnating
Environment', in H .B. Thorelli (Ed.), Strategy + Structure = Performance (Bloomington:
Indiana University Press) pp. 249-58 .
Starkey, K., M. Wright and S. Thompson (1991) 'Flexibility, Hierarchy and Markets',
British Journal of Management, Vol. 2, pp. 165-76.
Steers, RM. (1975) 'Problem s in the Measurement of Organisational Effectiveness',
Administrative Science Quarterly, December, pp. 613-29.
Stewart, J.D. (1984) 'The role of information in public accountability', in A. Hopwood
and C. Tomkins (eds), Issues in Public Sector Accounting (Oxford: Philip Allan) .
Stewart, R (1967) The Reality of Management (London: Macmillan).
Strong, P. and J. Robinson (1988) 'New Model Management: Griffiths and the ~ .H.s : ,
Nursing Policy Studies Centre, University of Warwick.
Swieringa, RJ. and Weick, K.E. (1987) 'Management Accounting and Action'
Accounting Organisations and Society. Vol. 12 ~o . 3 pp. 293-308.
Taylor, F.W. (1911, reprinted 1947) Scientific Management (New York, Harper and
Brothers).
Taylor, S. and J. Crocker (1983) 'Schematic Bases of Social Information Processing', In
E.T. Higgens, c.P. Herman and M.P. Zamna Izanna (eds), Social Cognition: The
Ontario Symposium (Hillsdale, ~J : Earlbaum).
Teubner, G. (1983) 'Substantive and Reflexive Elements in Modern Law', Law and
SocietyReview, Vol. 18, no . 2, pp. 291-30l.
Thode, S.F. (1986) 'The trouble with Divisional Hurdle Rates', Business Horizons,
Jan-Feb., pp. 62-6.
348 Bibliography

Thompson, J.D. (1967) Organizations in Action (New York: McGraw-Hill).

Thompson, J.D. and A. Tuden (1959) 'Strategies, Structures and Processes' in J.D.
Thompson (eds), Comparative Studies in Administration (Pittsburg: University of
Pittsburg Press).
Thompson, S. and M. Wright (eds) (1988) Internal Organisation, Efficiency and Profit
(Oxford: Philip Allan).
Tinker, T. (1988) 'Panglossian Accounting Theories: the Science of Apologising in
Style', Accounting, Organizations, and Society, Vol. 13, No.2, pp. 165-89.
Tocher, K (1970) 'Control', Operational Research Quarterly, June, pp. 159-80.
Tocher, K (1976) 'Notes for Discussion on Control', Operational Research Quarterly, June,
pp .231-9.
Tolman, E.C (1949) 'There is More than One Kind of Learning', Psychological Review,
Vol. 56, pp. 144-55.
Tomkins, C (1973) Financial Planning in Divisional Companies (London: Haymarket
Publishing).
Tosi, H. (1975) 'The Human Effects of Managerial Budgeting Systems', in J.H.
Livingson (ed), Management Accounting: The Behavioral Foundations (Columbus, Ohio:
Grid).
Trist, E. (1976) 'A Concept of Organizational Ecology', AustralianJournal of Management,
Vol. 2.
Trumpington, Baroness (1987) 'OR in government', Journal of the Operational Research
Society, Vol. 38, pp. 907-12.
United Kingdom Government (1991) The Citizen's Charter (London: HMSO).
Urwick, L.F. (1947) The Elements of Administration (London: Pitman).
Van Gusteren, H.R (1976) The Quest for Control (London: John Wiley)
Vancil, RF. (1979) Decentralisation: Ambiguity by Design (Homewood, Ill.: Richard D.
Wiley).
Veblen, T. (1953) The Theory of the Leisure Class (New York: Mentor Book, New
American Library.
Vickers, G. (1965) The Art ofJudgement: A Study of PolicyMaking (London: Methuen).
Vickers, G. (1967) Towards a Sociology of Management (London: Chapman and Hall) .
Waldron, L. (1988) 'Accounting for CIM: The New Yardsticks', EMAP Business and
Computing Supplement), February, pp . 1-2.
Watts, RL. and J.L. Zimmerman (1979) 'The Demand and Supply of Accounting
Theories: The Market for Excuses' , Accounting Review, Vol. 54, April pp. 273-305.
Watts, RL. and J.L. Zimmerman (1986) Positive Accounting Theory (New York:
Prentice-Hall).
Watzlawick, P., P. Weakland and J.H. Fisch (1974) Change: Principles of Problem
Formation and Problem Resolution (New York: W.W. Norton).
Weber, M. (1922 translated 1968) Economy and Society (New York: Bedminster Press) .
Weber, M. (1948) 'The Social Psychology of the World Religions', in H.H. Gerth and
CW. Mills (eds) From Max Weber (London: Routledge and Keegan Paul) pp. 267-30l.
Weick, K.E. (1976) 'Educational Organisations as Loosely Coupled Systems',
Administrative Science Quarterly, Vol. 21, pp. 1-19.
Weick, KE. (1979) The Social Psychology of Organising, 2nd Edn (New York: Random
House) .
Weick, KE. (1983) 'Managerial Thought in the Context of Action', in S. Srivastva (ed.),
The Executive Mind (San Francisco : jossey-Bass).
 Bibliography 349

Weiner, N. (1948) Cybernetics (Cambridge, Mass.: MIT Press).

Weitzman, M. (1976) 'The new Soviet incentive system', Bell Journal of Economics and
Management Science, Vol. 7, pp. 251-7.
Welch, J.B. and Kainen T.L. (1983) 'Risk adjusted multiple hurdle rates: Better capital
budgeting', Financial Executive, May.
Whyte W.F. (1948) Human Relations in the Restaurant Industry (New York: McGraw-Hill)
Wildavsky, A. (1975) Budgeting: A Comparative Theory of Budgetary Processes (Boston:
Little, Brown).
Williams, K., J. Williams and C. Haslam (1989a) 'Do Labour Costs Really Matter?',
Work, Employment and Society, Vol. 3, No.3, pp . 281-305.
Williams, K., J. Williams and C. Haslam (1989b) 'Why take the stocks out? Britain v.
Japan', International Journal of Operations and Production Management, Vol. 9, No.8.
Williams, K. and C. Haslam (1991) 'How far from Japan?', Critical Perspectives in
Accounting, Vol. 2, pp. 145-69 .
Williams, K. J. Williams and C. Haslam (1992) 'Against Lean Production', Economy and
Society, Vol. 21, No.3, pp. 321-54 .
Williamson, O.E. (1970) Corporate Control and Business Behavior (Englewood Cliffs, NJ:
Prentice-Hall).
Williamson, O.E. (1973) 'Markets and Hierarchies: Some Elementary Considerations',
American Economic Association, Vol. 63, No.2.
Williamson, O.E. (1975) Marketsand Hierarchies: Analysisand Antitrust Implications (New
York: Free Press) .
Williamson, O.E. (1985) The Economic Institutions of Capitalism (New York: The Free
Press).
Wilson, R.MS. and W.F. Chua (1992) Managerial Accounting: Method and Meaning
(London: VNR).
Wolk, H., Q . Gerber and G. Porter (1988) Management Accounting (Boston: PW5-Kent).
Woodward, J. (1958a) Management and Technology (London, HMSO).
Woodward, J. (1958b) Industrial Organisation: Theory and Practice, reprinted 1965
(Oxford: Oxford University Press) .
Wrong, D. (1961) 'The Oversocialized Concept of Man in Modern Sociology', American
Sociological Review, Vol. 26, April, pp . 183-93.
Index

absorption costing 83-4 traditional methods 103-4

accountability accounting rate of return 81,82-3
choice of units of 252 Ackoff, R. 120
and control 165-7 action plan 7
education service 275 activity-based costing (ABC) 55, 113,
hard 299 148,227,283-4,312
meaning of 322-3 described 290-1
regimes of 299 reasons for 292-3, 294
role of accounting 321-3 strategic and behavioural 291
soft 300 in a UK bank 291-3
accountable person 323 Acts
accountants 225 Banking (1987) 288
conflicting roles 315 Building Societies (1986) 63,289,295
Japanese companies 232-3 Companies (1985) 63 ·
accounting 23 Education (1944) 271
and accountability 321-3 Education (1986) 274
activity-based see activity-based Education Reform (1986) 65,69
costing Education Reform (1988) 274,278,
base 17 280
based on cells 318 Financial Services (1986) 289,295
control 18-19,113 National Health and Community Care
cost 227,231 (1990) 64,65,69
criticised 311-13,316-17 Partnership (1890) 63
decline 307-8 adaptation 11,42,48,57
definition 12 agency theory 33-4,155-6,158-9,
incorporation of market in 234-6 165-6,302-3
information 77 Ahmed, M.N. 226
limitations 327-8 Air France 240
in manufacturing 306-8 airlines 238-54
policies 218 competition 240
rise of 306-7 costs structures 241
role in manufacturing 299-323 deregulation 238-41
and social structures 42-3 nature of the business 238-41
strategic 7 seealso lCA
symbolic 184 Alchian, A. 159
systems and controls 77-94 Aldrich, H.E. 22

350
 Index 351

Allaire, Y. 23, 186 environment 287-8

Allen, M. 224 overexpansion 288
Allen, SA 95,97, 101-2 Barc1aysBank 287, 288, 289
Alvesson, M. 181 Barnard, C. 5, 8, 301
Arney, L.R. IS, lOS, 106 Basle Agreement (1988) 288
analogies 129 Bator, F.M. 154
Anderson, B. 128 Beer, S. 9,12
Anderson, C. 119 Benbassat.L 93
Andrews, K. 120,124 Benefits Agency 168,172
Ansoff, H.I. 119,120,124 Berry, AJ. 10-11,14,44,78,190,255-6,
Anthony, R.N. 7,13,17-19, 119, 124, 264-7,297
125,139,168,224,271 Beyer, J,M. 129
Aoki, M . 157, 158 Beynon, H. 304
appropriability 159 Bhimani, A 108, 236
Argyris, C. 88,89, 129, 130, 132-3,258, Birnberg, J,G. 185, 187-9
301 Blau, P.M. 179
Armstrong, P. 94,276,303,304,305, Boeing 57
306-7,309,311 bonus schemes 90,91,113,250-1,280,
Arnold, J. 78,86 315
Ashby, E. 10,48 bottlenecks 313
Ashley Smith, D. 164 Boulding, K.E. 12
Ashton, D. 226 boundaries 4,160-1
Association of British Travel Agents country 141
(ABTA) 67 legal and institutional 140
Athos, AG. 194 ownership 142
Atkinson, J. 310 social and cultural 140,151
atmosphere 33 technology 143
attention directing 77,85 bounded rationality 33,34,128, 161,252
audit, internal 101, 103 Bourn,M. 25
Audit Commission 165,167, 172, 174, Bower, J.L. 105
176 Braybrooke, D. 54
Australia 198 British Airways 240
authority 159 Broadbent, J. 38,69,273,275,279-80
charismatic 70 Bromwich, M. 108,218,236,290
rational-legal 29 Brownell, P. 89
role and function 160 Bruns, W.J, 89,136,226
traditional 70 Brunsson, N . 24
autonomy, versus interdependence Buckley, W. 15
47-8,56 budget 30
at individual level 264-5
Baiman, S. 33,165 as authorisation 85-6
Baird, J. 199 as cash limit 261,269
balance sheet 215 defined 86
Ballew, V. 34 incomprehensible 262, 267, 268
banks 283-98 Japanese companies 231
activity-based costing (ABC) 3-4, 28, operational 126
290-3 participation 89,91
changing culture 295 profit 216-17
changing organisation 295-7 sales 215
control systems 284-5 role of 85,259
cost awareness 287-8 standard costs 87
diversification 288 strategic 126
352 Index

budget-continued Collinson, D. 319

top-down and bottom-up 87 Committee of Vice Chancellors and
zero-based 50,52,54-5 Principals 164
budgeting see programme planning and communicating and co-ordinating 86
budgetary systems (PPBS) Community Charge (poll tax) 68
budgeting process, ICA 243-5 complexity 41,44-57,144
Burawoy, M. 299,305 ,323 conflict 22, 132, 133, 145, 193,329
bureaucracy 45,51-2,180,206,300-1,305 contingency theory 22-3, 95, 284, 286--7,
see also hierarchy 293-4
Burns, T. 22,31,97,136,301 see also environment
Burrell, G. 14 contract networks 302-3
contracting out 38,144
Cadbury, Sir Adrian 309 contractual theory 155
Campbell, A. 96,98,102,114-17,127, control
158,204,253 and accountability 165-7
capital accounting 18-19,113
cost of 1O~ action 20,40
see also investment administrative 20
capital adequacy 284,288-9,293,295 anticipatory 10
capital asset pricing model (CAPM) 107 on behaviour 21
Capps, T. 68-9 budgetary 84-90
cartels 66 clan 25,35,40,157-8,180-1,274,303
Carter, N . 168 coercion 21
cash flow 215 collaboration 146,147, 149, 150, 152
cash limits competition 146, 150, 152
at individual level 264-5 cost 283-7
budgetas 261,269 and culture 186--9
NIiS 257,258,261,264-6,268-70 cybernetic 9-11
cells 318 direct and indirect 143
chains see networks in divisionalised structures 100-3
Chandler, A.D. 9~, 97, 120 economic assets 21
Chandler, R. 164,301 in economics 153-6
Charity Commissioners 63 embedded networks 143-8
Chartered Institute of Management and environment 286--7
Accounting (1986) 307 and equity 166--7
Chartered Institute of Public Finance and evaluation 277-8
Accountancy (CIPFA) 164 of expenditure 262
Checkland, P. 9, 13,14, 16 financial 100, 115-16, 148-9,228
Choi, F. 232 formal 190,196,316--17
Chow, C.W. 185 implementation 125
Chua, W.F. 7,19,78,86,326 informal 317-20
citizen, power of 167 information for 83-4
Citizen's Charter 164-5,167 interactive 137,196--7
clans 25,149-51,157,158,180-1 ,303 internal and external 61-6
Clark, B.R. 182,192 market-based 258, 273, 276--7
Clark, K. 225 normative values 21
co-ordination and objectives 287
and unity 305 operational 7
vertical 49,50 output 40
Coase, R. 32,156 personnel 20, 40
coercion 71 political and managerial 167-8
cognition 128-31 positive and financial 254
Cohen, M. 135 processes 17-18
 Index 353

control-eontinued self-perpetuating 199

progranrrnrred 137,196-7 currency risk management 218-19
quality 254 cybernetics 119,125,155,169
relevance 125 concept of control 9-11
requirements for effective 285-6 contribution to theory 15
results 20,40 defined 9
self- 20,23,35-6 see also feedback
separated from decision-making 41, Cyert, RM. 5,19,57,.89
56,59,86
social 20-2, 35-6 Dalton, M. 179
and strategy 7,113,116-17,119-38 Daniel 203
structural 27-40,100 Deal, T.E. 186
and surveillance 305 Dearden, J. 91,109
systematic and systemic 154 decentralisation 30-1,205
task 17-18,27-8,30 ambiguous 99
see also domination: power conditions for 117
controller 4 and environmental complexity 95
Cook, P. 164 functional and federal 96
Cool, W.R 103 geographical 210
Cooper, D. 133,284 to embedded network 144-5
Cooper, R 312 and uncertainty 137
core business, weakness 253 decision-making process
cost accounting 243-4 and market information 213
Japanese companies 231 separate from control 41,56,59
limited value 55 delayering 311
standard 87-8 demand management 239
technical problems 227 Demirag, I.S. 203,214,217
cost centres 92, 313, 315 Demsetz, H. 159
cost control 283-7 Dent, J.F. 25, 118, 136, 137-8, 184,203,
costs 287,293,294,295
fixed and variable 79-80 dependence, versus autonomy 47-8,56
measurement 292 deregulation 284,287-8,289-90, 293,
reduction 229,230-1,234,245 295
reporting systems 245-7 Dermer, J.D. 96,118,119,127,133,134,
transaction 32-3, 35 135
crisis management 197-9 Dess, G.G. 124
Critical Theory 25 Dexter, A.5. 93
Crocker, J. 128 differentiation 96
Cubbin, J.5. 177 discounted cash flow (DCF) 83,104,
culture 23-5,303 106-7,229
changing 195-6,293,294-5,297 disequilibrium 154
and control 186-9 divisionalised structures 209-10
definition 23 control in 100-3
embedding and transmitting 196-9 counterproductive 99-100
emergent view 24 division autonomy 101, 102
and environmental perception 187-8 division interdependence 101-2
national 185 division size 101
organisational 179-200; described environment 101
181-3; nature inferred 189-92; information flow 101,102-3
obstacle; to change 193; strong internal audit 101,103
192-4; subcultures 183-5 domination 5,68,145-8,150,152
as organisational variable 186 Dow, G. 159,160-1
role 68 Drucker, P. 20,86,277,316
354 Index

Drury, C. 78, 86, 104, 224 evaluation techniques, and performance

Du Point 97 82
Dugdale, D. 320 excellence 23-4
exchange rate changes 218
Eastern Europe 37 Ezzarnel, M. 25,94-6, 100, 102, 103, 226,
economic reason, inappropriate for 315
caring profession 278-9
economics, control in 153--6 Fama, E. 159,315
education service 37,85,164-5,271-82 Fayol, H . 28-9,300,324
accountability 275 feedback 155,19
control evaluation 277-8 first-order 41
delegation of management 274,275-{j negative 9-11
historical background 271-2 seealso cybernetics
inspection 272 Ferlie, E. 24
output measurement 273 Financial Initiative (FM!) 273
parents choice 277 financial measures 216
performance indicators 278-9 financial services sector 283-98
performance-related data 277 restructuring 283
strategic control 272 financial statements
testing 274 balance sheet 46
Edwardes, M. 310 cash flow 46
efficiency and effectiveness 5-7, 166, profit and loss account 46
224,238 sources and application of funds
Elgar, T. 310 statement 46
embeddedness 139,142,143-52 Firsirotu, M.E. 23, 186
emergent problems, and horizontal first order control loop 78
communications 157 Fiske, S.T. 128
emergent properties 12 Fitzgerald, L. 290
Emmanuel, CR. 23,78,85,91,98,99, Flamholtz, E.G. 186
286,314 flexibility 37-9
employment relations 310 flexible specialisation 308-9
empowerment 304-5,319,322 FMC Corporation 111
seealso power Follett, M.P. 301
Enthoven, A.C 36 Ford Motors 141,304
environment 15,22-3,61-73,94 Fordism 304
banks 287-8 forecasting 86-8, 89
and control 286-7 Forester, J. 145
and decentralisation 95,97 formula funding 274-5,276
divisionalised structures 101 Foster, G. 223-5
interaction with organisations 71-2 Fox, A. 132,303
and management control 293-4 franchising 144-5
new manufacturing 299,308-11, Freeman, J.H. 22, 62
321-2 Friedman, A. 318
and organisational change 295 Friesen, P.H . 130,295
perception of 128-9,187-8
seealso contingency theory Gagliardi, P. 181,192,193
equity 166-7 Galloway, D. 313
errors 11,58 gaming 174
Etzioni, A. 8,21,47,70,180 Ganley, J.A. 177
Eurocorp 136-8 Geertz, C 179
European Community 49,65,211-12, General Electric 112
240,296 General Motors Corporation 29, 97
 Index 355

General Systems Theory (GST) 9,11-15 Hilton, K. 96,315

Giddens, A. 62, 68 Hiramatsu, K. 232
Gioia, D. 129 Hirst, M.K. 92
Glaister, K. 122 Hofer, CW. 120
goals 5,19,49,57 Hofstede, G. 9,47,88-9
cultural and symbolic 7 Hogler 302
individual 78 Hood, C 273
local 47 Hope, T. 78,86
multiple 106, 108-9 Hopper, T. 94, 100,297,306,309,311 ,
organisational 78 319
see also action plan; targets Hopwood, AG. 20, 26, 90, 92, 118,225,
Coffman, E. 68 237,256,280,302,314,316
Goldratt, E. 80,312-13 horizontal communications 157
Goold, M. 96,98,102,114-17,118,119, Horngren, CT. 86, 105,223-5,315
127,158,204,253 Horovitz.j.H , 127
Gordon, L.A 286 Hoskin, K.W. 300,317
Gorz, A 278-9 Hrebiniak, L.G. 126
governors, board of 168, 272, 275 Hunt 302
Govindarajan, V. 92, 113,203 Hurst, E.G. 125, 126
Grabner, G. 139, 142 Hyman, R. 128,310
Gray, J. 171,223-5
Gray, R 63 IBM 185
Greer, P. 168 ICA (Inter Continental Airlines) 241-54
Grinyer, J. 107, 130 bonus scheme 250-1
groupthink 132,195 budgeting process 243-5
Gulick, L.H . 301 competition 254
Gupta, AK. 113 cost management practice 245-7
Guthrie, J. 273 cost and profit responsibility 243
cost reporting systems 245-7
Habermas, J. 25,62,69 cost saving 245
Haka, S. 82 key result areas (KRAs) 249-51
Hall, W.K. 112-13,117 management account systems
Hambrick, D.C 203 evaluated 251-2
Hammond, T. 189 management structure 241-2
Handy, C 303 monthly accounts 247-8
Hannan, M.T. 22,62 operating schedules 244
Harrison, R 44-6 profit index 248-9
Hart, H. 100,102,103 responsibility accounting 242-3
Hatherly, D.J. 319 revenue and cost centres 242
Hawthorne Effect 29 route profitability 248-9
Hayes, RS. 223 sales function 242-3
headteacher, role of 275 idealisation 192
Hedburg, B. 295 ideologies, weak and strong 24
Hertz, D.B. 82 Imai, K. 64
Heskett, J.L. 193 incentive schemes 90,91,113,199,219,
Hewlett Packard 182-3,193 250-1,280,315
hierarchy 12,28-9,300-2 income statements 215
education service 271-2,273 incrementalism 54
move to markets 273 individual, versus society 68
versus market 32-6, 64, 95, 97, inertia theory 295
149-50,153,155-6,160-1,226 information 33, 213
seealso bureaucracy for control 83-4
356 Index

information--<:ontinued Jensen, M. 155-6, 157, 158, 159,315

divisionalised structures 101, 102-3 job design 310
flow reduction 99 . Johnson 304
Innes, J. 284 Johnson & Johnson 136-8
Inoue, S. 218 Johnson, G. 119,124,130
Institute of Cost and Management Johnson, H.T. 55,84,94,97, 109,226-8,
Accountants (972) 307 284,307-8,312,317,322
Institute of Cost and Works Accountants joint ventures 38
(lCWA) 306-7 Jones, T.e. 149,306,310,320,321
integration 42, 57 Jonsson, S. 295
intelligibility 234 Joyce, W.P. 126
interactive management control 117-18 judgement, subjective 13-14
interest groups 133,134 just-in-time OIT) 195,309,312
internal rate of return (lRR) 81
International Air Transport Association Kahn, RL. 136
(lATA) 240,244 Kainen, T.L. 107
interpretive analysis 59 kaizen 231-3 ,310
investment Kaplan, RS. 55, 84, 94, 96, 97, 103,
decision-making 320 109-12,136,176,225,226-8,284,
erroneous rejection 105 307-8 ,312,317
Japanese companies 213-14,229-30 Katz, D. 136
long-term 39 Kelly, G.A. 128, 255, 310
see also capital Kenis, L 91
investment centres 314-15,319 Kennedy, AA 186
Itami, K. 64 Kenwood 221,230,236
Ivanevitch, J. 91 Kerr, S. 174
Keynes, J.M. 154
Jackson, P. 169 Khandawalla, P.N. 286
[alland, RM. 48,59 King, P.78, 105
Janis, LL. 132,195 Knights, D. 319
Japan 58,94,150,157,189,193-4,309 Kornai, J. 174
Japanese companies Kotter, J.P. 193 ·
accountants 232-3
approach to MA 221-37 labour
budgeting 231 productivity 234
case study 204-5 share of costs 224
cost accounting 231 seealso workers
currency: choice 217-18; risk man- labour process school 159
agement 218-19 Langfield-Smith, K. 203, 303, 326
European co-ordination centres Langlois, A 158
210-12 Laughlin, RJ. 25,63,69,275,280,
financial control 228 295-7
investment decisions 213-14,229-30 learning 11,58
management responsibilities 205-7 barriers to 133
organisational forms 205-7 double-loop 129, 130, 137
performance measures 214-19,228-9 organisational 138
in UK 203-20 single-loop 129, 130
use of MA 221-37 legislation 63-5
Value Engineering for the Customer failure 68
(VEC) 205 seealso Acts
Jaworski, B.J. 174 legitimacy 70
Jayson, S. 304 Leijonhufvud, A. 154
Jelinek, M.L. 186 Lerwick, L. 301
 Index 357

Levi-Strauss, C. 179 management control

Lewin, A.Y. 89 an interactive process 293-4
life world values 69,280 choice of style 253
Lilienfeld, R 14 criticised 293-7
Lillis, A.M. 105 defined 3-4,253,285
Lindblom, C.E. 54, 64 multi-disciplinary approach 324-6
Lindkvist, L. 181 and strategy 294
Llewellyn, D. 289 systems (MCS) 14-16
Lloyds Bank 283,287, 288, 289 managers
local government 165 changing view of 319-20
Local Government Comparative Statistics cognitive style 93
164 responsibilities 205-7
local management of schools (LMS) role 28
271-82 Mann, L. 132
Locke, EA 88 manufacturing, separate from sales
Loft, A. 96, 97 212-13,219
Logan, J.P. 120 March, J.G. 5,19,33,57,133,135
Lorange, P. 125, 126, 127 Marglin, SA 160
Lorsch, j.w. 95,97,101,102 market 180,305
Lowe,E.A. 7,11,19,89,125,326 failure 279
Lucas,R.G. 133,134 ideal 32
Lufthansa 240 information 213
Lundin, R.A. 295 internal 52, 54, 302-3
Lyne, S. 313 move from hierarchy 273
position 215
McCosh, A. 15,254 pseudo- 315,319
McGregor, D. 301 in the public sector 36-7,38
McGregor,1. 310 versus hierarchy 32--6,64,95,97,
McGuire, A. 257 149-50, 153, 155--6, 160-1
Machin, J.L. 124 Marquis 304
Maciariello, }.A. 15 Martin.], 184
Macintosh, N .B. 93 Marx, K. 68, 305
McKee, L. 24 materials and components 224
McNamara, R 50-1,53 matrix management 31,45,51-2,54,
Maeve, RH. 300,317 151,205,210
Malinowski, B. 179 Matsushita 193-4
Malmberg, A. 318 Maxwell, R. 71
management Mayo, E. 301
by exception 77 Mead,M. 179
contradictory roles 305-6 meaning 68
reorganisation 311 measure fixation 173-4
structure 241-2 Meckling, W. 155--6,157,159,315
styles and philosophies 114-18 Meek, V.L. 181-2
top-down 258 Merchant, K. 20-1,26,40,92
training 206-9 Meyer, S. 304
management accounting 78-84 Michael, D. 57-8
critique of 222-8 Middaugh, J.K. 286
evaluation of 251-2 Midland Bank 287
irrelevant reforms 226-7 Milani, K. 91
the literature 225-6 Miles, RE. 98, 113, 122
long-term decisions 80-3 Miller, A. 124
short term decisions 79-8 Miller, D. 130, 131,286,295
versus engineering theses 223-5 Miller, E.}. 146
358 Index

Miller, P. 317 Norton, D.P. 96, 103, 109-12, 176

Mintzberg, H. 118,119,120,134,136-7,
295 objectives, and control systems 287
misinterpretations 174 OFFER 67
misrepresentation 174 OFCAS 67
mission statement 191-2 OFTEL 67
Mitchell, F. 284 OFWAT 67
Monden, Y. 222,228 O'Leary, T. 317
monetarism 154 Olsen, J. 135
money, time value of 80 OPEC 66
Monte Carlo simulation 82 opportunism 33,160-1
morality 68 organisational structures
Morden, T. 124 in growth and complexity 96-8
Morgan, C. 11, 14,45, 131, 132, 182, 183, J-firm 157
193 M-form 34-5,50,90,95-118
Morrison, I. 288 mechanic and organic 96
motivation 85, 88-909 U-form 50,96, 131-2
multinational companies 185 organisational structures, and technology
Munro, R.J.B. 319 96
Murdock, R. 310 organisations
Murphy, D. 127 boundaries 140-1
Myddleton, D. 224 change 295-7, models of 296
myopia 173 coercive 8, 47
myths and stories 182-3 control a universal activity 324
defensive routines 132-3
National Audit Office 167 instrumental 47
National Coal Board 190-1 interaction with environment 71-2
national curriculum 274,276,277 labour-managed 157
National Health Service (NHS) 25,36-7, learning 131,196-9
64,164,166-8,174,176,255-8 mechanistic 22
budgets 259-62,264-5,267-8 normative 8, 47
cash limits 257,258 ,261 ,264-6, organic 22
268-70 person-centred 45-6
clinical directorates 260 pluralist view 132, 133, 134
Diagnosis Related Croups (DRCs) 88 power-centred 45,46
financial data and responsibility 263 processes and objectives 169,170
financial training 262 purposive 5-7
operation plans 270 as social structures 52, 328
reason for reforms 256-8 utilitarian 8
ward sisters 266 virtual 150
ward unit managers (WUMs) 258-70 ossification 175
National Westminster Bank 287 Otley, D.T. 10-11,44,47,78,89,92,99,
natural selection theories 158-9 255-6,264-7,286,314
NEC 221,229,232,233,235 Ouchi, W.C. 21,25,35,37-8,70, 149,
Nelson, R. 159 156,158,161,175,180-1,187,189,
Net Present Value (NPV) 81-3 271,300,302,303,305,326
networks output measurement 35, 38
contract 302-3 overhead allocation techniques 216
embedded 139,142,143-52 ownership, theory of 159-60
Newman, W.H. 120
Nissan 221, 230, 233, 234 Paine, F. 119
Northcott, D. 82 Palmer, B. 169
 Index 359

Pan Am 241 at individual level 264-5

paradigms 130-1,134-5 compact 48, 59
Parker, L.D. 108-9 complexity critique 47-8
Parsons, T. 42,45 dialogue 49
participatory or bargaining models financial 46-7
157-8 strategic 17-18,114-15,209-10,253
Pascale, R.T. 131, 137, 194 see also programme planning and bud-
Pask, G. 9 getary systems (PPBS)
payback 81,235 Porter, M.E. 120
discounted 82 power 21-2,133,159,329
three year 229,236 changing distribution 51
payment by results 28 citizens' 167
Payment Services Group 291 coercive 21
perception 123, 128-9, 187-8 normative 21
seealso psychology renumerative 21
performance 98,315-16 see also control; empowerment
evaluation 82,85, 88-90 , 92 Preston, A.M. 189,294
seealso incentive schemes; targets pricing decisions 80
performance indicators 103-4 privatisation 329
benchmarking 177 privileges, government-provided 64
benefits 169-72 product development 229-30
client satisfaction 176 production, small batch 233
costs 169,172-5 production model 168-9,171
exam results 278-9 production unit, contribution 80
mitigating dysfunctional consequences productive capacity 145,224
175-7 professional associations 66
public sector 163-78 profit 312-13
reasons for variation in 169,171 budgeting 216-17
roles of 165-9 contribution target 230
sensing and response mechanisms and design 205
173 index 248-9
performance measures maximisation 98, 100
currency choice 217-18 as performance measure 104
discounted cash flow (DCF) 104,106-7 targets 86
Japanese companies 214-19,228-9 profit centres 92,211 ,314-15,318,319
labour-centred 232 programme planning and budgetary
problems with 91 systems (PPBS) 50-2, 53-4
profit 104,216-17 psychology 19, 255
quantitative and qualitative 108-9 seealso perception
short-term 106, 108 public sector
and strategy 108 markets in 36-7,38
seealso residual income (RI); return on new public management model 273
investment (ROI) output indicators 161
Perrin, J. 257 performance indicators 163-78
Perrow, C. 19,97,301 performance measurement 169,170-1
Peters, T.J. 23,24,37-8,50, 156, 183, 186 see also education service; National
Pettigrew, A.M . 24,119,134,135,181, Health Service (NHS)
192 purchaser/provider relationship 168
Pfeffer, J. 19,130,182 Purdy, D.E. 37,256,257, 260
Pike, R. 82 Putterman, L. 158,159--60
Piore, M.J. 308 putting-out debate 160
planning 41,42-4, 86-8 Puxty, T. 19, 125,326
360 Index

quality control 246-7 Scholes, K. 124

quangos 67 Schon, D.A. 57, 120-30
Quinn, J.J. 119,120,127,134-5 Schonberger, R 222
Schreyogg, G. 127
Radclife-Brown, AR 179 Schwenk, e. 119,134
rationality scientific management 28,87,317-18
bounded 33-4,128,161,252 scorecard 77,85,90-3
formal 302 balanced 109-11
substantive 303 Scott, B.R 97-8,102,116
recipe knowledge 182 Seal, W.B. 158
reciprocity 180, 181 self-control 20, 23, 35-6
Reece, J.5. 103 self-interest 158
reflexive law 72 self-organisation 135
regulation self-regulations 9
EC 65 Senge, P. 145
external 63-7 Sephton, M. 291
non-governmental 65-7 set-up and changeover times 224,311
why it works 68-70 Shah, E. 310
see also Acts; legislation Shaiken, H . 309
Reitsperger 203 Shapiro, S.P. 302
residual income (RD 90-1,104,105-6, Shaw, RW. 89
314-15 Sheldon, A 130
resource allocation, efficient 98-9 Shillinglaw, G.D . 104
Resource Allocation Working Party Siehl, e. 184
(RAWP) 257 Simmonds, K. 108
responsibility, and accounting data Simon, H.A. 7, 33, 47, 77, 128, 133
242-3,263,265-7 Simons, R 112-13,118,136-8,196,198,
return of capital employed (ROCE) 83 203,287,294
return on investment (ROn 90-1,104-5, Sims, H. 129
215,312-13 Sinclair, A 199
Ricketts, D. 223-5 Skinner, n.c. 103,104
Roberts, J. 99 Sloan, AP. 30
Robinson, J. 257-8 Smircich, L. 23,186
Ronen, J. 174 Smith, A 154
Rose,M. 29 Smith, e. 309
Rousseau J.J. 48 Smith, P. 164, 171, 173,285-6
Roy, D.F. 179 Smow, e.e. 98
rules 180,300-2,305 Smyth, P.S. 13
Snodgrass, e. 185, 187-9
Sabel, c.r. 308 Snow, e.e. 113, 122
Sadan, S. 174 social structures 42,57-8
Salaman, G. 303, 304 stability 52, 54
Salanick, c.a 19 social typifications 182
sales, separate from manufacturing socialisation 23
212-13,219 society, versus individual 68
satisficing 7 Solomons, D. 103-4, 105-6
Scapens, RW. 99,225,226 Sony 221,230-1,232,234,235
Schein, E.H. 181,189,192,193,196 Soviet Union 174,178
schemata 128-9 SPC Fruit Canning Company 197-9
Schendel, D. 120 Spender, J.e. 130
Schiff, M. 89 spending limits 85-6
Schoderbek, P.P. 15 Spicer, B.H. 34, 291
 Index 361

Sproull, L.S. 182 Sundem, G. 223-4

stability 39 survival 7
and change 145 systems
Stacey, R.D. 121-4 adaptive 136
Stalker, G. 22,31, 97, 136, 301 boundaries 4
Starbuck, W.H. 295 hard 13-14
Starkey, K. 37 maintaining 136
Steers, R.M. 91 open 15
Steinmann, H. 127 root definition 13
Stewart, J.D. 165 of society 69
Stock Exchange Regulations 66 soft 9,13-14
stocks 223,233
Strategic Business Units (SBUs) 112-13, targets 86,169,172,219
117 cost 215-16
strategic change process 134-5 financial 216-17; versus physical
strategic control 7,113,116-17,119-38 231-2
education service 272 setting and monitoring 315
theory / practice gap 127 see also goals
strategic cost management 56,108 Taylor, FW. 27-9,276,300
strategic decision 95 Taylor, S.E. 128
strategic management, role of 136 technology 157,287,293,321
strategic momentum 130 advanced manufacturing (AMT) 309
strategic planning 17-18,114-15, Teubner, G. 72
209-10,253 Thatcher, M.H. 68,310
strategy theory of constraints (TOC) 312-13
business 112 Thode, S.P. 107
and control 113, 124-5,294; seealso Thompson, J.D. 5-6,16,97
strategic control Thompson, S. 156
conventional wisdom 123-7 throughput accounting (TA) 312-13
corporate 5, 112 Thwaites, D. 122
definitions 120 time horizon 43-4,79-83,146-7,209,
development of 210 214,254
discovery, choice and action 123 Tinker, T. 158
divisional 109 Tocher, K. 9
emergent 134, 135 Tolman, E.C. 128
evaluation and formulation 124 Tomkins, C. 103, 106
from organisational struggles 119 Tosi, H. 88
as a game 122-3 total quality management (TQM) 195,
implementation 125-6 312
long-term and short-term 126 Toyota 221,229,231,232,233
and perceptions 123 trade associations 67
and performance measures 108 trade unions 67
position and process 120-2 training 199
response to change 126-7 management 206-9
split/ fit 131, 137 NHS 262
Strong, P. 257-8 transaction, identification 161
structuration 68 transaction costs framework 155, 157-8,
structures 159-60,180,301,307
divisionalised 30-1 transorganisational workgroups 150
formal 189-90 Trist, E. 62, 67
suboptimisation 173 Trumpington, Baroness 166
subsidiarity 49-50,96 trust 303-5,320,322
362 Index

tunnel vision 173 Weber,~ . 29,70,300-1,302,303,305

TWA 241 Weick, K.E. 12-13,128-9,195
Tylecote, A. 203,214 Weiner,N. 9
Weitzman.M. 174
uncertainty 42,57-9,92,97,133, 137, Welch, J.B. 107
144-5 Western Electrical Company 29
United States 49, 96, Ill, 150, 189, 240, Whyte, W.F. 179
307,309 Wildavsky, A. 53-4, 256
Department of Agriculture 53 Williams, J. 222,223,231
Department of Defense 50-1 Williamson, O.E. 33,34,44-5, 95, 97, 98,
Urwick, L.F. 300 100, 102, 103, 104, 107,112, 118, 156,
159,301 ,324,326
values and beliefs 69-70,280,303-5,327 Wilson, R~.S . 78, 86
Van Gunsteren, H .R 47 Winter, S. 159
Vancil, RF. 103,315 Wolk, H . 223-5
variance analysis 313-14 Woodward, J. 31,97,301
Vickers, G. 11, 13, 14,59 work organisation 310
Virgin Airlines 240 workers
changing view of 317-19
wage decisions 219 view of themselves 319
Waldron, L. 312,313 seealso labour
Ward, T. 291 Wright, ~. 156
Waterhouse, J.H. 89 Wrong, D. 305
Waterman, RH. 23,24,37-8,50,156,
183, 186 Young, D.W. 168
Waters, J. 119, 120, 134 Young, S .~ . 174
Watts, RL. 315
Watzlawick, P. 195 Zimmerman, J.L. 315